CVT swap? (Into 2005 Civic, replacing 4-spd auto)

[Ecky]

I've seen maybe 1200rpm in my stock Insight and 1000rpm in my K swap. Higher idle, yes, but certainly not 1800. VSS is still not connected to the ECU?

[EcoCivic]

Originally Posted by Ecky

I've seen maybe 1200rpm in my stock Insight and 1000rpm in my K swap. Higher idle, yes, but certainly not 1800. VSS is still not connected to the ECU?

The VSS is connected to the ECU, I reconnected it because it wasn't going into decel fuel cutoff or activating VTEC with no speed signal. When the ECU isn't receiving a speed signal, the engine idles properly while coasting since it doesn't know the car is moving.

[Ecky]

Yeah, that's a tough one.

[cRiPpLe_rOoStEr]

Originally Posted by EcoCivic

it seems to idle a little higher with the AC on

I have never seen it, even though some cars have their cooling fans on full blast whenever the AC is on.

[EcoCivic]

Just plugged the ELD back in. The engine now idles at around 1200-1300 RPM while coasting as long as no accessories are on. If I turn the AC on, it still idles at around 1800 ish RPM. Not sure what exactly that means. I will need to do more testing to determine whether electrical loads also affect the idle RPM while coasting or just the AC.

I wish I could see the target idle RPM and IACV position, but my Scangauge doesn't show those things and this car is likely too old to report that information anyways.

[Ecky]

I wonder how the ELD works. Maybe it's a simple varistor? It sounds like when it's unplugged (open circuit) you get the highest idle, and plugged in with no load, the lowest idle. That would suggest to me that, if it is in fact a varistor, the higher the resistance (or the higher the voltage, I suppose) the higher the idle. A dead short on the sensor, or fooling it with a higher voltage, may lower your idle further.

Just a guess.

Edit: this might be helpful: https://www.civicforums.com/forums/3...-detector.html

[EcoCivic]

Originally Posted by Ecky

I wonder how the ELD works. Maybe it's a simple varistor? It sounds like when it's unplugged (open circuit) you get the highest idle, and plugged in with no load, the lowest idle. That would suggest to me that, if it is in fact a varistor, the higher the resistance (or the higher the voltage, I suppose) the higher the idle. A dead short on the sensor, or fooling it with a higher voltage, may lower your idle further.

Just a guess.

Edit: this might be helpful: https://www.civicforums.com/forums/3...-detector.html

That's a great question that I don't have an answer to. What I do know is that the voltage at the ECU's ELD signal pin reads around 4.5V with the ELD unplugged and drops when the ELD is plugged in or when electrical loads are applied. The higher the load, the lower the ELD voltage.

Now I need to figure out how to keep the ELD plugged in and functional while still disabling low charge mode. Perhaps removing the alternator control wire from the ECU connector would accomplish that goal since low charge mode is activated by shorting the control wire to ground. But whether that method would set a code and/or create another issue I have no idea. I guess there's only one way to find out.

[EcoCivic]

I'm not sure what else to try. The 1000-1200 RPM idle while coating doesn't seem excessively high, but the engine isn't particularly happy idling at that RPM and it feels quite rough and annoying. I can live with that though. Now the problem is that if I turn the AC on the idle increases to around 1600 RPM and the engine smooths out, but the unnecessarily high idle caused by running the AC reduces my coasting MPG significantly. I wish it would just idle at ~800 RPM all the time like when I'm stopped.

I thought of a couple things I overlooked when I checked for vacuum leaks- the brake booster and PCV valve. The brake booster will hold its vacuum assist for hours after the engine is shut down, so it seems alright. As for the PCV valve, I tried disconnecting it and plugging the vacuum line to eliminate it as a possible problem and that made no difference.

About the only other thing I can think of is to find a way to use the stock D17 IAC instead of the slightly larger one on my K20Z1 RSX-S throttle body, but I don't think that will make any difference since the ECU seems to be in control of the idle speed. I wonder which other inputs it uses to determine idle speed.

[EcoCivic]

Update- I changed the K20 RSX IACV for the proper Civic D17 one, which converted the constant high idle to an intermittent high idle while coasting.

After changing the IACV Saturday, I took it for a test drive and the idle was perfect- steady 750-800 RPM sitting still or rolling whether the AC is on or not, which is the first time I have seen that with this transmission. Awesome! The perfect idle continued until I got to college Monday.

Then on the way home, it started idling high again while rolling but not as high- 1K to 1500 RPM. Sometimes turning the AC on increased it like before and sometimes it didn't.

Thinking perhaps something in the brake booster was intermittently causing a vacuum leak, I disconnected the hose and plugged it off with a bolt to completely eliminate it as a variable.

On the way to college this morning the idle was perfect, but on the way home it started acting up again and sometimes idling high while I'm rolling.

I'm starting to wonder if the ECU is screwed up.

[cRiPpLe_rOoStEr]

Originally Posted by EcoCivic

Thinking perhaps something in the brake booster was intermittently causing a vacuum leak, I disconnected the hose and plugged it off with a bolt to completely eliminate it as a variable.

Gotta hate how some of those Japanese ECUs relied so heavily on vacuum input to adjust. Sometimes a small pipe or hose out of specification becomes a PITA.