EDIT: Compushift came up with a solution to this issue. See Post #7 for an update.

Hoping someone has seen this problem before or has some ideas about how to diagnose it better. Sorry for the long post, but this is a bit complicated and I want to make sure I provide a lot of info.

Short version: Autometer electronic speedometer needle bounces around and twitches erratically at idle and also when the ignition is on but the engine is off. Speedo is accurate and operates normally above idle. The interference seems to be related to the new transmission’s pressure regulator solenoid and/or the signal being sent to the solenoid by the throttle position sensor (TPS).

Long version with more info: Spent the better part of this year installing a new engine and transmission in my ’73 along with EFI, new fuel system, etc. Engine is a 383 SBC built by Smeding Performance. I installed a GM serpentine belt kit with all new drive accessories. EFI is Edelbrock Pro-Flo 4 XT. Transmission is 4L60E from Silver Sport Transmission. TCU is Compushift Sport. The TCU and EFI ECU communicate with each other via a CAN Bus so the only TPS on the engine is the one for the EFI system.

Just recently finished getting everything up and running and have been test driving the car this week. Everything else seems to be operating well. No obvious electrical issues with the engine, transmission or either of their controllers.

When I first turn the key to on with the engine off, the fuel pump primes and after it shuts off, I hear a loud buzzing. At first I used my stethoscope and found the noise appeared to be coming from the TPS. It must just be some type of weird resonance as the noise is actually coming from the transmission pressure regulator solenoid. It’s very loud. Compushift says this is normal and is a function of the way they operate the solenoid for better pressure control. At this point, if I give it just a bit of throttle, the buzzing gets fairly quiet and with just a bit more throttle, goes completely away. Compushift says the pressure regulator solenoid is controlled by the throttle – more throttle equals more pressure.

While the pressure regulator solenoid is buzzing is when the speedometer needle behaves erratically. When I give enough throttle to stop the noise from the solenoid, the speedo needle goes to zero and stays there as it should. Compushift thinks there is either a grounding issue or a wiring issue but they are stumped as they’ve not seen this before.

Nothing else appears to be wrong electrically with the car and there are no apparent issues with the TCU or EFI ECU so I’m not sure about a grounding problem.

I checked all the grounds while the engine and transmission were out, replaced and upgraded the battery cables, replaced and upgraded the charging circuit, installed an accessory fuse block, and beefed up many of the grounds including the engine grounds.

The sensor wire to the speedometer from the TCU runs in the same wiring harness bundle that contains many of the other wires from the EFI ECU and TCU. It does not run in the harness that extends to the transmission itself so it is really not close to the wire that runs to the pressure regulator solenoid except where both exit the TCU.

I did some testing with the ignition on and engine off while watching the speedo needle move erratically by starting and stopping various electrical loads on the car – headlights, dash lights, interior lights, AC blower (compressor is not connected yet, still need to charge), stereo, etc. Nothing I did seemed to have any effect on the needle except one thing. When I run either power window, the speedo needle goes to zero and stays there while the window is operating (either up or down). Once the window is no longer operating, the needle starts behaving erratically again. However, the power windows do not seem to have any impact when the engine is running - the speedo needle is still erratic at idle.

I have an email into Autometer but they were closed for the day when I sent it yesterday so I’m hoping for a response on Monday.

Appreciate any help on this one.

DC

 

Double check the ground on the back of the Auto Meter speedo?

 

How are these parts grounded? The ECU, TCU and gauges? I'd think you want the ECU and TCU grounded to the engine only and the gauges just grounded under the dash.

The PCS making that much noise is rather strange. It is PWM controlled so it might be running at an audible frequency, but still that loud is strange when you consider GM factory stuff you can't hear it all all.

 

I don't really want to take the dash apart to verify the ground just yet as that involves removing the seat, steering column and is not a pleasant chore. The speedometer and other gauges share a ground and all are operating normally except for this noise issue so hopefully I can figure something else out before diving into the dash. Plus, I had another ground wire to connect as a test - see comments below.


Agree that the noise is strange. It was a little quieter after the engine cooled down after yesterday's test drive. Be interesting to see if it's back loud again when I start it up the next time.

I wired out an accessory fuse block under the hood. 12V+ and ground to that fuse block are connected directly to the battery (actually via distribution blocks mounted in the battery compartment). Accessory fuse block has a ground stud that received the ground from the battery. The fuse block has a relay for the ignition hot circuits and the trigger wire for the relay is connected to the ignition terminal on the factory fuse block. ECU & TCU get their power from the accessory fuse block and both are grounded to the ground stud on the fuse block. Gauges also get their power now from the new accessory fuse block but are grounded under the dash.

I tried various things with a test lead including connecting various other ground points (engine, bell housing, radiator, etc) to the accessory fuse block ground but nothing made a difference. Just for grins, I tried connecting the ground wire for the fuse block to one of the mounting screws for the TPS but again it made no difference.

The speedometer originally got it's signal from a Hall Effect transducer mounted on the speedo output of the transmission. That transducer had a ground and a 12V+ that were connected to the same ground and hot for the speedometer. A separate wire sent pulses to the speedo. That pulse wire was reconnected to the speedo output of the new transmission controller. When I removed the transducer, I originally just taped up the two extra wires. Yesterday, I pulled the extra hot wire under the dash and coiled it up out of the way of other wires as I couldn't get to the other end to remove it completely without taking the dash apart. I also tried connecting the extra ground wire directly to the new accessory fuse block ground as that would tie the speedo ground to the TCU ground but it made no difference.

One thing I'm not sure about. The power windows have always operated without having to turn the ignition on. I have been assuming that the previous owner bypassed the relay somehow but last time I was in the console, I didn't notice any wiring that looked Bubba or out of place. This may not be related to the noise issue at all but I suppose I should diagnose the power window circuit just in case.

DC

 

Does the speedo receive the VSS signal from the ecm directly? And then the ecm gets the VSS signal from the speed sensor on the 4L60E? On one of my LS swaps with a completely new 4L60E with all new electrical components had a bad VSS sensor. Speedo was going crazy. Didn't shift right. Local guy that does tuning took about 5 minutes to figure it out. Replaced the VSS sensor and it ran like a million bucks.

 

Yes. There is a VSS output on the tail shaft of the transmission which connects to the TCU. The TCU has a speedo output that gets connected directly to the speedometer. Transmission shifts fine. That's the crazy part to me - there are no other issues apparent at this time.

DC

 

Thought I would provide an update for this issue. Autometer had me run some tests that pointed to the interference being introduced onto the signal wire rather than being EMI/RFI related. Compushift's Director of Research and Development got in touch with me and came up with a few ideas to resolve the issue. There was an extra pin output (Pin #14) on the Compushift Sport controller that provided a 5V signal. He sent me a wire to plug into that pin and a 5.1K Ohm resistor and had me wire it as follows:





Here was his explanation: "The problem here is that this speedometer is very sensitive to oscillations about ground. With the ground for the controller being very slightly different than the ground for the speedometer, the ripple voltage on the controller ground was triggering the speedometer. The pull-up resistor moves the floating voltage away from ground, and the speedometer doesn't trigger any more."

Interesting problem and solution.

DC