spark advance

Hi everybody,
is there a tool like trimalyzer for the spark advance table?

I could convert my amc 304 engine to tbi.When I used the original spark advance table from BPSA.bin the car didn`t run right. I have set the table according to the spark advance curves from the jeep-manual. Now it runs good when rpm is higher than 2000.At lower rpm and when accelerating i have to do that very smooth otherwise engine is bogging and trends to stall.I raised up spark advance in lower rpms slightly but no success up to now. When I take an other bin (bhmc) from a 5ltr engine, engine will not accelarate at all.
Attached my last logfile where ve tables are not yet set by trimalizer.

trimalyzer does have a knock analyzer built into it but knock analysis is as good as it gets, no tool will automatically tune spark (too dangerous, needs a practiced hand)

spark advance is tricky for sure

keep working on your custom table based on the jeep manual and don't forget that timing isnt' everything, maybe fuel has something to do with it too. where it bogs, add a few more degrees and see what happens. in areas of light load and low rpm its harder to cause damage.

Hi Steveo,
thank you for the quick answer.
Yes i have already realized the Knock table in your trimalizer.I have 4 big knocks in there.When I analyse the log.csv in excel, there are 5 knock retards.
Do I have to decrease my spark advance or increase it at those cells?
When I have the history running in tunerpro-rt I can see the knock retards in those cells but they are overwritten next time when no knock was detected.
Is there some manual about spark advance?

if it's knocking definitely remove some spark advance. if it still knocks there remove some more. if it still doesnt' go away you might have false knock (mechanical noise or knock filter no good for engine) or it's lean knock.

i'd focus on the areas with groupings of knock events, those are probably real.

honestly i can't imagine your knock sensor and its filter are proper for your amc engine so you might not be able to trust it? you'll find out when you try to tune it back a bit

Hi Steveo,
thank you for the quick answer.
Yes i have already realized the Knock table in your trimalizer.I have 4 big knocks in there.When I analyse the log.csv in excel, there are 5 knock retards.
Do I have to decrease my spark advance or increase it at those cells?
When I have the history running in tunerpro-rt I can see the knock retards in those cells but they are overwritten next time when no knock was detected.
Is there some manual about spark advance?
As well i saw that my park brake switch is on all the time. I know I havent connected the pins because I could start the engine with and without connection.Does that matter on the behaviour of the engine?

Hi steveo,
your answer was too quick, I wanted to add this question: As well i saw that my park brake switch is on all the time. I know I havent connected the pins because I could start the engine with and without connection.Does that matter on the behaviour of the engine?

park brake or transmission park/neutral?

they're very different things.

i've never seen an oldschool GM ecm where the park brake switch does anything

park/neutral or shift position switch definitely could if you're running an automatic bin.

Hi steveo
it is an 16196395 and my bin is set to manual transmission! In the csv it was the park brake. So I think this signal will not work on the programming.
In the meanwhile I had a look to John Wilson`s "How to tune a GM EFI System" He wrote he has an excel sheet what calculates spark advance based on spark advance out of the manual given at 100% MAP.He would sent it on demand.
Unfortunatelly he is no longer reacheable on the internet.Does somebody have his spreadsheet or his new email adress and is willing to provide this to me?

I'm noticing the same spark retard in the same 100 kpa area usually when I stomp go pedal rolling or from a dead stop. Took some spark out in that area but won't know the results until next play time (not my car). Don't own a wideband but do notice on the tip in that the O2 starts way down low and works it's way up to ~.950mv. When that occurs, knock retard diminish until gone. Idk if it's me expecting instantaneous feedback from the log or log is really lagging. I'll fatten it up if reducing spark advance doesn't help.

if your o2 voltage is below switch point at anything but cruise or closed loop, you're lean. need to add some more fuel.

maybe spark isnt your problem at all?

Thanx for the insight! Trying to keep this on topic, went back two weeks ago and watched some wot datalogs that didn't exhibit any knock retard in the 100kpa area, so two steps back (bin) compared to what I changed. Definitely took out fuel where it matters! I got too caught up in seeing 1000+mv and pulled some fuel out.

