|
|
LIN
IO-Slave/LIN Switch
Subscribe To Updates
LIN General Description
The continuous growth of electronic content in cars allows improvements in comfort, safety and fuel economy. Integrating these functions requires the application of multiple bus systems to guarantee that the applicable modules can communicate at the required speed and safety levels. The vehicle manufacturers have developed different bus network standards to minimize cost while maximizing the performance.
Melexis concentrates its efforts on development of all kinds of applications with LIN interface. The LIN bus is accepted on all new development platforms (local interconnect network) and focuses primarily on the comfort area. With LIN it is possible to affordably interconnect the steadily increasing number of comfort functions. Additionally LIN can be implemented in applications that were previously impossible to network due to the prohibitive incremental cost of the networking features.
New demands require new solutions. A pure IC development for smart LIN slaves now becomes a complete system development.
LIN Features and Benefits
- Application configurable uniROM
- Ready-to-use Firmware Implementation
- No customer software development needed
- Easy to configure I/O and LIN messages
LIN Environmental Information
This text will be replaced
LIN Frequently asked Questions
Programmable Hall Sensors
Question
Most capable analog engineers adapt pretty quickly to Melexis digital products. If you have specific axample of what you are trying to compare I am sure our applications and Field Applications engineers can provide a comparative study or assist in your setting up a bench test for your own comparison. A good example would be an analog Linear Hall IC. USe an example application note from allegro's A3506 or other analog Hall IC. Look at the example of a DC current sensor. With the analog part you need to adjust # of windings or external applifiers & trimmers to maintain an offset and gain arrangmeent that resuls in full output signal span depending on the input current sense range desired. using a Melexis 90251 it is a simple matter of reprogramming such characteristics to match each case.
For sensor signal conditioning the www.melexis.com website has many good application notes to help get you started. inaddition many evaluation boards and demo boards for specific roducts can help jump start your bench testing and trials. It would be helpful to know what you top priority is so that a specific product can be recomended as a first case study.
A good analogy is to think of replacing all those trimmers in your analog cicuits with digital bits controlled through a one time programming interface. It generally means you can more easily achieve the best performance even when your signal input varies from one unit to the next without a lot of "twiddling pots" and trial and error. Usually a process can be established to allow a PC to check the span, offset and thermal behavior , predict a set of digital compensation factors, prgram the sensor, then verify and iterate 1x or 2x until the desired performance is achieved.
Question
Thanks for the info. I agree the evaluation kit is a good idea for an old dog like me. I also read a bit of info on Melexis's sensor interface, and even though there is alot of digital talk, it seems there is an analog input and an analog output, which is what I am used to dealing with. I think this "out of the box" eval kit might be just the "bridge" I need to get me across the technology gap. A similar exrcise can be done for a strain guage pressure sensor with the MLX90308 compared to a series of instrumentation amplifier cirucits that would correct for the non-linearities of the pressure transducer over span and also due to temperature.
Linear Hall!
Question
Is it possibile to measure field up to 300mT with MLX90215EVA? And what about preprogrammed linear hall sensor MLX90215LVA-BC03(10mV/mT)?
Hall chip weight on plate sensing
Answer
If there is application when you need only two states (on/off) then you can use hall latch like US1881,2881,...
Then you have two possibilityes. You can use permanent magnet and adjust distance betwen hall switch and magnet, or you can made a coil but you need very accurate current source. In this case your distance can be constant and you adjust sensitivity with current!
The first option is cheaper. Current source can be made with a few transistors for second option.
The third option is that you took pre-programmed linear sensor like MLX90215LVA-BC03 (10mV/mT). In this case you must build external hysteresis circuit(one op-amp)and you can adjust it with two resistors(loop)! Here you can also use permanent magnet!
TH71101 Sensitivity
Answer
The LNA is just a cascode bipolar amplifier with very low feedback capacitance. The gain is set by the load impedance and the LNA current. I have never seen any instability in the application circuits, so I am not sure how high you can make it. It will be limited by the largest inductance you can tune with the approximately 2pF output capacitance of the LNA stage plus and capacitance coupling to the mixer. You will not get much improvement by making the gain much higher because the overall receiver noise figure is determined by the standard cascade noise figure equation. The current receiver overall NF is about 8dB. The single conversion receivers have virtually no image rejecton without a SAW filter on the input. If you use the TH71102 dual conversion receiver, it has about 5dB image rejection and slightly better sensitivity. The inage rejection can be increased to about 30dB by adding a double tuned circuit between the LNA and mixer
Frequency Deviation
Question
I wish to acieve as good a SNR as possible in an RF link. One way is I slow my baud rate down and limit the bandwidth of the receiver demodulator. What effect does changing the frequency deviation have? Currently I am using about +/-25kHz. Do I get any benefit from reducing this and using a narrower band IF filter? Any suggestions?
RFID
Question
Hello sir,
I am working on 13.56 Mhz RFID (MLX90121) project. I am little bit confuse on using symbol transmission method. T CK and Din lines are used to transmit the symbols, but I still not clear about the following:
1. Is that the width of the CK pulse be concerned?
2. Does the microcontroller need to provide the addresses of the symbols to be transmitted ?
3. Does the microcontroller send any neccessary signal through Din line during the transmission time(request)?
4. To transmit sym0, does the microcontroller need to send any pulse clock?
5. According to the MLX90121 data sheet(Mar/04), page 13 of 25, the three start bits are used to transmit sym0 to sym6. I really confuse because sym1, sym2 and sym4 all have one CK pulse what make the hardware encode the right symbol to transmit? is that the address involved to encode this symbols?
Question
Melexis is selling a development kit called DVK90121 which comes with the basic ISO commands.
I would recommend that you use this kit to help with the development of your application.
Here are the answers to your questions:
1. The maximum frequency of the CK pin is 1MHz (page 4),
2. Yes. The address should serialy given on the DIN pin using the rising edge of CK (page13),
3. See 2.
4. Sym0 can be encoded with only 1 bit on DIN, requiring only one CK pulse if you use only 2 symbols; or with 2 bit on DIN, requiring 2 CK pulses if you use up to 4 symbols; or with 3 bit on DIN, requiring 3 CK pulses, if you use up to 8 symbols (page13).
5. The initial 3 pulses on CK are used to intiate the transmission. Then you use as many CK pulses you need to encode the address of the symbol you intend to send out.
TH8082 Pullup
Answer
To switch off the internal pull up resistor you must enter the sleep mode (high->low transition on the EN pin). But in this mode LIN bus communication is not possible.
The TH8082 is designed according to the LIN specification Rev 1.3 and 2.0. In this specification the maximum of bus nodes is limited to 16. All pull up resistors and bus capacitors are designed for this maximum configuration.
Do you want to use our transceiver for a standard LIN network? If not, maybe it's better to use another transceiver from Melexis.
Answer
it's a little bit difficult to use special bus tansceivers for more then one field bus system. In my opinion it will be better to use two different transceivers maybe the TH3122 for K-Bus and the TH8082 for LIN bus. So if you place the pattern of both transceivers to your PCB and select the monitor type at the assembly of your module you will get 100% compatible monitor modules for both field bus systems.
