Rpm signal loss

[seattlejester]

Got my high rpm signal figured out, but having trouble with my low rpm.

Using a drill or hand the Ecu can pick up a signal around 150 rpm.

We tried doing this while cranking to see if the power drain was causing issues. Still picks up signal fine.

It is a VR sensor, so I took the cover off and adjusted both potentiometers counter clock wise and then slowly turned clockwise. No luck.

This is a PCb 3.57. From the updates it says full sweep is 5 full turns. Not getting anything from adjustment in the range.

Any insight would be appreciated.

Edited March 12, 2014 by seattlejester

[seattlejester]

Sorry I may not have been clear. This is during cranking. Ecu won't pick up a signal. With a drill it reads rpm fine,

[AkRev]

I am sorry if this sounds too basic, but just to be sure. Are you saying that you tested this with the drill, at the same time you cranked the car?

 

My first thought would be loss of power when cranking, but I am also assuming that you would notice that by losing connection to the megasquirt from your laptop.

So my next thought would be what changed? Does the VR have the same gap when you test it with a drill vs in the car? What type VR sensor and trigger wheel are you using?

 

Your car should be cranking at about 200 rpm. If you feel your car is turning over slower than 150 rpm, that might indicate that you need to take a look at the battery/wiring as a weak spot. An easy way to test if it is just cranking speed, pull the spark plugs so the engine can turn over faster (don't forget to disconnect power to the coil.)  

 

Hope this helps....

[skirkland1980]

Just wondering if you've set your cranking rpm in megasquirt. Not sure if that will help but maybe. I haven't tinkered with ms for a while.

[seattlejester]

Yes at one point we did crank while running the CPS off of a drill. We had the injectors and the spark turned off, we were just worried that maybe during cranking the ECU couldn't pick up on the VR sensor.

 

Yes, same gap. Using the stock toyota CPS 24-2 opposing teeth I.E. 12-1 crank. 

 

I made a video and marked the CPS and timed it, it is cranking 60RPM on the cam, so 120RPM at the crank.

 

I am running a shrunken BT20 battery hooked up to a 30amp rapid charger, thinking it might be the battery, I took the deep cycle optima battery from my other car, and had the charger set to engine start mode (80amps). Engine cranked a little faster but still no rpm signal.

 

Thanks for the suggestions.

 

Cranking RPM was set at 300, we played around with it by setting it down and up a bit and even changing the start location of the fuel map, but I think it turns out the VR sensor was the culprit.

 

The VR sensor was just super super super picky about the gap. I noticed that I could get megasquirt to read a screw driver passing by it all the way down to 5 rpm, so playing around with the distance the screw driver was to the sensor got the problem. I got it to pick up an intermittent signal while cranking by basically having the sensor plate touch the toothed wheel and machine a new gap between them. 

 

Now just to go over my other settings, hopefully I won't be posting many more of these help me threads for a while.

Edited March 14, 2014 by seattlejester

[seattlejester]

AkRev: sorry I totally missed your suggestions/observations. I've edited my post above to reflect my responses.

 

So I ran the tooth logger for my intermittent signal at diyautotune's suggestion. Megasquirt was picking up maybe 1 or 3 teeth every other second at cranking speed. With a drill it was picking up a consistent signal, above 500 RPM but was not reading the missing tooth. The voltage/signal intensity spikes were really all over the place as the rpm's fluctuated. I think the stock toyota VR sensor does work at higher rpm, but at low rpm I think it is having problems communicating with the ECU.

 

After 2 days of cranking I am throwing in the towel. I bought a 6 3/4 36-1 crank wheel from DIY auto tune and a halls effect sensor. I am thinking I can tap my crank pulley bolt for the center and add a bolt to keep it from loosing center on one of the grooves. Hopefully we will have good news next week.

[OldAndyAndTheSea]

 

After 2 days of cranking I am throwing in the towel. I bought a 6 3/4 36-1 crank wheel from DIY auto tune and a halls effect sensor. I am thinking I can tap my crank pulley bolt for the center and add a bolt to keep it from loosing center on one of the grooves. Hopefully we will have good news next week.

