AdrianE

I have an E11v2 on my 1995 RX7 (FD) which has a Garret GT35R, a mild street port and 550/1000cc injectors. I had a lot of trouble getting decent sparks with the standard ignition system, and ended up fitting a Haltech S4 Hall Effect sensor / custom aluminium timing wheel as well as 2 MSD ignition boxes (a Digital 7+ for leading [in waste spark mode] and a DIS2+ for trailing [not waste spark]).
See the thread https://www.rx7club.com/haltech-forum-62/haltech-output-msd-cdi-system-663604/ for some of the history on this.

I had the timing wheel made up to accept 3 countersunk Haltech rare earth magnets and machined with 3 grooves that corresponded to 20, 10 and 0deg BTDC. The trigger angle was set to be 45deg BTDC *when* there is 90deg between the HE sensor and the mark on the engine timing cover. See attached diagram of the wheel.

Now that the engine is up and running after a break of 18mnths I managed to do a basic road tune with the help of my PLX R500 wideband/datalogger. Initially it was difficult to take the fuel map that I had on my previous ecu (Microtech LT8S) and modify this map to suit the E11v2, but now I have the AFR's pretty much where I want them. I run low 12's at idle, mostly 12-13.5 in vacuum (apart from around staging at -22KPa where it still hits low 10's where the 1000cc secondaries kick in), and low 11's on boost (which is limited to approx. 10psi).

However I am still having problems with the ignition mapping, hence the high EGT'S I'm getting with the exhaust manifold runners, turbo and downpipe all glowing orange after a brief run on boost! The engine temp is still low 70-80degC plus the V_mount intercooler and intake runners are cold to touch, so this doesn't make sense to me. Why isn't engine temp rising?

I cannot be 100% certain where TDC is until I can accurately measure the angle between the HE sensor and the mark on the engine timing cover (this is assuming that the aluminium timing wheel has been machined *exactly* to my spec). What I've done so far is gradually increased the trigger angle from 51deg to 41deg (ie. effectively added more ignition advance) to see where the engine felt most comfortable. So while I check these measurements and use my timing light again, can somebody please have a look at my ignition timing map and tell me if there any problems there? Crispeed ... and all you other rotary guru's?

Also, is there any danger in using multispark on trailing ie. if the leading/trailing split is large?
I changed the split to be 10deg up to 2500rpm and 8deg from thereon, because the engine felt smoother and picked up better than when using a gradual mapping from 15deg at idle to 10deg at redline.

I even tried running the engine on leading plugs only by switching off the MSD DIS2, but the high EGT's still remained.

Thanks in advance,

Adrian

C. Ludwig

iTrader: (27)

By increasing the trigger angle you are retarding timing, not advancing it. Further, without a way to know exactly where TDC or any crank angle reference truly is you're wasting your time and assuring engine damage at some point.

Your fuel is rich and if timing is retarded it's no wonder the exhaust is hot.

You need to find true TDC. Once you do that your sensor should be positioned so that the #1 magnet aligns with the sensor at approximately 50-70* BTDC #1. That 50-70* will be the value you enter as your trigger angle.

Also, it appears your trigger wheel drawing is flawed. You say that all angles are relative to TDC and that corresponds to the groove on your stock wheel. The groove on the stock wheel is not TDC. It's 20 ATDC. The marks to the left of the indicated TDC at 10 and 20* relative to TDC are ATDC not before. The marks would need to be on the right of the TDC mark for them to be BTDC.

Claudio RX-7

iTrader: (1)

Chris is right, you're chasing your tail here. My suggestion would be to put the stock crank triggers back in there and just use an RA10. Make sure you zero your timing correctly to begin with or your engine will certainly grenade it self soon.

Also, your ignition settings are odd. And finally, are your MSD units configured with DIP Switches or rotary dials?

AdrianE

My bad, I meant decreasing the trigger angle from 51 to 41*, so this is advancing the timing

Is this because the fuel is being ignited too late and is still burning as it is being forced out of the exhaust ports?