If you can't do this the TH3122 is closer to your specification than the TH8082. But keep in mind that the TH3122 is a K-Bus transceiver and not 100% compatible to the LIN bus. For instance it's possible that you get problems if you want to sent LIN messages with the TH3122. For a 'monitor only' application it should work.
Question
I want to use the TH8082 LIN-Bus Transceiver for a customer project.
I'm not sure if I can meet the customer requirements, because of the following problem:
From the datasheet I know that there is an internal 30kOhm pullup resistor on the Bus line.
In my application I need to switch this pullup on and off.
Can I do that using the Enable pin?
If yes, what means a switched off 30k pullup? Does it mean the Bus pin is grounded?
Another question:
What do I need the external 1k pullup in a master node for?
Question
thanks for the information.
Actually I want to use it as a K-Bus transceiver to build a K-Bus monitor. But probably later the application should also be usable as LIN-Bus monitor.
Do you have a chip which can handle both?
TH7122 Adajcent Chan Selectivity
Question
What is the maximum achievable Adjacent channel selectivity in a narrowband application at 173MHz ? Although using single conversion with a low 455KHz IF (with 2 cascaded filters) would reduce coupling between pin 32 and pin 1, presumably the ACS is still limited by VCO phase noise. Is a 60dB ACS spec achievable ? What is the expected phase noise at +12.5/25KHz ?
EVB71102
Question
I have the EVB71102 and want to use it for wide band FM (200KHz) with good linearity (+-5%), what are the changes that I have to do at the EVB?
Answer
Make sure the discriminator phase shift capacitor between pins 15 and 16 is relatively large, i.e. 12pF and you may also use a small cap of about 2 to 4pF across the ceramic discriminator to lower the Q.
Alternatively you can use an LC discriminator tank. Then you have all the freedom to optimize the circuit in terms of lineraity by resistive loading of the tank.
EVB72032 Antenna Gain
Answer
The loop dimensions are already given in the EVB description: http://www.melexis.com/prodfiles/evb72032_rev006.pdf . Nevertheless there are some postings and links regarding this issue: "TH72012: need more power" & "TH72002".
TH7107 FSK frequency
Answer
the component values should be in between the values for the 315 and 434 MHz versions, but we don't have a tuned EVB for 369.5MHz. This is because we don't have quartz crystals for each non-standard frequency available. You should replace the existing crystal on the EVB (or your board) by a 11.54688MHz crystal. The rest is only fine tuning of the matching/antenna network and the pulling caps.
TH7122 matching network
Answer
For matching please refer to chapters 7 - 9 of the TH7122x cookbook.
The cookbook can be downloaded here for example:
http://www.melexis.com/relinfo_detail.asp_Q_id_E_374
Antenna for MLX90121
Answer
Thanks sharing to me the antenna design. I just like ask some questions regarding the design. Not all the values of components are given, does this mean that I have to assume values for the unknowns until I achieve the value of 50 ohms and Q factor of 14? Are capacitors for this antenna serve as coupling or what? I'm also wondering if the thickness of the antenna trace would have a significant effect in the antenna system? Please enlighten me on this matters.
RoHS
Answer
The plating of the US1881KUA is pure matte tin, so therefor it is ROHS compliant.
More info:
Melexis Lead Free Statement
Melexis Lead Free (Green Packaging) FAQ
TH72035 Parasitic Oscillation
Answer
the TH72035 contains an open-collector output stage. Therefore the output stage needs to be loaded at the desired operating frequency. Assuming the collector load as a RLC parallel tank, the residual load will be only resistively at resonance. Since the collector current is well controlled by the IC, the voltage swing at the collector (output) node is directly related to the real part of the load impedance. To avoid instability, the collector potential must not drop below the base potential of the output transistor, so that the output stage operates fairly in linear mode. For the TH7203X the real part of the optimum load impedance is approx. 250ohms which gives a stable operation over the whole supply range (1.9 .. 5.5V). For the TH7200X & TH7201X this is approx. 300ohms. If the lowest provided supply level guaranteed by the application can be e.g. 3V (instead of 1.9V), the available voltage swing at the output node will be higher so that the load resistance can be increased which res
MLX90121 antenna circuit example needed
Answer
The schematics of the EVB can be found in the associated documentation which can be downloaded from the MLX WEB.
We don't provide the schematics of the antenna on line, as we had the belief that it was too applicaion specific.
However, as we indeed have more and more requests for assistance in this area, we decided a couple of weeks ago to include the schematics of the antenna, as par of a general upgrade of our documentation.
As far as FCC/EN compliance, we have nothing that could be made publicly available. But again, if the market calls for it ...
Question
I need to design a board directly into our system. So getting a EVB does not help me a lot, it would just take me longer to get the task done. Couldn't you just put the diagram of the antenna section online? That would speed up things a lot.
Concerning EN & FCC, I know it's my duty. But as an example, your colleagues at Philips recommended in an AN an antenna circuit for their IC, so that we could easily pick up from there and knew what to do to make it complient. After all the information is available in your company, so why not communicate it to your users?
Convinced ? :-)
MLX90109 read problem
Answer
First please check if the transponder replies by measuring with a scope the load modulation on the COIL pin.
Then place the transponder at a distance of 2-3 centimeters to avoid strong coupling which may disturb the communication.
To finish, please have a look to the configuration of the chip (SPEED, MODU and MODE pins) and be sure that the Clock and Data signals are externally pulled-up.
audio and data transmition in one th7122
Answer
This combination is possible. Please have a look at the cookbook:
http://www.melexis.com/relinfo_detail.asp_Q_id_E_374
TH7101.2 PLL noise sideband question
Question
We are having issues with Canada's reinterpretation of occupied bandwidth (OC) with this chip now 'failing' the 0.25% occupied bandwidth requirement at 433.92MHz (99% of TX power must fall within 1.08MHz or so). The people we are talking to are bent on getting rid of the whole mask/general limit go-no go gage and are simply summing the power from the transmitter to come up with the OC.
Modulation is AM-OOK and the spectrum distribution matches the curve found on your rev 8 datasheet, however the test stipulation is that you have to use the actual modulated pattern, not simply CW, causing a loss in the S/N and losing more margin on the OC calculation. Do you have any PLL filter or XTAL oscillator optimizations/changes to rectify this issue, or is the noise inherent in the divider/PFD architecture (right now the latter is my bet, in which case do you have a better chip available)?
While this doesn't affect our current product mix, new designs will have to deal with it.
evb90267 & mlx75007
Answer
i assume that you have installed the EVB normally and included the development libraries.
The driver (SYS and INF) is then installed and all you need to do is to copy the CCAPI.DLL and CCAPI.H files from the development directory to your project directory. Include the H file to your software to have access to all the high level functions. Those functions are described in the user's manual together with an example code in C.
If you link the DLL statically, we also provide a LIB file in the development directory.
MLX90217 AS CRANK SENSOR..
Answer
In principle, the MLX90217 could be used for an active crankshaft position sensor.
It will be able to deliver an electrical signal from the first-edge ("Zero-speed") up to 10000 RPM (i.e. 10 kHz input signal w/ Crank target 60-2 teeth) regardless of the direction.
