Have an LS3 in my 65 vette, and I am having issues with the electonic speedometer that I purchased. The issue I am having is the speedo for sure is getting interference from one or more coils causing the pulse reading portion to be erroneous. I am sure this has been an issue for someone previous. A search of the forum has turned up using a pull up resistor, just curious if anyone else has had this issue and what they found for a solution
Thanks in Advance
G

 

Gerry - need to put your equipment, speedo by?ECU by? Coils by?

 

Its all pretty standard, 6.2 LS, the coils are AC Delco, ECU is a stock modified chev unit,dialed in for aftermarket use. The speedometer unit I purchased through Corvette Central, other vendors like Zip carry the same unit. I cannot find out who the manufacturer is as it is not listed on any of the descriptions. GM part number 77115669. The signal wire from the ecu is giving me a signal from the engine through interference I believe. When I go to calibrate, the speedo is fired up, reading something other than the speed as the car is standing still. I think a pull up resistor may work but I am a little worried about frying the unit. Was hoping someone had a solution.

 

Shielding
twisted pairs

both don’t fry units

no idea where you’re doing this jumper with out pictures and wiring diagrams

somewhere you got the idea which means there could be details around that

 

Are the speedo wire(s) near the coils? Reroute the speedo wires or how about putting shielded wires on the speedo?

 

Let's apply a little physics here. Twisted pairs work great for balanced signals but this sounds like you are using a single wire thus it can't be balanced. The first thing I would try is adding a shield wire that extends the full length of the signal wire and spirals around it. It should be grounded on one end only and the other end should end in a small loop around the signal wire. Grounding both ends may sound like a good idea but it's not because that can set up ground loops which can make the problem worse. The idea behind the shield wire is to reduce stray fields which seems to be what you are seeing. Surprisingly, the spiral wrap (maybe 1 cm. between loops) looks like a solid shield to anything except very short wavelengths. With only one end grounded, it can't hurt anything and it may well help. Beyond that, without schematics there's very little to go on.

 

Something like this possibly.

https://www.mcmaster.com/products/em...ding-sleeving/

 

Ok so purchased some shielding for the signal wire, pulled it from the ecu and routed it independently from the rest of the wiring harness, grounded one end only, still getting lots of interference. Then tried a pull up resistor which is essentially a 12v supply with a 1K resistor tied into the signal with the hopes of eliminating some of the noise, again no luck.
Tried an independent ground as well , there just seems to be a lot of noise associated with a modern system and a fiberglass body.
I purchased this unit from one of the vendors and it is available on a number of sites, description makes it sound like a plug and play unit, cannot beliieve no one else has purchased this unit for a C2 with an LS3.
Any help would be appreciated

 

Without an oscilloscope or spectrum analyzer attached, it's almost impossible to pin down what's going on. The noise may not even be coming over the signal wire but rather the power lead. As a couple of shots in the dark, your next cut at this might be wrapping a few loops of your signal wire and power lead around ferrite cores. To the greatest extent possible, your signal wire should only cross other wires at right angles to minimize coupling. You could also try a terminator network on your signal wire rather than tying just to the +12V side. The idea is that you replace your single resistor with a pair of resistors of usually equal value connected in series with one connected to +5V or +12V and the other connected to ground and the signal wire is connected to the middle. I'm really surprised that Corvette Central's support can't contact the manufacturer's support and give you some ideas.

It would also help (a little) to know if your get any speed readings when your sitting still. If it's the ignition coils causing the problem, you might try disconnecting 2 or so of them and see if the problem gets better then try another pair to narrow down the hunt. The other approach might be to punt and install a GPS-based speed sensor to feed this thing and don't overlook the notion that your unit is simply bad. What bothers me is that this thing may need nothing more than a capacitor on the power lead but only the guys that built this thing can tell you. There has to be some way to get some level of support.

 

Go after corvettecentral

It appears you have a hybrid of original metal gauges and somehow modern bits and pieces

Did you buy it recently?

If yes, it’s pretty clear what they are selling

did you read the description?

