The Top DSM Community on the Web

For 1990-1999 Mitsubishi Eclipse, Eagle Talon, Plymouth Laser, and Galant VR-4 Owners. Log in to remove most ads.

Please Support JNZ Tuning
Please Support Rix Racing

1G Tach Question

This site may earn a commission from merchant
affiliate links, including eBay, Amazon, and others.

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Man if ANYBODY would ask, I have the cars here to show the answers usually.
Appreciate the thanks! Just being as helpful as possible for the whole community! :)
If you need pictures of connectors (like I do for a auto temp sender pigtail) @brads is where I always go and is always super helpful!
Thanks @steve, you have contributed a LOT of information to this thread! Really appreciate that!
Marty

@steve, what would happen if someone just connected the filter wires together?
Your thoughts on this are important.
:idontknow:
 
Last edited by a moderator:
I'm curious if anyone has any feedback on my variation on this question. I have the same issue (tachometer doesn't work, Dsmlink does show the rpm). However my setup is a touch different. I have a 1988 chysler conquest with a 1990 ecu and 91+ wiring harness (very heavily modified) and a pair of evo 8 coil packs. As a result I do not have a PTU and pin 109 is empty. As a result, I took ecu pins 54 and 55 and built a NOR gate to hook them up to. The output of this NOR gate is hooked up to my tachometer. This does not work. The reason I put a NOR gate in is because it appears that's how the factory PTU works based on the factory wiring diagram.

You must be logged in to view this image or video.


I think it would be good to be able to replicate the factory PTU in a more compact manner.
 
The output of this NOR gate is hooked up to my tachometer. This does not work. The reason I put a NOR gate in is because it appears that's how the factory PTU works based on the factory wiring diagram.

I think it would be good to be able to replicate the factory PTU in a more compact manner.

I have no idea except, which PTU would you want to replicate? The 1990 or the 1991 and later?
The one you have pictured is the 1991 and later.
 
Thanks Steve! This thread will be a ton of advice, assistance, and facts. If I had a "spare" filter, I'd open that thing up and see what the hell is going on inside. I'm curious as hell.
Appreciate your input!
But Steve, the coil pack has a "Tach Filter" attached on a 90, what does that dam little extra metal box do on the 90 harness?
IDK is an honest reply :)
 
Thanks Steve! This thread will be a ton of advice, assistance, and facts. If I had a "spare" filter, I'd open that thing up and see what the hell is going on inside. I'm curious as hell.
Appreciate your input!
But Steve, the coil pack has a "Tach Filter" attached on a 90, what does that dam little extra metal box do on the 90 harness?
IDK is an honest reply :)

Marty, don't open up that little 284L02901 that you showed in post 4.
If we want to figure out what's inside, we could do it non-destructively with a multi-meter that measures capacitance.
I think. I'm kind of assuming there are only passive components in there, like a capacitor and a resistor.
Do you have a multimeter?
 
I forgot there are two things on the bottom of the 90 coil pack. The Tach Gate and the Capacitor on the power feed.

If the Tach Gate is really an OR Gate on the 90 and a NOR gate on the 91+ PTU's as the diagrams show then the two signals are inverted.

You must be logged in to view this image or video.


I can't find the diagram I had for the noise filter so I can't say if it's a high pass or low pass filter for the signal the tach gate generates before that signal goes into the cluster and ECU but it sends part of the signal to ground and the rest on.
 
I suppose it may help to look at the conquest wiring diagram in my case to get a feel for what the tachometer is expecting, although I would think if it was a matter of having a nor gate instead of an or gate the tachometer would show half or double the actual value instead of zero like what I'm seeing. Steve, do you know what the signal should look like? I'm guessing a square wave with a 50% duty cycle and twice the frequency of the engine rpm. A good starting point for me might be verifying that my circuit is creating the output I expect it to.
 
Steve, do you know what the signal should look like? I'm guessing a square wave with a 50% duty cycle and twice the frequency of the engine rpm.

Not off hand but I doubt it looks like that. This is connected to the low side of the coils. So if it's a NOR gate it's going to be high only for as long as the coils are being charged.
 