You will not be able to meet the repeatibility specification required by the engine management control unit and strategy. This is linked to the working principle of the MLX90217 i.e. a fully digital track-and-hold circuitry.
For Crankshaft application, we recommend the MLX90254. This circuit is an AC-coupled differential Hall sensor. It can be used together w/ a back-bias magnet at the front of a ferromagnetic toothwheel. It can also be used at the front of a magnetic wheel.
The AC-coupled differential circuitry allows the MLX90254 to meet extremely tight jitter specification compatible w/ the engine management requirements.
However, it is not a zero-speed sensor.
TH7122 Standby mode
Answer
indeed that sounds strange. The difference in current consumption between RX and Standby mode is approx. the target operating current in RX mode (8mA). Do you share some control signals with another IC (e.g. micro controller) which operates on a different supply level? Is there a slight, linear or exponential increase in current consumption when you change the supply level, e.g. from 3.3 to 5V?
DC motor IC:PWM out and Hall inputs
Question
I saw a data sheet for a motor controller with an internal microcontroller and power supply and LIN communication. Also had Hall sensor inputs. I can't find it on the site. Anybody know this part?
Question
Melexis makes many circuits that are not on display at Product Selection Guide. There are only pure standard ASSP (Applicarion Specific Standard Circuits) We sell many ASICs that are not on display. Some times we make agrement to sell ASICs as ASSPs with some restrictive exclusivities from the customer who we made the ASIC for.
This circuit you have seen the datasheet from, you have to ask Melexis closest distributor (this link) http://www.melexis.com/contact.asp If they do not know about that circuit they will pass the question to one that knows about these circuits
There exist also data sheet for our pure ASICs and they are keept confidential from Melexis.
Please make a contact to our sales organisation and they will help you.
IR wavelength & Fresnel Lense
Answer
The wavelength range is from 6 to 15um.
We have never used fresnel lenses with good result. Be careful about stability and heating/cooling of the lens. You will have to recalibrate the sensor of course.
If you need a smaller FOV, I suggest you check our MLX90601KZA-CLA with a FOV of 40degrees. It has a PWM output instead of an analog though.
MLX10407
Answer
This is explained on page 8 of the datasheet.
The principle:
- binary value 0000000000 corresponds to 0 degree
- binary value 1111111111 corresponds to 89.8 degree
For other values you do a linear interpolation, f.e.:
- 45.5 degree is 1000000000
- 30.0 degree is 101010101
MLX90127 - Datasheet
Answer
I am looking for the datasheet for the MLX90127.
Also I need to purchase the following part number for my sample design.
MLX90109 - I can buy this online
MLX90124 - I can not buy this online
MLX90125 - I can buy this online,
I noted that the MLX90124 is the transponder, which I need to make it work with MLX90109 & MLX 90125.
Answer
Thanks for your interest in our products. The MLX90124 and the MLX90127 were both discontinued last year..
You can indeed buy on line both the MLX90109 and the MLX90125. But I would recommend you buy the development kit DVK90109 which comes which suitable tags for the MLX90109.
MLX90603 datasheet?
Question
Hello
Could you please provide me a link or send me the datasheet for the product MLX90603. I've been looking for it in this site without any luck. Thank you for your time.
TH7122EVB Antenna
Answer
1) The EVB7122 contains a 50ohm SMA connector for RF in/output which can be used for a lambda/4 monopole antenna. This antenna is more efficient than a PCB loop antenna. A distance of 350m line-of-sight should be achievable, but it depends on several environmental conditions.
2) It should be ensured that the TX center frequency is equal to the RX center frequency. Of course also RX and TX should operate on the modulation scheme. Pls. also check the output power of the TH72011 in comparison to the MC33593. If the TH72011 antenna network is properly tuned there should be an improvement of approx. 3dB in comparison to the MC33593.
It is also important that the data rate selection bits (DR0,DR1) of the MC33593 are set according to the TX data rate. I don't know if it's necessary to send a kind of sync word to the RX in order to start-up data decoding - this should be checked by your side.
FM with all TH71xxx /TH72xxx
Answer
Since the FSK modulation in our transmitter IC's is done by quartz crystal pulling, it is possible to pull the crystal by means of an audio signal by using a varactor diode. The audio signal should be pre-filtered in order to limit the occupied bandwidth. A circuit example can be found in the TH7122x Cookbook pg.6 http://www.melexis.com/relinfofiles/AN7122x-Cookbook_rev-003.pdf which is essential the same circuit that can be used for the TH7201x. I would suggest to use the 3rd generation of our transmitters (TH720XX) because there are much more power efficient. The TH7107x transmitters are not recommended for new designs.
Answer
according to the data sheet the load capacitance of the quartz crystal should be 10 .. 15pF. That's because the equivalent capacitance offered by the oscillator is ~18pF (pin ROI). Since the CX is a series connection between ground and crystal, the capacitance seen by the quartz crystal can never be greater than 18pF. When you apply a crystal with a specified CL of 20pF the frequency will be above the specified. If a varactor is used, the total capacitance seen by the crystal is the series connection of CRO, CX and CVD. Pls. have a look on the data sheet for further information regarding crystal pulling.
I would suggest to use a CL of 12pF.
MLX90109 Question
Question
Now Why must I make the tag cling to the antenna to communicate.
another question:Can you customize the 100mm diameter antenna sample according the MLX90109?And pls provide the 100mm diameter antenna parameter for us.
TH7122 theory of operation
Question
1) All digital control signals of the TH7122 can be driven by any hardware which fulfills the specification of the TH7122. The hardware (algorithm) which evaluates the received signal from OUT_DTA must be able to extract the expected data from the stream.
2) The isolation between the RF output and the OUT_DTA is not that important because during receive the PA is normally off. Also the base band and the RF are fairly far off. If you intend to use an external RF amplifier (e.g. RF2117) a good isolation between the VCO tank of the TH7122 should be provided by the board. That includes separate VCC traces for the components with good de-coupling.
Question
The frequency range of the RF2117 is typically between 400 and 500 MHz. Do you know of external, signal and power, RF amplifiers that have been tested with the th7122 at approximately the 27 to 150 Mhz range?
Also, with respect to the statement:
"If you intend to use an external RF amplifier a good isolation between the VCO tank of the th7122 should be provided by the board. That includes separate Vcc traces for the components with good de-coupling..."
What exactly do you mean by the phrase, "That includes separate Vcc traces for the components with good de-coupling". What specific operating conditions did you have in mind which required separate Vcc traces and can you elaborate upon the term "de-coupling". The term, "de-coupling" has more than one meaning depending upon the circumstance to which it is applied. Can you define the term with respect to the components and circuit conditions?
Answer
I guess there are a lot of RF amplifiers available - probably the RF2114 is one of them. We don't have tested a solution containing an external amplifier.
The separate pcb traces means that every VCC (VCC_DIG, VCC_IF) of the IC as well as each building block which is supplied by VCC (Tank of the LNA, Tank of the VCO, Tank of the PA and possibly an external PA), should be routed as close as possible to the supply. If e.g. a long trace supplies the PA and in between the connection of the loop filter is located, it may happens that large current transients caused by the PA, produce a voltage drop over the trace which causes disturbances in the VCO control voltage (VLF). As a result the VCO contains more noise than normal.
