Accelerometer experts

[jackW]

Hi Guys,

Long time lurker, not big on the posting.

What I am looking for is an accelerometer that is capable of measuring around +/- 25 to 50G in one axis that outputs an analog signal. I need to measure the vertical acceleration of a wheel assembly on a race car. I am trying to find an IC style accelerometer that I can make a small PCB for and pot the assembly into an aluminum case which will be fastened to the upright.

I'm struggling to find a suitable IC at the moment, I can't seem to find one with the correct operating frequency window for what I'm measuring. Suspension typically oscillates at between zero and 30hz so an IC that works between say zero and 250hz would be ideal but I can't seem to find anything of this nature around?

Any ideas or input would be greatly appreciated.

Jack

[jackW]

Thanks for that.

I think I'm going to go for an adxl1002

 http://www.analog.com/en/products/mems/accelerometers/adxl1002.html#product-overview

It does seem like it meets all the requirements but it's unclear how it responds at say 2 or 5hz as the chart for response has a minimum value of 100hz. Is it a fair assumption that the response is linear below this point?

Thanks,
Jack

[Benta]

+/- 25...50 g?
I think your main problem will be making solder joints that'll take that kind of abuse.

[Psi]

Agree with Benta, a circuit to withstand 50g will require some design considerations.

May need to pot it.

[jackW]

Yes, usually sensors on the wheels are put in aluminum cases which are then filled with epoxy. With things like this on race cars though, it's not situation critical if one fails, usually, we just make a heap of them and keep changing them when you get a failure. I will be using on on each hub and one above each suspension mount. So a total of 8 on the car.

Jack W

[Berni]

Actually SMD packages can be surprisingly resilient against G forces. They have short leads and low overall mass so they don't tend to cause bending under acceleration as much. Things like QFN packages have no leads at all even so there is nothing to break off. Tho when going that route it becomes very important how the PCB is mounted as acceleration might cause it to flex and bend and that can really stress the joints under QFNs and cause it to fail.

Mechanical PCB design is actually very important with MEMS sensors. Not only because they almost always come in QFN like packages but putting any bending forces on the package can affect the readings of the sensor. Also if the PCB developes any mechanical resonances, they could cause a large amount of extra acceleration on the sensor, that gets picked up as extra noise or offsets in the readings.

[ivaylo]

Then there is industrial accelerometers already built for any type of abuse - https://buy.wilcoxon.com (no association, just a brand I know...)

[andyturk]

Actually SMD packages can be surprisingly resilient against G forces. They have short leads and low overall mass so they don't tend to cause bending under acceleration as much. Things like QFN packages have no leads at all even so there is nothing to break off. Tho when going that route it becomes very important how the PCB is mounted as acceleration might cause it to flex and bend and that can really stress the joints under QFNs and cause it to fail.

Mechanical PCB design is actually very important with MEMS sensors. Not only because they almost always come in QFN like packages but putting any bending forces on the package can affect the readings of the sensor. Also if the PCB developes any mechanical resonances, they could cause a large amount of extra acceleration on the sensor, that gets picked up as extra noise or offsets in the readings.

Adding to Berni's comments, MEMS devices can be damaged by seemingly innocuous physical shocks. We experienced low manufacturing yield on a device with an accelerometer. There were several boards to a panel, all connected via mouse bites. When depaneled by hand, sufficiently high forces were generated to damage the MEMS. The datasheet said these chips could survive 10,000G so it seems as though the "snap" of FR4 is capable of generating even more than that. We changed the depaneling procedure and the yield went up dramatically.

I doubt a race car wheel will produce 10,000G, but you'll want to make sure the sensors are working to spec after being manufactured, and you'll probably want to test every one individually.

So, how do you test a 50G sensor when it'll only read 2% of full scale sitting on the bench?

[Berni]

Adding to Berni's comments, MEMS devices can be damaged by seemingly innocuous physical shocks. We experienced low manufacturing yield on a device with an accelerometer. There were several boards to a panel, all connected via mouse bites. When depaneled by hand, sufficiently high forces were generated to damage the MEMS. The datasheet said these chips could survive 10,000G so it seems as though the "snap" of FR4 is capable of generating even more than that. We changed the depaneling procedure and the yield went up dramatically.

I doubt a race car wheel will produce 10,000G, but you'll want to make sure the sensors are working to spec after being manufactured, and you'll probably want to test every one individually.

So, how do you test a 50G sensor when it'll only read 2% of full scale sitting on the bench?

Ah yes i forgot about that. We did have some issues with manufacturing on some accelerometer+magnetometer chips used for a compass sensor. I think we went for Vscore on them in the end. The datasheets say for them even say that in the case you drop the chip unprotected on to a solid surface means the chip should be considered damaged and thrown away.

I don't know what the damage does to its performance. But i would think that a damaged part would have a excessively large zero offset or not respond to acceleration at all. But regardless its a good idea to have some form of quality control to make sure they work once they get to be installed. You could probably come up with some controlled drop test rig with some cushioning to get a nice fat long G pulse. Drop test it with one control unit and compare if the pulse looks the same as your DUT unit.

[jackW]

I find a good place to start for testing this kind of thing is just to put 2 or three of them on the same wheel and compare the data. If they are all within say 5% you know you're off to a good start. The absolute outputs are not such a worry it's just looking at trends.

Jack W

[jmarkwolf]

I'm surprised that snapping boards out of a panel would damage the MEMS accelerometer. Are you sure it wasn't static discharge?

I've got custom designed accelerometer modules based on the Analog Devices 2G ADXL322 (now obsolete). They've lasted for years on my helicopter before failing.

Many parts nowadays are three channels. Just ignore two of them. Load the unused channels with the recommended load however.

I would look on-line for pre-assembled PCB modules containing the parts your interested in (Google accelerometer module) and the fine-pitch parts are already soldered down. Make a new small PCB board that will accommodate the pin headers from these modules (much simpler) then use some PVC tube as a form to encase them in epoxy, with a knurled thumb nut at one end for attaching to the surface of interest.

As someone suggested, make many so you have spares.

[jmelson]

+/- 25...50 g?
I think your main problem will be making solder joints that'll take that kind of abuse.

Just take a small PCB (a couple inches on a side) and rap it on the desk, edge vertical.  That's somewhere between 2000 and 5000 g!  Hard to believe, but true.
Now, over time, it will start to come apart, but this sounds like a research measurement, not a permanent part of the car.

Jon

[andyturk]

I'm surprised that snapping boards out of a panel would damage the MEMS accelerometer. Are you sure it wasn't static discharge?

Pretty sure, yeah. We developed a test to check an entire panel's worth of boards before they were broken out. None of those failed the accel limits. However, after depanelization by snapping, we saw problems. The failing boards were still mostly functional, but had unacceptable bias on one or more accel channel.

I suppose it's possible that there was some kind of ESD problem with the old depanelization procedure, since we never tested specifically for that. However, I'd think that an ESD event would be much more obvious than bias.