Using the Stock Tachometer in 260z-280z (3 wire tach) with RB25 (or Z32 ecu)

[mtnickel]

Hey guys,

So I tried to read around for a definitive word about how to get the stock tach to work with the RB25, but came up a little empty.

Some have said this:

1) It will work if you use a simple pullup resistor

http://forums.hybridz.org/topic/109455-rb20det-utilizing-stock-gauges/?p=1023058

http://forums.hybridz.org/topic/69786-tach-issue-after-swap/?p=674261

 

2) It will work with an MSD tach adapter

 

3) It won't work no matter what.

 

Well I tried the pullup resistor and had no luck.

As a little background, the stock tach triggers off the coil negative signal. The coil negative signal actually looks like this:

reference (and a good read): https://linnbenton.edu/auto/scope/ignite/pri.html

 

Before point 'a' the coil will show 12v (because the other side of the coil is always supplied with 12v).

Then when either the distributor points meet (or electronic ignition triggers) it gets shorted to ground. For how long it sits on ground is the dwell time (the time time current is allowed to build in the coil). It's when the coil negitive is released that you get a massive spike to B. This can often be hundreds of volts. It's due to a property of coil's (or inductors) where current in them wants to continue flowing and when it doesn't voltage rises (very short explanation).

Once the spark completes the voltage settles to 12v again.

 

It's from this graph that we learn why the options may not work. The stock tach does not trigger of the rising or falling edge of 12v, it actually triggers when the large spike is measured.

 

Jim from zdriver took it apart and found this:

"It consists of some protection circuitry and an RC network to differentiate the pulse from the (-) terminal of the ignition coil."

reference: http://www.zdriver.com/forums/240z-260z-280z-performance-technical-124/74-260z-tach-grounding-problem-17555/#post328041

 

So now we know we need to give the tach the same pulse type signal. The megasquirt guys I believe came up with the gist of the idea:

 

Here, the ECU (MS PCB) gives a signal to a transistor (2N5551) to flow switch current through a relay coil to simulate that voltage spike. Although it says relay coil, it could be any appropriate inductor coil, relay coils just happen to be readily available.

 

So great, now we all we need to do is make a similar circuit using the nissan's tach output signal.

I actually built the above circuit, but it did not fire the tach. Most places I read stated the nissan's output signal was a 5v Square wave. If this was true, the 5v through the 1k resistor should be enough to turn on the transistor.

I didn't have a scope to measure it myself, but the readings on the multimeter made me second guess if it was really a square wave. After doing a bit more internet research I've found that the ECU itself likely just switches between an open circuit and ground. Those who measure a square wave may do so when the ecu is connected to the tach which possibly has a built in pullup resistor.

So now I simply added a pullup resistor directly to the output of the ecu and ran the same circuit, and voila working Tach!

 

circuit looked like this:

 

 

Design thoughts (for beginner circuit designers like myself).

Some of the resistor values may change depending on what transistor you use.

1. You want the transistor to operate in a saturated state. To do this the following must be true: H_FE > I_C / I_B

The transistor has a spec called H_FE. This is current gain. Try to find approximately what the current gain is under the load we’ll have.

 

Our load is the 14v supply through the automotive relay. My relay was about 80 ohms, so the current into the collector (I_C) was roughly  180mA. On the spec sheet, it showed a minimum current gain of 120 at 200mA.

 

Using this H_FE­ of 120, it’s a formula that I_c/I_B be > 120. Since I_c = 180mA, I_B= 1.5mA. We also want a safety margin so perhaps make the base current even higher. One site suggested 30% more which would yield about 2mA. I had originally built the circuit assuming the 5v signal so this value was close, but upon switching to 12v, it was higher, but only 5mA, so lots of extra turn on current.

 

I have read some cheaper NPN transistors can have much lower H_FE gains, so take note to adjust accordingly

 

With 2mA of current into the base:

V = IR.

14V = .002 * R.

R = 7000ohms.

So R1 and R2 need to add to less than 7000 ohms. I'd say make sure each is at least around 1k, but I just happened to have these 2 values handy so used them as shown. 680 is about as low as you’d want to go for the pullup resistor; 1) to limit current running through the ECU, 2) it’s at about ¼ watt limit of the readily available tiny resistors.

 

Finally, my 260z had a 2.2k tach resistor in the passenger foot well. Now i'm guessing the stock coil put out a much higher voltage considering the current going through the stock coil, so it was likely added to tame that voltage down slightly. We dont' want to do that since our relay coil probably isn't firing voltages quite as high, so bypass this resistor out, or simply remove it and connect the new tach signal to the other side.

 

Lastly, a std 12v relay makes a great template for buidling the circuit. it has all the terminals you will need (4 - 12v, ground, tach in, tach out), it has the relay coil, and they often have a built in diode you can repurpose to protect the ecu output.

Though this isn't mine, mine is very similar. I will try and get actual pics up later.

