3P's TCCS Disassembly/Analysis
- Thread starter 3p141592654
- Start date
First off I'd like to say good work in here and count me in for a daughter card when they are available. Get that GUI working!!!
What does the Vf do during a FPU starting scenario? And is there anyway to tell if the tccs is in FPU hot start mode, without your daughter card, other then checking the sig wire out of the tccs? I've been getting some occasional super lean hot starts, stumbley and stalling and I'm beginning to think the the FPU vsv is faulty because it goes away with some pedal and time.
What does the Vf do during a FPU starting scenario? And is there anyway to tell if the tccs is in FPU hot start mode, without your daughter card, other then checking the sig wire out of the tccs? I've been getting some occasional super lean hot starts, stumbley and stalling and I'm beginning to think the the FPU vsv is faulty because it goes away with some pedal and time.
There is no way to tell if its active except by observing the FPU signal.
Vf will behave like it normally does. It will be 0V when open-loop, then switch to reflecting the status of the fuel trim variables once the ecu switches into closed-loop. Since the engine will be running rich, it will sink to 1.25 and then 0V and then FPU will be terminated. VF is actually the sum of the short term and long term fuel trims, plus a small correction term that is a function of air flow.
Vf will behave like it normally does. It will be 0V when open-loop, then switch to reflecting the status of the fuel trim variables once the ecu switches into closed-loop. Since the engine will be running rich, it will sink to 1.25 and then 0V and then FPU will be terminated. VF is actually the sum of the short term and long term fuel trims, plus a small correction term that is a function of air flow.
I only ask because during cold start the Vf is set at 0v during the cold start time. As soon as I see Vf spike from 0v to happy, ~2.00v for me, AFR's slowly lean out to 14.7 from ~12.5. I thought Vf might react the same during FPU hot starts but I can't get the symptoms to repeat so I can see what my Vf is doing.
What symptoms?
Hot starts can be weird. Here's a datalog of one that behaves the way I would expect. This is a restart after a few minutes from a hot shutdown. Note that it essentially in closed loop right after startup and cross counts quickly. In the video VF Trim is a high resolution version of the diagnostic VF signal (100% = 2.5V). Short Term Fuel Trim shows how the ECU is managing fuel to get cross counts at the desired rate.
[video=youtube;sXZJ9JVT3bY]http://www.youtube.com/watch?v=sXZJ9JVT3bY[/video]
Next is a heat-soaked warm start, about 30 minutes after shutdown. It's not hot enough right now to get the FPU to engage, but this startup was followed by a few seconds of rough low idle with the ECU open-loop (VF trim = 100%). Note after the ECU decides conditions are right for closed loop operation it starts adding fuel to force the OX sensor to go rich, but it won't go and fuel trim hits the rail at 180%. The only way this will start cross counting is to put it in gear and go driving. I'm not to sure what is going on here, but you would see VF at 2.5V even though STFT is maxed out because VF trim only get updated on the lean to rich transition.
[video=youtube;n5pi09c7LMA]http://www.youtube.com/watch?v=n5pi09c7LMA[/video]
Hot starts can be weird. Here's a datalog of one that behaves the way I would expect. This is a restart after a few minutes from a hot shutdown. Note that it essentially in closed loop right after startup and cross counts quickly. In the video VF Trim is a high resolution version of the diagnostic VF signal (100% = 2.5V). Short Term Fuel Trim shows how the ECU is managing fuel to get cross counts at the desired rate.
[video=youtube;sXZJ9JVT3bY]http://www.youtube.com/watch?v=sXZJ9JVT3bY[/video]
Next is a heat-soaked warm start, about 30 minutes after shutdown. It's not hot enough right now to get the FPU to engage, but this startup was followed by a few seconds of rough low idle with the ECU open-loop (VF trim = 100%). Note after the ECU decides conditions are right for closed loop operation it starts adding fuel to force the OX sensor to go rich, but it won't go and fuel trim hits the rail at 180%. The only way this will start cross counting is to put it in gear and go driving. I'm not to sure what is going on here, but you would see VF at 2.5V even though STFT is maxed out because VF trim only get updated on the lean to rich transition.
[video=youtube;n5pi09c7LMA]http://www.youtube.com/watch?v=n5pi09c7LMA[/video]
Nice progress guys!
I recently got my hands on a JDM 1G-GTE ECU wich features a blackbox and 2 dials for remote setting for (I guess) fuel injection duration & ignition advance. (possible "BuddyClub Brackbox"?) I've opened it up but the mystery components were all covered in white goo. I can take some pics if anyone is interrested. Here's one of the blackbox & controller:
I recently got my hands on a JDM 1G-GTE ECU wich features a blackbox and 2 dials for remote setting for (I guess) fuel injection duration & ignition advance. (possible "BuddyClub Brackbox"?) I've opened it up but the mystery components were all covered in white goo. I can take some pics if anyone is interrested. Here's one of the blackbox & controller:
Last edited:
Looks like one of the factory TRD ECUs used in the early 90s by various racing teams. Apparently Toyota did not allow any of the TTE teams access to the ECU code until after 92, according to LordDigital, who pointed this out to me a while back.