Isn't that 20 ADTC with respect to trailing? So assuming a 15* split (used by the OEM ECU), the mark would be 5 ADTC with respect to leading. The Haltech's ECU trigger angle is also supposed to be with respect to leading. So when I use the timing light this means I just have to subtract 5* from the trigger angle I arrive at if I line up the "TDC" mark on my wheel with the mark on the engine timing cover.

Yes you're right, but I only wanted them to guage how much I had to adjust the trigger angle by. If I had to get the wheel made up again I wouldn't have bothered with them, as the MSD's multisparking over 20* makes it a bitch to know which mark is which! I did find a way around this problem though, by making a small diagonal groove joining on to the "TDC" one.

But I realise the only way I can find out if my timing wheel has its "TDC" mark machined at the true 5* ATDC is to take it off, put the standard timing wheel back on and line up the mark here with the one on the timing cover. I wish I'd done this in the first place ... but I was in a hurry to get the engine up and running again and I trusted the machine shop when they said the marks on the wheel were accurate to within 1*

C. Ludwig

iTrader: (27)

Ok. If you know for sure the marks on your wheel are accurate with respect to TDC of the actual crankshaft here is your procedure to zero the timing.

Lock the timing. Set leading lock at 10 and trailing lock at -10. This will fire both leadings at 10 BTDC and the trailing at exactly TDC. So T1 should align with the TDC mark on your wheel. Attach your timing light to T1 and start the engine. Adjust the trigger angle value until T1 aligns with the TDC mark on the pulley. Trigger angle effects both leading and trailing the same. It's only purpose is to synch the ECU's ignition timing output with the actual crankshaft angle. Once you have T1 lined up with the TDC mark you can attach the timing light to L1 and you should be able to eyeball that it is firing at about 10 BTDC. Would be nice if you had a mark there but eyeballing it is really close enough. You just want to make sure it's not firing 90* out or something wild like that.

Do all that and get back to us. Again, if you're not 100% certain that the marks on the wheel correspond to the actual crankshaft orientation you're wasting your time and asking for a broken engine.

AdrianE

OK I verified that my timing wheel's "TDC" mark is *close* to the original wheel's 5* ATDC mark by putting the original wheel on again lining it up to the timing cover mark, then putting my timing wheel back on again and measuring the difference (ie. while the engine's crank was locked in place by leaving the car in gear). It is only 1mm out, when comparing the centre's of each of the marks, which works out at just over 1 deg. So, effectively my wheel has just another 1 deg to add to the 5deg from the 5*ATDC mark on the original wheel. This is the good news!

But here is the bad news... When I have my timing light connected to L1 or T1 and then point it at my timing wheel, with both the ignition angles locked via the Ignition Setup screen in the Halwin v1.89 software, then as I increase engine rpm the "TDC" mark moves in a clockwise direction!

The "TDC" mark is easier to follow when using T1 because the multispark operation from the MSD DIS-2 stops at about 1700rpm, as from then on it's just single spark. So comparing its position at 2000rpm to its position at 5000rpm there's a difference of at least 5*. Also the mark becomes blurred as rpm increases. By 6000rpm it's really bad and the engine sounds rough too. So my guess is that the ignition is firing within a small range of angles and/or intermittently.

But what could be causing this? I'm not using ignition retard on either of the MSD boxes. Maybe its some sort of signal timing problem between the E11v2 and the MSD boxes? I'm not using the stock FD3S igniter between E11v2 and the MSD's (I already tried this and the engine was misfiring so badly I couldn't drive the car, as well as getting some weird effect with the ignition angle between L1 and T1 growing by over 5* when I checked this with the timing light).

Or would a problem with the signal from the Haltech HE sensor cause this rpm-specific type of problem? I have the sensor mounted so it faces the outside edge (circumference) of the timing wheel, to which the magnets are countersunk. The gap between sensor and wheel edge is 1-2mm. Any ideas?

AdrianE

I found what was causing the ignition to retard, it was a *noisy* signal from the Hall Effect Sensor. I didn't manage to get a 'scope on it, but I went on the Haltech Forum and saw some examples of what can happen to the square-wave signal from HE sensors.

