rx7locost

I am having trouble with missing and erratic spark. For a bit of background, I am using a modified 13B engine in a Lotus Seven replica. It is a standard S5 N/A block (rebuilt by me) using a stock CAS and stock coils/igniters. It has a custom intake using a Ford throttle body. The MS is a MSII on a v3.0 board. I am using the 2.1.0 released firmware and talking to the MS using TunerStudio MS Beta v 0.991.1 The MSII seems to work well using the JimStim on the bench. But that doesn’t test the CAS function.

Building this car has been a multi-year project. I am not quite ready to actually run the engine yet. That is about a month off. I am in the process of checking out all the circuits etc. So far so good. I have the Intake Air Temp and Coolant Temp functioning. I have installed a 14point7 NAW_7S wideband and it is talking to the MSII. I can command the BAC open and closed. The Ford TPS is calibrated and works smoothly. I am using a single point ground technique, not the RX7 wiring system. Mine fans out from the battery cutoff switch installed between the B- lead and chassis ground. I believe it to be a clean ground implementation because I see NO changes in any of the gauges while I crank. The Air Temp, Coolant Temp, TPS and Air/fuel gauge all are rock solid. I can read RPM on both the TS screen and the RX gauge cluster while cranking.

I moved on to check the ignition system. I stabbed the CAS per the manual at 5 BDC. I gave a quick ignition test by using a spare plug sequentially on each of the 4 HT wires and laying it against the block. All 4 wires have spark, not a consistant, rythmic spark, but fairly erratic. I then set the timing at “fixed” in TS and hooked up the timing light. It too, showed missing pulses and maybe adding a few too. So I checked and rechecked the wiring of the 2nd sensor circuit. I am using the original factory 4 conductor shielded wiring from the CAS to the MS with the shield grounded at the MS end only. I know I could be better off using two separrate shielded wires. Do most installers use the factory wiring?

I have tried to add the .01uF cap to the input of the 2nd trigger circuit that Ken recommends. I found where MS-Extra instructions suggest using a 300 ohm resistor instead. Neither of these fixes the problem. I have tried adding a bit of hysteresis to the standard detector circuit to no avail.

Right now I am just trying to get a stable spark. I’m relatively sure it is a noise issue. What have I missed and what can I try?

Chuck

muythaibxr

On the lm1815 circuit, are you running it with pin 14 going directly to ground?

How many ground wires are you running from the v3 board to wherever you're grounding?

Can you lay out your whole grounding system for me?

So for example, on my car, I have the MS grounded to the engine using 4 wires. Then I have the engine grounded to the chassis off the transmission bellhousing mount, and again near the starter. Then I have the battery also connected to the chassis.

The sensors will not fluctuate during cranking unless you have REALLY bad wiring for their grounds. Normally their grounds go all the way back to the ECU so I'd be surprised if there's a problem there. However, the CAS is a lot more sensitive to such things, hence my grounding method, and the way I did my lm1815 circuit.

Could you post an msq and a datalog as well? That would help me tell if there is any sync loss, and would let me see if any settings are incorrect for this engine.

Ken

rx7locost

Pin 14 on the LM1815 is direct to ground.

For grounding the MS, I have 6 pieces of ~18 AWG wires bundled together and going to chassis ground through about 12 inches of wire. This chassis ground is about 2 ft away from the origination of my "Star" ground and not in any high current path.

For a grounding scheme, I have the battery "neg" terminal going through about a foot of cable to a cutoff switch. The opposite terminal of that switch is bolted directly to my frame. The car has a space-frame. From that same terminal, I go through about 30 inches of cable that terminates on the top of engine's rear iron at the clutch inspection cover bolt. In this manner, any high current grounds are not running throught the chassis.

I got to thinking about the sensors not fluctuating. I agree. All the sensors I mentioned are 2 and 3 wire sensors grounded at the MSII so there wouldn't be any offset during cranking.

I have attached a couple of photos showing the grounding, Also a zipped MSQ. This is the latest file that should have the timing fixed for checking with the timing light.

I'm not familiar with datalogs. Are you wanting the datalog from the top menu or the Ignition logger form the Diagnostics tab? I'll try both later this morning and post again.

Chuck

muythaibxr

I was thinking of starting with the datalog from the top menu, but both would help...

Ken

rx7locost

Here is the Datalog. This is with the .01uF across the input of the 2nd VR circuit. I have also added a .22uF across the +5V in the proto area.

muythaibxr

I'll have a look after work.

rx7locost

Thanks, Ken. I know you do this in your free time to help us all out. I am grateful for any help you give.

