[SJS]

I got the tacho alive, but I'm having trouble that once the tacho is connected the engine shuts down after a some seconds, also I have trouble getting it started when tacho is connected. Engine can run some seconds without problems, but then something goes wrong, usually it backfires when that happens and engine shuts down. I'm using the speeduino 0.4.3 and one of the four HC outputs for the tacho. And after the HC output I have the following setup:
https://photobucket.com/gallery/user/sc ... cGc=/?ref=
Not my photo. Originally from here: https://www.mr2oc.com/printthread.php?t=418284&pp=50
And I found it through here: http://www.msextra.com/forums/viewtopic ... 01&t=57445

The tacho is an original Opel Manta GSI tacho, which is originally connected to coil -.

What could be wrong in my setup?

[PSIG]

Most Speeduino boards use HC MOSFETs that have a lower breakdown voltage (V_(BR)DSS), and are unsuitable for driving tach flyback voltages. I don't know which transistor you used (STP?), but it should be rated 100V minimum and 100ma minimum when used with a typical Bosch-style 30A/40A relay. A TIP122 would barely make it to the voltage minimum. There are many TO-92 NPNs of this type out there (e.g., KSP43, MPSA42/43, 2N6517, 2N5550/51, APT27, etc), but using a lower max/breakdown voltage will cause two things: the breakdown threshold will limit the flyback pulse to that voltage (33V for the STP60/75NS04), possibly too low for tach triggering (and we have seen that issue here); and flyback voltage may enter the controller, frying or causing mischief.

What I usually see when building the entire circuit with an NPN is two things — first is that the 1k resistor can simply be between the TachOut and the gate or base of the transistor (see diagram below). You need to verify which way your TachOut signal works; either from float to high signal, or from float to ground. Doing it this way also helps to protect the controller output from direct voltages through the NPN transistor. This isn't a problem if using a MOSFET or IGBT, though most FETs have an internal body diode that can cause its own issues.

Finally, most relay coils have a built-in resistor or diode to snub flyback pulses — the opposite of what we want. However, you should try yours with the snubber intact (especially if you have a minimal breakdown voltage), but if the voltage is held too low, simply cut the snubber out of the relay. Most other inductors that can also be used for this (solenoids, old injectors, little vacuum valves, etc) usually do not have this flyback snubber. Hope that helps your issue.

David

Tach_adapter_circuit.png (57.69 KiB) Viewed 10429 times

[SJS]

Thank you PSIG for the answers. I'm using TIP122 in the circuit. Isn't that with higher voltage rating than the TIP120 that was in the example? I wonder how guys have got it working. But clearly I'm having trouble and your explanation cleared me why. I happened to have this TIP122 so that basically was the reason why I went to this circuit and not the one you showed (I found similar to that also). I will purchase one of the transistors you listed and do the circuit. I definitely like the fact that it is more safe to the controller
I believe the relay does not have a diode nor resistor, at least there wasn't one in the circuit drawing on the casing and I didn't see one visually when I opened it.

[PSIG]

I'm using TIP122 in the circuit. Isn't that with higher voltage rating than the TIP120 that was in the example?
…
I believe the relay does not have a diode nor resistor, at least there wasn't one in the circuit drawing on the casing and I didn't see one visually when I opened it.

Yes, the TIP 122 (not the TIP 120 in the photo) is better suited for the purpose. As I said; much depends on what output signal the ECM gives, as well as the required pulse the particular tach needs. I've seen working tach adapters using 2n2222s, but wouldn't suggest it would work or last long in every installation.

Your relay may not have a resistor or diode snubber, but that is unusual in automotive relays, as the snubber reduces voltage spikes in the wiring and back to the controller. Most automotive relays have them for this reason, which is why we often ask when there are random noise spikes in the system, if there are any relays operating. It is also why the standard wiring for Bosch-style relays is 86=V+ and 85=GND, so if the snubber is a diode it won't be destroyed. Snubbers are typically 1/4W resistors or similar-sized diodes spot-welded between the coil terminals.