De-coupling means that each VCC trace should contain a high Q capacitance (300pF .. 1nF) to ground as close as possible to the circuit. The ground (VEE) should be preferably a very conductive layer within the pcb (bottom side of a double laye
TH7122 operation frequency
Answer
The tuning ratio (fmax/fmin) of the VCO is approx. 1.135 at VCC=3V and 1.19 at VCC=5V. Using an additional hyper-abrupt varactor diode this ration can be increased up to 2. In your case the required ratio is 5. It's therefore impossible to operate in that huge frequency range without any change of the external tank circuitry. A Multi Band Switching circuit is explained in the TH7122 and TH71221 Cookbook http://www.melexis.com/relinfofiles/AN7122x-Cookbook_rev-003.pdf
Answer
Since the VCO operates on the same frequency as the intended RF, the VCO range has to cover the frequency band of interest plus the separation of the IF in RX mode. For RX low side injection (default) it would be fmin_vco + IF .. fmax_vco. For example with a tuning ratio of 1.1 and an IF of 10.7MHz, the frequency range of the TRX at a center of 915MHz band is 877 .. 953MHz. This can be achieved with a proper tuned tank circuit. If you intend to operate at another frequency band, the tank circuit has to be changed. This can be done by means of an band switching circuit mentioned in the cookbook.
For ASK operation the comparison frequency (fxtal/R) should be as high as possible (e.g. 1MHz) in order to allow a faster correction of disturbances caused by load pulling effects in the ASK-modulated PA. So the bandwidth of the loop filter is usually higher than for FSK. The higher comparison frequency means also that the channel spacing is higher since comparison frequency == c
I made a mistake...
If you only want to address a fixed frequency within a certain band you should make fxtal/R ~ 1MHz and the bandwidth of the loop filter 100...200kHz.
EVB71102 external Reference frequency
Answer
the C_RO is only a coupling cap. The real part of the RO input impedance is usually negative - it depends on the applied amplitude level. So there's no need for attention. The generator level should be adjusted to provide an amplitude of approx. 200mV.
TH7122 Antenna
Answer
pls. find at:
www.inf.fu-berlin.de/inst/ag-tech/scatterweb_net/ews/downloads/433_helicals_antenna.pdf and
www.inf.fu-berlin.de/inst/ag-tech/scatterweb_net/downloads/868MHz_helix.pdf
some examples of small helical antennas. Note that there are no simple numerical expressions for a small helical antenna (circumference << wavelength). For resonant helical antennas (circumference ~ wavelength) some design curves can be found at: http://www.poynting.co.za/tech_training/snec_designcurves.shtml.
Unfortunately we cannot give more support on this issue.
TH7122 VHF narrowband, 2-point modulation
Answer
When the channel spacing is 12.5kHz the max. data freq. should be 125..250Hz for xtal pulling. In that case the bandwidth of the PLL should be adjusted to 1.25kHz. This results in a minimum data freq. of 12.5kHz for direct VCO modulation. Your data freq. is DC..600/1200Hz. Therefore it is necessary to reduce the PLL bandwidth so that significant spectrum contents can be applied directly to the VCO. For determining the PLL bandwidth the average PSD of the modulation signal should be analyzed. It should be located in a minimum side-lobe of the spectrum (if exists). Then the 3dB LP/HP frequency should be 10 times lower/higher. You can try to decrease this factor stepwise (maybe also non-symmetrically). For the quartz crystal modulation I would suggest to use a varactor diode as depicted in the TH7122 Cookbook.
Answer
the two-point modulation approach is quite sophisticated. We don't have a solution ready. Pls. check if it's sufficient to have a modulation frequency from DC to 10kHz. In that case the normal "crystal pulling" approach can be used.
TH7122 OUT-PA bias voltage
Answer
the best way would be to have a battery of e.g. 4.8V (2x 2.4V). If this is not possible I would suggest to use a LDO regulator to provide a supply level of 5V. Also it's better to have the whole IC at this level, because the VCO tuning range is better because of more headroom for the VCO control voltage (VLF). I've also doubt concerning switching regulators, but it could be tested using an efficient de-coupling.
TH8061 as LIN master
Question
I am inquiring if there are any reasons I cannot use an TH8061 as a LIN master node, provided I apply the correct termination. Please advise.
Answer
The TH8061 LIN transceiver can be used for master and slave nodes. If you are using it for a LIN slave node you don't need any additional external termination, because of the integrated 30kOhm slave termination.
If you want to use it for a LIN master node a additional master termination to Vbat consisting of a diode and a serial resistor of 1kOhm is necessary.
Source for the EVB7122X-433-FSK-C?
Question
Can you tell me where I can buy the EVB7122X-433-FSK-C evaluation board? I have not been able to locate it from distributors or the online store here.
MELEXIS 90217 Operating Bias
Answer
If the Rev. CA is working OK, then the only reason to explain why Rev. CC is not working should be linked to the hysteresis... It means that your application shows a marginal field modulation. Still OK for the typical CA but not good for the typical CC and therefore not good w/ the worst case CA/CC at 10 mT.
Questions
=========
=> Are you sure about your 1.28 T ? How have you measured this ? At the magnet surface or at the operating airgap Magnet/IC ?
=> Have you made a mapping w/ a Gaussmeter to evaluate the modulation while the ferromagnetic target is running... What value do you have there at maximum airgap ?
MLX90308 D High Temp linearity
Question
Has anyone noticed a change in the high temperature performance of the new version D of the MLX90308/90314? I'm observing a non-linearity in the low range (0-10% of full scale), at temperatures 110C - 140C. I'm not seeing this with our older "CA" chips. The sensors we use aren't the source of it (I checked). If I'm the only one with this problem - maybe I should reevaluate my design. Do you have any thoughts on what pins might be sensitive to component selection (sensor ~1-2kohm, 0.1 to 4.7uF bypass capacitors, ESD protection, etc.)?
I am in absolute mode. The output load is a 0.1uf cap, 11V zener diode, and DVM in parallel, all to ground. I used the same application circuit for CA and DA, but now I've removed the 5k resistor recommended in your application circuit - this helped remove some high-temperature non-linearity.
Answer
If you keep a PD of about 10k and that you calibrate the output offset to a value higher than 150mV you should not have linearity problems at high temperatures.
MLX90217 Short Pulses
Answer
If I read you correctly, the target you are trying to sense is a small tab of 3 mm width linked to a rotating shaft. What is the width of the so-called valley and what is its depth ?
It would be good if you could provide the mechanical profile of your target w/ all the dimensions. The phenomenon you describe is of course strange and is likely linked to how and in what range the magnetic flux density is modulated. It can be either linked to the magnet or the target. It would be good if you could map this w/ a gaussmeter placed nearly at the same position of the IC and the target rotating (measurement step of 5 or 10 degrees).
If you work in the improper conditions, everything can happen such as some marginal effects over temperature,... That is may be what you see w/ the fact that the part operate first OK then start losing its function...
PS: Have you check the polarity of your magnet ? At what supply voltage do you work?
Programming the MLX90215 manually
Answer
Is it for evaluation in the lab or to be used in a real application around a micro-controller (for example)?
