Honda D Series Forum banner
21 - 36 of 36 Posts

·
Premium Member
Boosted 90 crx si
Joined
·
1,113 Posts
A little late to the party, but starting with the engine coolant temp sensor would be on the very top of my priority list.

Credit to Jerry about the different dizzy mounting points on the head, vs the 88+ zc heads. Just a quick glance and I could tell it's completely different.

Since your on a chipped ecu, is it set for datalogging? If so, datalog on a COLD engine, koeo, iat and ect should be withing 1 degree of eash other if they are functioning correctly.

When the ect sensor goes bad, and shorts out, it tells the ecu it's -40*f, so it's DUMPING fuel and pulling timing. Excess and un burnt fuel causing misfires.
 

·
Registered
Joined
·
28 Posts
Discussion Starter · #23 · (Edited)
Hey theotherchris, The second half of your post, I recognize some of the letters. But it sounds way too hard for me, I did look into burning chips etc, but it seems even proffesionals struggle with it, no way I could make it work - IQ too low.

so I checked the whole ECT circuit and everything seems to be in order. Connections to the ECU are good, and resistance reads 5kOhm at 5 Celsius, which is in line with the table https://www.d-series.org/attachments/a-jpg.28898/

So currently the last thing remaining on the list is ignition timing. If that doesn't work, then I'll consider following options:
  • set the whole damn thing on fire and drive it off a cliff
  • swap in a Toyota engine (because I want something that's actually reliable, unlike this Honda rubbish)
  • convert it to carburetor
  • park it somewhere and wait until there are cheap and available batteries for EV conversion
But realisticaly, I think it's some electrical gremlin in the wiring. The thing was hacked together so it can run full throttle and last one season of autocross, then be scrapped. So it's my fault for trying to make it what it's not.

EDIT:
after driving it some more, I maybe figured out what's wrong with the transmission. It should have been obvious, but it's probably the clutch.
I had the flywheel skimmed, but it was done badly, so it's now wobbly and shudders. An even though my clutch grabbed pretty high, the wobble probably made it drag a little bit. When I adjusted the clutch super high, shifting improved. So that's probably that. A little embarrasing, but at least one thing had a easy solution.

About the engine, I think maybe what if I gotten rid of the codes - maybe it's running in some safe mode and that makes it run like crap.
I have to research how to fool the ECU into thinking there is a electric load sensor and vehicle speed sensor.
 

·
Registered
Joined
·
28 Posts
Discussion Starter · #24 · (Edited)
In the meanwhile, I'd like to trick my PM7 ECU into not throwing codes 17 and 20.

17 - Vehicle speed sensor:
On my car this ecu wire is simply cut, and there's no actual sensor (it's a 1986 car). According to this article, the VSS signal to the OBD0 ECU pin is a pulsed short to ground: Honda VSS Engineering Analysis
From reading the article, I thin I might fool the ECU by shorting the pin to ground with some big resistor (so I don't melt the circuit).
Can anyone confirm this is an ok idea?
Alternatively, I could make up some simple electronic switching circuit, or fabricate and actual speed sensor out of a magnetic reed switch (getting too difficult here)

20 - Electric load detection:
I have no idea about this one. Anyone knows what the signal on the ECU pin is supposed to look like?
Possibly short it to ground via a 820Ohm resistor as well?
Honda Electrical Load Detection (ELD) Bypass - Honda Accord Forum - Honda Accord Enthusiast Forums
 

·
Registered
Joined
·
28 Posts
Discussion Starter · #25 ·
I think the symptoms could also be bad MAP sensor.
Does anyone know if all Honda MAP sensors work the same? Can I replace my 1986 sensor with one from let's say 1998 Civic? (because I have one on hand)
 

·
Registered
89 EF 4wd sedan
Joined
·
1,441 Posts
In the meanwhile, I'd like to trick my PM7 ECU into not throwing codes 17 and 20.

17 - Vehicle speed sensor:
On my car this ecu wire is simply cut, and there's no actual sensor (it's a 1986 car). According to this article, the VSS signal to the OBD0 ECU pin is a pulsed short to ground: Honda VSS Engineering Analysis
From reading the article, I thin I might fool the ECU by shorting the pin to ground with some big resistor (so I don't melt the circuit).
Can anyone confirm this is an ok idea?
Alternatively, I could make up some simple electronic switching circuit, or fabricate and actual speed sensor out of a magnetic reed switch (getting too difficult here)

20 - Electric load detection:
I have no idea about this one. Anyone knows what the signal on the ECU pin is supposed to look like?
Possibly short it to ground via a 820Ohm resistor as well?
Honda Electrical Load Detection (ELD) Bypass - Honda Accord Forum - Honda Accord Enthusiast Forums
I've had both of these codes before and neither affect how the car drives.
 

·
Registered
Joined
·
28 Posts
Discussion Starter · #27 · (Edited)
Ok, I think I have a breakthrough - it's the damn MAP sensor! Yes, the one that never goes bad.
I plugged in a different one and the car drives so much nicer, much better lower end, doesn't jerk as much and stinks less!! (considering no cat)
I have to drive it some more to really confirm, but so far this looks promising!!!
Still doesn't idle silky smooth, but much better than before. I still haven't adjusted the dizzy, so that might sort it.

