Crank Sensor

-
redlined

redlined

Well-Known Member
Joined
Jan 5, 2011
Messages
209
Reaction score
16
Location
Odessa Tx
Posted this in fuel/air too, trying to increases my chances lol.....my apologies.


I have a TBI LA360 with a non-stock compression ratio, need to take the timing control away from the PCM. To do that I need two separate signals, one to trigger the ignition, one to the PCM so it will trigger the injectors. The factory dizzy is a 5v trailing edge square wave with a #1 reference, I want to replace it with an advance type, and I also want to use parts that are available at just about any parts house, no having to wait 3-5 days for mail order replacements.

Considering a '92-'95 5.2 Magnum flex plate and sensor. Anybody tell me the voltage and type of signal the OBD-I Mag sensor puts out? It's a 3 wire sensor, not sure if it's 5v or 8v, the #1 reference etc.

I've spent hours digging the net, the MS site, the MSD site and others until I'm cross eyed, can't seem to find the specifics of the sensor signals to compare the two system's inputs.
 
The best I know without certainty is the flex plate sensor is a Hall sensor. They are a 3 wire sensor with open collector output. The output is pulled up with a resistor (4.7k), to 5V or what is required at PCM. The supply is often wide range, if it has internal regulation. Might be best to bench test at 5V, with current limited supply at 50 ma, steel target, output is active low.

Going mechanical advance is a step backwards.

You may consider multi-point injection and electronic advance with mega squirt or other ECU.
 
Thanks for the info. Why is a properly curved mech/vac distributor a step backwards?

As for the aftermarket stuff, first, it's too expensive, second, I already have a factory controller and all the wiring in place, and third, try finding an MS or other controller in the Middle of Nowhere USA. I carry a spare $30 boneyard PCM in the toolbox, may never need it, but it beats sitting beside the trail or walking.
 
A mechanical distributor is two independent 2D curves. A 3D timing table is much better. There are many desired tuning capabilities, not possible with mechanical distributors.

A 3D table lets you control vacuum advance at throttle release at high RPM, or coast down light throttle conditions. The ported throttle vacuum source is a poor control. There are many more improvements, related to starting, temperature controls, mixture interactions, knock controls ...

There are likely ways to tune the OEM PCM by reflashing or other means. I developed my own ECU, so never tried.

I have never failed an ECU in over a decade of working with them. Things like TPS or other sensors likely go first, based on my experience of repairs on OEM systems.
 
As I understand, you want to install an earlier 1970's distributor for spark and thus need another crank sensor so the PCM will continue controlling fuel, without knowing the difference.

Unlikely a TBI engine would use sequential injection, so all the fuel controller needs is probably just an rpm signal. That is how Holley Pro-jection works. They just monitor coil- for the rpm signal. You might simply need a 5 V "tach" signal. Look at aftermarket tachometers, or maybe using the GM coil that connects to their 8-pin HEI module would work (wht wire is tach).

If you do install a new crank sensor, it might be hard to use a 5.2 Magnum flexplate w/ "toner ring" unless your transmission bell housing has a mount for the sensor. Also, if like my 3.8L, the pickup slots are not uniform, but rather ~12 fairly wide slots of varying width that the computer decodes. I doubt that signal would be at all like the one from your factory TBI distributor. A 36-1 wheel might work better. I put one on my 65 small block, w/ a sensor off the water pump bolts (behind pulley). I made brackets for both a Ford VR sensor and a Mopar Hall-effect. See my post w/ photos (haven't taken it further yet).

You need to see what pickup wheel and signal is in your TBI distributor and make a crank sensor that gives a similar signal. If just 8 notches like a 70's distributor, then I would think simply using a tach signal off the coil would work the same. Anyway, I agree with Kit that if above is your plan, it is going backwards in technology.
 
Bill hit it with a 36-1 wheel off the balancer hub. The worst place to put a crank trigger is off the distributor!(via crank to timing chain to camshaft to oil pump drive, all have slop) Then you can run a megajolt or microsquirt using only the spark out, or even an EDIS.
 
Hey Bill. You're correct on what I'm wanting to do. The shutter wheel in the distributor has eight slots, one of which is offset for the #1 reference. Been looking at various Magnum flywheels, the '92-'95 won't work as is, has eight slots, four too many unless I weld up four and offset one of the remaining four, and I'm beginning to suspect that may be the best/easiest/cheapest solution overall, and here's why.

I've considered the eight pin GM module conversion, piggy backing the 5v reference to the PCM, pin R, to pin E, which would probably cause a voltage drop problem to deal with.....or simply running two modules, one switching the 5v square and providing a clean signal to the PCM, and the other to trigger the coil.

