Direct Electronic Boost Control
Nov 14, 2006 | 03:48 PM
#1
Direct Electronic Boost Control
Boost control on our cars seems to me as one of the most challanging aspects of tuning - people go through endless troubleshooting to resolve issues like boost spikes, taper, BOV flutter, etc. I've only owned a turbo car for about 2 years now, but after doing a lot of reading I really have started to wonder why boost control is so complex?
I think the answer is (at least in part) a combination of two factors: (1) the OEM boost control system is designed with cost-savings at heart, and (2) after-market boost control elements (e.g., boost controllers, WGAs, BOVs, etc.) are all meant to compensate for the short-comings of the OEM system, but not replace it. My question is: why try to shore-up a design solution that was driven by economic considerations, and try to turn it into a performance solution? Given the cost of after-maket boost control hardware, why not wipe the slate clean and start with a design that is from the ground up meant to perform?
From the factory, boost is essetially controlled (as far as I understand it) with a pneumatic, analog system, paired with some ECU control for error correction in the analog system: the waste-gate and BOV are open/closed by pressure differentials, diaphragms, bleed lines with restrictor pills, etc., and the boost solenoid changes the flow rate in the system based on some electronic control.
MBC's remove error correction from the ECU and put it on the "driver-brain" instead (cold weather - > turn down the boost). EBC's move error correction from the ECU to a separate computer and solenoid, but the system essentially remains pneumatic/analog, only the error correction has changed. Why not do away with the pneumatic system altogether?
Here is what I am suggesting:
1) use an Electro Mechanical Actuator (EMA) to open/close the WG
2) use an electric solenoid valve to replace the BOV
3) use a dedicated electronics box (analog or digital) to control both
You'd need at a minimum two inputs: TPS and boost pressure
1) Trigger the "BOV" based on TPS rate, if the throttle is slammed shut fast, open the BOV. If it is opening or closing slowly, leave it closed. Say good bye to leaking at high boost, poor partial throttle response, fluttering etc.
2) trigger the WGA/EMA on TPS value, and boost pressure. Set a desired max boost pressure at WOT, and scale target boost at partial throttle with the TPS signal. If boost pressure is below target for the current TPS value, the WG is closed; if the target boost is reached, the WG opens. If there are problems with "overshoots" also monitor the derivative of the boost pressure to damp the response. No more boost taper, fastest spool up possible, nice linear throttle control (or any other curve you care to define).
Obviously, I have not tried this - but I would imagine this type of boost control system to be much more predictable and reliable than the OEM pneumatic system, with or without its aftermarket electronic crutches. EMA's aren't that expensive these days (~$800 ?), and solenoid valve's are dirt cheap. The approach seems so "duhh" simple, I wonder why nobody has tried it (especially racing teams where cost isn't an issue)? What (stupid) mistake am I making in my thinking?
I found one article online that actually talks about this approach for an electronic BOV (http://www.autospeed.com/cms/A_2188/article.html), but I have not found anything that talks about EM WGAs.
I think the answer is (at least in part) a combination of two factors: (1) the OEM boost control system is designed with cost-savings at heart, and (2) after-market boost control elements (e.g., boost controllers, WGAs, BOVs, etc.) are all meant to compensate for the short-comings of the OEM system, but not replace it. My question is: why try to shore-up a design solution that was driven by economic considerations, and try to turn it into a performance solution? Given the cost of after-maket boost control hardware, why not wipe the slate clean and start with a design that is from the ground up meant to perform?
From the factory, boost is essetially controlled (as far as I understand it) with a pneumatic, analog system, paired with some ECU control for error correction in the analog system: the waste-gate and BOV are open/closed by pressure differentials, diaphragms, bleed lines with restrictor pills, etc., and the boost solenoid changes the flow rate in the system based on some electronic control.
MBC's remove error correction from the ECU and put it on the "driver-brain" instead (cold weather - > turn down the boost). EBC's move error correction from the ECU to a separate computer and solenoid, but the system essentially remains pneumatic/analog, only the error correction has changed. Why not do away with the pneumatic system altogether?
