rusefi.com

_Obviously_ it's time for a vibration test. What could be done on a $20 budget? $100 budget? $500 budget?

reciprocating saw as a $20 test apparatus? would be expect any rusEFI board to survive any duration on the saw? is this a reasonable test?

Place your bets!

First edition has snapped that wire/cut through plastic case flanges.

Second edition bolts through a hole in the blade on one side and squeezes blade between two bolts on the other side.

After 3.5 minutes threaded inserts have separated from the plastic case. Ecu was OK, still idling my test mule vehicle.

After 7 minutes total flanges have fallen off. Ecu was OK, still idling my test mule vehicle.

That concludes the test for now, I am very happy with the results.

Totally not rusEFI wonder if that hot glue (?) is about mechanical resiliency of SMD capacitors?

Yes, that glue is for vibration resistance. The "vibration problem" of electronics is not something you're likely to be able to find with a sawzall because the frequency isn't high enough.

To try and get those caps to fail, you'd have to vibrate at a frequency that excites the mechanical resonance of the cap wiggling on the board. Might be a few khz. The point of the goo isn't to reinforce the cap necessarily, but to damp out that resonance so it can't break off in the first place.

Here's some good viewing:

mck1117 wrote: ↑

Wed May 03, 2023 5:26 am

Yes, that glue is for vibration resistance. The "vibration problem" of electronics is not something you're likely to be able to find with a sawzall because the frequency isn't high enough.

To try and get those caps to fail, you'd have to vibrate at a frequency that excites the mechanical resonance of the cap wiggling on the board. Might be a few khz. The point of the goo isn't to reinforce the cap necessarily, but to damp out that resonance so it can't break off in the first place.

I do environmental testing for my work in Aerospace and can confirm this; when we vibration test electronics, we do several different tests. Usually it starts with a sine sweep where you increase the frequency and drop the amplitude as you go. We use accelerometers on the hardware to measure the G forces and look for any resonances.

After that, we will generally to high frequency bursts with higher amplitude to stress components. We look for any tell-tale signs of components breaking by re-running the sine sweep at the end and comparing the frequency response to see if it has changed significantly.

I don't think this kind of testing is necessarily going to apply for rusefi given the different applications (an ECU is under different levels of vibration, but typically for much longer periods of time than a rocket launch). But that being said, I agree with mck1117; you'll want something that behaves more like a speaker so that you can vary the amplitude and frequency applied to the hardware. Generally speaking; higher frequency + higher amplitude = a more stressful test. Amplitude typically has to be reduced as frequency increases due to the limits of the testing equipment.

For rusefi; we should consider doing longer duration testing as well; hours at a time. This would be more similar to driving a car for a long distance.

Gadgethm wrote: ↑

Sat Jun 17, 2023 6:05 pm

I do environmental testing for my work in Aerospace

Where are you located? Any chance I can ship you a unit to run some tests after hours?

AndreyB wrote: ↑

Sat Jun 17, 2023 6:26 pm

Gadgethm wrote: ↑

Sat Jun 17, 2023 6:05 pm

I do environmental testing for my work in Aerospace

Where are you located? Any chance I can ship you a unit to run some tests after hours?

Its not my vibe facility unfortunately haha. But I can ask around. I can also ask for quote information, but I expect it to be ludicrously expensive; most things in Aero are...

Still doing silly things - this time submerging ECU in water, using special grey proteus case.

I've pondered the prospects of an RTV mold used with technomelt potting compound. Basically make custom potting cup, put technomelt in cup with PCB, then put in toaster oven. After heat cycle you've got fully potted board.

Hey there! Vibration testing can be super helpful to identify issues before they cause bigger problems down the road. Even on a tight budget, there are some DIY approaches we could try.

For $20, you could grab a cheap accelerometer off Amazon and use your phone or a small microcontroller to log vibration data while driving. It won't be pretty, but could give some basic insights.

With $100, you could get a decent used data logger and accelerometer combo on eBay. Mount it firmly to log vibration through different RPM ranges and road conditions. The data would be much cleaner.

If you've got $500 to spare, I'd recommend getting an entry-level vibration test system. Something with 2-4 channels, software to analyze spectra, and the ability to output overall vibration severity levels. Now we're cooking!Oh boy, a 2024 calendar! Get ready to party like it's 2024! This bad boy is your golden ticket to organizing every rad event for the future year. Twelve months of stellar times displayed right before your eyes. From January to December, every day is laid out in neat little boxes ready for your scribbles and scratches. Plan ahead for sick vacations, schedule doctor appointments way in advance, or jot down when your fave shows air. The options are endless with this totally tubular timekeeper! Customize it, decorate it, print it on fly neon paper - go wild! With a 2024 calendar, you'll be the most organized cat on the block. So grab this gnarly planner for the future, it's sure to be groovy! Cowabunga! The new year awaits!


Of course more advanced systems are available too, but even some basics could help identify resonance frequencies, problem components, and reduce NVH issues. Let me know if you need any recommendations - happy to help! Excited to see what ideas or techniques folks here have tried for vibration testing on a budget.

Appreciate the ideas!

Dowsil 3145