What I don't know is what effect the lack of a #1 reference will have on the PCM. Maybe none, maybe a seizure. I've kicked around the idea of widening one lug on the reluctor to delay the mag pickup's voltage drop, hence the trigger point and creating an offset signal that way, but that would give one cylinder late timing, so that's not an ideal solution.

The other thing to consider is the movement of the advance mechanisms screwing with the PCM since it's expecting a stable reference point. It may read the movement as an increase/decrease in RPM and do weird things with the fuel. Or not......

The only way I'd know would be to try it, but....right now the new engine's in the truck but not hooked up, and all the sheetmetal is still off too. If I'm going crank sensor, now's the time. Another consideration......welding the slots in the flex plate may cause the pickup signal to be "dirty" if the metal thickness and magnetic properties aren't uniform. Overall it's an interesting conundrum, but if I wanted boring and easy I'd drive a Chevy.
 
The '92-'95 Mag flywheel. May have been used in later applications as well. The reason I'd like to go this route as opposed to a balancer mounted setup is for some protection of the sensor/ring since this motor's going in a 4x4. One errant piece of pea gravel around the balancer and I'd be walking across the desert with the buzzards circling overhead.
 

Attachments

  • 318 flywheel.jpg
    42.3 KB · Views: 497
Here are some other ideas. It is possible to add tab and Hall sensor reference to standard distributor. It is possible to use a $1 microchip, and write simple program to decode flex plate, or distributor sensors, and make the output signal what you want.

I have done this to make engine management systems for various engines, retrofit PCM for different engines, make PCM test equipment and injector flow bench controlls. It involves using a timer to measure sensor edges, and state machine and timer abilities to generate desired logic outputs. Deleting every other pulse is trivial, however without reference sync, it will mess with fuel sync in MPI system, but work with bank fire.

I am not selling my service, but willing to help. I can often figure out what needs to be done and write code in a couple hours. Snap me a picture of the OEM distributor trigger wheel, and crank trigger if it exists. I am a bit confused with what year of PCM you have.
 
It's an '89 TBI setup. Computer is supposedly from a Dodge van, the rest is a mish mash of '88 318 and '90 360 parts. It's run flawlessly for 6 years or so on an '82 360. I've looked at adding the shutter wheel to a std. distributor, would require several custom machined parts. A spacer/mount for the advance plate and canister, and a removable upper shaft assembly with the mech. advance attached.

I recently built the '90 360 with a mild Hughes EFI roller, mildly ported 308's and .005" below deck flat tops which yield a 9.3 static compression ratio. The dynamic is just a touch under 8.5:1. The open chamber heads and compression ratio is the reason I want timing control, the PCM's curve is programmed for EGR and weak suck compression.

Can't get pics to load, I'll try again later.


The TBI is basically a bank fire in the respect that it triggers both injectors, once per Hall effect input. Really not much more than a computer controlled fuel leak.
 
Computer is a Mopar Reman, # 1-R4557391. 1989 B,D series, 5.9 EFI, LD, A/T, Fed......and a pic of the shutter wheel in all it's grubby glory. #1 reference tab/slot on the upper right.
 

Attachments

  • 0602141932.jpg
    38.8 KB · Views: 376
I see the reference tooth, the falling looks to be where the rising edge would be for normal tooth. The early rise is detected by comparing to normal gaps.

I think can make that signal with micro, and looking at flex plate signal an a single reference tab on distributor. You would need to add that tab and sensor. Adding to one end of advance slot limit plate. There is a Mallory, I think that has a dual trigger distributor with reference and 8 trigger. I think it costs more than you want to spend, then more $ for electronic advance ECU. Makes MS look cheap.

I see a bunch of work, little gain by trying to use mechanical distributor. It might be easier to piggy back and alter timing signals on the OEM system with micro. Or find a way to alter OEM timing tables. I doubt tuners or chips are available.
 
Kit, I appreciate the offer of the chip, the help and the knowledge, truly much appreciated, but the main thing I'm after is to use electrical parts that could be found on a Saturday morning in Dead Dog Arizona if the need arose. That means no mail order aftermarket spark boxes, injection systems or one off custom stuff. I keep a spare fuel pump and a known good PCM in the tool box, the rest of the stuff is readily available just about anywhere.

The factory left pretty much zero wiggle room in the timing curve for variables, such as no EGR for instance. On the low compression, well worn '82 360, a couple degrees one way or the other caused problems, a tad too slow and it was doggier that usual, 3-4 degrees above spec noticeably helped performance, but it rattled horribly at cruise. Rattled a little anyway on spec if the ambient temps were high.

I screwed up on the new motor by not physically checking the deck height before I ordered the pistons, the decks are a tad short. At the time I was shooting for 9.0 on the compression with a different cam, with the given that it would be a premium fuel engine. The actual ratio came in at 9.3 and the shorter cam I decided on of course raised the dynamic a little.