If they are no help, make it public where it counts

 

I think I found the installation instructions for your unit and I assume that you are using the green wire to connect to the ECM output and leaving the yellow (S1,S2) wires taped but unconnected. It's not clear how this thing knows to ignore the yellow wires in this scenario. I would guess (and it's just that) that those yellow wires are acting like antennae and they are the ones that need shielding or some form of termination. The instructions also mention interference problems and want you to add foil-wrapped insulation to your firewall. This thing sounds like a device that works in spite of itself and that a Corvette is a less than ideal environment for it. Again, you need support not SWAG's.

 

What transmission are you using? Sounds like you might need to get the signal off of the VSS on the trans. I have a LS3 with TKX, but I'm using AutoMeter speedo. I used to have the electronic one that you have. On the AutoMeter, I think that I put a 1 ohm or something, it came with the AutoMeter LS conversion kit, and it made it work great. I had two wires switched at the Harness connector at first. Once I figured that problem out, it has worked flawlessly.

 

Yup, hard to determine if we are dealing with conducted (through the wires) or radiated (through the air) interference.

Some ferrite snap on cores may help as mentioned. Also verify that the ground to the gauge goes to a good ground, not just bundled in with other ground wires. They can be long and also introducing noise.

Not knowing what the signal is, or should be, you may try a small value capacitor (like 1000pf or 0.001uf, same thing) on the signal line to gnd near the speedo. Too large a value may distort the signal so it may take some experimenting.

As was mentioned you could try some terminations. I've had good results with putting a series resistor and shunt capacitor (to gnd after the resistor) inline with oscillator signals to "soften the edges" a bit which helps reduce harmonics. I'd try a 10 ohm or 100 ohm in series with 100 or 1000pf caps.

I understand the typical mechanic does not have these parts readily available. Having spent years designing products followed by doing EMI testing in various labs, I kind of have a bunch of parts laying around.......

Tom

 

A quick review of the installation may help as well.

All ground wires should be short and tied to a solid metal ground (i understand we are talking about a plastic car....). As an example, a 3 ft ground wire for your lights may work just fine but a 3 ft gnd on an MCU may just become a big radiator of noise. A good DC ground (or return) may not always be a good RF ground. Best to keep them as short as possible.

 

Just a shot in the dark here, but have all your tests been taken with the speedo on the bench? Is it possible that the case needs to be grounded ?

 

Thanks for the input, I do have a TKX tranny with the two wire sensor so I think i will try and use that for the source, run the wires up through the shift boot, tie into the yellow wires, shield the wires, try an independent (shielded) power wire and short ground. Will call the vendor tomorrow and see if they have any suggestions. Of interest is the odometer reading, when in calibration mode, and the engine is running starts counting pulses when it is supposed to be reading the sensor which would be static as the car isnt moving. To me this suggests I am getting interference through the ignition system, and with the metal dash pod, wiring etc may be very hard to determince where the signal is getting in. Instructions say to shield the fire wall, not sure that is even feasible, would the steering coulum, pedal assembly, etc not be signal conductors. Again thanks for the suggestions.

 

I have the same speedo. It worked perfect on my TKX vss output. I installed a 4L60 with a GM ecu it won't work. I'm still fiddling around with it, but in the meantime I'm using a GPS signal to that speedo and it works fine. As a side note, when I went to the GM ecu I had to use a pull up resistor to get the tach to work.

Just connect the two yellow wires to the TKX plus the 12V keyed and a ground and your in business. Calibration is easy and I found the unit very accurate. The only thing I don't like is the odometer, hard to read.

 

Nice to hear it worked for someone that is enough encouragement to try and resolve the issue.
I for the most part enjoy challenges like this, but taking off or putting on the oil sender fitting along with making sure all the bulbs and plugs stay in place without scratching something makes the job annoying

 

Again with no schematics to go on, I strongly suspect that the two yellow wires carry differential signals so now a twisted pair from the speedometer to the sensor is a good idea. Nothing really special is required just connect the wires and twist them around each other. The idea is that the current in one direction cancels the current in the other resulting in no magnetic field. If you then want to add a shield wire to the the twisted pair, so much the better.