In the first case, you should buy a MLX programming unit (e.g. PTC-04). It is delivered w/ all the necessary tools to program this part as well as other MLX programmable Hall sensors. It is a quite universal tool.
In the latter case, Melexis does not recommend this mode of operation as it was not the intention to let the device work permanently in RAM. This RAM mode is only for calibration purpose in conjunction w/ the programming unit i.e. in a well under-controlled environment. There is not enough check to guarantee this mode in a not-controlled environment (e.g. EMI perturbations,...). Furthermore, the programming protocol is based on "exotic" waveforms on the supply line (up to 15 V) and the output line. You will need to add dedicated H/W to make this happen.
In summary, we do not disclose the programming protocol for the MLX90215 and we recommend to use the ptc-04 in all cases.
Question
It's for a real application in a lab :-)
I'm part of a group of physics students spatially profiling magnetic fields by scanning them with a Hall probe, and we chose this device because
a) the position and the size of the Hall plate is specified
b) we were hoping to be able to reprogram the sensitivity/range with the device in our apparatus
Unfortunately our budget is quite limited. We were prepared to build a small amount of dedicated hardware to facilatate the programming, but if it is only possible with the PTC-04, I guess we will be using a selection of pre-programmed devices.
TH8080 trouble
Question
i am experimenting with a couple of LIN transceivers, TH8080 and TH8061. The TH8061 works all as expected but with the TH8080 i have trouble with the RX pin, it doesn't respond at all.
i have Tx and the LIN bus up and running but no sign of live from the Rx pin.
it is neither GND, 5V or 12V but seems to be floating.
I have tried to exchange the with another TH8080 (i have a few) but it is still the same result.
i have double checked my wiring and it all seems correct to me.
any ideas on what the problem could be?
Question
i have tested my circuit with the pull up and it all works perfectly now, so thanx again.
the application circuit in the TH8061 datasheet (page 26) shows a slave-node with TH8061 and a master-node with a TH8080. anyway there is no pull-up in the TH8080 circuit. maybe someone should point that out to the author of the datasheet.
i have a general question about LIN. It seems like most LIN developers prefer the 9600 baudrate dispite the loss in response time. what is the main reason for that?
TH71221 problem!
Answer
as indicated in the data sheet, the VCO tank must be adjusted to the target frequency of operation. Since the VCO in your application runs around 250MHz the PLL will never lock at 138MHz because the PLL settings has to fit into the VCO tuning range. When you measure the voltage level at the LF pin you will find a permanent high level. That indicates that the PLL try to lower the frequency but without success because of the de-tuned VCO tank. You should increase the inductance of the tank to lower frequency. For further informations pls. have also a look in the TH7122 cookbook which can be found at: http://www.melexis.com/relinfofiles/AN7122x-Cookbook_rev004.pdf
About MLX90320
Question
Sorry if this sounds like a dumb idea, but if I understand your setup correctly, you need to move the sensor (and calibration equipment) to another location within your building to complete your calibration. Is this correct?
Could you put your programmer and associated equipment on a rolling cart? If so, you could use a computer UPS (Uninterruptable Power Supply) to keep the equipment operational while you move the cart. UPS systems are fairly cheap and provide AC line filtering as a bonus.
Just plug everything in with the UPS in-line and start your calibration, then pull the AC plug and roll everything to the high temperature place, plug the AC cord back in, and you should be able to complete your calibration.
MLX90316 Sensor Specs and Availability
Answer
MLX90316 is commercially available and you can get it through our sales network. If you want to know the nearest one from your location, please browse to the following link:
http://www.melexis.com/contact.asp
Please refer to the datasheet to get an overview of the working principle of this Triaxis IC:
http://www.melexis.com/prodmain.asp?search=mlx90316&family=MLX90316
If the Integrated Magneto Concentrator (IMC) is not saturated, it will not introduce any significant non-linearity error. However, if you saturate it, the output of the IC will show a large non-linearity error (the signature of this error is 4 periods over a full 360 degrees revolution).
The saturation starts above 70 mT. In other words, you need to make sure the magnetic flux density at the IC level is less than 70 mT to avoid saturation effect.
th7122 SCI and comp. interfacing
Question
The reason why I had posed the question to you was because the "wording" in the data sheet is not clear. A definite reply to my question will leave no doubts as to how the th7122 inputs data.
If I send 32 bits in some bits will be ignored if I have a 24 bit register. The 32 bits which I am sending in, data stream from the parallel port line to the SDTA line of the th7122, has an equivalent hex value of C000F35C (C=1100, F=1111, etc.) The rightmost C (=1100) is going in first, then the 5 and then the 3 and so on till the last leftmost bit goes in. My question is how will the th7122 input the data? Which register will be loaded, what data will be be contained in the register. If you sent a reply with 22 bits of data (and the name of the register loaded) in zeroes and ones (22 zeroes or ones) and labelled what the least significant bit(s) of your answer and most significant bit(s) of your answer will represent in the workings of the TH7122, there will no room for interpretation.
I re-read the question and I want to make some modifications so you will understand my question:
If I send 32 bits in, some bits will be ignored and some bits will be entered into the register since I have a 24 bit register. The 32 bits which I am sending in, serial data stream from the parallel port line to the SDTA line of the th7122, has an equivalent hex value of C000F35C (C=1100, F=1111, etc.). The rightmost C (=1100) is going in first, (the zeroes of the C=1100 go in first) then the 5H and then the 3H (H=hexadecimal value) and so on till the last leftmost bit goes in.
Answer
Just for imagination: The TH7122x contains a 24-bit shift register which shifts at the rising edge of SCLK. A rising edge at SDEN load one of the register words. The register word which will be loaded is defined by the first two bit's (23 & 22) out of 24 you have shifted in. With a rising edge of SDEN only the last 24 bit's which you have shifted in previously will be loaded. In your case: You have sent 32 bit's, so the first 8 bits doesn't have any effect (5C is ignored). The good thing here is that you can share the SDTA and SCLK line with some other devices - you only have to separate the SDEN.
mlx90255 max speed
Question
I use linear sensor MLX90255 in the SMD8 package. I was surprised to read in a recent datasheet (the Rev008 of november 05) that the max frequency is 500 KHz. The old SMD8 sensors (datasheet rev 006) are assumed to work at 1024 KHz! Why are performances of the new products weaker than the old sensors? And how will I do when I will have used all my old sensors. I really need this max speed of 1024KHz!
MLX90316 programming sequence
Answer
The best way to start programming the MLX90316 for your application is to use the user interface/ solver procedure.
Use the following sequence for a 2 point calibration:
- New device
- Set settings ( output , rotating direction,..)
- Characterize ( in case of analog output)
- Search RG
- Move the magnet to the first calibration point.
- Set point A ( angle point A= (360-100)/2=130deg; level point A = 10% )-> clamping low = 0..130; clamping high = 230..360 deg.
- Move the magnet 100 deg to the second calibration point.
- Set point B ( Level point B = 90%)
- Finish program cycle.