 

 

 

Anyways, this is getting very very long, but I wanted to get it out there before I lost my enthusiasm of the great victory of getting it to work. All to often i move on and forget to share in my success.

 

Regards,

Mark

 

[mtnickel]

As a side note:

 

I happened to have a spare 260z tach I was testing with and when I couldn't get the adapter to work initally i was going to just build my own tach driver just like Jim.

the white and black wires in the tach are just the leads of a current meter, and that meter is the tach needle. it registers more or less linearly to 6000rpm with 21mA of current though it. For full sweep, it is probably roughly a 0-25mA current meter. Keep that in mind if your one shot multivibrate hitachi unit fails and you want build your own tach circuit. There are many designs around the internet on how to do so, but keep that final current value in mind for your design.

Mark

[pat1]

I have always used a dropping resistor and never had a problem with the stock tach and aftermarket tach.

 

It's very possible that you damaged the ecu trying to attach a tach. I believe the stock tach uses a positive signal and a RB ecu sends a negative signal or vise versa I can't remember.

Edited January 13, 2015 by pat1

[mtnickel]

The stock tach has at least 8k ohms of input impedance. Extremely unlikely I damaged the ECU regardless of whether there was 12v or ground reference signal inside the circuit.

 

 

Here's more people who experienced same signal type for reference:

"the RB20 tacho uses a 0-5V square wave signal. The ECU only appears to switch pin 7 between open and ground, so i'm assuming the tacho provides it's own 5V connection to pin 7 through some internal electrics."

"yep, that is exactly what I found. The ECU appears to just invert the #1 coil signal with a low side driver."

http://forums.nicoclub.com/info-on-using-stock-tach-t132119.html

 

A pullup resistor, yes, should always solve issues with aftermarket tachs not working. Aftermarket tachs typically accept the 0-12v signal easily.

 

I can't explain why a simple pull-up resistor worked for your stock tach and not mine. I can assure you I've read of many people who can't get the stock tach to work however. Perhaps the Rb26 and Rb20 ecu are different? Perhaps your stock tach was in better condition? Maybe the 12v square wave you send is on the cusp of triggering and the hitachi module is finicky of how strong a signal it gets.

What I can state with certainty, is that emulating the signal that the tach originally received (the coil negative type) is likely your best possible chance at getting the stock tach to work. Sure try the simple pullup resistor first, but if that doesn't work and your Z's tach was working before, then I'd give a circuit like this a chance.

[Blob1620]

I'm currently swapping an rb25 into my 78 280. I'm not embarrassed to say, your knowledge of circuitry design is far better than mine, and this whole thing made my head spin a little.

 

That being said, I really am hell bent on getting the factory instrumentation to work as best i can with the swap, to keep the interior looking factory. Would you be willing to manufacture a relay such as this for me to purchase?

Edited January 15, 2015 by Blob1620

[mtnickel]

I'm currently swapping an rb25 into my 78 280. I'm not embarrassed to say, your knowledge of circuitry design is far better than mine, and this whole thing made my head spin a little.

 

That being said, I really am hell bent on getting the factory instrumentation to work as best i can with the swap, to keep the interior looking factory. Would you be willing to manufacture a relay such as this for me to purchase?

I'd try the traditional approach of using a pullup resistor first. Pat seems to think it worked fine on stock tachs. If that fails, I'm sure I could throw one together for you.

If you need specific notes on what the "pullup resistor" method is, I can outline it.

[Kash]

Someone should make these available for sale.

[mtnickel]

Someone should make these available for sale.

The MSD 8910 should be internally similar to this circuit, though it will lack the pullup resistor. Maybe some have tried just the pullup resistor, and then just the 8910 or 8920, when really the tach adapter would need the pullup resistor for it to see the correct signal.

 

On another note, if you're wondering how an ecu would put out such a circuit (with no positive wave), it would be due to it's internal construction looking like this (a "low-side driver/switch" since the low-side (ground) is the one being switched:

 

If that Yellow region is not hooked up, then the output will only change from being connected to nothing, to being connected to ground. In case you don't know how a transistor works, it is basically a switch. When the switch turns on, the top and bottom wires allow current to flow through easily. When the switch is off, current is not able to flow through Q1 from top to bottom.

[Bishopsrock]

Thanks to all who are contributing to this thread - i am amazed that after all these years we have finally got a thread on this subject where progress is being made by someone who understands both electronics and motors.

 

My question relates to the tach that many of us will have and that is the plain old induction tach as fitted to the 240Z.

 

I am like the rest of us in that I really want to retain the stock gauges. At the moment I am using a RB26DETT tacho mounted in a 240z tacho pod but this has left me without my turn signal lights which are mounted within the face of the stock gauge.

 

What I actually want to do is run the stock gauge.