All things considered, it's either a carb and flooding when wheeling, or keep the TBI and build a custom distributor. The custom parts in the dizzy being non-wear items such as the shafts, looks like that's the best option.
 
The 92-95 flexplate slots look similar to the ones in my 2002 3.8L. Even there, the slot design varied sometimes year-year, so tricky to decode the slots. I think it would be much easier to work off a signal from the distributor. A 70's distributor would give you 8 pulses per rev, similar to the 89 distributor. You need to convert that signal (variable reluctance) to a 5 V TTL, similar to what your 89 distributor produced. An O-scope would help you see, and you can get cheap USB ones on e-bay. I suspect a simple circuit with a few capacitors, clamps, and digital chips could convert to 5 V pulses, w/ no microprocessor needed. There may even be something for tachometers and such. Look at MSD and others. You could use either the raw VR signal, or the spikes on coil-. Tachometers use the later.

I don't know how the reference tooth was used in 1989. I doubt for spark since the distributor rotor takes care of clocking to #1 cyl (unlike later distributor-less ignitions). My guess it phased the TBI injectors to the valve opening, i.e. "sequential injection". Not essential for TBI and probably minimal value (useful only at idle even w/ MPFI). The question is if the fuel controller would still work without it. My guess is it would pulse fine from equal-spaced pulses. You don't need an engine to test it. Spin the distributor by hand and power the controller w/ 12 V. You don't even need fuel, you will hear the injectors clicking. Indeed, before you even fool w/ the 5 V circuit, you could test that. You can generate a stream of 5 V pulses from a PC audio output, using freeware on the internet. See if that satisfies your fuel controller.
 
if your running a TBI engine, 1988-1991 LA series then it does not use a crank sensor, it uses a special distributor that has a hall effect sensor much like the one that the magnum uses except it mounts different

TBI won't run off a stock electronic ignition in mopar, unless you use all chebby parts

to clarify TBI Uses 2 injectors mounted on the throttle body, is that what you have?
 
An HEI module {$20 at any parts house} will generate the 5v square from the Mopar magnetic pick up, but the lack of #1 reference, and the movement of the advance mechanisms all contrive to cause issues.

I need two separate signals, one to fool the PCM, and one to trigger the coil using the curve of choice. Sounds simple when you say quick it doesn't it? :D
 
redlined said:
An HEI module {$20 at any parts house} will generate the 5v square from the Mopar magnetic pick up, but the lack of #1 reference, and the movement of the advance mechanisms all contrive to cause issues.

I need two separate signals, one to fool the PCM, and one to trigger the coil using the curve of choice. Sounds simple when you say quick it doesn't it? :D
Click to expand...

I'm confused..........since when does TBI need a "no 1 reference?"
 
Ditto 67Dart273.
As I suggested, try it and see if the injectors pulse. If so, you are good to go. Even if the factory phased the injector pulses to the valves, you don't have to. That was a very minor emissions tweak. I am sure it will run fine, as long as it gets the correct time-averaged fuel flow. I could disable the fuel and stand there squirting shots of starter fluid down the TBI throat and the engine would run.
 
The reference sensor comes in handy for dual plane manifold fed by a 2-barrel TBI. The operation of the injectors is alternating at low demand, then switches to both for heavy loads. The alternation provides less fuel time in the ports, for less fuel dropout. TBI injectors have high flow, so running half at the low end increases the fuel control resolution. Otherwise for full bank, low pulse widths do not even fully open the injector, and fuel delivery is nonlinear at those short intervals. This information comes service training manual of the era, but I have not verified. I plan to use similar strategy if I inject the Barracuda.

After reading the various posts I like the 66PlyValiant idea, "use chevy parts".

You can get there with nearly any of the strategies if you learn to interpret signals with a micro and change as needed. With today's integrated micro's it is an easy task. Much better than diddling with discrete logic.

Wish you lived closer.
 
Hell I'll give it a shot, can't hurt, might help. I had originally considered an HEI module anyway, but the #1 deal and continuously, and sometimes rapidly, varying signal input gives me pause. But let's assume that will run and have a negligible effect on pulse width and driveability and take a look at the 8pin module and mag pickup.

Am I seeing this correctly? The module uses magnetic pickup output to generate 5v on pin R via the AC to DC converter? If so that would only require the use of pins P, N and R on that side of the module by leaving the bypass relay de-energized.

Here's the kicker.....or maybe not. I've read that the GM style mag pickup makes more juice than the Mopar design since it has 8 poles that align at once, while the Mopar "AC generator" only aligns one pole at a time with the magnet. Whether that's true or not, can't say. If all of the above is correct, it raises the question of whether or not the single pole Mopar pickup puts out enough juice at low RPM to get the job done.
 

