Help! Supra doesn't run right

Piratetip

Far From Maddening Crowds
Staff member
Super Moderator
Dec 30, 2005
1,177
69
48
38
MKE, WI
Start methodically unplugging every sensor while the engine is running.
AFM
TPS
Idle Air Controller
Coolant Temp
ect...

Test one, plug it back in, go to the next.
May give you a clue which sensor is culprit.
 

Abe's 1987

Member
Sep 5, 2017
236
5
18
Houston,TX
Did that and only thing that made a slight difference was the afm which once I unplug and revved up it stalled. But it still acted the same revving up as it does with the afm plugged in other then stalling. But I didn't test the coolant temp though. Could that cause this problem?
 

Abe's 1987

Member
Sep 5, 2017
236
5
18
Houston,TX
I opened the ecu and didn't see anything out of the ordinary other then stains in a few places as if some liquid was dropped on it.
 
Last edited:

Abe's 1987

Member
Sep 5, 2017
236
5
18
Houston,TX
I just thought of something! So my headlights stay on unless I unplug the battery(will fix that problem later). Leaving the battery unplug, doesn't that erase the error codes? And wouldn't I need to drive the supra to get the error codes to come back on? Cant believe i didnt think of this until now. If that's the thing then it will be fun for me to get the codes back as the supra won't drive.
 

plaaya69

87T Supra
Nov 18, 2006
947
7
18
Lake County, IL
Yes that will cause a problem and the engine light might not come on right away. On my Supra lately I been messing with the ECU Masters standalone and going back and forth to the OEM ECU as well. What I notice when I do this is once I plug back in the OEM ECU back in it needs to relearn and it will not run idle good when coming to a stop and will stall out/die but once you turn it off and start it back up it runs fine. If I were you pull the fuses out of the fuses box under the hood for your headlights and get some driving going on with some on/off key cycles also. You most likely will get some codes then.
 

Abe's 1987

Member
Sep 5, 2017
236
5
18
Houston,TX
Yes that will cause a problem and the engine light might not come on right away. On my Supra lately I been messing with the ECU Masters standalone and going back and forth to the OEM ECU as well. What I notice when I do this is once I plug back in the OEM ECU back in it needs to relearn and it will not run idle good when coming to a stop and will stall out/die but once you turn it off and start it back up it runs fine. If I were you pull the fuses out of the fuses box under the hood for your headlights and get some driving going on with some on/off key cycles also. You most likely will get some codes then.

Will codes come if the supra sits and idles for ahwile? In the state the supra is in it wont drive but it can idle.
 

Abe's 1987

Member
Sep 5, 2017
236
5
18
Houston,TX
It might but some codes need to exceed certain values or engine loads in order for the light to come on.

Ok gotcha, so as an example a different voltage range of one part could set of error codes.

So I went ahead and temporally fixed the headlight issue so I can leave the battery plugged in. Will start her up later today and hopefully at least drive a block or so if she let's me and leave the battery plug in and do this over the next couple of days. Hopefully I can get an error code that can lead me in the right direction.
 
Last edited:

suprarx7nut

YotaMD.com author
Nov 10, 2006
3,811
1
38
Arizona
www.supramania.com
Ok gotcha, so as an example a different voltage range of one part could set of error codes.

So I went ahead and temporally fixed the headlight issue so I can leave the battery plugged in. Will start her up later today and hopefully at least drive a block or so if she let's me and leave the battery plug in and do this over the next couple of days. Hopefully I can get an error code that can lead me in the right direction.

In my experience, one common thing that does not throw a code is a vacuum leak. It's probably a silly question now, but I don't suppose you might have a vacuum leak somewhere?
 

Abe's 1987

Member
Sep 5, 2017
236
5
18
Houston,TX
Ok so update. Started the car today and let it get up to operating temp. Coolant temp connection seems to be messed up as the gauge didnt work and then after 15 mins of ilding it jumped to mid range and stayed there. Turned it off and car insistently had no power to anything. Check volts on battery and it had 12.4 volts but no volt output to anything. Unplugged negative on battery for 5 min and plugged back in and everything had power again. I did notice that the volt on the battery did not rise up when the car was running as if the alternator was not charging. But tested the volt output on the charge fuse and it was at 14 volts. When the car gets to operating temp the revs started fluctuating between 1100 to 1500. I hear a loud hissing sound from the air filter. I don't think there are any vacuum leaks. Waiting for the ecu as it just shipped out today.
 
Oct 11, 2005
3,815
13
38
Thousand Oaks, CA
This car was designed in the mid 80's, that was the era of floppy disks and the IBM PC was state of the art with 8 MHz CPU and 64k RAM. Get the picture. There is not a lot of computing horsepower in the ECU, and it has to do a lot of things. So the diagnostics are mostly rudimentary.

The ECU only sets a code for the AFM if the Kv signal is shorted or open. If its giving bad data it will never set a code. The ECU does not have a way to compare against other sensors like a modern ECU would do and figure out something is wrong with the AFM.

So, for example, if someone has "cleaned" the AFM with TB cleaner and screwed the optics up, chances are high the ECU will never set a code, but it will run like crap because the sensor is FUBAR.

That is why it would be helpful to measure the frequency of the Kv signal during operation. You will know immediately if its bad or you need to search elsewhere. Relying on codes is certainly good practice, but not enough to always figure out what's going on.
 

Abe's 1987

Member
Sep 5, 2017
236
5
18
Houston,TX
This car was designed in the mid 80's, that was the era of floppy disks and the IBM PC was state of the art with 8 MHz CPU and 64k RAM. Get the picture. There is not a lot of computing horsepower in the ECU, and it has to do a lot of things. So the diagnostics are mostly rudimentary.

The ECU only sets a code for the AFM if the Kv signal is shorted or open. If its giving bad data it will never set a code. The ECU does not have a way to compare against other sensors like a modern ECU would do and figure out something is wrong with the AFM.

So, for example, if someone has "cleaned" the AFM with TB cleaner and screwed the optics up, chances are high the ECU will never set a code, but it will run like crap because the sensor is FUBAR.

That is why it would be helpful to measure the frequency of the Kv signal during operation. You will know immediately if its bad or you need to search elsewhere. Relying on codes is certainly good practice, but not enough to always figure out what's going on.

See this explains a lot. What would be the frequency level I would need in the multimeter? I'm guessing bigger then 50 hertz?
 
Oct 11, 2005
3,815
13
38
Thousand Oaks, CA
At idle its 30-60 Hz, and it can go up to several kHz at high load. For diagnostics, you want to see that it has a nice steady 50 Hz or so at idle and ramps up cleanly with engine load.

Also, following up on another post, some codes are "instant on" (codes 51 and 53) but most are two-cycle. Meaning they need to occur for two ignition cycles in a row to be set. So yea, if the ECU power is cut after each cycle you will not catch the majority of codes.

The TSRM lists available codes and their meaning.
http://www.cygnusx1.net/Supra/Library/TSRM/MK3/manual.aspx?S=FI&P=26
 
Oct 11, 2005
3,815
13
38
Thousand Oaks, CA
Yup. That's it. Probably easiest to access Ks from the ECU connector. There's a bunch of cheap meters (e.g. harbor freight) out there that read frequency. I don't know how good they are. My expensive one works well.