Classic FAST notes

TomShea

Minnesota Grand National & T-Type Association
Joined
May 10, 2002
I am starting to refamiliarize myself with my Classic FAST sequential box. I thought I would post a few observations. It has been quite a while (Many Years) since I acquired this unit and I have not taken the proper time to tune my car. It mostly sits while I allow myself to be distracted by other seemingly more pressing issues. It was and still is, always something else that gets the time and attention instead.
In the ensuing years between the initial install of this box, I have had the engine out & back in a couple of times and made some changes. This was due to to some problems that I was trying to tune around, in the previous set up. This resulted in a very improper tune for the engine set up that is in there now. This time is like starting over for me.
I have managed to spare a few moments in the past few days to tinker with it a bit. I have completely forgotten some things, and others are vaugely coming back to me through the mists of time.
Since this is a sequential box I will have to assume that there may be differences between mine and others. I have heard that the physical box itself was not quite capable of holding all of the programming that was available. Kind of like 10 pounds in a five pound bag.
I have heard that some of these classic FAST boxes with sequential were not programmed with the knock retard feature. My knock retard seems to be working, in response to both in shop and road testing. The knock detector (alerter) seems to be way more sensitive than the FAST box. By this I mean a separate aftermarket box or gauge. I assume the ESC retard, in response to the actual knock signal may be buffered, in addition to the lag time for the operation of the unit itself.
I do have a cut out wired in to the circuit between the sensor and the ESC sender. I tried to cut out the signal between the FAST box and the esc when I first installed a (Bank to Bank) FAST box to replace my old gen 6 DFI which was batch fire. The bank box required the esc to be connected to run at all. I don't know if this sequential has that same requirement. This is the reason the cut out is in place bewteen the actual sensor and the esc module, instead of the esc and the ecm. The ecm still recieves the esc circuit signal but can be disabled without disrupting the operation of the ecu. The cut out is controlled by a rocker switch on my dash. I like having the option to easily disable it at will.
I noticed that the (Target A/F) does not show up as a trace line or in the dash (shows 0) on the VE table in a data log, but the 3d target A/F screen can be used with a trace line. This is not untenable but does cause a little more difficulty when going over a log.
I noticed the usual lag time between the (Actual A/F) and the correction. I assume this is quicker with the new XFI box. I have heard that the XFI boxes are several times quicker over all. I have observed XFI data logs, but there is no practical way (for me) to compare the boxes head to head.
I have attempted to increase the response speed of my correction by using the Gain control. I have increased my gain setting from 20 to 30 for now. I think the default for the gain may be 20.
I have increased my minimum step from 9 to the maximum allowed, 10. I think the default might be 9.
I will try some more tuning and try to get the base fuel table (VE) to more closely match the target table before forming an opinion on where the gain should be set. It will likely be a compromise between too slow and overshoot.
My engine and injectors are probably appropriately size matched, so I don't think the gain needs to be very far from default. My engine is 274 CID and my injectors are 84s.
I'll post follow up notes when I have some more time to tinker with the tune.
I hope this will be useful to somebody for reference. If there are obvious mistakes or inaccuracies with my method or data, please don't hesitate to comment. I can use all the help I can get.
 
