dmjames

I'm working my way through diagnosing a long-crank/very occasional no-start/stall issue, and the next thing up is the speed/reference sensors. I've seen lots of discussion of 'tach bounce' while starting, but can't get a sense of how much 'bounce' is to be expected.

Is this a normal amount of bounce, or is it possible I've got a sensor putting out a weak signal?

https://goo.gl/photos/3dEpr3gvCZZwVHhu9

Also, apologies for the portrait video- I know, I know... It's what I could get to stay steady in a pinch.

dmjames

Obviously, this wasn't a long-crank start, but the minimal 'twitching' on the tach is comparable to what I get on the long starts, so I think this is still representative. I'll try to remember to take another quick video at some point to show the long crank.

mytrplseven

From my experience, that 1/16" bounce you saw prior to the engine start was accurate.

Tom M'Guinn

It's really more of a twitch than a bounce, and from what I could see in the split second before it fired, it was clearly twitching. Since the motor fired right up, you have to assume that that level of twitch/bounce is good to go. If it's twitching like that when it won't start, it's probably not the speed/ref sensors, but jiggling those wires while someone cranks never hurts (along with the fuel injector harness). Trust you've checked for fuel and spark when it won't start, and the DME relay, alarm, and other common no-start culprits?

dmjames

Thanks, guys- I just wouldn't have termed the 'twitch' a 'bounce', so I was skeptical! Good(ish) to know that it's likely not the sensors.

Yeah, I haven't really dug in at this point, since it *always* starts up for me, even if it does periodically take two attempts. The weird part is that whenever it gets in the 'long crank' state, it's like I can either hold the key for a long time and it eventually catches, or I can try for a few seconds, stop, then try a second time, at which point it catches extremely quickly.

In the last week, though, I've had a couple times where the car dies when I come to a stop at a light or stop sign, and a few cases where it would die if I didn't give it a bit of gas all the time. Thought I might take a look at the aux air valve and the adjustments for the throttle body and position sensor- if the air valve wasn't operating correctly, I could have the throttle body set screw and the idle adjustment screw set incorrectly, resulting in who knows what behavior.

I've replaced the DME relay, all the vacuum lines, electrical ignition switch, distributor/rotor/plug wires, spark plugs, re-sealed the throttle body and fuel injectors, and tested the fuel system to confirm I'm getting appropriate pressure in the system. I've also checked the alarm system, which seems to be behaving correctly (and has been for the past year - no changes). It's not all been part of this process, but mostly just replacing stuff as I go through refreshing the car.

The biggest issue with figuring this out is that it's not reproducible enough to be able to do much- it only does the slow start once, then it has to sit for some time before it'll do it again (hence the check valve thought), and the 'won't run or start w/o giving it gas' part is extremely random and has only happened maybe 3 or 4 times.

dmjames

I haven't really looked into the possibility of a failing DME, either. No water infiltration issues, but I do wonder if maybe there are solder joint issues or questionable wiring harness conditions. I've been cleaning grounds as I find them, too.

Tom M'Guinn

That description would make me think your air-fuel ratio is off for some reason. Could be as simple as the idle set too low, not allowing in enough air to run, but the fact that its intermittent, and that you can crank a little, stop, and then fire it right up, sounds like it sometimes needs the unburned fuel from the first crank to get it going. The first crank might be acting like a primer. Pure speculation, but I'd lean toward checking the sensor inputs and connections to see if the DME is occasionally getting a bum signal from the AFM, temp sensor, TPS, etc. Oh, and as always, check for vacuum leaks!

dmjames

You, sir, are awesome.

I'll give it all a look! When the car needed throttle to stay running, it smelled 'off'- not 100% sure it was a fuel smell, but my sense of smell is pretty much worthless under normal circumstances. So that goes right along with your thought of the mix being off.

I am using the standard workaround of adjusting the throttle stop to never let the TPS click 'closed' so there's not the idle 'shudder' that comes with running an early DME. That didn't actually seem to affect starting before, but I'll set it back to normal just to eliminate variables.

joes

I think that anyone that uses "tach bounce" to accurately diagnose a problem is going to have a hard time. There many other ways to check sensors correctly. My favorite is a 1.5v led inserted in the plug at the end of the sensor or at the DME.