Thrasher 92 Timing Curve

[Scott231]

Re: thrasher 92

Originally posted by johnz
I have emailed scott about 3 times foe a chip and he never responds

You need to learn how to type then...I've never recieved an e-mail with the word "Thrasher" in it that I haven't responded to. Now if you are requesting a Thrasher chip for your Ford pickup truck, those go straight into the Deleted mailbox.

 

[gnjones231]

That sounds about as trustworthy as "Trust Me" from a complete stranger on the internet... If you have the answer, and know you are right, then why not share it with all of us?

Well, I'm just trying to help. Someone asked what the timing in a Thrasher was, and like I said, I'm the only one here that can do that. You see, I read the EST signal that comes out of the ecm. So no matter what kind of crazy things people do in the chip, the end result shows up on the EST. I'll bet you didn't know that there is a fudge factor in the chip that is slightly off and this skews the actual timing that the engine sees from what is programed in the chip tables. No scan tool in existance can read the actual timing either. I'll bet you also didn't know that I did some testing for Tom Chou when he was developing this chip. Gee, I wonder if he still remembers me?

So if anyone wants to know the timing in a Thrasher or any other chip, you need to send me the bin file. Might be interesting to see what that 18* or 20* chip has for timing, you may be surprised! You think I'm getting free chips? Nope, I do my own chips and you couldn't pay me to run anything but my own. I also don't do chips anyone else, so don't bother to ask.

Here is some data on a stock chip and a MAG4 chip. Conditions are: LV8=255 Engine Temp=190


rpm 87 87 mag4 mag4
prom actual prom actual
3600 22.1 25.2 34.1 35.9
4000 22.1 25.8 35.8 38.0
4400 22.1 25.9 35.8 38.2
4800 22.1 26.5 35.8 38.4
5500 22.1 27.1 35.8 39.5
6500 22.1 28.2 35.8 38.1
7500 22.1 32.2 35.8 35.1
8446 22.1 35.7 35.8 33.7


As you can see, the actual engine timing doesn't follow the chip table.

So there yo have it. No BS, just the facts!

Dave
gnjones231@mchsi.com

 

[VIN7]

gnjones231,

Could the difference between prom and actual timing be because the prom takes into account the crank reference angle? Just like on the Felpro, user sets the ref angle otherwise the timing on the Felpro doesn't match actual timing.

Mike

 

[Scott231]

Originally posted by gnjones231
Well, I'm just trying to help. Someone asked what the timing in a Thrasher was, and like I said, I'm the only one here that can do that.

Well, you still haven't asnwered the original question...

So if anyone wants to know the timing in a Thrasher or any other chip, you need to send me the bin file. Might be interesting to see what that 18* or 20* chip has for timing, you may be surprised!

So, why are you the only one in TR-land that knows how to do this? Why has this never been mentioned before by chip gurus like Bob Bailey, Scott Mueller and/or Carl Ijames?

Here is some data on a stock chip and a MAG4 chip. Conditions are: LV8=255 Engine Temp=190

rpm 87 87 mag4 mag4
prom actual prom actual
4800 22.1 26.5 35.8 38.4
5500 22.1 27.1 35.8 39.5
6500 22.1 28.2 35.8 38.1
7500 22.1 32.2 35.8 35.1
8446 22.1 35.7 35.8 33.7

Uh, where did you get a stock chip with data tables that go beyond 4800 RPMs...up to 8,446 RPMs? Better yet, how did you measure the EST at 6,500; 7,500 and 8,446? Can you easily explain how to read data from the EST circuit?

 

[blackbuick87]

Originally posted by gnjones231


Well, I'm just trying to help. Someone asked what the timing in a Thrasher was, and like I said, I'm the only one here that can do that. You see, I read the EST signal that comes out of the ecm.

Like Scott said, if youre just trying to help, why not just share the info you have with everyone? You have already stated you make chips only for yourself and have nothing to gain monetarily.

Why are you the only person in the world that can read the EST signal?

I do not want to start a big battle, but we are all here to learn and if you have something to share that is what this board is for

 

[tminer]

I think if you measure the time between the ECM's pulse and the default 12 degree pulse at a known rpm you can calculate the real timing.