Attachments

  • gm-ign-module-schm.jpg
    43.3 KB · Views: 561
It will work. That is what TrailBeast uses for his system. It is more than a AC to DC converter, it is a voltage compliant comparator. When cranking at 200 RPM the signal can be as low as 2V p-p, but at 6000, it can be over 100V. The reluctance changes result in generation of the voltage signal. The sensing means is called variable reluctance, VR.

There are chips like LM1815, MAX9924, and NCV1124, that with a few components convert VR signal to TTL signal. I used the NCV1124, a locked rotor lean burn distributor, and my ECU, to make an electronic advance ignition. I did some rigorous testing and data logging, and found the MOPAR VR is not too good, unless you consider a few degrees of timing variation acceptable. The reluctor is small, and the roll pin offsets it on the shaft enough for runout related timing variances, related to varying signal strength changing timing. In the motion control industry VR sensing, is not used for position measurement, but is used for tachometer. It was an inexpensive way, but Hall or optical are now less expensive and better. I worked at MicroSwitch, they developed and manufactured the MOPAR Hall sensors, they had some early reliability problems, today they are robust.

I made a dual output optical distributor, ready for coil on plug, no rotor ignition. I wrote the code and well bench tested, but have not tested on car. I am working on a bucket list, to get 3 project cars completed and sold. That limits my play time. But if I clean my mess of cars and get space, I will get a dyno and have real fun.
 
redlined said:
, it raises the question of whether or not the single pole Mopar pickup puts out enough juice at low RPM to get the job done.
Click to expand...


There are many many of us using HEI modules fired by Mopar distributors. This difference does not seem to be an issue. You DO have to get polarity correct, or rotor phasing will be wrong

Somebody, don't remember, drew up this nice diagram

attachment.php


after I drew my real ugly one, LOL

zu5qn8.jpg


In any case, the male end of the dist. connector must be as shown on the 4 pin box to obtain proper phase. TrailBeast can show you on the multipin
 
So where's the 5v source?

From what I've read, {true or not?} on a GM pin B is powered 5v in the run position only. During cranking or below 400 rpm the ECM opens the relay bypassing the EST signal from the computer allowing base timing. That would require killing power to pin B, which seems to indicate the reference is powered from another source, the 12v input. The ECM would still require the reference signal to read rpm {400 or less command} regardless of pin B's on/off status.
 
I expect that the internal diagram of the HEI chip is incomplete, and for understanding only. I expect it generates 5 V from the 12 V source (BATT+). I understand the purpose of the Bypass is to make it revert to "base timing". In that case, it ignores the "B4" spark advance signal coming back from the computer. Simply don't connect the bypass pin and it will run in base mode. The "5 V out" label looks wrong, should be "5 V in to enable advance".

The advance signal simply time shifts the 5 V reference signal, as shown. A very similar thing is done in Ford's EDIS module, termed the SAW signal. Read more on both on the megasquirt site.

The 8-pin HEI module is simplest to wire. Grab a GM "external coil" and cable that goes with it and a simple plug & play. Also steal the connector for the pickup. I also grabbed a knock sensor and module (my Holley Commander 950 can use that). Connect to your Mopar distributor, spin it and watch it spark the coil. Twist the pickup wires or you can get "positive feedback" or "self-excitation". Best to drill a hole in an old dist. cap and shine a timing light to see where the rotor is when it sparks. You have a 50:50 chance of getting polarity right for the pickup wires. The only pin you need on the 4-pin connector is the ref pin. You will see that TrailBeast doesn't use that connector. My Holley controller does (w/ GM adapter harness).

As mentioned, then send the 5 V Ref signal to your Mopar fuel controller, power it with 12 V, and see if it pulses the injectors as you spin the distributor. If so, you are about done.
 
To check pickup polarity it is simple to place grounded plug and cable at coil, then use timing light on that cable to verify that spark occurs when reluctor teeth are aligned with pickup. Incorrect polarity will show between teeth, a horrible, iffy place to trigger a spark.

One way to get the near to the EFI trigger you need is to add a current transformer, with conditioning circuit, as sensor for #2 plug spark. A clamp on timing light uses such a sensor. There were a few crank triggered cars in the mid 80's that used such a signal for phase reference. The #2 signal, then lets you replicate, the offset tooth for #1, by stating with the normal electronic distributor signals. Yes it will involve a small 8 pin micro, but the task is easy. It avoids adding a reference tab and sensor to distributor. The timing variation due to mechanical advance may not create significant problems, but lack of correct reference pulse may force the ECU into limp home strategies.
 
You must log in or register to reply here.
-

Back
Top