What does the boost do on wastegate only - zero boost solenoid duty? What type of solenoid, what rating of wastegate spring, are you using two ports on the wastegate? You may be compounding a few issues which are giving you oversensitive boost control.

Looking at the log above: It is at 20-30% DC and 34psi - that looks over sensitive and potentially a wastegate flow priority issue showing at higher rpm depending on the answers to the queries above. With such low DCs and resolution - you may struggle to get any good boost by TPS control as its appears super sensitive.

Agree with Tran above about mapping DC versus MAP versus RPM including zero DC for building accurate base boost solenoid DC tables. It may also highlight any issue with the plumbing setup and WG priority. Tran's plot shows a pretty linear relationship so most likely decent wastegate flow, backpressure possibly building at top end and higher boost so boost falling off slightly. All looks normal.

 

Yea the link is weird when you change units to psi. subtract 14.5psi from those numbers. It starts at 14.5psi as atmospheric less than 14.5 is vacuum ~100kpa. The target wgdc are correctly setup. 26.1 is essentially 12.6psi of boost. Setting 1 in the boost target table is just basically what the mechanical boost controller by itself puts out. It still creeps up to about 16-18 psi on the stock 6psi straight gate spring pressure only. which I was also kinda surprised by this as I thought I would be able to run lower boost than I could get out of my old internally Gated setup. I don't think I could get down to the 330 HP number I use to run for NASA TT2.

 

Thanks for the explanation. I think there's 2 issues here,

1. Boost by TPS ought to be such that 90% TPS always gives (or at least tries to give) less boost than 100% TPS
2. You are making 16+psi with spring only, which means the boost control has limited authority.

For point 2, I suspect there has to be something not right with the setup, since I'd at least have hoped that the straight gate could give a much bigger authority than a poppet type. Are you plumbed in like this?



In my opinion, work out what is going on with #2 and once that is sorted, try doing RPM sweeps at a locked throttle TPS, to work out what the authority of the boost controller is able to do at part throttle.

I haven't got my mapping laptop at the moment but have got the excel sheet I used for some of my calculations and it should give you an idea of what to go for. Running this EFR6758 0.64 with a max target of 200kPa (same as in the graph above), I also measured the boost threshold per gear and it looked like this.(X axis is RPM, Y axis is Gear and Z axis is the max achievable boost in a pull, forgive the formatting, it's from a calculation sheet). Being autocross, I guess first, second and third are the main ones you want to have mapped well, and you can see the response is very different....

 

The boost control is doing exactly what is programmed to do it hits boost target and stays there once its beyond the low rpm threshold where it takes noticeably and annoyingly more rpm than it did before and comes in like a switch. The issue is throttle control. It will hit boost target weather the throttle plate is a little over 40% open or 100% open makes little difference. entertaining yes. fast and controllable to drive not so much.Im sort of seeing that ~60% of throttle is the same result as 100% when the turbo spins up. Traction control is saving our *** 14% slip is where we are currently at. and it mostly can keep it there with ignition timing.

 

putting in a really nonlinear throttle curve helps allot to improve drivability in boost. makes it feel even more dead out of boost. I think the TPS boost target Trim table might help but I'm curios how hard its going to be to get stable with boost target moving rapidly and constantly. Still sort of shocked by what the turbo change did to the behavior of the car.

 

This is the point of TPS control. Target is a function of tps angle, so hitting boost target at a given throttle angle is what you get. Of course with a 16psi spring, you will never get less than 16psi, agnostic of target setting. If you want 16psi at 40% throttle, that is what the controller commands when the throttle is at 40%.

If the DC table above is accurate, and you actually have a 16psi spring, you are probably on a 4 port config, which is the hardest to control and really only should be used if you really need the dynamic range. A 3 port configuration is superior for applications where boost target is anything between about 1.3x spring and 2.25-2.5x spring. If you aren't using a very low solenoid frequency, you probably have no solenoid control at duty cycle less than ~25-30%. You would know if the frequency was that low on a 4 port, the boost would noticeably flutter. If link has an option for a tps vs rpm table for target, that is the way to go.

Extrapolation of tables in controllers like this (targets outside of the range covered by the table) is typically static at the boundary. It is advisable to ensure the table covers the entire operable range.

Edit: so looking again at the screen shot above, if 40 psi is your wot target, but you want the largest dynamic range with a 4 port, my recommendation would be to use a lighter spring 8psi will be a 5x gain with 4port or 10psi for 4x gain with 4 port. Extend your open loop table to at least cover spring pressure with 0% duty with a second row at around 1.4x spring also set for 0%duty. Tune the rest of the table over the desired range starting with a min duty of 25% in row 3. Set your tps modifier (wtf is that anyway?) To be -40 for everything below say about 60% tps then taper to 0 between 60% and 90% (0). Retune the open loop table. You can also set all non-zero rows in the open loop table to 100 below 4krpm (based on dyno above) for faster spool if link doesn't have a spool option. Finally, the dyno above has a max boost of something in the low 20s, and if this is really your taget then a 3 port with 10psi spring is the way to go, with all the listed changes above.

Edit (again): that wastegate looks effin awesome, wish that was around when I was building cars.