I also believe that the ECM can make multiple tweaks to the main timing table value before it arrives at the final timing value.

Sorry, I am too lazy to look up the specifics and too old to remember them off the top of my head.

Tom

 

[ijames]

Tom changed the base timing table offset from stock, which is why when you use DirectScan it looks like you are running 30-40 degrees of timing. Basically add say 10 degrees (I'm making that up, I don't think it's what Tom used) to the offset and subtract 10 from every entry in the main timing table, but since DirectScan assumes the offset is still the stock value it reports a wrong number. If you go through all the changes he made, you wind up at about 18 degrees of advance at wot over 3600 rpm, along with some other changes at lower rpms to enhance spooling. All of this has been posted before both on the gnttype list and in the chips forum here.

As for measuring the est signal directly, that's a neat idea but I have to wonder if the mag4 chip is really giving 38 degrees of advance at wot. About the most I've heard of anyone using even with a stock turbo is 30 degrees, with C16. Also, what were the coolant and mat timing trims under your conditions with the stock chip? Do they add up to the difference you saw between the 22.1 deg in the main table and the measured 25 or so degrees?

 

[bruce]

I'm not convinced you can read the EST line, and be more accurate the reading the RAM location like DS does.

Judging from how linear the timing error is, it looks like there is a constant timing error, ie fixed timing error, and since the GN code doesn't use any spark latency corrections, I doubt the error is in the code or ecm.

 

[gnjones231]

Well, it's already late, and don't have alot of time to answer too many questions or post the data just yet. So I'll answer a few questions.

Could the difference between prom and actual timing be because the prom takes into account the crank reference angle? Just like on the Felpro, user sets the ref angle otherwise the timing on the Felpro doesn't match actual timing.

Going from memory here...The crank reference angle in the stock chip is set to 70*. If this is changed, the EST will also change by the same amount. This would be a global change. In other words, if you change it to 69*, every value, for any load or combination will be off by 1*. I wonder if DS can pick this up? Anyone wanna verify this?

So, why are you the only one in TR-land that knows how to do this? Why has this never been mentioned before by chip gurus like Bob Bailey, Scott Mueller and/or Carl Ijames?

As far as I know I *AM* the only one who can do this. It's not that tough to do. I guess you need to ask those guys you mentioned why they haven't measured the actual timing, I can't answer for them.

Uh, where did you get a stock chip with data tables that go beyond 4800 RPMs...up to 8,446 RPMs? Better yet, how did you measure the EST at 6,500; 7,500 and 8,446? Can you easily explain how to read data from the EST circuit?

I've got a bench ecm. I've had one for close to 13 years already. It's real easy to 'spin the motor' to 8 grand To read the EST, I made an electronic interface between the crank signal and the EST signal and wrote a program so my laptop can read it. I took some readings way past what anyone can run just to see how the ecm reacts.

Like Scott said, if youre just trying to help, why not just share the info you have with everyone? You have already stated you make chips only for yourself and have nothing to gain monetarily.

I will share what I've found. A few days ago I didn't have a thrasher chip to check. I've got one now, actually three, and I've already tested them, and made some graphs. The data is coming...keep your shirt on...and there is some interesting stuff going on...possibly a bug in the ecm program

Why are you the only person in the world that can read the EST signal?

Because I'm probably the only one who has tried! I don't see anyone who has posted here saying that they have, and until someone pipes up, I'm still the only one!

I do not want to start a big battle, but we are all here to learn and if you have something to share that is what this board is for

I'm trying to share, and have shared similar information in the past, I guess people just forget. There will be no battle here, you are unarmed By this I mean, if you have no data to support any claims, then you can't enter the fray with any credibility. Chest beating is not allowed

Well, that's it for tonight. I'll continue tomorrow. Sorry about not showing who the quotes belong to, you'll just have to reread the posts to figure that out.

Dave
gnjones231@mchsi.com

 

[bruce]

Originally posted by gnjones231

I've got a bench ecm.

Ahhhhhhhhh,
the magic of actually seeing what's going on, LOL....
I thought I was the only one, bothering with one.