I want to see this work.
Hell, if that just plain works, lets write it up so all the 90 owners can have that info. :thumb:
Those little filters are hard to find and nobody knows whats inside, LOL.
You can actually just bypass that filter and the tach will work fine.

To add a little more info on this subject the 90 and 91-94 tachometers are different. the receive different kinds of signal to operate, the 90 being from the tacho interface and the 91-94 which comes from the power transistor. to tell the difference looking at them the 90 has the speedometer in increments ending in 5 where-as the 91-94 is in 10's. I'm currently dealing with one that's giving me issues if someone could shed some light. Car is a 90, has the engine harness, ecu, transistor, coi pack, and cluster out of a 91-94. Tach still does not work.
 
Last edited by a moderator:
If you try to use NOR or OR gates on those 90 wires you will probably let the smoke out. Possibly in a spectacular fashion. The low side of the ignition coil switches to ground to dwell, and spikes up to several hundred volts when the coil fires. Most OR and NOR logic gates are not set up for that. They may use a logic gate symbol, but in most cases it is a wired NOR, which is basically using NPN transistors.
As I recall from designing a tach adapter(and it's been a long time since I did that so my memory may be wrong), the 91- PTU isn't just looking at the coil signal 0-12V. It actually looks for that high voltage spike, and ORs those together. So a misfire would be indicated to the ECU. That is why aftermarket ignitions require you to either add an adapter to simulate that spike, or bypass the PTU and run a tach wire from the aftermarket ignition to the tach output wire going from the PTU to ECU.

In the case of the Conquest rewire here, since you are wiring the NOR to the gate of the ignition coil transistor, you will be seeing low voltage, and probably safe. But even then, tying each of those signals to the gate of a separate N channel transistor, tying the output of those together with a pullup resistor to 12V to make your signal, should get it working.
 
I'll have to boot up my desktop to find the circuit diagram for the nor gate, but I did use mosfets to build a nor gate, not an off the shelf nor gate as I couldn't find any that function at 12 volts. I also just pulled the gauge cluster out on my conquest and hooked it up to a DC power supply. By tapping the positive output of the DCPS on the ignition input for the gauge cluster the tachometer moves, so I'm fairly confident the duty cycle of the signal or wave form doesn't impact the result. I imagine a 1g dsm is the same. So for me the next step would be to get an oscope on my nor gate to see what it's producing as well as verify that I haven't crossed any wires anywhere. I imagine I picked some nice big resistors to ensure I don't starve the ignition system of any current so I may be just producing a really weak signal.

Based on the wiring diagram you are correct that the 91 PTU looks at the high side where as I am looking at the 12 volt side. In my case the RPM isn't showing on the gauge. I'm not sure if the ECU is also not seeing the tach signal but I'm not getting misfire CELs. Maybe if I play around with dsmlink more I could see what the ecu is seeing on that pin for the tachometer input.
 
So I grabbed a signal generator and my conquest gauge cluster along with a 1g and 2g gauge cluster. The 2g gauge cluster is really easy to get to work, pretty much any signal of at least 10 volts with any duty cycle or wave form will result in the tachometer responding really cleanly. It's simply frequency (hz) * 60 / 2 = rpm. I wasn't able to get either the 1g or conquest tachometers to work under any conditions. My signal generator can do up to 20 volts and sign, square, or saw tooth waves. I took the conquest tachometer out of the housing in case any of the reed switches or diodes were causing issues and it still didn't work. Then I swapped the positive and signal wires and it worked poorly. The tachometer wouldn't go past about 3500 rpm and it seemed very inconsistent. I'm going to try one more test where I swap the signal so that instead of having the low side attached to ground and using the high side as the signal I use the low side as the signal and attach the high side to the high side of my DC power supply. Not expecting it to work though.