 

 

You and me both. The optical sensor in my dizzy died when I decided to retest my cranking RPMS. I too am going 36-1 crank mounted trigger wheel. I have to wait for tax returns to do it though...We'll see what happens. I'm definitely not a fan of distibutors after this ordeal.

 

AND no more timing the stupid oil pump driveshaft!

Edited March 14, 2014 by OldAndyAndTheSea

[madkaw]

Well don't feel bad guys, it seems to be the major culprit in with ECU issues-OEM and aftermarket. I'd like to brag and say my set-up is working perfect, but if you have read my thread you would see I am still stuck at this point. The direction I went of drilling the flywheel was an attempt to avoid all issue like this, but.....

I think once I get my MS wired correctly for my sensor I might be okay.

Just got off the phone with a buddy who had all kinds of issues with his Jeep-CPS. It was replaced by the dealer but installed wrong by the technician, so thinking that was not the issue, he spent money trying to fix other things. Turned out that the EXACT procedure for install was not followed which varied the gap ever so slightly and caused miss-fire codes. 

[AkRev]

I must admit VR sensors are a pain. I am very happy with the 36-1 wheel from diyautotune, and am confident that you will get it to work. One frustrating things with Megasquirt in general is it is so picky on noise. So many times did I run into dead ends to find out that the wires needed better shielding, or I wired up the VR sensor backwards, or my personal favorite I had the gap wrong on the sensor. lucky for me, it was well worth it. My buddy claims I get giddy like a school girl every time I find and learn a new feature (latest discovery was the non-linear MAT corrections and how they work with the crappy weather of Utah)

 

On suggestion, you mentioned that you would add a bolt to keep it from losing center. I would hope you add two, 180 degrees apart to keep balanced. Also, keep a watch on the condition of the rubber in the crank pulley. I have seen one start to slip after the addition of a wheel... it was mostly the fact that it was 30 year old rubber. 

 

PS I do not know if this will help, with my set up, my potentiometers ended up at 1.25 turns for the bottom one, and .75 turn for the top. 

[seattlejester]

Andy: I definitely hear you on the oil pump, I spent maybe 2 hours checking the clocking on mine only to find I didn't have an ignition signal. Probably a good 10 times of checking tightening, loosening, ugh, definitely unpleasant. I was actually considering ordering a few more to play around with (different methods of attaching to pulley), if your tax return is far off, I wouldn't mind sending you one. 

 

madkaw: I tried quite a few configurations with gaps definitely finicky about it's location. The toyota sensor actually has 3 sensors in it 2 for the 1 tooth wheel and 1 for the 24 tooth cam wheel. Turns out all three sensors are the same, so I may play around with the other two and see if it was maybe an old sensor. Granted the other two sensors would be just as old...but it should give me something to do over the weekend while I wait for my new trigger wheel.

 

Akrev: I'm pretty much thinking that the VR sensor just outputs too small of a voltage at low RPM to read through the wire harness or something of that nature. I bought additional shielding so I will definitely shield the bit that goes to the new halls effect sensor. Thanks for the potentiometer settings. I literally tried 1/2 turn increments for a good 5 turns, bottomed it out and tried again, so I think that it may be better to just cut the cord after switching out the sensor tomorrow. The 36-1 wheel will have better resolution anyways .

 

 

The DIY on the diyautotune sight showed one bolt to steady, although they may have just been demonstrating one hole. I think the 7mgte pulley is just like the L-series pulley in that it is not used to balance but instead just used to run accessories off of, but I may just add the second bolt for reassurance. The current problem is that I have a full radiator and intercooler, so it is kind of tempting to actually just weld the trigger wheel to the pulley, although I'm not sure how I would get rid of the center portion. I think it would be better to remove radiator and intercooler, and just drill into the pulley bolt for the center bolt and then space it out with a few washers and secure it on the grooves in 2 locations 180 apart, unless the pulley is really easy to drill into, in that case 4 holes at 90 apart.

