quote Originally posted by Will: Crankshaft in a lathe...

Probably the easiest way to test it if he has a spare. The VR sensor should be magnetic, maybe it has a weak pull and is, for lack of a better term, floating and not providing a clean signal to the converter.

I assume you're able to log the signal while the engine is running, does one signal appear longer than the others? can MS display in that kind of resolution?

Edit:

The LX9 3500 uses a hall effect sensor in the stock 3.1/3.4/3100/3400 position, it requires +12V, but would also allow you to eliminate your converter box.

here's the pinout for the 3500 sensor.




Edit again: I haven't confirmed the 3500 sensor works in the early block, but it's probably worth a try.
------------------
"I am not what you so glibly call to be a civilized man. I have broken with society for reasons which I alone am able to appreciate. I am therefore not subject to it's stupid laws, and I ask you to never allude to them in my presence again."

cognita semper

http://www.fiero.nl/forum/Forum2/HTML/119122.html

[This message has been edited by ericjon262 (edited 04-29-2020).]

A (normal) lathe won't turn fast enough to spin a crankshaft to its redline.

I think I would be able to do something with a computer soundcard and a step-up audio transformer to simulate a VR sensor signal. Not as good as a true end-to-end test, but it would be inexpensive!

********************************************************************************

I was too anxious to troubleshoot things, so I jumped ahead and started working on solutions:

I switched to the optoisolated input on the MS, and I removed the low-pass "Dave" capacitor to remove any possible time-smearing that could merge two notches into one.
After doing this, I still lost sync at 3k RPM.

Then, I removed the crank sensor bolt, and I pulled away the sensor from the engine block by a smidgen. This also increased the sensor to notched wheel gap by the same smidgen.
Bam, I was able to rev to 4k RPM without a hiccup. Since the engine was cold, I didn't want to rev it any higher, and I didn't have the time before supper to let the car warm up.

With magnetic sensors like this, you want a small gap to maximize the signal level. However, when the gap is too small, in addition to seeing the teeth, you also see the imperfections much more. So the best signal-to-noise ratio will exist at some certain distance; not the closest gap possible.

I suspect that these imperfections are confusing the MAX9924, and that the ideal gap with the MAX9924 is greater than with the DIS brick.

Closed black-boxes such as the MAX9924 are the manager's dream... you can just buy a pre-built solution, and nobody needs to spend time developing an in-house solution. It's all good, until you have problems... you end up wasting time trying to reverse-engineer the black box and guessing at what it's going to do. An in-house solution on the other hand, intimately understood by people in-house, is far easier to understand and debug. In this case, I fell victim to the manager/outsourcing syndrome.

In this case, I will probably test different shims to space the sensor away from the machined mounting pad on the engine block... I'm hopeful that I can find a working solution this way.

********************************************************************************

My understanding is that the MS tooth interrupt is edge-triggered, and only by rising or falling edges (the user chooses which edge is the most accurate reference edge). So ONLY the reference edges are logged.

Here is an example showing a missing double-notch:


The dizzy (cam) sensor is represented by the green trace; we only see every other edge in the log. In reality, it is a square wave with approximately a 50-50 duty cycle.

********************************************************************************

I studied the possibility of using a 3500 crank sensor.

Here are pics from my evaluation:
https://www.fieromontreal.c...86.msg33228#msg33228

The early and late sensors interface with the engine block in the same manner.

However, the reluctor wheels are of a very different design:
Early: smaller major diameter, flux through the concentrator depends on presence/absence of a notch.
Late: larger major diameter. magnetic flux is directed forwards or rearwards, and the sensor detects this.

So, using an early crankshaft, I have to use an early sensor. I did not study the possibility of using a non-GM sensor with the GM block/crank.

I liked the idea of a 3-wire interface to the ECU, with a robust digital link. Since my sensor didn't do this, I did the converter box instead.

[This message has been edited by pmbrunelle (edited 04-29-2020).]

------------------
"I am not what you so glibly call to be a civilized man. I have broken with society for reasons which I alone am able to appreciate. I am therefore not subject to it's stupid laws, and I ask you to never allude to them in my presence again."

cognita semper

http://www.fiero.nl/forum/Forum2/HTML/119122.html

[This message has been edited by Will (edited 05-03-2020).]