How many Rex/REGINA flex plate patterns are there?

[DET17]

Well I know for sure there are at least 2 of them for the RB tractor engines.

I have a supposed 94 engine on the stand; it has 19 windows, 1 missing window (completely cut out), 1 solid steel blank. (this pattern TWICE on the flex plate)

A spare I was given (labeled as a 93 ?) has 20 windows, 2x solid steel blanks. (this pattern TWICE on the flex plate)

Both styles would have a total of 44 divisions on each device. I'm going to check the BENDIX green book that I have on the Rex/REGINA ECUs and see if they detail what styles work with what year of engine control.

 

[NLRegina]

This is interesting since to my knowledge, all regina ignition components (save for maybe the head vs block mounted distributors) can be cross-swapped between years, including the ecus. Is it possible that the crank sensor reads the blanks the same as the missing windows? I'm assuming the flywheels are identical other than one having a missing window and one having a blank in its place.

 

[mikep]

Probably. The transitions are what the computer is counting. The 2 compression events per crank revolution are what it adjusts spark and fuel to.
Not having the two in front of me to compare, i am guessing that the slope transitions mostly match and are either constantly transitioning, or are not transitioning, at the same crank degrees.

 

[DET17]

I'm betting that BobXYZ will weigh in here in a bit.

I'm no EE or software geek; however the standard windows are certainly pulses to the sensor and the ECU's. However the full space of BLANK will take the sensor to one end of it's scale (output) and then the FULL space right next to it will send the signal to the opposite extreme.

I think Bob post some months back that the CPS on the Bendix machines is different than that used on LH. Since as we all know on these tractor engines, everything happens on 180* increments, perhaps that is the IGNITION trigger, and the 19/20 holes it just uses for RPM calculation. As old JohnV told me years ago, "2 up and 2 down". Perhaps all REX needs to see is that blank window.... when the sensor hits the FULL window edge, it pulls the trigger on ignition (adjust of course per the tables built in. However, if that were true, then the 20/2 FULL pattern will not see the same waveform.... just the 20 blips and then I presume the signal goes HIGH when it hits the 2 steel FULL sectors.

Perhaps I'll add some pics to this to help visualization.

 

[Khrrck]

Perhaps I'll add some pics to this to help visualization.

That would be helpful, I am having trouble visualizing the patterns you're describing. What you describe sounds like this, which would give two different lengths of the "long" pulse:


Perhaps the computer doesn't care about the length of that "plateau" - just cares that there's 20 "short" readings (would be either the edges between holes or the edge between a hole and a gap) and then 1 "long" reading (the blanked out holes). If that's true then it sounds like both flexplates would match that pattern.

But the way you've described it, it sounds like one flexplate is 21 spaces on each side (19 + 1 + 1) and the other is 22 (20+2). Are the holes the same size on both?

 

[bobxyz]

Can you hold the extra flexplate up to the stand and take a picture with both of them together? Try to align the 1st small hole on the right of the weird teeth.

Being the best 1980s technology, the EZK/Rex ignition box probably only looks at one edge of each tooth - either the transition from open-to-metal, or from metal-to-open. It may be that the difference between the 2 flexplates doesn't matter because the difference is NOT the transition that the Rex box is looking at.

 

[DET17]

That would be helpful, I am having trouble visualizing the patterns you're describing. What you describe sounds like this, which would give two different lengths of the "long" pulse:

View attachment 32289
Perhaps the computer doesn't care about the length of that "plateau" - just cares that there's 20 "short" readings (would be either the edges between holes or the edge between a hole and a gap) and then 1 "long" reading (the blanked out holes). If that's true then it sounds like both flexplates would match that pattern.

But the way you've described it, it sounds like one flexplate is 21 spaces on each side (19 + 1 + 1) and the other is 22 (20+2). Are the holes the same size on both?

Yes, you've nailed it from my description (I did it correctly!). I'll take pics today and add them.

Your waveform sketch is what has me thinking the REX ignition control must be different, OR the entire pattern is clocked 1 tooth different on the flex plate. When I pull the supposed 94 unit off (PENTA rear crank seal must go in), then I'll probably temporary them both on and see if the solid BLANK edge is in the same place for #1 at TDC.

 

[DET17]

Here are the 2 types of patterns.

First, the "supposed as bought" 94 R/R engine:

and next, the spare flex plate which I believe was stenciled "from 93 R/R":

With the engine precisely at #1 TDC, these "trigger points" are roughly in the RB starter hole...... just as the LH system is.

The puzzling one for me is the 94 version......19 small windows (pitch is roughly 8.1* each), then complete OPEN window, then FULL steel.
Based upon the sensor output, I'm betting that Bob can explain this.

 

[bobxyz]

While those flexplates do look much different, I can understand how they'd work the same for missing tooth detection. The Rex box must be using just the open-space-to-metal transitions from the sensor, call this the start of the tooth. The metal-to-open-space transitions, the end of the tooth, is ignored. To get the crank alignment, the Rex box looks for the first start of tooth after the missing tooth region, i.e. the first open-to-metal transition after the gap.

Here's a picture showing the 2 flexplates in approximate alignment, and the first tooth edge after the gap:


I don't know which pattern is better. I'd probably just pick the flexplate with the best starter teeth.

For reference, here's a picture of the rare Penta AQ171 flywheel (Rex/Regina compatible)