Here's a similar looking Group A 7M ECU.
http://www.supraforums.com/forum/sh...rs-and-head.&p=3826299&viewfull=1#post3826299
Here's a similar looking Group A 7M ECU.
http://www.supraforums.com/forum/sh...rs-and-head.&p=3826299&viewfull=1#post3826299
So, just how advanced do you think you can go with this? Program for 550, 680, 720, etc etc injectors? Fuel cut a chosen point?
Last edited:
"Competition Parts" control boxes were also used on 3s ECUs (link) and 2JZ ECUs (link). It was never clear to me what the potentimeter control was for... most of these ECUs came from Japan junk yards and the potentimeter modules were missing ,but I highly doubt that Timing was one of the controlled paramaters.
ah nice! It seems this is something sililar to the techtom boards.
I'm recieving an 4A-GE buddyclub ECU with exactly duch a blackbox and flat calbe as well.
Is seems to me there is one manufacturer for various tuning companies on this one, just like with the Techtom boards.
I'm recieving an 4A-GE buddyclub ECU with exactly duch a blackbox and flat calbe as well.
Is seems to me there is one manufacturer for various tuning companies on this one, just like with the Techtom boards.
Nick M;1832567 said:
I've been delaying responding to this, because it really needs to have a thorough description of the fuel calculations to make any sense. As it stands, the factory program converts the AFM frequency signal to a load value, after factoring in the HAC input. The load is a 16 bit number that currently triggers fuel cut when it gets to C800h, which is 78% of the maximum value.
That means that we can move fuel cut higher by only 22% before we hit a mathematical limit. Obviously that is not going to make anyone happy since even a switch to 550 injectors is a 25% increase.
The solution is therefore to scale the load value and keep fuel cut at 78%. This is basically what the Lexus AFM mod does, and the problem is that the ignition and fuel enrichment tables, along with a whole lot of other tables that use load as an index are then being scaled inappropriately, and you get too much ignition advance and other problems as we all well know.
So, the strategy I am working on involves allowing the AFM input to be scaled, and then recomputing all critical tables using load to correct for the scaling so that things remain the same for load ranges that match the factory region, and then allow customization for loads greater than that. What is not addressed by this is that there is a maximum flow for the AFM above which it becomes inaccurate, and so we cannot just crank up the boost and hope the AFM will manage to handle the extra air. I am not sure how to address that as the Lexus AFM is the largest one out there that I am aware of.
At the other end, we can scale injectors pretty easily, since the final computed injector duration goes through a scaling routine that is presently set to 1, and that can be set to different values to handle different size injectors.
By the way, Techtom code disables fuel cut, allowing load to be as high as >100%. That's not really a solution at all, as fuel cut is there to protect the engine and that protection is gone with their approach. Since the tables all stop before 78%, the ECU just stops scaling after 78%and uses whatever max value is in the table!
I have a 7M computer just sitting here. Let me know if you would be interested in it. It says RCTS on it with a daughter board.
3p141592654;1837858 said:
I have a complete orginal 1UZFE housing minus electronics you can have if you think it would help. The bypass is set without the adjustment. It is not a scaled up 7M meter. Well, it is, but the bypass is noticably different. What I really want is a larger mass air flow sensor of the analog variety like Hitachi made and others duplicated. The Lex meter works great for me with my power levels, and that will only change a little bit in the future.
3p141592654;1837858 said:
I'm working on a hardware approach to this. I figure that for less than $50 in electronics parts and a few hours soldering I can make a board to step up or down the frequency of the AFM signal by converting it to a DC volt signal, trimming the level up or down, and then converting back to an FM signal that is linearly proportionate in frequency to the original signal, less noisy, and equal in amplitude.
Would this be a good solution? I have a Lexus AFM. Initial tests suggest that the native lexus electronics do indeed put out a Higher air flow value than the supra sensor and electronics. I don't see any reason the lexus sensor should be more accurate at high flow rates, but maybe it would be. I just started building the frequency altering board to safe money on electronics as well as learn more and have fun doing so. Now you have me worried about throwing off other tables and functions, so I will have to check the latest version of your cracked ECU code and see what you mean.
For air flow accuracy, i was thinking of adding another channel for air to bypass the AFM, other than the set screw's channel. I already started with 660 injectors, thinking that most people put the set screw mostly all in, or buy a longer screw, to trim the AF ratio properly with the Lexus Mod. I figured the surface area of the bypass channel, as well as the overall surface area of the lexus vs supra AFM, and It seemed like 650cc to 700cc injectors would better compensate for the larger AFM with the bypass channel open.
To use a modified stock ECU, such as what you are designing, I'm thinking I can't go wrong by making the engine tune as close as it can be with low tech mods, and rely on tweaking the software as little as possible. Just for the reason you said: there is a limit to how much the software can alter parameters, and how much the hardware can respond to alterations in output from the ECU.
It sure would help to have a modified ECU to play with at this point. I figured you would be successful two years ago and planned my rebuild around that assumption. I won't nag you about how soon, you know we are all in awe and gratefully awaiting the fruits of your labors.
3p141592654;1837858 said:
But if you can tune the ignition table then it really doesn't really matter if you use any scaling, right?
Maybe just an option to disable fuel cut on the initial release and then work for an update with your scaling approach?
My feeling is that we have to scale the AFM input, and then fixup the tables. I don't see any other solution. For air flow beyond a single 1UZ meter, the solution is probably the MAF translator (not the 'pro' version) hooked to a GM 3" or 3.5" hot-wire MAF.