If you readback the EEPROM - Linear setup parameters, you will see the following results :
A_X = 130 deg; A_Y=10%, S0=A_S=(90-10)%/100deg=0.8%/deg;
B_X = 130+100deg; B_Y=90%; B_S= A_S
C_X= 360 deg ( not used)
TH7122 or TH71221 at 151MHz
Answer
frequency operations below 300MHz requires an external varactor. Please find related informations in our Cookbook which is available at: http://www.melexis.com/relinfofiles/AN7122x-Cookbook_rev004.pdf
Best regards
Rumburak
ABOUT 90217’leads
Answer
I personnally do not see where you get the information that you should not arc the leads toward the rear of the package.
Many people do this lead-forming operation and when it is well done, there is not any risk nor issue.
However, please note that lead-forming needs to respect rules to have no impact on the quality of the parts. Manual lead-forming without appropriate tools (clipping the parts,...) is not recommended for hi-quality mass production. Lead forming of any IC can create latent reliability problems which do not manifest as failures until exposed to thermal cycling and aging effects. Please consult with Melexis Application Engineers for advice on lead forming techniques.
TH8100 & LIN API
Question
I'm using the TH8100,but i can't find the documentation: "MLX LIN API specification".Can you tell me where can i get it?
By the way,it seems that the TH8100 can't be used as the LIN master.Can you recommend me the other similar controllers that can be used as the LIN master.Thanks!
MLX90306
Answer
The figure 3 (Temperature Linearity Correction) of the datasheet will help you to understand what your engineer wants to do. Let me assume that the temperature range of your application is -40degC till 140degC. Your engineer wants probably to use 75degC as T1 (see figure 3 of the datasheet) and find the optimal coarse and fine settings at this temperature than use 0degC to find the appropriate values of GNTC1 and OFCT1 and eventually go to 130degC to find GNTC2 and OFTC2. In this case you will have at the end 2 temperature gaps: the first will be for a temperature lower than 75degC and the second for a temperature higher than 75degC. In the software you have to fill in the following values: - Coarse Gain, Coarse offset, Fixed gain, Fixed offset found at 75degC - Temp Gap1 is the temperature code at 75degC -OFFTC1 and GNTC1 found at 0degC -OFFTC2 and GNTC2 found at 130degC -Temp Gap2 will be in this example equal to 1023.
Reception with MLX90121
Answer
I developed a reader circuit similar to that of the circuit in the data sheet of MLX90121 with a loop antenna for near field communication. I'm using a DS89C420 uC to control the communication. This will be used to implement the Tag-it protocol for TI tag and extract the SID number. Now, please fill me in if there's something lacking in what I did or correct me if there's something wrong.. At first I verified if I can configure the chip by enabling and producing an output on the XBUF pin be it a 6.78MHz or 13.56MHz. To make it really convincing I configured the DecTimeRef also and verified all of it in a oscilloscope.. In transmission, I configure the chip to 100% and I would suppose it is in ASK for direct mode. Then do a field activation(carrier freq presence) for 1ms and transmit the request using pulse width modulation as stated in the protocol. By the way, the request that I used is the example stated in the documentation of the Tag-it protocol for SID poll request. I verified my all timing and all are fine.. One thing that I noticed is that I can see 13.56MHz carrier in the TX pin as well as in the spectrum analyzer by coupling the my loop antenna with another antenna connected in the spectrum analyzer.. From what I saw there is no ASK modulation taking place. I tried to simulate a 13.56MHz, ASK, 100% using a signal generator, the waveform is different from that of my reader. Next while transmitting, I can see that there were signal output from DOUT pin, is this really something to expect during transmission. When I combined reception, which requires to wait 311us for the tag response and using dual subcarrier, I could still see the same signal from the DOUT pin regardless of the presence of the tag and I don't think it's the expected response from the tag. I also used majority voting reception but still same scenario except that DOUT signal changed to the timing of DecTimeRef.
I would like to ask what's the proper sequence for configuring and setting up the chip for direct transmission and reception? What's the best method to use during reception? My reception is my biggest problem.
Answer
Using internal decoder (means majority voting MV) will provide you a synchronization signal DSYNC to easily get the filtered data from the MLX90121. Consequently, the load of your microcontroller will be reduced.
Nevertheless, the MV needs to be started by the microcontroller by rising the CK signal at the beginning of the TAG reply. Then, the microcontroller needs to poll DOUT line to "find" the beginning of the TAG reply.
The decoding of the data from a TAG (manchester, NRZ, ...) needs to be done by the microcontroller as well as the protocol level (SOF, EOF, start bit, stop bit, CRC, ...). As the microcontroller controls the protocol level, it can easily recognize from SOF of ISO15693 and TAG-it protocols.
rohs compliance
Answer
there is only general information about RoHS on our website: http://www.melexis.com/relinfofiles/lead_free_statement_may05_1.pdf
For Melexis, leadfree = ROHS-compliant.
Because in the ROHS-directive, there is mentioned a banned substances list: for Melexis only the lead must be banned from our products (all the other substances are not included in Melexis products).
So, if a product is leadfree, it's also ROHS-compliant and vice versa.
Noise Filtering for MLX90215
Answer
Low pass filter can only help and the lower the cut-off frequency, the better results you will get.
At the sensitivity range you are thinking i.e. 140mV/mT or 14 mV/G, the internal sampling frequency of the chip is already pretty low (4 kHz) and therefore a cut-off frequency of 500 Hz (or less) for your filter will be adequate to eliminate the noise components which are associated to the internal sampling.
You can also work w/ MLX90251 as this IC has also an on-chip filter you can set through the programming unit (e.g. together w/ offset and gain settings). Using the strongest filter (i.e. Filter = 15), you will already get a serious noise reduction and you may not need an external low-pass filter: a decoupling cap will be enough. Advantage of MLX90251 is that 140 mV/mT is not at the top end of its specification while MLX90215 gain settings is somewhat limited to this value.
In any case, all those filter (internal or external) considerations need to be checked vs. your bandwidth requirement for your application. If you do not have specific "speed" requirements, you have the complete freedom. If you have speed/response time/bandwidth requirements, then you need to consider those while dimensioning your filter.
Problem on PTC04 Programmer
Question
My company purchased PTC04 programmer tool with Hall sensor 90251 from your Melexis three month ago. The "Melexis Programmable Toolbox" (User Interface Software) can not find the PTC04 programmer hardware any more. When I use the menu selection Tools->Search Devices->All, a warning box pops up which shows that "No Melexis Devices were found during Scan". But When I used this tool before, it could find the devices very easily.
I have checked my operations. Power is on and connection Cable is linked. I have tried two kinds of cables.At first, I use RS232 cable, it did not work. Then I tried USB Cable. Computer could recognize the PTC04 hardware and installed the driver for it. But, in the "Melexis Programmable Toolbox" software, I still can not find the devices.
How is the resolution of 2SA-10?
Answer
It is an analog output i.e. the resolution is theoretically infinite.
In practice, it is of course limited through the noise of the amplifier.
If you use decoupling caps on both signals without adding a filter, you can achieve a resolution between 9-10 bits depending on the environment (noise on the supply lines,...).
However, if you add a filter (1st order i.e. simple RC or 2nd order...), then you can reduce the noise, increase the resolution while slowing down the whole system (reduced bandwidth). There, you can reach 12 bits.
