Fuel pump priming on K20a DC5 with PRC ECU

[DrPhyzx]

Greetings,

I'm new here, but will probably be back from time-to-time as I dig into understanding things on the EM side. I've got a custom car - Race Car Replicas Superlite Roadster - I acquired a couple of years ago from the second owner as a weekend toy. It's essentially a replicar of an Ariel Atom with very similar specs which was sold for a couple of years by RCR/Superlite Cars. Power is from a K20a DC5 with a PRC ECU. The car was built in a couple of stages, first the basic engine and chassis in 2009 (found the build thread on k20a.org), and then sold to the gentleman I bought it from who subsequently completed the work to register it for road use in CA in the 2010-2011 timeframe. Sometime subsequent to that, a KPro was added to the ECU to get around the speed limiter: I'm aware of this because the previous owner also passed along to me an HKS plug-in device that was meant to do the same thing, but which he never installed because he added the KPro instead: obviously a good choice! However, I don't know exactly when the KPro was installed or what generation it is: I know I'm supposed to include that info here, and I will when I can get it out/open and identified, but I don't think it's necessary to know for my underlying question.

After understanding nearly everything else on the car, I've arrived at sorting through the electrical systems. The original build work appears to have been done quite professionally, while some of the work to make it street legal (auxiliary electrical) seems a bit hacky to me (I'm a perfectionist, but I digress....) One element of that original build work that I'm questioning is the fuel pump setup. Instead of being wired to the ECU, the return line from the FP relay coil is grounded so that the fuel pump runs all the time when the ignition is on to get around the lack of an immobilizer. Having done quite a bit or research, I understand that this has often been recommended practice, and doesn't harm anything per se, but it does kind of annoy me and I'd like to make it work as designed if that's possible. This is where I find widely varying advice because I gather the solution depends on what ECU you are running and perhaps also whether you have a KPro and maybe even what generation it is.

As it stands, the ground return from the FP relay coil is connected to a wire that runs back to the ECU, but is terminated with a ring lug that is attached to one of the ECU mounting screws. If it were really as simple as connecting that wire to the correct pin on the ECU (E10?), that would be incredibly easy. However, it's not clear to me whether this is supposed to work with the PRC ECU, what the role of the KPro is here (if any), and whether a specific version of the KPro is required. So, the basic question is: Is there a way to make the fuel pump priming work "correctly" with this engine and ECU? If so, what are the requirements?

FWIW, I'm a Mac (previously Linux) guy but have just purchased a cheap Windows laptop so I can run KManager. Undoubtedly that will soon lead to more questions :-)

Thanks and Cheers,
Tim

[Hondata]

I'll just answer one point now - people see the ECU fuel pump output and think 'oh great, that is what I need to connect to the output side of the fuel pump relay'. They do that and the ECU fuel pump relay output IC disappears in a puff of smoke because it sinks current to operate the fuel pump relay. Usually the person then does the same thing to another ECU and then to their friend's ECU as well. As a consequence many (most?) of the ECU available have the fuel pump output burnt out and people wire the fuel pump directly to ignition. So your ECU many not operate the fuel pump.

Otherwise the main relay and fuel pump relay output from the ECU is simple. E1 grounds the fuel pump relay. F7 grounds the main relay. Both coil sides of those relays should be constant +12V because that is how the ECU gets it 'keep alive' current - through the relay coil. O2 sensor relay needs to be on switched ignition power otherwise it latches on and the battery will go flat quickly.

[DrPhyzx]

Thanks for the quick reply. I'll be going through everything carefully so all good advice is appreciated!

Yes, I'd read your discussions of the FPR issue here and elsewhere, so I am of course concerned I'll find that the ECU is damaged. However, there have been other places on the car where I have found low hanging fruit - a capability just waiting to be used - where for some reason it had been bypassed when originally assembled. So, I want to at least understand what I'm dealing with before deciding what actions to take.

Your answer leads me to some more questions, to make sure I'm understanding, to consider next steps, and perhaps to leave a clearer trail for others with similar questions, since I found it very difficult to find clear and authoritative information on how this works when there is no immobilizer in the car.

E1 grounds the fuel pump relay.

So, how is the immobilizer logic bypassed? Is that something that the KPro does by default? Also, what are mentions of E10 about for JDM ECUs in various forums: misunderstanding turned into bad internet advice? A different ECU?

F7 grounds the main relay.

This confused me, but I think you mean E7. Yes... I hadn't even noticed that the main relay is also controlled by the immobilizer logic because it doesn't have such obvious symptoms. I just checked, and in my case, the main relay coil return is indeed connected to E7, so apparently that is working. However, I have the same question as above: is the KPro the thing that bypasses the immobilizer logic?

=================

I'm not sure what to make of the above. It may suggest that a knowledgable person went to wire it and found the FPR relay output on the ECU fried, since they wired the main relay correctly, and decided to just ground the FPR coil return. However, it's puzzling that the main relay control circuit never got fried in that case - by whomever did it wrong at some point - so I'd like to make sure I understand what the situation is before deciding how to proceed.