I keep on trying to get a composite log but seem to be having some issues with learning its operation.

I also am running 2.1.0. Should I be using -b,-c or -D version?

muythaibxr

Use the d one, although I don't think any of the fixes that went in after the original 2.1.0 release would affect rotary operation.

Ken

rx7locost

I will try to add update the firmware sometime in the next day or so.

I have done a lot of experimenting today. I continue to have problems using the Composite Logger. A lot of the time, it will not register anything. Sometimes it will. When it does capture something, I have problems saving it to a data log file. So, I'll try to explain what I have seen.

Earlier today, I got one composite capture that showed a few seconds of capture. The pulses were very sporatic. I thought about what Ken said earlier about grounding to the MS to the engine and not the chassis. SO I lifted the MS ground wires from the chassis and hooked a clip lead from those wires to the point where the engine/chassis ground wire attaches to the engine, at the clutch inspection cover on the rear iron.

I retested and now have a very consistent pulses in the composite logger(when I can get it to capture. Most of the time nothing shows up on the screen).

I compared it to some others I have seen on the net and found that my 2nd wheel signal looked inverted from other's. So I checked and rechecked and found that my Trigger Wheel setup was set for Falling Edge rather than Rising Edge. I corrected that and now I have a very stable (and correct?) Composite Log. I still can't get the logger to work most of the time and cannot get it to save almost all of the time. below, I have a screen capture of how it sits now.

Now I have a fairly stable timing light at the pully. The problem I have now is, the timing light is about 40 degrees retarded. If I understand correctly, when I set "timing" at fixed and 0 degrees , this should override all other timing settings in the setup, no? My thinking is that I should be able to use the timing light and get close to the 5ATDC that I stab the CAS at no?

muythaibxr

Your tooth #1 angle should be set to 5 degrees BTDC. Is it? I'll take a look at your settings for myself tonight.

Ken

rx7locost

The tooth angle was at -5, at least in the last ms file. I restabbed and now it is very close to what I expected.

I manged to get a couple of datalogs out this afternoon. one for the wheel and one composite.

If you could, let me know if these look reasonable.

rx7locost

Don't know why the file didn't upload. Here it is (I think)

muythaibxr

Please also post the working msq.

Thanks,

Ken

muythaibxr

OK, so a couple more notes based on your old config:

1) If you have VROUT running to TSEL, and you're using an lm1815 circuit, you'll want both the primary and secondary input capture settings set to rising.
2) Other than that, your trigger wheel settings looked fine... 5 degrees BTDC is the proper tooth #1 angle for this engine with the proper CAS stab.
3) I usually run a bit less dwell with the stock 2nd gen coils.
4) The ignition table will probably need some work if it's basically a stock NA engine without much porting or anything like that.

Triggering on the falling edges of the CAS signals was likely causing your weird erratic ignition issues though.

Ken

rx7locost

1) I have the VROUTINV wired to TSEL and one LM1815N circuit. Primary input is set to falling, 2nd set to rising. Should I reverse TSEL wiring and the primary input to rising? Is there a difference in the way the software acts on them?
2) OK
3) I have seen both 3.0 and 3.6 dwell recommended in various places. What value do you recommend?
4) I would guess that my engine should be considered stock. No porting at all. Can you recommend a better start point? I'm sure that it has nothing to do with this problem but a better start point would be good.

Attached is the current MSQ file.

muythaibxr

That's an odd way of doing it, but it should work fine since you've changed the edge your triggering on AND inverted the signal.

I've always used about 2.2-2.6. I've never seen any need for more. The stock 2nd gen coils actually increase the dwell from what you command anyway. When you command 2.6ms or so at 12v, it actually increases the dwell to near its current-limit. Any more dwell than that will just make the coils hot and can potentially generate noise. On my engine, if I go more than about 2.6 I start getting sync loss.

I just think that at full throttle, you'll want a couple more degrees of advance. I'll have to go back and look at it again here in a few to give any other recommendations, but that was what stood out. I think I'm running 25 or 26 degrees at full throttle on an S4 NA. The S5 NA moved the leading plugs up a little higher, which seems to me like it would cause you to need a bit more advance than the S4, and in my experience the S5 engine does actually like a bit more advance than the S4. This is all anecdotal though since I've never had either on the dyno. We were using the "butt dyno" to tell, but 23 was a lot weaker than 25 on the S5 I originally tuned years ago (the first 2nd gen with a fully stock ignition system controlled by the MS).

Ken