Answer
It is difficult to compare as the 2SA-10 delivers sine & cosine signals while the MLX90316 provides directly an angular information from 2 raw signals which are comparable to the sine and cosine signals of the 2SA-10.
Those raw signals have comparable accuracy although it should be mentioned the MLX90316 offers a larger compensation capability.
If you process the 2 signals from the 2SA-10 w/ an external circuit (ADC + CORDIC algorithm + linearity compensation), you should get similar results than the MLX90316 for which the signal processing is done on-chip in combination w/ EEPROM space for compensation.
In terms of figure, it is always worth separating the linearity error (at room temperature) and the thermal error.
The linearity error you can get from the MLX90316 is governed by the intrinsic linearity error (max. 1 Deg. over 360 Deg.) and the linearity error linked to the magnet position w/ respect to the IC. The MLX90316 provides a multi-point calibration which will lower the linearity error of the output transfer characteristic (the efficiency of this multi-point calibration is higher when the angualr stroke is smaller).
The thermal drift is mainly governed by the thermal offset drift. It will induce an angular drift of 0.3 Deg. Taking into account other contribution, the thermal error is limited to less than 0.5 Deg.
As mentioned earlier, this can be achieved w/ 2SA-10 using off-chip signal processing !...
Answer
Datasheet is correct. Above 80 mT, the linearity error linked to saturation of the magneto concentrator will start increasing.
We usually recommend to work between 20 and 70 mT.
we can say that the target should be 45 mT in order to get a quite symmetrical tolerance window i.e. +/- 25 mT.
In this case, you should be safe over airgap, magnet, temperature tolerances...
Do not forget that a magnet gets always weaker at high temperature and stronger at low temperature...
90316 serial protocol
Answer
The 90316 responds only if all requirements are fulfilled. Please check the following:
- EEPROM parameter SPI is set.
- Check application diagram as described in datasheet
- Check capacitors on out1 ( MOSI/MISO) , out 2 (SCLK) and out 3 (/SS). Typical 1nF. To big capacitors will jeopardize the timing specifications.
- Check signals at pins 90316 on scope. ( /SS = low, SCLK = 80 Clocks and AAFFh as startbyte during falling edge of clock.) Note that your uC has to send 5500h on mosi because of the transistor.
- Check signal levels ( 0..5V ).
- Check timings according to specifications in datasheet. Verify EEPROM if 90316 is set to low / high speed mode.
In case you do not get any respond (angle nor error code), then I would advise to create a scope plot of the 3 signals + screendump off EEPROM parameters and send it to me.
Question
I took some data from the MLX90316 serially but checking bits after positive edges, when I recieved your reply,I tried the falling edges but no data is displayed, so I returned to my old method.But I have some errors in it, the errors are not repetitive (the time between erronous data is not constant, the magnitude of errors it not constant)
so if you have any idea about this phenomena please support me with.
Answer
The positive going edge of the SCLK shifts a bit to the Slave’s output stage ( =90316) and the negative going edge samples the bit at the Master's input stage. This means that the value is only stable during the negative going edge and will toggle during the positive going edge.
Please check the serial protocol mode of 90316, as described in the dataheet on page 21, and the timing diagram on page 22.
You mentioned that during the falling edges no data is displayed. Can you verify that the uC Mosi is low during the reading of the data? It would help if you can send me a scope plot of the 3 signals. A scope plot is the only helpful tool to check the signals and timing of the SPI communication.
MLX 90316 eccentricity
Answer
- The typical non-linearity error of the 90316 is <1 deg and the signature of the error is the sum of a single and double period.
- In case of an eccentricity error between magnet and hall plates, the max non-linearity error is more or less equal to 320 x (offset)^2 / 4D^2. (D = diameter of magnet)
This means that angular errors due to a given off-axis misalignment are becoming smaller with increasing diameter of the magnet. The signature of the error is a double period over 360 degrees.
- If the signature of the error is a quadruple period over 360 degree, than we identify this as a saturation error and is caused by a too strong magnetic field > 70mT.
Based on the shape (number of periods) you can identify the nature of the error.
Question
hmmmm... perhaps the answer is found in your statement that the calculation is based on the offset "to the Hall plates".
Am I correct that there are two Hall plates, and they are not centered at the same spot in the chip package? If true, would there not be a fundamental difference between the two outputs, given some magnet offset?
This is the real question: With a redundant chip, a fundamental requirement is to be able to compare the output of one chip with that of the other for safety purposes. Sometimes this is called interlinearity.
Given a range of radial offsets (both in magnitude and direction), will BOTH outputs respond to the offset in exactly the same way? That is, will the two output signals still track (correlate) with each other, or will you see a greater error or perhaps a difference in the arithmetic sign of the error between the two outputs, depending on the direction of the radial misalignment?
Answer
- You are correct that there are 2 different positions in the redundant version, and they are centered in only 1 axis, and a small difference in the other direction. (See package information in datasheet.
- The redundant version is 2 x the 90316 die in a TSSOP package and this offset difference is compensated during the calibration step in our production. The nonlinearity error for both outputs can be +/- 1 deg over 360 deg, but the interlinearity is then in worst case +/- 2 deg
- In case of an eccentricity error, both outputs will have an error but not necessarily exactly in the same way. The signature of the error is for both a double period over 360 degrees, but the magnitude / 0 deg error crossing on the X axis can be different..
Question
I would make an example to right understand the formula on the top:
320 x (offset)^2 / 4D^2. (D = diameter of magnet)
I have eccentricity d=0.5mm and an Magnetdiameter of 10mm so the amplitute of the non-linearity errror is..
320 x (0.5mm)^2 / 4*(10mm)^2 = 0.2%FSR(Full Scale Range)
Is this correct?
MLX90217
Answer
Hi, sorry for the late reply.
Prior answering your specific question, I would recommend you to look at the other Posts linked to the MLX90217 as there are precious information there.
From your message, I have the impression you plan to use the MLX90217 at the front of a rotating magnet. It can work that way of course even if the most common application is w/ a so-called back bias magnet (magnet at the back of the IC) and a ferrous rotating target at the front of the IC/magnet.
The IC senses the magnetic flux density and from the modulation of this flux density, it derives a digital output signal. The output transfer characteristic is based on track-and-hold circuit and the hysteresis for the peak detection is the value in the specification i.e. 1.8 mT up to 10.0 mT.
The modulation of the flux density can be generated by a spinning magnet (like you are doing apparently) or a spinning ferrous target at the front of the IC w/ a back bias magnet.
Answer
This item has already been discussed on the forum. Please have a look in some previous posted messages about the MLX90217.
To make your life easy, I have inserted it again below.
Any kind of magnet (SmCo, NdFeB, AlNiCo, Ferrite) can be used as long as the amplitude of the magnetic signal modulation (due to the rotation of the wheel i.e. succession of teeth and valleys => AC signal) at the IC level is more than 100 G (10 mT). This should be verified at maximum Airgap (i.e. max. distance between sensor and target wheel). Pay also attention to the magnetic bias itself (i.e. DC component of the magnetic flux density applied on the back side of the IC): it should be maintained below 4000 G (or 400 mT) w/ the proper polarity.