@steveo
which tables do I have to observe ? Engine is running good over 2000rpm but on accelerating spark advance falls back to zero.In open loop this is not that much but in closed loop engine is nearly stalling.I have to accelerate very smooth when rpm gains ,spark advance gains again and engine gets power.
Or should I increase spark advance generally in spark advance at idle table?

who may analyse my xdl?
Here I have the monitor at a certain point where knock retard occurs. In none of my spark tables this stellation occurs
MAP about 100
rpm about 1000
spark advance 15.8
what can cause this spark advance?

Watching this log and the previous posted, something seems amiss. Take it you're not running vss as I don't see any movement for mph. Some have alluded that this pcm needs vss but I don't know what negative impact it would have. Couple of observations, certain areas tps cycles up and down rapidly within a span of seconds. 1st log seen rpms go from 4500 to 0 jump to 6300 back to zero w steady tps. When it's wonky seen bpw go as high as 20.0 way beyond static then drop to 0 along with idc. Iac pegs to 150. Also yes there's periods of times spark would be hovering at 31* then drop to 0*. Suspect grounds? I would check tps ground and w key on engine off operate throttle while observing tps operation. Make sure volts within idle/wot specs and increase smoothly.

Hi,
yes I changed some parameters between first log and this one.Mostly in the spark advance tables and had run trimalizer a few times for the ve tables.I have only connected the DRAC but no speed sensor. The speed sensor I could get has three pins red, black and white but the DRAC has only two input lines.I checked TPS signal while engine wasn`t running today , TPS and throttle angle reacted both in tunerpro. But I could find these abnormalities you mention when I examined log file in excel too.Will continue my search .
Thank you for your comments.

Sorry before I choke on my foot which adx you using? Assuming $0E???

hi stew
no, its $0D.
Attached bpsa14.bin and bpsa15.bin.I have changed spark advance in bpsa 15 completely but not driven it yet. My problem-zone is idling up to 2000 rpm, then the car is running.

Hi stew,
i read too fast. I thaugt you speak about the bin. The bin is $0D. The adx was $31 , i have improved it to §31_new.adx

Sorry for the delay just got back from a business trip and didn't bring my tuning laptop with me. Gonna try to cut to the chase...I opened two sessions of tunerpro and ran your logs side by side using a $0D.adx and your $31.adx. Good news is the numbers are purdy much dead on the same in the history although there's some disparity in knock retard in same reported areas. However when using a $0D.xdf and $0E.xdf...line address do start in different areas in the bin between the two areas I've looked at. Spark shows all negative numbers in the $0E.xdf. VE in $0E.xdf has areas where values random zero with areas in high 90s slightly off idle. Someone correct me if I'm wrong but if you're going to use a $0D.bin then you should be using a $0D.xdf and $0D.adx. Your screen shot a few post back, you don't find it odd that you have ~dozen peaks and valleys in the TPS that occurs in a few seconds? Or is that your foot feathering the throttle in that area???

Hi stew,
thank you for looking to my files.I will have a look for a 0D-xdf and a 0D -adx.I ordered an other Speed sensor and will wait until this works, then it will be much easier to to compare all the values to the momentary speed.I think the peaks and valleys were from feathering the throttle to prevent engine from dying.In the meanwhile i had copied the spark table from BHMC.bin to my bin but driving was worse. I took my spark table again and decreased spark advance at high MAP values and car ran better but still with problems below 2000 rpm.My question could I delete one table (spark advance near idle) or should I set all values to zero there in?

Quote:

---

Originally Posted by mecanicus

Hi everybody,
is there a tool like trimalyzer for the spark advance table?

I could convert my amc 304 engine to tbi.When I used the original spark advance table from BPSA.bin the car didn`t run right. I have set the table according to the spark advance curves from the jeep-manual. Now it runs good when rpm is higher than 2000.At lower rpm and when accelerating i have to do that very smooth otherwise engine is bogging and trends to stall.I raised up spark advance in lower rpms slightly but no success up to now. When I take an other bin (bhmc) from a 5ltr engine, engine will not accelarate at all.
Attached my last logfile where ve tables are not yet set by trimalizer.