I am starting to refamiliarize myself with my Classic FAST sequential box. I thought I would post a few observations. It has been quite a while (Many Years) since I acquired this unit and I have not taken the proper time to tune my car. It mostly sits while I allow myself to be distracted by other seemingly more pressing issues. It was and still is, always something else that gets the time and attention instead.
In the ensuing years between the initial install of this box, I have had the engine out & back in a couple of times and made some changes. This was due to to some problems that I was trying to tune around, in the previous set up. This resulted in a very improper tune for the engine set up that is in there now. This time is like starting over for me.
I have managed to spare a few moments in the past few days to tinker with it a bit. I have completely forgotten some things, and others are vaugely coming back to me through the mists of time.
Since this is a sequential box I will have to assume that there may be differences between mine and others. I have heard that the physical box itself was not quite capable of holding all of the programming that was available. Kind of like 10 pounds in a five pound bag.
I have heard that some of these classic FAST boxes with sequential were not programmed with the knock retard feature. My knock retard seems to be working, in response to both in shop and road testing. The knock detector (alerter) seems to be way more sensitive than the FAST box. By this I mean a separate aftermarket box or gauge. I assume the ESC retard, in response to the actual knock signal may be buffered, in addition to the lag time for the operation of the unit itself.
I do have a cut out wired in to the circuit between the sensor and the ESC sender. I tried to cut out the signal between the FAST box and the esc when I first installed a (Bank to Bank) FAST box to replace my old gen 6 DFI which was batch fire. The bank box required the esc to be connected to run at all. I don't know if this sequential has that same requirement. This is the reason the cut out is in place bewteen the actual sensor and the esc module, instead of the esc and the ecm. The ecm still recieves the esc circuit signal but can be disabled without disrupting the operation of the ecu. The cut out is controlled by a rocker switch on my dash. I like having the option to easily disable it at will.
I noticed that the (Target A/F) does not show up as a trace line or in the dash (shows 0) on the VE table in a data log, but the 3d target A/F screen can be used with a trace line. This is not untenable but does cause a little more difficulty when going over a log.
I noticed the usual lag time between the (Actual A/F) and the correction. I assume this is quicker with the new XFI box. I have heard that the XFI boxes are several times quicker over all. I have observed XFI data logs, but there is no practical way (for me) to compare the boxes head to head.
I have attempted to increase the response speed of my correction by using the Gain control. I have increased my gain setting from 20 to 30 for now. I think the default for the gain may be 20.
I have increased my minimum step from 9 to the maximum allowed, 10. I think the default might be 9.
I will try some more tuning and try to get the base fuel table (VE) to more closely match the target table before forming an opinion on where the gain should be set. It will likely be a compromise between too slow and overshoot.
My engine and injectors are probably appropriately size matched, so I don't think the gain needs to be very far from default. My engine is 274 CID and my injectors are 84s.
I'll post follow up notes when I have some more time to tinker with the tune.
I hope this will be useful to somebody for reference. If there are obvious mistakes or inaccuracies with my method or data, please don't hesitate to comment. I can use all the help I can get.
I have a sequential box and I'm a little nervous about a crude spark retard set up. I guess I'll have to start tapping with my hammer to see if the system reacts.I will be beginning to install my fast to see if I can get it to work.
 
When shop testing for knock, you must set the max TPS for idle to lower than what you TPS actually is at idle. The KR will not work at idle. This setting determines whether or not the idle functions work, so don't forget to return it to a proper setting. A proper setting for this is 1 or 2 points above your max TPS at idle. There is a wire on the ESC box on the right inner fender that can be rapidly disconnected and reconnected to simulate the knock signal. I don't use this first method, although I have tried / tested it and it works. Thanks to John Spina for that info.
I just use a long 1/2 inch socket extension bar and tap near the sensor / sender, near the back of the intake manifold.
You may probably see and or hear your knock detector / alerter going off during this process.
I just set the log to run for 10 to 20 seconds and make sure the esc settings are correct. The ESC in mine was set for 5 on the attack rate. It will retard to about 3.5 with that setting. I now have it set fo 10 on the attack rate and have left the decay at 1.5. With 10 on the attack rate the KR goes to about 7. The spark BTDC should correspondingly drop. Mine does not exactly match the 10 set in the attack rate but it doescome close to the 7 shown in the ESC retard. The retard will show on the very bottom line of the screen. It's difficult to see unless there is some retard shown. The trace line is pretty much exactly overlayed on the bottom line of the window, except for the small spikes up like triangles. These should slightly preceed the resulting down spikes of the spark BTDC trace line.
I have observed that my detector/alerter seems to go of alot prior to any KR occurring. This make me a little nervous.
I have been told by a few that I respect, that there are plenty of things that will set the detector off falsely. This false signal will result in the same reaction from the ECU, KR! The ecu does not have the ability to discern the difference between real and fasle knock signals. That's why I have the option of disableing mine.
I suggest you assume the detector is reporting actual knock unless you are very sure it's not, and are willing to risk the consequences.
Some of the possible culprits for a spurious/false knock signal could be Forged pistons, different block with slightly different metal composition or casting density, roller lifters, high lift cams, solid or mechanical lifters, rollerized valve train components like rockers, double roller timing chain (these will be used without the tensioner and come with a metal gear instead of the stock nylon toothed one), exhaust pipes rattling against other components like the frame even the transmission shifting hard or tires slipping, and lastly even bumps in the road.
 
Top