I re-checked HE sensor's wiring and found that the grounding of the braiding of the HE sensor's cable wasn't good enough. I also routed this cable well away from the HT leads and wires that connect the MSD's to the Coils.

Now, from idle there's just an initial retard of about 1-2deg up to ~2000rpm, and it just stays there as the rev's climb to 7000rpm. The "TDC" mark on the pulley still blurs as the rev's climb, but maybe this is caused by the timing light I'm using. The engine sounds smoother too.

Also I found a slight vacuum leak from the intake elbow. On a road test today, my boost guage showed a maximum of 12psi now instead of the 10psi I expected!
The Blitz SBCiD boost controller is switched OFF, so it should be running from the 8psi w/g spring. Maybe the w/g is sticking for some reason? It's a Turbosmart 44mm w/g and I used to get 1-2psi of boost creep before (but not 4psi !!). I'll check the vacuum lines to the w/g first, but maybe I'll have to replace it.

C. Ludwig

iTrader: (27)

The mark on the stock FD trigger wheel is 20* ATDC not 5*.

AdrianE

Yeah it's 20* ATDC, I kept thinking in terms of when L1 fires when the Trigger Angle is set to a certain value, assuming a 15* split between L1 and T1.

But I locked the leading timing at 10* BTDC and locked the trailing timing at of 20* ATDC, set the idle to just under 2000rpm (to stop the multispark on trailing), then adjusted the Trigger Angle until T1 lined up to the timing cover mark. So now it's done, the turbo is not heating up like it used to and all I have to worry about is a couple of degrees of timing drift due to the HE sensor.

Thanks,
Adrian

C. Ludwig

iTrader: (27)

Sounds good.

Enable the advanced trigger angle setup and you can program a variable trigger angle versus RPM. That means no drift when properly setup.

AdrianE

I didn't know you could do that, there's nothing in the manual on that subject. Sweet! It should save me time as now I don't have to modify my ignition map. Thanks again.

AdrianE

I moved the HE crank sensor so that I could use use a larger trigger angle (ie. to give the ecu more time to work out when to 'fire'). It's now 74* (at idle) instead of 34*, and the engine runs much smoother now.

But, the timing mark still gets kinda blurry (spreads out) when the revs are above 4000rpm. This happens on both L1 and T1. I don't know whether this is caused by the timing light, or a form of ignition breakup. Could this be down the MSD boxes feeding standard FD coils?

AdrianE

I'm thinking that the MSD Digital-7 and DIS-2 ignition boxes I'm using with my setup (as described at the beginning of this thread) are still causing some very high EGT's. When I run the car at night I can still see the exhaust manifold / downpipe / turbo housing glowing orange after a quick run down the road in 2nd and 3rd gear revving up to 7000rpm.

It's not as bad as it was before I changed the crank sensor position and then zeroed the timing using the 20*ATDC mark on my crank pulley, but it is still bad enough to make me worry about losing the engine! I think my wideband sensor is overheating as it just registers the minimum AFR of 10.0 when I hit 4-5000rpm on full boost (which is still only 10psi). Then after a few seconds of cruising it starts reading 'as per normal' again. I also get a knock reading of 3-4V (max is at 5V on the PLX data logger) at 4-5000rpm when just cruising, and over 4V on boost.

The engine doesn't sound harsh but I still don't know if I'm getting ignition breakup of some kind, due to
(1) using standard FD coils (which don't appear to be arc'ing and seem OK at idle)
(2) maybe some strange effect on the E11v2 ECU caused by RFI noise from the MSD boxes (which are both mounted in the footwell space opposite the ECU , about 2ft away, because of space limitations in the engine bay).
(3) power drain by the 2 MSD boxes (even though ECU voltage drops to min of 13V, and is usually around 13.8V)

When pointing the timing light at the pulley, first locking the timing and then gradually increasing the rpm up to 6000rpm, the timing mark spreads out according to engine rpm. But it spreads out much more than I would expect (ie. its approx 2mm at idle, and then maybe 8-9mm at 6000rpm). Is this evidence of 'spark' breakup?