---

Don't waste your time using any of the stock Chevy timing maps. To be honest, I know nothing about AMC engines and what they like for spark timing. My approach to this would be to do a Google search for a AMC 304 spark timing table. There was one hit that I saw that they were using megasquirt. Perhaps you could experiment with that. https://grassrootsmotorsports.com/fo.../118096/page1/

Arghh...just reread it's referring to a AMC 360...

everyone needs to stop interchanging the letter O with the number 0 in mask names.

masks are hex, there's no letter "O" in hexidecimal

this will help you when searching for information about masks like 0D and 0E (also help others that might be searching for threads and posts like this one)

Hi steveo
you are right, so I wouldn`t mind if an administrator would correct this!

Two things that may help:

• A VSS is necessary in most all applications because there are many IAC positioning and other decisions made by the ECM code based on whether the car is moving or not. This applies to both manual and auto cars. Does not affect spark, but does many other important items.

• Playing hit or miss with spark is difficult. Best to just build a table based on your requirements. This may help: https://www.thirdgen.org/forums/diy-...ml#post6301288

One other general note. For those who don't have the ADX for your mask, it's always helpful to post a .csv or .xls* file that you export with your ADX rather than posting the XDL. More people might be inclined to review the log and help if this if done.

After some tries and errors, i could build a spark table what seems to work on the engine. But engine is not running round and when having a look to the graphs, I see some spikes in rpm. Could the ICM cause these spikes or should I search for an other fault?
The txt(csv) file is only a small part because its real size was too big for the upload.

Quote:

---

Originally Posted by mecanicus

Hi steveo
you are right, so I wouldn`t mind if an administrator would correct this!

---

I have corrected the offending messages in this thread.

Now I have changed the icm with brandnew one but the same spikes and interruption in ignition.
What to change next ? ignition coil ?
What other options do I have?

Those RPM spikes are odd. Are you using a duraspark dizzy/hall sensor? I wonder if your ignition (coil and spark plug) wires are so close to the hall sensor wires that they are interfering with the signal...?

No, it is the original AMC/Ford distributor , mechanical and vacuum regulation eliminated. The spikes are not that regular that they will be caused by interference but I will check that too. I had the car conected to an old motor-test with osciloskop and couldn't recognize afault in the curves. I tend to change the coil too.
Other ideas ?
Attached my last logfile.If I convert it to txt it has 6000kb and is too big tofor attaching!

Quote:

---

Originally Posted by mecanicus

No, it is the original AMC/Ford distributor , mechanical and vacuum regulation eliminated. The spikes are not that regular that they will be caused by interference but I will check that too. I had the car conected to an old motor-test with osciloskop and couldn't recognize afault in the curves. I tend to change the coil too.
Other ideas ?

---

Copy (Duraspark dizzy with advance removed). When you removed the vacuum advance, and welded your mechanical advance, did you screw down the hall sensor so it cannot move?

My understanding, in our situation (I have a 401 with a similar configuration), The computer gets the RPM signal from the Ignition module, the ignition module gets RPM signal from the hall sensor. If you have spikes, then the fault is somewhere in that chain. Do you have an extra ignition module you could try? Are you running the Ford wide cap or stock AMC tall cap? The hall signal is pretty sensitive so if your coil wire or one of your plug wires runs near it, EMI could distort the signal (especially if you ignition wires are acr'ing). I am also currently fighting really crappy ignition wires (arc'ing to each other and ground). I am not sure that failure would show similar results but might be something worth checking.

Hi,
it seems you were right.I moulded the cables with wire mesh and grounded this. I remember we had this on gasoline driven cars in the army. At first glance (had no time to drive it) it looks good.
Attached a view pictures

I hope that does it! Looking forward to hearing about it after you drive it.

Hello, up to now its running without spikes in ignition.Here is how i did the isolation.
First i had a used water hose from the plummer.

next I cut both ends

when you press the mesh together you can easily pull out the inside.

Put a piece of a round wood inside the mesh, then you can put on some solder to the mesh.
next post!

solder a piece of wire to the mesh, it will be the earth point.
put some shrinking tube on the mesh
put the cable inside the mesh
shrink the shrinking tube on the ends of the mesh
reddy cable
build in with earth point