So surprisingly, that actually worked. The conquest tachometer seems to be rather particular about duty cycle, 50% seems to work best, so I'm not sure how that works. I think the problem I was seeing earlier regarding the signal dropping out was because I was testing with 20% duty cycle. As the frequency gets higher I think the duration of the pulses are just too short for it to detect them. That being said, I can't figure out the equation for the signal and think this tachometer may be damaged or something. It seems like it reads about 15% lower than the 2g tach at all frequencies. I'm just going to try and retrofit the 2g tachometer into the conquest gauge cluster. I've been trying to avoid that since there are other idiot lights integrated into the tachometer that I'll effectively lose unless I do some Frankenstein stuff. I may just remote mount those idiot lights anyways.
 
Last edited by a moderator:
Good info. I've found in the past that using a signal generator to emulate a signal can sometimes have unexpected behavior with whatever signal conditioning is inside the target. Sounds like that was the same reason you changed around your test wiring. Using the signal generator to trigger an nFET, with a pullup resistor(or better an inductor) would make it look more like the signal off the coil negative that is what the tach normally reads.
 
Just thought I would post this for anybody who has one of these little 1990 "tach filters" out where you can get at it with a multimeter that measures capacitance. This is the thing Marty showed in post #4.

This is how your meter should react to it, if the thing is a resistor and a capacitor as proposed in post #21.

You would make these 6 measurements: 3 resistance, and 3 capacitance.

I did this with a known resistor and a known capacitor put together as shown in the diagram, and wrote in what the result of each measurement was.

So anyway, this is how we could probably find out what's inside without taking the thing apart and probably wrecking it. I would measure mine but I can't even see it - hiding somewhere under the intake manifold I guess.


You must be logged in to view this image or video.
 
are you talking about the tach filter shown in post #4 or the other thing steve showed in post 33?
 
are you talking about the tach filter shown in post #4 or the other thing steve showed in post 33?

I'm talking only about the little module that Marty showed in post #4. Not the one Steve showed in #33.

For whatever it's worth, I'll show part of this 1990 4g63 wiring diagram that a lot of us have been using.
Where it says "This wire goes to your tachometer" - there are 2 little items in that wire from the coil.
The first one is labeled Noise Filter (tach resistor). That I would think is probably what Steve shows in post #33, the little black squarish module with "9818" printed on it.
The second one isn't labeled at all and the symbol used there is nothing I know of as a standard symbol. But it is shown with a ground connection going off to the side. That one might be the one Marty and I are talking about, and it looks like the guy who made this diagram didn't know what it was either!

You must be logged in to view this image or video.
 
Last edited:
That symbol is a shielded wire symbol, I believe. It just means there is a shield wrapped around the wire, and grounded at some location.
 
The tach filter does nothing but try to cancel interference out of the radio. It has nothing to do with the car running.

As for the tachs, like the sifter cable setup it all has to match. 90 tach, 90 ptu, 90 coil pack. Since people switch out clusters and gauge face at this point hard to tell, but there is a part number on the top of the cluster you can reference provided no one has touched it.
 
The second one isn't labeled at all and the symbol used there is nothing I know of as a standard symbol.

If you mean this?
You must be logged in to view this image or video.


It IS an indication of a shielded wire.
 
So I had my car running the other day and checked the tacho. It wasn't working. I'm pretty sure it's because I had assumed the ECU pin 54 and 55 go high when the ignition fires when in reality, they go low. Here is the circuit diagram I've now just converted my system to:

You must be logged in to view this image or video.


The LED is the output (white wire that goes to the tach, doesn't actually connect to ground in my circuit) and the two switches are pins 54 and 55 of the ECU (again, these don't actually go to ground as the ECU is acting as the ground). I'm debating whether to decrease the size of the 10k resistor as the voltage drop on the test bench is about 1.5 volts across the system. the mosfets I'm using are both IRF3205's.

For anyone who's curious this is the website I use to build test circuits: https://www.falstad.com/circuit/
 
Support Vendors who Support the DSM Community
Boosted Fabrication ECM Tuning ExtremePSI Fuel Injector Clinic Innovation Products Jacks Transmissions JNZ Tuning Kiggly Racing Morrison Fabrications MyMitsubishiStore.com RixRacing RockAuto RTM Racing STM Tuned

Build Thread Updates

Vendor Updates

Latest Classifieds

Back
Top