If you take a "cubic" magnet of 5 mm x 5 mm x 5 mm NdFeB and you place directly at the back of the IC, you should be fine.
For the gearwheel, if you take valleys w/ 4 mm depth, you should be fine.
Again, to do a good job, once in the design phase you should measure the field strength between magnet and target with linear hall sensor or gaussmeter and plot this to be sure the signal is adequate for reliable operation over temperature and life.
TH72015 - crystal
Answer
of course you can also use an external oscillator like a TXCO. If you go to the TH72015 DS, pages 3 and 4, you will find the schematic of the crystal oscillator. You just have to input the signal to pin 4 (ROI). Make sure the signal is AC-coupled and the level is between 100 and 500mVpp (try a bit until you get the optimum suppression of PLL sidebands).
But what is your problem with the xtal? It should work quite easily. Please have a look at the DS, page 8, where we provide the recommended xtal specs. Many xtal suppliers have those xtals needed for our RFICs on stock. Note that FSK modulation via xtal pulling requires an xtal!
Answer
well, as I said, in this case the modulation must come from the TCXO. Some of them can be FM-modulated. Can you tell me the TCXO part number and manufacturer, and provide a link to the data sheet? I would like to have a look at it.
Question
The link of data sheet :
http://www.kdk-group.co.jp/pdf/58%7E59_nto.pdf
Can you give me your mail address. I can send circuit to you.
MLX90316 questions
Answer
Yes, the field range that the 90316 should see is 20..70mT, over the complete temperature range. Lower fields will influence the performance; higher fields will saturate the IMC and induce additional nonlinearity errors.
The 90316 will automatically adjust the gain in the application, based on the measured field.
TH7122 FSK signal polarity
Question
I am using TH7122 as FSK transceiver.we have two board,one transmitter,one receiver.While senting BARKER code, the result on receiver port was wrong,the polarity is just invertted. Crystal pulling Modulation. DTAPOL is '0' as default.If set as '1',the polarity is correct.Why?To me,it is reasonable that the DTAPOL is setted as '0'.
TH71102 at 27MHz
Answer
thanks for considering the TH71102 for your 27MHz application. OK, the xtal freq. can be calculated by REF = (RF+IF)/2 = (27+10.7)/2 = 18.85MHz. Other passive components that should be tuned from 40MHz to 27MHz are L2, C4, L3, C6, C7, C8, L4 and L5. You can use the well-known equation for a parallel resonant circuit for the single tank combinations of L2/C4 and L3/C6. This is f0 = 1/(2pi*sqrt(L*C)). The same eq. applies for the combination of L4/C8/L5 but you use just 1/2 of the C-value as there are two Ls in the circuit. C7 defines the coupling between the two LC tanks. I would try something between 27 and 33pF. You may use the same 100nH inductors and just tune the capacitors a bit up to resonate at 27MHz. That's it.
Possible error in antenna booster?
Question
For the last few days I was in trouble with the antenna booster AN from Melexis as can be found at [link] http://www.melexis.com/Asset.aspx?nID=3881 [/link]
I made the exact schematic as layed out on page 1 of the above mentioned AN.
I've set the duty cycle as the AN describes using two multi turn variable resistors of 2k ohm.
I installed the N and P fet, and hooked up the power supply.
The mosfets got really hot.
So my conclusion was that the mosfets were into a conducting state at the same time (which results in a short, as every one knows) since my schematic is the exact same as the Melexis, there must be an error in the AN’s schematic.
I solved it by connecting resistor "R4" to +5 volts and "R2" to ground (which in my case were the leeds of the respective variable resistors.
Some other solution would be to put the NFET into the ground path (instead of in the +5V path as it is with the Melexis schematic) and the Pfet in the +5V path switching speed may get lower, to be frankly I didn’t test this yet, but it should work.
Is there any one else who is experiencing/ has faced the same trouble?
Or did I make a mistake without seeing it.
MLX90316: Realtionship between angular poistion and linearity error
Question
I´ve got questions about the MLX 90316.
In the front-end calibration of MLX90316 is a figure (page 7) which shows the relationship between the angular position an the linearity error.
First question: Exists a formula to calculate the relationship between the field strength and the linearity error. For e.g. a field strength of 100 mT causes an linearity error of 7%
and a field strength of 120 mT causes an linearity error of ????
Second question: The conversion from linearity error to angular error. For e. g. a non-linearity error of 1% leads to an angular error of 0.15 degree
and a non-linearity error of 7% leads to an angular error of ????
Third question: What is under a field strength of 20 mT. You told me that the noise to signal ration will increase. Does also exists a relationship between the field strength and
the non-linearity. For e. g. a field strength of 10 mT causes an linearity error of 1% and so on. Exists a formula to calculate this relationship and to converse the non-linearity into an
angular error?
Last question: Can you send me the back-end calibration of MLX90316. I searched on the melexis homepage, but I didn`t find anything.
Answer
When the field is increasing, the concentrator starts to saturate in the center, and the saturation will be visible mainly in the Y-axis when applying a flux in the X-axis and vice versa.
This is complex phenomena that we have characterized by measurements of the X an Y signals, and not by using a formula. Based on the measurements we know that the typical relationship of the nonlinearity ( NLE) vs the field ( B ) is a expositional function [ NLE = 0.0004 x e^(0.0973 x B) ]
- The relationship between the max angle error and the nonlinearity was simulated based on the knowledge of the characterization measurements ( NLE vs field) and confirmed by measurements. The signature of the nonlinearity error is 4 period over 360 degrees.
The typical relationship of the max angle error vs field is [angle error = 6E-05x e^( 0.0961 x B) ]
- In case of a 10mT signal , the nonlinearity and angle error is negligible small. But because of the low field , the gain is increased to a very high level and therefore the noise and offset will become the main players of disturbing the sin and cos signals.
- The AN for the back end calibration is under development.
TH71081 replacement
Answer
We can suggest following replacements:
TH72031 Transmitter - 868/915MHz, FSK (SOIC8)
TH72032 Transmitter - 868/915MHz, ASK (SOIC8)
TH72035 Transmitter - 868/915MHz, FSK/ASK (10L QFN 3x3 Dual)
For an overview of the main parameters you can download our quick section guide:
http://www.melexis.com/Asset.aspx?nID=3752
For pricing, sampling and purchasing I would like to point you to our local partner in Slovenia:
IC Elektronika d.o.o.
Vodovodna cesta 100
Ljubljana.
1000 Slovenia
Phone +386 1 5680111
Fax +386 1 5689107
E-mail prodaja@ic-elect.si
Web www.ic-elect.si
TH7122 at 40.68MHz
Question
I'd love to have schematic for 37MHz.
Right now I'm drawing something like this:
- design with additional varactor diode BBY65
- modulation via direct VCO injection
Please correct me if I'm wrong - chip alone is rated for quite large voltage range - but as far as I understand - voltage changes during work (I'm not talking about voltage drops etc. - let's say it's powered from alkaline battery - and voltage changes during work - from almost 5 - down to 3V) will detune whole circuit???
I'm considering using some LDO for powering modulation circuit (to obtain same modulating voltage level). Or maybe it's ridiculous since proportional vcc drop will cause modulation change in some way?
Buy LIN
|
|