Another MAX9926 VR board

[ivan141]

So, I've got 16 boards in this first batch with the shunt inside the current limit resistors.
I have enough 9926's to assemble about 5 boards if I don't mess up and I have enough
stock of the required passives to last me quite a while.

Any takers are welcome to inquire, bare PCB or fully assembled, whatever tickles your fancy.
Do note that I am yet to test the first board, have to think up some kind of testrig that allows
me to test it up to some serious rpm. I'll feel better selling these if I know for a fact that they're
in good working order.

I think Fred's price of 25USD for an assembled BrickRPM unit was fair, so that's what I'll be asking too.

Here's a pic of the first assembled board, baked in my flavorwave turbo reflow oven:

[Fred]

115v 50Hz is a good test, and it's what I was doing. Mine failed that with a bad signal, Jean's nearly caught fire :-D

Max you're likely to see frequency wise is about 60 teeth * 8000 RPM / 60s = 8kHz as higher revving stuff almost always uses lower tooth counts.

Home built the price is fair, I'll jack it up a little for the tax man, if/when I resume this stuff.

Re shunts, leave them out, install externally, no drama. :-)

[ivan141]

Thanks for the input Fred.
Cant test with 115v unless I use a transformer, we have 230V mains in Europe.
Was thinking more along the lines of a real VR sensor, a toothed wheel and a drill.

[Fred]

I was in Europe at the time, and we have 230V here too. I bought a crappy little transformer from one of the two old school electronics places in BCN and used that. Another alternative (that I will probably use to destruction next time) is a variac, 0-250V out from 230V in. Super handy, maybe you can find one second hand somewhere locally for an OK price? I have 5, but that's a long story... lol. They can drive quite a load, too, so you can do stuff like this: https://www.youtube.com/watch?v=SgUWnIPi4hA

The shops were called Diotronic and Onda Radio (in case anyone reading this happens to need a capacitor in Barcelona).

EDIT: For the higher frequency, you can just feed one half 5V and oscillate the + input back and forward from 0 to 12v.

[ivan141]

The beginnings of a test rig have been assembled.. now for a quick and dirty way to mount everything solid enough and I'm good to go.
Crank sensor is a Renix VR sensor off an 98 Renault Megane, virtually identical to the one my volvo 360 has. Tested it with the multimeter
and I get a solid 0.2V when I give the pictured 4AGE cam gear a whirl near it. I am in dire need of an oscilloscope, but since those rarely
come cheap I think I'll just wire it up to an Arduino pin interrupt and see what RPM I can measure.



On a side note: I ordered a stencil for the qsop16 chip to get better results straight out of the reflow oven.

[Fred]

For testing this I'd recommend a saleae or, for the less morally inclined, a china clone of one. More info in this thread: Scrub that, can't find the post I wanted. I'm sure you can find them yourself, though.

A scope would be 100% necessary to look at the input wave forms if you get an unexpectedly bad output signal for any reason, but if you know the spin rate of the wheel, and the tooth count, then you can just view the LA wave form and give it a "yes" or "no".

What's that cam wheel off? I know they all look the same, but it looks VERY familiar.

Good move on the stencil. It's still a minor pain, but way less of a pain :-)

[ivan141]

That's off an '83 4AGE bigport, my AE86's original engine.

[ivan141]

Ordered some chinese knockoff logic analyzer, as a programmer I hate myself for it, but until I figure out how much use one of these
will be in my rare spurts of electronics tinkering I dont want to be spending big amounts of money.
Most knockoffs I've seen are based on Cypress SoC's, a far cry from the FPGA based design of Saleae, but in my experience reliable
enough and pretty versatile with it's programmable analog options (we have IO boards built around them at work for our burglar detection and access control system).
Perhaps be fair and try some open source software first (sigrok?).

Mounted the gear and sensor to an old cutting board, but I'm afraid I will be spinning the gear by hand for starters.

[ivan141]

My chinese knockoff logic analyzer came in the mail today so I hooked my VR board up and gave it a whirl (literally).
Installed sigrok instead of Saleae Logic, works just fine.
Here's the result:

This was the fastest I could spin the (36-0) gear by hand, it looks OK to me, but for a real performance test I suppose I'd need a way to spin it
longer and at many speeds. It does tell me the board does what it's supposed to, I'll settle for that.

[ivan141]

Hmmmm.. If my calculations are correct that was about 33 rpm I achieved... pretty shit as tests go.
My test contraption doesn't allow for serious speeds as is, I suppose it would be better to test that
on an actual engine of some sorts.
I just noticed what looks like a 'missing' tooth before the 3100ms mark, wonder what caused it.
Likely just wobble due to the axleshaft not being an exact fit.