Next you'll be telling me you've gone to the trouble of making a ROMless conversion......
That's when things get neat in my book, develope the actual source code, and then have some fun.

 

[Danster]

Being that you obviously know much about other chips and apparently don't like them much as you do your own chips, why would you not offer a chip that would be perfect for other TR owners? Great way to make mucho $$$ unless of course you already have it Do you share any of this info to other chip burners?

Curious though, you mention of a flaw in the ECM, can you tell us if this can be easily fixed? Does this flaw hurt the car's performance?


Dannyo

 

[gnjones231]

Well, I just spent 1.5 hours answering two statements from tminer and my windows crashed and lost it all. I'm now tired and mad as hell. It's awful quiet in here, is anybody still interested in this thread or am I wasting my time?

Dave (gettin' ready to beat the snot outta my pc)

 

[UNGN]

No, we're still interested. I feel you pain about your computer... at least once a month, sometimes twice.

 

[tim87tr]

Still interested here as well. I can follow at least half of what you are talking about, but can research and learn the rest . I just ran for the first time a Thrasher 97 with 100 octane Xylene mix and just traded a RA 93 for a Thrasher 100, so I am curious to know any differences in timing of the Thrashers. I have some basic knowledge of timing from my Trans + adjustments. Tim

 

[Danster]

I'm still interested. I don't know much on these cars but I can be a good student

Dannyo

 

[gnjones231]

I need to back up just a bit here.....some idiot posted....

gnjones231 wrote

Someone asked what the timing in a Thrasher was, and like I said, I'm the only one here that can do that.

Geeezzz, I can't believe I said that. That's not exactly what I meant to say. I didn't mean that noone but me can read the actual timing. In fact, anyone can read the timing on the est output if you have the desire, knowledge, and equipment to do so. What I meant to say was that I believe that I was the only one here who had taken the time to read the est signal and compare it to the tables in the chip and discover a descrepency. Sorry , my bad.

tminer wrote

I think if you measure the time between the ECM's pulse and the default 12 degree pulse at a known rpm you can calculate the real timing.

Well, you're on the right track. But let me explain how it's done. I have two ways of measuring the actual timing that the ecm puts out on the EST signal. The first is an electronic interface between the RPM signal ,the EST signal and my laptop. In order to read the timing you need to measure the amount of time the RPM siganl is 'low', by starting a counter on the falling edge of this signal and stopping this counter on the rising edge. By knowing the time and by knowing that the 'low' time also represents 60*, you get time/60*. At 4000 rpm, the low time will measure 2.5ms. Let's label this T1. Knowing this, you can calulate that there are 41.66us/degree. Now, if you start another counter that starts on the falling edge of the crank signal (rpm signal), and stop this counter on the falling edge of the EST signal, you can calculate the degrees in advance , which is the actual engine timing. The time measured this time will be labeled T2. The formula to calculate the timing is this...

actual timing = K1 - T2 / (T1/K2)

where:
K1 = the starting time in degrees. Since we start the counters on the falling edge of the crank signal and there is a 10* offset, and the width of the 'low' signal is 60* then K1 = 60*+10* K1=70*

T2 = time from crank falling edge to EST falling edge.

T1 = time from crank falling edge to crank rising edge

K2 = Number of degrees from the falling edge of the crank signal to the rising edge of the crank signal. There are basically 3 'hi' levels and 3 'low' levels of this signal. That makes 6 level changes in 360 degrees of roatation. 360/6 = 60* per level change. K2 = 60


The other way I use to measure the EST signal to get the actual timing is with my scope. By having two different means of measuring the same thing, you get correlation. Correlation is very important, because without it you can get misled very easily. I'm sure *correlation* will be mentioned again later on.

tminer wrote

I also believe that the ECM can make multiple tweaks to the main timing table value before it arrives at the final timing value.