David

Here is an outline of the two types with snubbers:

Relay_snubbers.jpg (252.76 KiB) Viewed 10382 times

[SJS]

Ok, so the setup I have is even better than in the instructions and still problems... I no doubt I will order parts for the another option. I have 2n2222 trainsistors laying around, but I would like to get it right once

Like said the circuit diagram on the casing doesn't have any evidence of any extra components between 85 and 86. I checked once again visually the relay internals, no evidence of any extra components. It is a cheap 30A automotive relay. I measured 89 ohm from the coil.

[PSIG]

Like said the circuit diagram on the casing doesn't have any evidence of any extra components between 85 and 86. I checked once again visually the relay internals, no evidence of any extra components. It is a cheap 30A automotive relay. I measured 89 ohm from the coil.

That's fine, and you should get a solid higher-voltage spike from it when you test it. If you want to confirm this without a multimeter or oscilloscope; power-up your tach, wire the relay coil to VBAT, and connect the relay coil ground to a metal brush, comb, or file. Take a wire from battery ground and stroke it across the brush or file (creating a series of short pulses through the relay coil), and you should see the tach needle jump.

Assuming that all goes well and using a v0.4 example; you are ready to add the NPN to your proto area and connect the base to resistor and on to the pin-49 output, collector to your IDC output pin choice (leading out to your relay pin 85), and the emitter to proto ground. Configure TunerStudio for your setup. Done!

Note: If you still wish to use the HC output, no problem — just remove the HC MOSFET and replace it with your NPN transistor.

David

[SJS]

Thank you PSIG once again for help. I have now ordered 2N5551 transistors so I can make the circuit. I will do it on the proto area. I will report the results once it is done.

Another very different issue is that I ordered at the same time components for another V0.4 board, I had only some of the components missing. The supplier didn't have the Panasonic ERZ-V14D220 available (U1 in the schema), but they had S14K14 (SIOV) which I understood is equivalent. Could someone confirm this just to be sure?

[PSIG]

I didn't dig into each device's spec's, and they are not equivalent, but it would appear that both are usable for Speeduino transient protection. Be sure to use a 1A (preferred) or 2A automotive fuse in the Speeduino +12V power feed for supportive protection.

David

[Experion]

That's fine, and you should get a solid higher-voltage spike from it when you test it. If you want to confirm this without a multimeter or oscilloscope; power-up your tach, wire the relay coil to VBAT, and connect the relay coil ground to a metal brush, comb, or file. Take a wire from battery ground and stroke it across the brush or file (creating a series of short pulses through the relay coil), and you should see the tach needle jump.

Hello David,

If I understand that correctly i should also connect the tacho to the ground side of the coil in the test?

Im running a 1988 Toyota Starlet and i think i need a simillar circuit to the one posted above.
Ecu wise Im running a Core4 which switches its tacho signal to 12v. Does that mean I can connect the tacho output to the positive side of the coil, connect the negative side to ground and also the tacho and be done with it?

should I maybe also put a diode between the ecu and the relay to prevent ecu damage or is it protected for that internaly?

Thanks in advance,

-Gary

[PSIG]

…If I understand that correctly i should also connect the tacho to the ground side of the coil in the test?

Yes. As we are working with both ignition and relay coil references, please include (for clarity) which you are speaking-of. The test described is only a function test of the relay coil's ability to pulse a signal into the tach, and the tach respond to it. if it responds with twitching, then you can proceed with adding the transistor into the circuit for Speeduino control. It is a crude but effective response test.

Tach_test_for_function.jpg (93.47 KiB) Viewed 8812 times

Im running a 1988 Toyota Starlet and i think i need a simillar circuit to the one posted above.
Ecu wise Im running a Core4 which switches its tacho signal to 12v. Does that mean I can connect the tacho output to the positive side of the coil, connect the negative side to ground and also the tacho and be done with it?

The pulse must be directed to or through the tachometer, and with the ground connected (circuit not broken by the transistor), the pulse is dissipated to ground. Altering the circuit to pulse on the (+) side of the relay coil might possibly work, but you'd have to test it in order to see if the reverse-polarity pulse works on your tachometer design, and high-voltage reversed-polarity pulses may have undesirable effects.

should I maybe also put a diode between the ecu and the relay to prevent ecu damage or is it protected for that internaly?

I would check or verify with @Jama as to what protections may be available on the Core4.