So, the next question is what is the easiest way to check? There is currently no pin in E1: do I need to get a pin/wire in there so I can check the output or is there a clean/safe way to test the E1 output without a pin in that position? (Without putting a lot of thought into it, I'm assuming the E connector needs to be plugged in for the system to work so that I can actually check the E1 output: this is essentially what I am asking for verification of.) I guess I could borrow a test lead with a small socket (as for a scope) from work: I don't have any here at home. Since one has only two seconds to check whether E1 is getting grounded, it will be challenging to test without a good setup.

Following that, let's say I find that the FPR control IC is fried. Is there any recourse? Does anyone rework these boards? Is there a source for reliable replacements?

Happy New Years (Eve)!
Tim

[DrPhyzx]

BTW, this seems clear/authoritative and is an example of conflicting information:

https://ff-squad.com/wp/?p=682

Here, it notes that the PRC ECU has no immobilizer integration (since JDM cars didn't have that), so there is no need to bypass it: that makes sense. It also says the FPR coil return is on E10 instead of E1 for the PRC.

[DrPhyzx]

I guess I can get the ECU where I can power with the cover open in order to test, but that kind of thing - probing a live board even with decent test clips for the connector leads - always makes me nervous. Also, once I get KManager working (have laptop here now) I guess I could set fuel pump to run always, so I don't have the 2-second limit to making a measurement. But... still happy for any further suggestions.

[DrPhyzx]

For completeness:

IMG_6153.jpg

[DrPhyzx]

OK, good news. So, it looks like it is correct that the pin for the FPR Coil Return line is E10 for the PRC ECU. With the ECU repositioned I was able to safely get a meter connected to test continuity between E10 and Ground at power up:

https://vimeo.com/898953865?share=copy

The meter is off-screen, but you can hear it beep for two seconds after the click from the main relay and then go silent. So, it looks to me like someone didn't know how to get this working correctly (perhaps they had no PRC experience and tried E1) and just grounded the FPR Relay Coil return line. So, I think I only need a connector pin to make this work correctly. BTW, I wonder if it's possible to get the ECU schematic somewhere - or even just PRC-specific pinouts, which could come in handy someday. I looked pretty hard, but don't find any trustworthy PRC-specific pinouts/diagrams.

I did check the wiring of the main relay: the coil does see battery voltage with the ignition off. I didn't check the O2 sensor relay yet, but it has sat for periods with the main switch on (it has an emergency main switch) and battery drain has not been an issue.

I think the final question will be whether there is good reason to upgrade to KPro V4. I'll try to educate myself on the differences while I dig around in KManager. Thanks again: if nothing else, this gave me the kick in the pants I needed to just dive into this.

[DrPhyzx]

I thought it would be good to close this out with more complete and correct information from checking everything I could think of before going ahead and hooking things up. Hopefully this will help someone else at some point.

First, I find on my ECU that E1 and E10 are shorted (< 1 Ohm). I don't know if this is something that the KPro installation accomplishes, since docs for the US ECU seem to indicate that E10 isn't connected to anything. I don't see any signs of a jumper or rework on the top side of the board, but cannot rule out something on the back side. Maybe the PRC has them shorted by design?

Second, I wanted to make sure things are clear about the ability of the ECU to deal with voltage and sink current for the FPR coil. When the main relay comes on it does put 12V on the FPR coil which connects to E1/E10. If there is no current flowing there (e.g. after the two second prime when the resistance of the ECU line to ground is large) then ECU components connected to E1/E10 are exposed to the full 12V, so in a technical sense it probably isn't the voltage that blows that circuit on the ECU if connected incorrectly to 12V, but rather the the power dissipated through the component that grounds that line for fuel pump operation if there is no relay coil resistance in series.

Here are measurements I took of the setup:
V_coil = 13V (open circuit)
I_coil = .163 A (return shorted to chassis through meter) ==> R_coil = V/I = 80 Ohms
R_E1 = 33 Ohms (resistance to ground measured during 2 sec prime sequence)

So, if you were to connect E1/10 to the battery (or a low resistance like the FP motor) without the relay coil, when it tries to operate the pump, the ECU must dissipate:

P=13V^2/33 Ohms = 5.1 W

That's a lot of power for any small transistor, especially if it's part of an IC.

But, if you connect E1/E10 to the return of the FPR relay coil (as you are supposed to), then:

I = 13V/(80 Ohms + 33 Ohms) = .115 A

P_ECU=I^2 R = .115 A^2 * 33 Ohms = .437 W

So, connected correctly, the ECU only has to dissipate 437 mW when operating the relay, which seems more manageable :-)

Anyway, after checking all this for sanity (and confirming online that 80 Ohms is reasonable for the FPR relay coil), I jumpered the FPR relay return to E1, and the car primes normally (and I assume will run correctly, but I didn't start it.)

ECU pins are on order!