Here is what I am suggesting:
1) use an Electro Mechanical Actuator (EMA) to open/close the WG
2) use an electric solenoid valve to replace the BOV
3) use a dedicated electronics box (analog or digital) to control both
You'd need at a minimum two inputs: TPS and boost pressure
1) Trigger the "BOV" based on TPS rate, if the throttle is slammed shut fast, open the BOV. If it is opening or closing slowly, leave it closed. Say good bye to leaking at high boost, poor partial throttle response, fluttering etc.
2) trigger the WGA/EMA on TPS value, and boost pressure. Set a desired max boost pressure at WOT, and scale target boost at partial throttle with the TPS signal. If boost pressure is below target for the current TPS value, the WG is closed; if the target boost is reached, the WG opens. If there are problems with "overshoots" also monitor the derivative of the boost pressure to damp the response. No more boost taper, fastest spool up possible, nice linear throttle control (or any other curve you care to define).
Obviously, I have not tried this - but I would imagine this type of boost control system to be much more predictable and reliable than the OEM pneumatic system, with or without its aftermarket electronic crutches. EMA's aren't that expensive these days (~$800 ?), and solenoid valve's are dirt cheap. The approach seems so "duhh" simple, I wonder why nobody has tried it (especially racing teams where cost isn't an issue)? What (stupid) mistake am I making in my thinking?
I found one article online that actually talks about this approach for an electronic BOV (http://www.autospeed.com/cms/A_2188/article.html), but I have not found anything that talks about EM WGAs.
Nov 14, 2006 | 09:06 PM
#2
Hmm...amazing concept. Why would you need a $800 EMA though? What about a servo? or linear actuator?
Please go into more detail about the WGA scale/map you described.
What if you did use a very fast acting EMA (or other actuator) and simply used peak desired boost. If you desire 22psi, then maybe at 21.5 psi the actuator slams open. Can the WGA move that fast without being damaged?
If you are using electromechanical actuator, then you could use a really strong spring and compensate with a really strong actuator. then at 21.5psi *blam* WG open
Maybe Im way off base...Im not an engineer.
Please go into more detail about the WGA scale/map you described.
What if you did use a very fast acting EMA (or other actuator) and simply used peak desired boost. If you desire 22psi, then maybe at 21.5 psi the actuator slams open. Can the WGA move that fast without being damaged?
If you are using electromechanical actuator, then you could use a really strong spring and compensate with a really strong actuator. then at 21.5psi *blam* WG open
Maybe Im way off base...Im not an engineer.
Nov 14, 2006 | 09:11 PM
#3
ok wait... you would need to constantly monitor boost pressure. You can't just slam the WG open because you would need to know exactly how far to open it to simply stop/hold the boost @ 22psi.
My theory above would have dumped all the boost out the exhaust...no good... The "black box" would need to be a microprocessor and the EMA or Linear actuator would need an encoder on it.
This just got complicated and Im really tired. Someone with a engineering degree chime in here. I'm really interested in this.
My theory above would have dumped all the boost out the exhaust...no good... The "black box" would need to be a microprocessor and the EMA or Linear actuator would need an encoder on it.
This just got complicated and Im really tired. Someone with a engineering degree chime in here. I'm really interested in this.
Last edited by RogueSTi; Nov 14, 2006 at 09:13 PM.
Nov 17, 2006 | 01:27 PM
#4
did somebody call for an engineer 
? On the boost control algorithm - you want something that scales boost with throttle position, otherwise it becomes really hard to control the car. The ECU controls boost baed on "load", which is mostly TPS but also some other stuff. I guess the proposed boost controller could try to read "load" from the ECU, but that may not be fast enough, so I suggested just using the TPS.
? On the boost control algorithm - you want something that scales boost with throttle position, otherwise it becomes really hard to control the car. The ECU controls boost baed on "load", which is mostly TPS but also some other stuff. I guess the proposed boost controller could try to read "load" from the ECU, but that may not be fast enough, so I suggested just using the TPS.