[AkRev]

I realized after the fact that my potentiometer settings will not help you due to the fact of how many variables there are in every setup. Hope things are going better with the DIY wheel. Here is a pic of how I drilled mine to the outside of the pulley. I will keep it mounted this way but I will move the sensor when it comes time to put the engine back in the car.  

[NewZed]

 I bought a 6 3/4 36-1 crank wheel from DIY auto tune and a halls effect sensor.

Not an expert.  But I think you need a wheel with magnetic teeth to drive a Hall effect sensor.  The Hall effect is what happens when you push a magnet through an electric field.  With the steel DIY wheel you need a VR sensor.  Which is what you already have with the 24-2 Toyota setup.  You'll just be replacing one VR wheel with another and installing a sensor that won't work right with either.

 

Edit - Never mind the above.  They must have developed sensors based on the Hall effect that will read a steel tooth's effect as it passes by.  They say it works.  Good luck.  Not sure why the sensor wouldn't work with your 24-2 setup though.

 

 

A VR sensor should be able to pick up tiny voltage pulses.  If you can't get the Megasquirt circuitry to do it directly, you could probably use a GM HEI module to amplify and condition the VR signal.  I think it's described out there somewhere.

Edited March 16, 2014 by NewZed

[AkRev]

NedZed, you are correct. The 36-1 are designed to work with the edis system. These are VR based. Good catch, I read over that the first time. 

[seattlejester]

Yea it is a bit confusing, I believe the setup you are referring to is called the flying magnet style of hall effect magnetic wheel. You wouldn't want a magnetic sensor trying to pickup a rotating magnetic field, it would be quite chaotic if that was attempted. 

 

I'm tired, so this may seem a bit off, but for anyone who is curious in the future, the difference to my understanding...

 

A VR sensor reads the flux in the magnet. The magnet is permanently attached to the back of the sensor and it is pulled at maximum when a metal object is close and a voltage is generated as it passes by. The spike can then be tuned with software. The problem with these lie in the fact that...one your coil has to generate enough of a field that the passing object will generate a detectable flux, and two the signal generated needs to be processed to get a signal for the passing teeth. This is preferred when the tooth wheel is small, wiring is difficult (no power wire needed), or the location demands maybe a hotter operating temperature.

 

Hall effect is a sensor with a magnetic field as well although most are now electrical coils or funneled through a crystal of some sort (my apologies my memory is failing me at the moment). When the electrons are knocked off course it basically fails to read. Thus hall effect sensors basically have on/off detection, either the electrons make it to the end, or they do not.

[NewZed]

I did a little more looking around and found that DIYA specifies that they're Hall sensor works with their square-tooth wheels.  Implying that they don't work well with the pointy teeth wheels.  So the sensor might not work with the 24-2 Toyota wheel (I have no idea what tooth shape Toyota has).  

 

http://www.diyautotune.com/tech_articles/trigger_wheels_index.htm

 

For anyone thinking about just getting a DIYA Hall sensor to solve a trigger problem.  There's more to it.

[AkRev]

I did a little more looking around and found that DIYA specifies that they're Hall sensor works with their square-tooth wheels.  Implying that they don't work well with the pointy teeth wheels.  So the sensor might not work with the 24-2 Toyota wheel (I have no idea what tooth shape Toyota has).  

 

http://www.diyautotune.com/tech_articles/trigger_wheels_index.htm

 

For anyone thinking about just getting a DIYA Hall sensor to solve a trigger problem.  There's more to it.

 That is interesting. That makes me wonder if the hall sensor is as easily susceptible to noise like the VR sensors are. In general they state that hall sensors have a high degree of positional accuracy over the VR.

