So my buddy has a problem with his speedometer not being accurate. The weird thing is, it reads right when logging, but it's reading 15 mph FASTER on the cluster. Adjusting the speed pulses per mile isn't changing anything, so now he's looking into the IP divisor constant, but needs to find out valid settings.
Can anyone suggest any tips?
It is typical of these gauge clusters to read too high. Something about too much resistance on them as they get older. Hopefully someone else will chime in that knows more about electronics.
Here's a quote from Shawn(GnatGoSplat) @ the w-body board:
"Anyway, there is a white ceramic 14-pin(?) chip. It's not really a chip, but just a piece of ceramic with a black resistor painted on it. If you look carefully, you'll see a cut line in the resistor. That's how it was fine-tuned, the deeper the cut, the less the resistance. Laser-etched resistor I think it's called. Anyway, my theory is over time, the resistor degrades and the resistance increases. The way I fixed it, was I attached a 1-megohm (or was it 10-megohm?) resistor across the original laser-etched resistor in parallel. Basically, one leg of my resistor I soldered to the pin connecting to the laser-etched resistor, and the other leg of my resistor I soldered to the other pin of the laser-etched resistor.
That gave me perfect accuracy. At first, I thought it might be better to use a tiny adjustable trimpot for fine-tuning, but the fixed resistor I used worked perfectly. I'm not sure if the same value will fix most peoples' dashes, maybe I just got lucky.
I'll see if I can dig up pics. I'd like to think I took pics of the actual mod, but not sure if I did."
This is referring to a factory TGP cluster of course...(not sure if you're referring to a TGP or a car with a TGP swap? Aren't you the one with a friend that has a TGP setup swapped into a J-body?)
I am friends with a couple J bods with the TGP swap and A L body too. This is for a L body.
This is how it went down.
The car was stock with the V6 and 3t40 trans and everything worked great. No mess ups in the cluster, idle was perfect etc etc. The person had used parts from the TGP and made it work on his beretta. Then he had help with the tuning by someone that is good, but not familiar with the L bodies. So After a tune, they realized the speed reading on the laptop during a log session was reading correctly and the speed on the gauge cluster was reading 15 mph faster.
:???:
Sounds like a cluster issue.. What are you using for datalogging and tuning..
The cluster has been swapped out with 5 other working units. I am not doing the tune, he is with someone else, but I do know he is using datamaster to log it. "AUAF" is what is written on the chip.
Someone once mentioned about using a buffer box...... Will this be needed? Have anyone used one? Also, IF one is needed, where would you buy one from?
I keep saying there is a way, just have not found it yet. BUT it's not me the one that needs it corrected, so w/e. heh
I believe the IP Pulse Divisor is different for each type of cluster. For my digital 88 cluster its 64. My Road Speed Sensor Constant is 18606. These read speeds that are off by about 2 mph at 80 on both the ALDL data stream and my cluster. YMMV
It's able to be dealt with 2mph off, but not 15.
So my friend was looking for this so called buffer box.... I couldnt even find one!
Who sells them?
Are the gauges off by 15 mph the whole way, or is it off 15 mph at highway speed and 5 mph at low speed
Good question, I'll have to ask that specific question.
Now I have to know, why you ask that? What are you getting at?!!?! lol
???
how does the tranny act? Does it shift weird, in and out of overdrive and is the cruise control messed up? Does the speed also fluctuate or does it stay constant?
It is a constant thing. Trust me he looked & I looked into the trans as an issue.. nope not that. For giggles he replaced the VSS and checked all wiring, not that. Swapped clusters from a working daily driven car , not that. We all suspect its in the tune since the car is now using the 8F when factory it was something else (I dont know what a stock 2.8 Beretta 7730 ECM is) .
IF what "GhrarhG" pointed out turns out to be true, what would should be focused in on?
Quote from: TGed on October 09, 2008, 01:49:01 AM
IF what "GhrarhG" pointed out turns out to be true, what would should be focused in on?
Here's an old post that touches on what GhrarhG mentioned:
http://www.tgpforums.com/index.php/topic,2708.0.html
I will forward this to the guy.
Thank you.
I think the cluster reads right now (thanks to that link) . Might be 2-3 mph off still, but the laptop should see corrected speedo still.
I guess GM never has put 100% effort into there development. lol
My friend said the guy went with a "factor of 9 which is decimal "32"."
Apparently it's still off, but only by 5 mph (cluster reading 5 faster than actual speeds) . The datalog is on par though.
Sorry about falling off the map. Been so busy I haven't even had time to work on tuning my car. There are two different constants one to get the ECM right, and another to get the cluster to match. There are different cluster constants for different clusters.
The reason I asked if it was off the whole way by 15 or if it was off by 15 at high speed and less at low speed was that if the error changes with speed it's a percentage error. As in you have a constant that is wrong. If you're off by 20% at 30mph it's only 6mph off, but at 75 the error is 15mph. If the datalog is correct you have the road speed sensor constant correct. Just try changing the IP Pulse Divisor to get the gauge to match. I had some trouble with that, as it seemed to change minimally and then jump.
Good Luck
Pretty much now his cluster is 5 mph off from 20-60 that he knows. I guess he'll try going 70-80 to make sure.
We thank you GhrarhG