 

I found this great item but need something more directly related to the 240z tacho as I don't understand many of the terms http://www.dinoplex.org/tachoconversion/

 

Does anyone know of any other way to run a 240z induction tach with an RB26DETT motor setup? or could anyone spell the stuff in the above link out for me as in exactly how I would connect this to a 240z tacho and an RB26DETT?

 

With thanks in anticipation.

Edited January 30, 2015 by Bishopsrock

[mtnickel]

Thanks to all who are contributing to this thread - i am amazed that after all these years we have finally got a thread on this subject where progress is being made by someone who understands both electronics and motors.

 

My question relates to the tach that many of us will have and that is the plain old induction tach as fitted to the 240Z.

 

I am like the rest of us in that I really want to retain the stock gauges. At the moment I am using a RB26DETT tacho mounted in a 240z tacho pod but this has left me without my turn signal lights which are mounted within the face of the stock gauge.

 

What I actually want to do is run the stock gauge.

 

I found this great item but need something more directly related to the 240z tacho as I don't understand many of the terms http://www.dinoplex.org/tachoconversion/

 

Does anyone know of any other way to run a 240z induction tach with an RB26DETT motor setup? or could anyone spell the stuff in the above link out for me as in exactly how I would connect this to a 240z tacho and an RB26DETT?

 

With thanks in anticipation.

 

Did you wire the car yourself?

 

The 4 wire tach I've read can be used without too much trouble.

 

First, here's a diagram of the stock configuration.

It's important to note that the coil does not get power (+) from any other wire. How an inductive circuit works is that when the coil is charging and discharging, there are bursts of current going through it's power wire. The 240z tach reads those current spikes with it's inductive pickup (since when current flows through wire, it creates an electric field).

 

Another member has written:

"You can also use the inductive (current sensing) style tacho if you power the RB coils from the same wire that provided power to the stock coil. This is what I am doing with my RB swap."

http://forums.hybridz.org/topic/80017-rb-zed-tacho/?p=759685

 

So what he is saying is that you need to run the power wire that goes to the coil-on-plug harness through the stock tach. Then it should be able to read the same current pulses. Again, I reiterate that you aren't just 'tapping' into the power line, but it needs to physically flow through the tach.

 

Something like this:

The only benefit to option 2 is so that power comes from the relay as it was intended, and not from the key. Perhaps you have specific fuses on coil power from there, etc. 

This should give you the right idea....i hope.

Edited January 30, 2015 by mtnickel

[Bishopsrock]

Mark, thank you so much for the straight forward and understandable explanation. Its great to find out that we can use the stock 240z tacho as easily as this. Yes, I did wire my car myself and I added a few bits into Chris Rummells wiring thread as I did it. I can see this tacho thread being very popular in the future though as all of this has been missing from the forum until now.

 

Just a couple of questions that im curious about and that you might be able to shed some light on:-

 

1 - The 240z uses a single coil whereas the GTR uses 6 coil packs. Is the rev counter still likely to be accurate with the GTR coilpacks?

 

2 - I have just discovered an old 260z tacho (amongst my hoard of parts) like the one featured in your first explanation at the start of this thread. Is using this likely to result in more accurate readings?

 

Cheers

[mtnickel]

Mark, thank you so much for the straight forward and understandable explanation. Its great to find out that we can use the stock 240z tacho as easily as this. Yes, I did wire my car myself and I added a few bits into Chris Rummells wiring thread as I did it. I can see this tacho thread being very popular in the future though as all of this has been missing from the forum until now.

 

Just a couple of questions that im curious about and that you might be able to shed some light on:-

 

1 - The 240z uses a single coil whereas the GTR uses 6 coil packs. Is the rev counter still likely to be accurate with the GTR coilpacks?

 

2 - I have just discovered an old 260z tacho (amongst my hoard of parts) like the one featured in your first explanation at the start of this thread. Is using this likely to result in more accurate readings?

 

Cheers

 

1) The single coil just fired 6 times as often. It used the distributor to determine which cylinder to fire to. So yes, it should be accurate enough. Note I have not tested the current sensing tach personally, so can't say 100% that it will work. If dwell times (the length of time the coil charges for), or amplitude of current (due to differences in the resistance of the stock vs RB coil) are different enough, it may not trigger the tach, but based on the other users success, I can't see why it would not work.

 

2) I cannot speak to the accuracy of the early vs late tach's. It would seem logical that the newer design was better, but tough to say. I feel like I've read the 240z tach tends to be 'bouncier', taking longer to settle. If it's accuracy you're after though, I believe both units have trim pots where you can tweak the calibration of the unit.

Good luck!

[Bryce Clinton]

What would your thoughts be on using a 5V Halifax sensor as the square wave signal input? I'm using a diesel engine and don't have an ecu for electronic tach output.  

[DuffyMahoney]

Any early tach works perfectly with coil on plug.  Sense pulses, works with zero modification. 

[Bryce Clinton]

ya  I understand that but I have a diesel in my car so therefore no spark plugs or distributor