There is a least one tweak that the ecm makes to the timing tables. As most of you know, the timing tables in the chip only go to 4800 rpm. How can the timing be affected after this rpm is exceeded? At address 164 hex is something called 'time domain correction to spark'. The value at this address is 0D hex. In the stock chip at 4800 rpm,LV8=255 the timing is listed as 22.1* In actuality it's about 26.5*. If I change the data at address 164 hex to 04, the actual timing becomes about 22.1* If I check the actual timing at 6000 rpm under the same conditions, at data=0
d hex the timing is about 28*, change the data at address 164 hex to 04 hex, the actual timing is now about 22.1* Just for grins I'll change the data at 164 hex to 20 hex. The timing is now 37.8*
So now we have a good news bad news thing. The good news is that there is a way to affect the timing after 4800 rpm. The bad news is that is skews the actual engine timing from the prom tables.

So far, I've got 2 hours in the above. and I'm getting tired sitting here at my pc. I can't believe it's taking me this long, and I still haven't posted the thrasher data. I guess when I'm trying to explain something in depth, I read and reread and try to make sure that it makes sense and has no errors. There's nothing worse than a bunch of BS that waste's people's time. I hope this doesn't fall into that catagory Tomorrow....


Dave

 

[bruce]

Originally posted by gnjones231

There is a least one tweak that the ecm makes to the timing tables. As most of you know, the timing tables in the chip only go to 4800 rpm. How can the timing be affected after this rpm is exceeded? At address 164 hex is something called 'time domain correction to spark'. The value at this address is 0D hex. In the stock chip at 4800 rpm,LV8=255 the timing is listed as 22.1* In actuality it's about 26.5*. If I change the data at address 164 hex to 04, the actual timing becomes about 22.1* If I check the actual timing at 6000 rpm under the same conditions, at data=0
d hex the timing is about 28*, change the data at address 164 hex to 04 hex, the actual timing is now about 22.1* Just for grins I'll change the data at 164 hex to 20 hex. The timing is now 37.8*
So now we have a good news bad news thing. The good news is that there is a way to affect the timing after 4800 rpm. The bad news is that is skews the actual engine timing from the prom tables.

I don't for some reason see it.
The factor at 0164 is usec., or entry of (D) is 14 or .000014 secs.
That ain't a whole lotta time.
I also looked thru the code and see it only mentioned once, and that's at Compute and output EST parameters to ECM. And it's used as a subtraction, rather then an addition.

I also tried using values of 0D, 04, and FF.
There were no reported differnece in reported timing.

I've been thru the code, and just don't see what your claiming, Ya there will be differnces in the events you mention as a function of timing, and dwell.

I just wonder if it's just not something factored in for difference in cam sensor timing due to slop in the timing chain or something.

Not argueing or anything, just trying to inderstand it.

 

[FlyinGN]

this is a *VERY* interesting thread.. Some good knowledge going on here

 

[ijames]

I thought the time domain correction to spark was an allowance for whatever delays there are in the ignition module between getting an est edge and firing the coil. 14 usec sounds reasonable for some digital logic and then a driver and then a big power transistor to switch, and it should be a subtraction since the est has to go low early so after the delay the spark is at the right time. That delay shouldn't change with rpm, but it will represent a different number of degrees as the rpm changes, which agrees with what you are seeing. Any chance you can hook up a module and coil and use an inductive pickup timing light to see if the actual timing at the plug is varying like the est output that you are looking at? I've done that at idle and maybe 2000 rpm in park with a delay light and what I read off the timing light agreed with what directscan was showing the timing should be.

Definitely keep this thread going. We were quiet (well, I was ) because I was waiting and hoping that you would post again.

 

[tminer]

gnjones231,

Are you positive that the crank signal is symetrical? I assume it is electromagnetic with some signal conditioning so I have to wonder. Is the falling edge the actual trigger? I have read elsewhere that it is the rising edge. I wonder what your results would be like if calculatiing secs/degree using trigger edge to trigger edge (120 degrees).

I do not see the difference in chip vs measured as being linear. It is thru 6500 rpm for the stock chip but then gets a bump. I am looking at degrees/rpm variance. If the tables only go to 4800 then I would think that there must be something else happening.

I could be all wet. My observations come from my own lack of knowledge as much as anything. I am just trying to understand. Obviously I don't know your entire setup and you may very well have all this accounted for. Heck, I end sentences with prepositions!

Tom