I guess I'll have to buy a new MAP sensor. Problem is the one for my car costs $250 shipped. But for a newer generation civic (1998) it's much cheaper, only $50.
Could that possibly be used? The both have 3 pins and both run on 5V in fairly linear curve, so it could work? I already ordered it, guess I'll find out the hard way.



EFB055
thanks, that's good to know! I'll try to get rid of them anyway, CEL makes me anxious.
 

·
Registered
Joined
·
9,980 Posts
Your temp sensors are easy enough to check. You cannot obviously check the whole range, but you can certainly measure the resistance, and compare to ambient temps on the honda charts in the tech manual

EDIT Map sensor, eh? Temporarily make a patch harness incase the new one doesnt work out.

As for the transmission, if the car has been sitting for a long time at any point, might just need some driving time to figure things out.

I prefer synchromesh. 3.5 liters is a lot for the transmission. Make sure you leveled the fluid out with removing the fill plug on the side with the car parked flat. Do not overfill.
 

·
Registered
89 EF 4wd sedan
Joined
·
1,441 Posts

·
Registered
Joined
·
28 Posts
Discussion Starter · #31 ·
Small update - I received my cheap MAP sensor, it is this one (079800-5410), I think out of 2001 D17 Civic, Oddyssey, Accord etc. I got it from Alien express, for $12 USD.
I rigged it up and as far as I can tell, the car runs well! I can't guarantee it's same as the expensive original one, but it works just as good as the 35 year old spare original from DOHC ZC.
Would have to be bench tested in lab conditions to confirm for sure.

Material property Auto part Adapter Cable Electric blue
 

·
Registered
Joined
·
28 Posts
Discussion Starter · #32 ·
Well that's annoying. I finally got a timing light and checked what I have - and the timing was way off. Maybe 4cm to the rear?
I had to make a entirely new mounting of the dizzy to be able to rotate it enough and get timing on mark.
Instantly the engine idled much better, now it finally sounds smooth.

But.

I immedieately got a Code 9 - cylinder position sensor.
Did some googling and it apperantly happens when the exhaust cam is one tooth off. Mine isn't, but can this be caused by rotating the dizzy? That's the only thing I changed.
So it probably means I have to rig up the cam position sensor to match? Rig up because it's not adjustable, so I have to adapt some brackets and rotate it roughly the same amount as the dizzy?
I must admit I don't like this idea very much. But I guess because the TD03u dizzy is wrong for my engine, that's why the original builder of the car made it like this (he put on the dizzy "wrong", but then changed the timing in software to compensate). Unfortunately he doesn't talk to me anymore, so I can't confirm.
 

·
Registered
Joined
·
28 Posts
Discussion Starter · #33 ·
Yep, just confirmed it. When I rotate the dizzy back to the wrong position, the CEL goes away.

So I'm guessing some sensor in the dizzy is oriented inconsistently with my engine's other cam.

Now the question is, do I rotate the sensor on the exhaust cam? Or just live with wrong ignition timing and shit idle?
 

·
Registered
Joined
·
28 Posts
Discussion Starter · #34 · (Edited)
Guys, check this out.
Apparently the exhaust cam sensor is modified so it's got different timing.
W. T. F.
Auto part Automotive lighting Personal protective equipment Metal Fashion accessory


This is engine at TDC. Notice the exhaust cam sensor is pointing to ... ????? Shouldn't it be right at the pickup coil?
Motor vehicle Automotive tire Engineering Machine Personal protective equipment


Notice here the end of the shaft has been reoriented from the rest. Wtf
Plant Gas Terrestrial plant Flowerpot Metal
 

·
Registered
Joined
·
28 Posts
Discussion Starter · #35 ·
Ok, I suspect the CEL is because of mismatch between TDC and CYL signals.
What the "builder" did, he put it together from whatever parts he found in the rubbish bin, eyeballed it, and then compensated for everything in software.
Because I can't do that, I have to align everything in hardware. If I'm reading this page from 1986 Integra manual correctly, the CYL sensor should be offset liek this. So on my previous picture it's roughly in the position it should be, but not quite. I guess I'll attempt to move it a little, and if can't, I'll cut some slots into the sensor body and make it adjustable.
Unfortunately i don't have any means to measure things accurately, so I'll have to just eyeball it and hope for the best.

Modified cars, am I right? I'm having so much fun I feel like punching myself in my balls.

Rectangle Font Parallel Slope Number
 

·
Registered
Joined
·
28 Posts
Discussion Starter · #36 ·
Ok, I'm getting somewhere.

When everything is oriented as it was originally, and I'm using the chipped ECU, the timing is spot on and I'm getting a smooth idle!

When I plug in a stock PM7 ECU, the timing mark is about 1.5 cm to the rear, and idle is crap.

Someone smart please tell me - what is likely different in the chipped ECU timing wise?
Is it possible he changed the coil firing timing in the software? How about the CYL sensor, is that changeable in software too?
Because if not, then that means mine is probably set correctly, and I only have to worry about the dizzy.
 
21 - 36 of 36 Posts
Top