Wastegate mod.....

[avos at cochlear.com.au]

     Jay,
     
     I have a friend who has a HKS Electronic Valve controller. His car is 
     a 4 cylinder turbo running about 300 hp at around 15 psi. The HKS EVC 
     is a great little unit, extremely accurate boost levels, and very good 
     at reducing lag, however its main aim in life is to hold the boost at 
     the static set level regardless of the engine load etc. This can be 
     extremely bad if you are trying to balance a car on the throttle 
     through a corner, as boost comes on very quickly, and your reaction is 
     too back off, which reduces boost extremely quickly, and then there is 
     a lag time before the boost comes on again really quickly, however 
     with a bit of practice you learn to do fine adjustments of the 
     throttle to avoid large fluctuations in boost, but still not ideal. I 
     agree that it is good to provide a dynamic set point for the boost 
     level, but what would it be mapped against? I have been thinking of 
     incorporating a 2D table that contains set points as a funtion of revs 
     and throttle position. This has the advantage that not only can you 
     adjust for throttle position as you have suggested, but you can also 
     control with revs, which may be good if you are having problems with 
     lean out at high revs/boost, as you can reduce the boost at high revs, 
     and still have good mid-range punch. If anyone who knows is reading, 
     what is the most common signal for a throttle position sensor output? 
     I assume that I will have to read revs from the ignition pulses.
     
     BTW- I don't suppose you want to swap cars?
     
     Regards,
     
     Adrian


______________________________ Reply Separator _________________________________
Subject: Re: Re[4]: Wastegate mod.....
Author:  diy_efi at coulomb.eng.ohio-state.edu at INTERNET
Date:    17/10/97 1:35


At 08:56 AM 10/14/97 EST, you wrote: 
>     Simon, Robert,
>     
>     In my experience, Electronic boost controllers (EBCs) can considerably 
>     improve the boost curve of a car, but my experience revolves around 
>     cars do not have the wastgate directly connected to boost pressure, 
>     but instead have a solenoid in the line controlled by the ECU. Many 
>     Japanese cars have this setup. The boost is controlled by the ECU 
>     outputting a PWM waveform to the solenoid. Most ECUs that have this 
>     setup do not have closed loop control, but instead calculate the pulse 
>     width from the revs and throttle position. This setup means that if 
>     exhaust/induction mods are made, the boost curve can change due to the 
>     control system being open-loop. Also, atmospheric conditions can alter 
>     the boost curve. Also most manufacturers will attempt to make the 
>     power curve of the car come on smoothly for a 'refined feel', which is 
>     not always optimal for the performance enthusiast who wants as much 
>     power as possible at any operating point. The EBC improves the bottom 
>     end boost of such cars, and also provides extremely accurate closed 
>     loop control of an adjustable boost level. I admit that many cars with 
>     an EBC can feel very peaky as torque can come on extremely quickly, 
>     but this feels great, and as long as you stay in boost revs, it is 
>     quite responsive.
>     
Snip...
     
I have been following this thread with considerable interest. I 
have a 93 rx-7 twin turbo that is a wonderful car but, at least in 
comparison to a friend's mr-2 supercharged, has less than ideal 
throttle response. This is particularly unnerving in corners when 
boost suddenly comes on, especially in the wet where traction is 
limited. One possible solution would be for a PID boost controller 
to incorporate the throttle position as the boost setpoint. For 
example, zero boost (atmospheric pressure) would be allowed at one 
third  throttle, with increasing boost levels allowed with 
increasing throttle position, up to maximum boost at full throttle. 
In this way some of the peakyness might be taken away, giving a 
more linear throttle response. This is clearly a complication to 
the PID controller circuit / code, but you might want to consider 
allowing for such a variable setpoint in the initial planning 
stages. What do you think: does this idea have any merit?
     
Jay