First, I'll agree that clamping to open loop at the first sign of boost will hurt low boost performance. I just don't like that idea at all. Initially the 2 stage approach sounds like it would work well. It looks like the MSD Programmable DIS-2 could handle the 2 stage setup your talking about. It has a Multi-Step retard feature:

<MSD_QUOTE>
Multi-Step Retard: Three separate stages of retard can be activated by either rpm or a separate activation wire. This is perfect for all nitrous applications. Programmable from 0°-15° in 0.1 degree increments within 800 - 15,000 rpm in 100 rpm steps.
</MSD_QUOTE>

A switch could be fashoned to activate the appropriate timing maps. You could have a map for open loop and a map for closed loop. This is where the conversation goes back to my idea...but I'll save my comments on this "switch" for the end of this post(bear with me here).

It also looks like the programmable DIS-2 would handle the different loads(like you were talking about earlier in the thread) that the different gears see with the Gear RPM and Retard feature:

<MSD_QUOTE>
Gear RPM and Retards: Program a different rpm shift point for each gear as well as a retard for every gear change.
</MSD_QUOTE>

Actually just reading about it, it looks like a great ignition. You can program the boost timing curve. It can handle up to 45psi. You can adjust individual cylinder timinig. It doesn't support knock sensative timing retard. I know jamesman has one. I think a couple of people have it. It's $420 at Summit.

Ok now back to the switch.

Lets call it a loop controller instead of MetalJim's idea...well we could call it that too, but I'll call it a loop controller. I've thought about it a bit more.

It still works exactly like I said before, clamping the TPS to 85% when it hits a particular boost threshold, forcing us into open loop until the actual throttle position goes above our clamp voltage(85%), where we would relinquish TPS voltage control back to the TPS sensor. This would give us the open loop's more "boost friendly" timing curve. It could be set to clamp at any particular boost level - I'll call that the boost threshold. The boost threshold on the controller would have to be adjustable. Maybe you would want to set it to 6. What if you're running a higher compression, like 11.5:1 or 12.5:1. With that kind of compression, you would definately want to pop into open loop at a lower boost, maybe like 1-3 pounds, or maybe the first sign of boost. I think this would be perfect for stock, DIS-2, Crane, or Safegaurd ignitions. It gives you "boost friendly" timing at a particular boost level. You can use the stock open/closed timing curves up until your loop controller's boost threshold, then it switches over to open loop(lets say 6 pounds in your case). You can take advantage of the safer open loop curve until you need to start to retard timing(lets say 8 pounds in your case). Then you can use an aftermarket ignition to retard the boost from there. AT THE VERY LEAST, you'll have safe timing at any throttle position AND if you use a ignition retarding unit - you'll always be retarding the open loop curve because the retard starts well above the loop controller's boost threshold. Its more like a safety device for situations like the one that you ran into where you went into closed loop at 10psi.

Well what about staging open and closed timing curves with something like the Programmable DIS-2 unit? Lets take your switch based solely on TPS. At what voltage do you switch over from open to closed loop? The switch would have to calibrated to be very precise. It would have to be at the exact moment of the switchover, otherwise its just useless. It's not like theres a wire on the ecu that tells us weather or not we're in open loop...or is there? I think the only way to ensure that the switch will give a correct open/closed loop signal is if it controls weather or not the car is in open or closed loop. Thats fine. My loop controller(as it stands) doesn't exactly work for that situation either. But lets make a more complicated loop controller, one that can tell us without a doubt that we're in open loop. This one is going to have to jump to open loop when we reach a particular throttle angle, lets say 75%. When we hit 75%, we just automatically clamp the TPS to 85%, throwing us into open loop. Then when the throttle angle reaches our clamping level we relinquish control to the TPS sensor. Also, as you back off below 85%, you have to keep the controller's TPS signal clamped to 85% till the sensor goes below 75% then you can relinquish control back to the TPS sensor. So the controller works exactly the same way, except that it will also put us in open loop right before the stock computer would normally do it. You'll still have a boost threshold and anything above that threshold will throw you into open loop. The loop controller would live up to it's name. It would control every aspect of switching between open and closed loop operation. At that point, you could have a wire running out of it that could tell something like the programmable DIS-2 unit weather we are in open or closed loop WITH AUTHORITY. I'm not sure how practical or pragmatic it would be to have this much feedback and control over the ECU's loop state. I think that with the simple loop controller(boost threshold activated only), you could get away with using a much simpler retard unit without having to worry about what state the ECU is in. I guess my argument is that you could use the programmable DIS-2 with the simple loop controller(the one described in the previous paragraph) and you wouldn't need to have different curves for open and closed loop operation. Just set the boost threshold right below your critical closed loop pressure(6 pounds in your case) and then pull timing at your critical open loop pressure(8 pounds in your case).

Ok, it sounds like I just convinced myself out of the usefullness of a more complicated loop controller. I think the simple loop controller is all anyone would need. I'm gointg to try to make my questions a little more specific this time. From some of the comments, it sounds like the loop controller shouldn't activate that much. Mostly I'm concerned with our computer's "learning" abilities. What if the boost threshold was lower(like for my high compression examples) and it does activate often? Does the computer learn timing curves like it does with fuel trim? Would there be any drawbacks(while using a loop controller) with the amount of time the computer is "learning"?


ride__________95 Eclipse RS : 5 speed
power________Gude cams : AFX UDP : ram air : test pipe : Thrush glasspack
suspention___BOMZ F/R upper strut bars : ES inserts
sound________Kenwood KDC 5000 : Audiovox 8" sub toob - no sub amp(damnit!)
shizzle_______AutoCommand remote start : dual air horns
incomming___Hurricane F/R lower strut ties : Jeep TB : Field SFC Hyper-R : GST catback

95 Eclipse RS : 5 speed
14.76 @ 94.72mph
Jeep TB writeup - http://www.dimensia.com:81/jimbo/JeepTBfor2gnt.html