I was using the msextra site as a bible. http://msextra.com/doc/ms2extra/MS2-Extra_Dizzy.htm Makes the claim about the hall sensor needing a flying magnet wheel. If the hall works with the diy 36-1 wheel, I might just get one to try out and see if that can cure some of the noise issues I can still see in my logs. 

[Metro]

I did a little more looking around and found that DIYA specifies that they're Hall sensor works with their square-tooth wheels.  Implying that they don't work well with the pointy teeth wheels.  So the sensor might not work with the 24-2 Toyota wheel (I have no idea what tooth shape Toyota has).  

 

http://www.diyautotune.com/tech_articles/trigger_wheels_index.htm

 

For anyone thinking about just getting a DIYA Hall sensor to solve a trigger problem.  There's more to it.

 

I use the same Hamlin hall sensor DIY sells with an '81 L28et crank wheel modified to 30-1. It has tiny teeth, more akin to nubs, about 2-3mm high with a 10mm gap between each tooth. I haven't had a single issue with it since I installed it, reading RPM up to 7000rpm. If you go by the documents, the 81 wheel is probably out of spec, but the sensor is very versatile. 

Edited March 18, 2014 by Metro

[seattlejester]

If you look at the specs it specifies a minimum of 2mm wide and 2mm gap between each tooth for the hall effect sensor to read. I believe it has something to do with a switching voltage. If the teeth are too pointy the flow of electrons can just be interrupted at high speed and cause the ECU to miss or ignore the signal.

 

I think you may be confused if you think I am trying to use the hall effect sensor with the toyota wheel. I ordered the DIYA 36-1 wheel to use with the hall effect sensor. Won't be using the toyota wheel.

 

The teeth on the toyota CPS are actually fairly flat so I actually wouldn't be surprised if just buying a hall effect and drilling into the housing may actually fix a trigger problem. If someone was truly lazy they could also weld the smaller 3 inch 36-1 tooth wheel onto the CPS wheel and add a hall effect sensor off the side of it as well.

 

I'll take a tooth log once I install it. I have a few tooth logs from my drill powered VR sensor so it should illuminate the noise question.

Edited March 18, 2014 by seattlejester

[seattlejester]

So hall effect sensor is working quite a bit better. RPM signal is a lot more consistent. I swear I can hear the motor catching once in a while, but still nothing promising yet. I've increased pulse width of the injectors increased the duration of the coils while cranking, but still no dice

 

Current msq

https://drive.google.com/file/d/0B21hrliT9P7pbUF2b2NFYnRtZ1U/edit?usp=sharing

 

Tooth logs and composite log

https://drive.google.com/file/d/0B21hrliT9P7pUmNORXlwcUJGc3M/edit?usp=sharing

https://drive.google.com/file/d/0B21hrliT9P7pSmZZRm1kSjEtYjg/edit?usp=sharing

 

 

[AkRev]

So hall effect sensor is working quite a bit better. RPM signal is a lot more consistent. I swear I can hear the motor catching once in a while, but still nothing promising yet. I've increased pulse width of the injectors increased the duration of the coils while cranking, but still no dice

 

Current msq

https://drive.google.com/file/d/0B21hrliT9P7pbUF2b2NFYnRtZ1U/edit?usp=sharing

 

Tooth logs and composite log

https://drive.google.com/file/d/0B21hrliT9P7pUmNORXlwcUJGc3M/edit?usp=sharing

https://drive.google.com/file/d/0B21hrliT9P7pSmZZRm1kSjEtYjg/edit?usp=sharing

It looks like you are flooding out the engine.

Are the plugs wet?

 

Reason I say this is your priming pulse (assuming 60 degrees) is 14ms and cranking is 200% of the VE table with 440 injectors. To be honest I use about 5ms cranking pulse with half the injector flow (my MS is pw for cranking and not ve% like yours)

 

You have flood clear set at 50% tps, Make sure your tps is calibrated and floor the petal to clear out and confirm. 

 

Good luck,

 

Thanks for sharing your results with the Hall sensor, that gives me a good direction with my car.

Edited March 22, 2014 by AkRev