coyoteboy
|
| posted on 6/12/11 at 10:47 AM |
|
|
Voltage ripple
Hi all,
In the hunt for my never-ending, here one day gone the next, idle misfire and consistent misfire (reads in logs a spikes or dips to +-anything from
100 rpm to 5krpm, at ~3200, 4000, 5000) I'm wondering if another electrical system (such as brushes on the alt) is creating the problem. The
stock ECU runs fine on the system, idles fine with the occasional miss but it's always had that and it's bearly noticable.
What sort of voltage ripple do you see "outside" the ECU and "inside" ? I've got some pretty spectacular noise on the
raw 12v as you'd expect but I've not scoped inside the 5v side yet. My timing pickup is a hall and uses shielded cable, all fed from the
ECU's clean 5v so I doubt it's that. Every time I've scoped that at idle it looks perfect (apart from the changing speeds from the
misfire).
|
|
|
|
|
scudderfish
|
| posted on 6/12/11 at 10:56 AM |
|
|
I had intermittent spikes on my RPM read (my ECU rather optimistically believed a RV8 could momentarily do 16000 RPM and did unnecessary things to the
spark advance) which I finally traced down to noise generated by my low impedance fuel injectors. Some 25W resistors on the +Ve to them sorted it
out.
Regards,
Dave
|
|
|
coyoteboy
|
| posted on 6/12/11 at 10:59 AM |
|
|
Cheers for that. Unfortunately my low imp injectors are already driven through a resistor bank (Toyota chose that route) so I'm not sure
it's that. But I'll see if I can get to them to check their condition in case they're going a bit dicky with age and heat.
Mine gets a bit unhappy when boosted at 1.25 bar I get a moment of twice or 3x the RPM and my timing usually gets shoved up the scale and advanced a
bit and I get massive knock triggered. IT's only a matter of time before I lose a piston, hence it's down at 6psi until I can sort
this.
[Edited on 6/12/11 by coyoteboy]
|
|
|
scudderfish
|
| posted on 6/12/11 at 11:30 AM |
|
|
Are you running your injectors with pulse width modulation? It's the rapid banging on/off that makes the noise.
|
|
|
coyoteboy
|
| posted on 6/12/11 at 11:40 AM |
|
|
Yep, it's hi-resolution Ms1 extra so I've no longer got the PWM option. I've no decoupling cap on the crank sensor so I've
just grabbed a few to add.
[Edited on 6/12/11 by coyoteboy]
|
|
|
MikeRJ
|
| posted on 6/12/11 at 05:45 PM |
|
|
quote:
Originally posted by coyoteboy
Cheers for that. Unfortunately my low imp injectors are already driven through a resistor bank (Toyota chose that route) so I'm not sure
it's that. But I'll see if I can get to them to check their condition in case they're going a bit dicky with age and heat.
Mine gets a bit unhappy when boosted at 1.25 bar I get a moment of twice or 3x the RPM and my timing usually gets shoved up the scale and advanced a
bit and I get massive knock triggered. IT's only a matter of time before I lose a piston, hence it's down at 6psi until I can sort
this.
[Edited on 6/12/11 by coyoteboy]
As boost rises the HT voltage required to create a spark rises. This also causes the primary ("low tension" side of the coil to have a
higher peak voltage, and you may see 200-300volts here which will be clamped by the coil drivers in the ECU. If there is any insulation weakness in
the wiring or grounding issues on the ECU, this is exactly when it's likely to make itself seen.
|
|
|
coyoteboy
|
| posted on 6/12/11 at 09:57 PM |
|
|
Unfortunately (fortunately?!) the driving is taken care of by the stock ignitor and dizzy which is fired by one of the LED lines (with the LED
removed) so nothing of the coil LT should be anywhere near the ECU in this configuration. I'm trying to remember what dwell I'm running, I
think around 3.5ms across the board, I dropped it in case it was overheating the ignitor and neither dropping or raising it to 5ms helped the
occasional blips.
But the misfire shouldn't affect the RPM readout unless it's causing resets, which it doesnt seem to be. Also, all analogue sensors are
reading very cleanly (temps etc) except map which varies by +_ 1psi at full chat.
