Syvecs Forums

I'd like to control a secondary throttle based on RPM and Boost level...... the main throttle is cable operated, can I use the drive by wire in-build option?

What are you wanting to do with this chap so i can see if the DBW strategy could run it?

I want to control a sort of throttle body in the turbo inlet, do you remeber the F1 setup in the 80's ?

Yes. Will have to have a think on that one.

Any ideas Pat?

I'm intrigued, what effect are you looking for by throttling the compressor intake?

Well , the idea is to have a dynamic intake restrictor. When on light load the throttle is closed (NEVER completly) and the compressor can spin faster allowing high boost at lower rpm. As the rpm and boost increase you open the throttle.

Not sure how brutal it can be on a "standard" sport (ish) car, but it was working very well on the F1s !!!

Cool, it will certainly allow you to manage the way a marginal compressor behaves when very close to the surge threshold.

It will be interesting to see if you'd run into output temp issues as a result of the increased pressure ratio that the compressor could potentially see?

The DBW strategy will probably not work because it tries to control the tps1 and tps2 parameters, which themselves are used to control things like acceleration enrichment with no means of switching those corrections over to ppsA/ppsB... you'de essentially want to put the ppsA/B sensors on the real throttle and then use the DBW to control the DPV throttle, however since the DPV curve won't match the real curve most of the computations will be on the wrong parameters....

Furthermore, DPV control needs to take into account other dynamic factors that don't necessarily manifest themselves on a manifold or port throttle, therefore a proper DPV strategy would need to be more complex than the DBW strategy and would almost certainly need hooks into the wastegate control strategy etc. Whilst it may seem relatively trivial to allow different controlled parameters for the DBW (eg introduce a dpv1A/B, dpv2A/B inputs and allow those to be the controlled parameters rather than tps1A/B, tps2A/B) this option hasn't been explored despite DPV control being on my "todo list" for quite some time now. Anything that is worth doing is worth doing right, no point in expending engineering time implementing a suboptimal solution when you could do it right in the first place

There are other interesting turbo control things that are also on the "todo list" such as VIGV control, but again deciding on how best to implement that is far from trivial, and there are still many more things that are higher priority because they will be more useful to the majority of users. That's not to say that they won't happen, just that it may take some time.

For now the DBW strategy seems the most promising but you will eat up inputs....

Take the wiper from the real TPS and connect it to FOUR inputs... ppsA, ppsB, tps1A and tps1B. Make sure that ALL cylinders are assigned to BANK1. Now connect the DPV sensor to another two inputs, tps2A and tps2B. Connect the DPV actuator to the dbw2 output. This way, ppsA,ppsB,tps1A and tps1B will naturally track, whilst dbw2 will try to track the dbw demand. If the pps->tps curve is nonlinear then the positioning system will detect an error on bank 1 and may shut down the DBW system, so a larger error window may have to be permitted. This is by no means an elegant solution, but it may get part way to achieving the desired effect, and at least there will be use useful feedback which may prove invaluable in determining what additional controls / compensations a true DPV control strategy requires.

Hope this helps,

Pat.