Ludwig, Crispeed, Claudio and any other tuning experts out there, can you please give me some more help on this problem?

13B-RX3

iTrader: (2)

How in the world do you do that. I have looked at the software for about 30 min and i cant find it anywhere!


Thanks!
Allen

BDC

Trigger setup, Allen. Geez.

B

13B-RX3

iTrader: (2)

I don't see it there

Havoc

here's a pic

13B-RX3

iTrader: (2)

Thanks! I guess I'm blind.

AdrianE

Advanced Trigger setup is definitely a useful feature...not sure how much you would use on the standard FC or FD CAS but I'm using up to 4.5* at 6000rpm with my Hall Effect Sensor.

I swapped my leading FD coils for 2 MSD 8230's running in parallel and I noticed a huge difference in smoothness and power. I guess the Digital 7 was too much for the standard inductive coils! They ended up causing a lot of misfiring even though they worked OK at first.

But there may still be some ignition triggering problems though. I hooked up an oscilloscope to the Lead trigger wire and did a few runs (at 10psi max boost) in 2nd and 3rd gears to 7500rpm. The 'scope logs show that some of the triggering events (ie. on rising edge) are being skipped and also a few occur at the wrong intervals (eg. > 60* too early or late).

The HE sensor signals seem OK with no load at 5000rpm, but I haven't seen what they're like under load though. The battery voltage under load does have spikes which could be occuring when each plug fires. So maybe the (MSD 8.5mm) HT leads need better routing or could it be the wires from the MSD boxes to the coils? I'm also taking off the rev limiter while I investigate as this could have be cutting in due to a bad trigger signal.

13B-RX3

iTrader: (2)

What sensor are you using?

AdrianE

I'm using a Haltech S4 Hall Effect sensor, which is triggered by 3 magnets (1 for home, 2 for trigger).

I'm not totally convinced that the ECU is sending leading ignition pulses at the wrong intervals, as there are some missing samples in my 'scope logs. But I when measured the ECU supply voltage with the 'scope I found a suprising amount of RFI noise. Huge spikes (+-5V and 50uS duration) when the engine's under load at high rpm, and smaller ones at idle. These are related to the firing of the Leading plugs. There are also much smaller spikes coinciding with the trailing plugs, and more noise which looks like it could be caused by the charging and discharging of the 4 coils.

This can't be helping the ECU, and some of this is finding its way into the ignition outputs. Quite how much is uncertain because the 'scope leads could also be picking up this noise. Now I have to figure out how its finding its way into the wiring...

I routed the MSD HT leads away from the loom wiring in the engine bay, but this had no effect! The MSD coil to HT lead connectors are not fully covered by the boots so I will fix this and see what this does. Also I'm wondering if the leads from the MSD boxes to the coils could be causing this noise as these boxes are mounted inside the car (even though I tried to keep these leads away from other wiring)?

13B-RX3

iTrader: (2)

How do you have the output of the Haltech setup. I initially had the Haltech setup like Haltech said (constant duty) and i was loosing spark randomly. I set it to Constant Charge (like MSD said) and have not had any issues since.

AdrianE

Yeah I've seen the posts about using constant charge instead of duty too. I have mine set at 2.5ms rising edge. I could try a range of 1.0-2.0ms and see if this makes any difference. I think CDI coils only need 1ms or so, but inductive (standard) coils need more charge time...

13B-RX3

iTrader: (2)

I am running two digital-6 and a DIS-2. I have mine set to 2.5 ms rising edge with a break time of 600 and have not had any issues.

C. Ludwig

iTrader: (27)

It's possible for any or all of the above to cause issues. CDIs can create a lot of problems. Are you running resistor plugs? If not try switching to those and see if it helps. Log "trigger at last home" and rpm and see what you come up with. If there are any spikes in the rpm or the trigger at last home is not a consistent 2 then you have a trigger issue which can easily be created by the noise the CDI boxes make.