Open to any ideas at this stage though as I'm properly naffed off with it!
(FWIW it's on an otherwise almost stock elec system on a '91 toyota. Plus are one grade colder which I considered might make the idle
miss but it doesn't miss with the stock ECU and colder plugs.
|
|
|
coyoteboy
|
| posted on 7/12/11 at 12:01 AM |
|
|
A bunch of bypass caps on the daughter cards, the crank sensor power supply as it hits the PCB, the MAP sensor (for giggles) didn't make a jot
of difference. The 3200 rpm issue is pretty consistent, the others are less so. Worth a shot but it was a shot in the dark!
|
|
|
MikeRJ
|
| posted on 7/12/11 at 07:54 AM |
|
|
Are you using the original crank sensor and mounting?
|
|
|
coyoteboy
|
| posted on 7/12/11 at 09:18 AM |
|
|
In a sense, but not in a sense. The original sensing is done via 24+2 (IIRC) on the cam. When I built the system there was no option for this layout
in the code and I couldn't for the life of me be bothered modifying it, so I bought another stock dizzy for a fiver and went at it with a saw.
It's now a 2x 12-2 on the cam (12-1 gave too small a missing pulse during low battery cranking) and instead of using the VR's that were
there originally I retro-fitted an allegro hall effect sensor and feed that directly in via the 0-5v hall circuitry (v2.2 PCB).
|
|
|
matt_gsxr
|
| posted on 7/12/11 at 10:06 AM |
|
|
If you upgraded the MS1 to an MS2 then you could run a tooth log during the time that you get this problem and see that the RPM glitch is caused by
improper reading of the trigger wheel. It might allow you to determine which wheel and such. The other thing with MS2 is that it might support the
stock wheel and sensor. These problems are hard to track down.
Just thoughts.
|
|
|
coyoteboy
|
| posted on 7/12/11 at 10:30 AM |
|
|
Yeah I'd been seriously considering the MS2 but the cost is prohibitive (what are they now, nearly 100 quid for the module?) when I've got
another project and another MS to buy and build. I could borrow an MS2 chip from someone I know but there really shouldn't be a problem for
this, it's effectively just a hall-based 12-2 crank wheel using pretty much an OEM harness so it's not like I've disturbed anything.
I've scoped the signal all day long but with it being so intermittent I've practically no chance to catch it and no way of tooth logging
until wrong like with the MS2. I've re-wired the ground for some of the bigger items with 6mm^2 back to the block so I've got sweet FA
ground differentials (2 small ones for the ECU, one large on for things like the wideband heater and larger switched items like relays .
Cheers for the suggestions. I'll be keeping my eyes peeled for a cheap MS2 daughter card. Do you happen to know if the outputs wired into an
MS1 CPU (jumpered direct to pins under socket as per manual) will be swappable onto the MS2 daughter card? I thought there were some slight config
differences.
|
|
|
MikeRJ
|
| posted on 7/12/11 at 12:11 PM |
|
|
No chance of the new sensor vibrating on it's mountings, or any resonance in the cam drive? This is exactly the kind of symptoms you can see
from CPS problems.
|
|
|
matt_gsxr
|
| posted on 7/12/11 at 01:10 PM |
|
|
quote:
Originally posted by coyoteboy
Cheers for the suggestions. I'll be keeping my eyes peeled for a cheap MS2 daughter card. Do you happen to know if the outputs wired into an
MS1 CPU (jumpered direct to pins under socket as per manual) will be swappable onto the MS2 daughter card? I thought there were some slight config
differences.
I think the pins are equivalent except there is a 12v thing that is different on one pin ( details here
http://www.msextra.com/doc/ms2extra/MS2-Extra_Conversion.htm ). Lots of retuning needed, so more than plug and play.
I had a loose wire from my regulator that gave me triggering problems, but it sounds like your electrical system is pretty well nailed down.
You could disconnect the alternator to see if that makes any difference. These intermittent problems are always hard to sort out.
|
|
|