 

It's a 6 psi spring not a 16 psi spring. I'm using the stock 6 psi spring the lightest they make. The butterfly valve on the straight gate is zero offset so only the pressure in the actuator moves it and not the exhaust pressure on the valve. The tuning window for the straight gate is is listed as over 5X base spring pressure so it should control well up to 30psi. It seems easier to control than anything Ive worked with. The wastegate is Y'd into the collector and has a fairly straight shot connecting it to the down pipe so the flow should be good even better than most of the pop off valve style waste gates. It still makes like 16psi on the 6psi base spring though I confirmed on the dyno the valve opens all the way. EBC valve is a MAC three port plumbed the standard way the way in the image Tran posted. I think my %DC1 table works good or wont requre much of any change for the 4 dash selectable boost targets I have. 0) Default 20.3psi 1) mechanical lower limit 11.6psi 2 note: creeps up to over 16psi 2) low boost 17.4 psi, 3)kill mode 23.2psi. The TPS boost target trim table should just move up and down my %DC1 table as needed to match the variable target I think.

 

I understand the conflict between DBW profile and boost control - two separate issues which may also affect each other.

With a 6psi spring creeping to 16psi you simply don't have good mechanical boost control. You have control for your higher target, but you will struggle to have the resolution to turn it down to get an effective boost/TPS map as the system is in runaway on a 6psi spring. I think you have to solve that first - then review the boost by TPS curve as you may also have a larger throttle area and so also see full boost at lower tps %. I've not mapped DBW but know the cable actuation is not linear so there is playing with DBW maps to get a linear feeling throttle also.

Your straight gate is massive with massive flow potential, the entry to it is massive from the manifold, but something is not working in the geometry/mechanical setup. Id expect that gate to be large enough to control your boost down to 3psi flat out on a 3psi spring if it had the priority of flow.

As an example I run a Mitsi evo TD05 turbo on my BP. On the earlier twin penny housing I could control boost down to 3psi flat out if wanted. I'm now on a later exhaust housing which is single penny with less wastegate cross sectional area. I'm not able to control boost as effectively, have some minor creep at top end - using a 12 psi spring goes to 14psi at top end on 0 DC, 14psi spring goes to 16psi on 0DC - but I still have a very wide range of control resolution as it is just creeping. With this I can still shut down boost at lower TPS to have effective boost by TPS.

With your setup I'd probably do a test like run it flat out with the straight gate held fully open with no spring installed to see what the mechanical setup can ultimately control down to. Can do it on the dyno with a loaded ramp run, or on the road in a longer gear to give it decent load. I'm not sure on the exact figure of an issue but if you creeped up to say 5psi or above at top end then I'd say you don't have mechanical control. Looking at your results it seems you may hit c.10psi in this config alone.

 

Ahh, ok that clears a ton of stuff up.

So to be clear, if you eliminate the MAC solenoid all together and plumb the wastegate directly to the outlet side of the intercooler, the turbo will creep from 6psi to 16-18 psi?

 

My old GT2971R with the wastegate ported as big as I could and still have the flapper door cover the hole and actuator removed and safety wired full open would creep up to ~14psi. The G25-550 once above the rpm threshold is way more boost happy than that turbo like I said it will boost through partially closed throttle like its not there. I haven't done the tests but I watched it on the dyno with mechanical boost control only and the wast gate was hitting full open you can watch the butterfly move. It is plumbed to the boost port on the compressor housing not after the inter cooler. I do think my manifold design does like to spin turbos. It's exactly equal length smaller diameter 1-1/4" schedule40 to maintaining pulse velocity with a really smooth collector into the turbine.

 

The throttle angle isn't really a determining factor for modern turbos. You probably should do a pull with the wastegate wired open and another with wastegate plumbed directly to turbo outlet with no solenoid if you haven't already done so.

 

Agreed.

It's a huge PITA to modify the manifold to find a better balance of flow priority between wastegate and turbine inlet. But it may be required to achieve the mechanical control to allow effective boost by TPS. Your turbine appears so hair trigger to boost that you could probably balance the priority way in favour of the wastegate and still have little performance loss.

Another other approach is to choke the engine with the DBW mapping profile. I've only seen videos of people doing this to attain lower boost than the mechanical minimum. Sounds like yours would have to be a strongly modified curve as it takes so little throttle angle to achieve full boost. Again like Tran suggested - you could do flat out runs in a longer gear like 4th with 10% TPS, 20, 30, 40 etc. to see what the boost by TPS profiles look like. You will get into a position that you may have one good TPS curve for one specific boost solenoid map. If you try to have different boost maps the inter-relationships of DBW map and Boost solenoid map may complicate simple changes and not give you simple, consistent changes.

I'd focus on achieving good mechanical boost control first. The overall control strategy to run different boost levels etc. should then be the easiest.

 

So are you saying that the straight gate 50 on your manifold can't flow enough to prevent boost creep? Hard to tell without seeing the inside, but it looks like exhaust gas from cylinders 1 and 2 would find it hard to make it to the wastegate. If this is the case, the proper fix is to modify the wastegate flow priority such that you can reduce the boost on spring only, but you may be able to achieve a similar effect by having the throttle close instead, though note this isn't ideal from an IMAP vs EMAP perspective, since you've effectively lowered the boost pressure by making the turbine work against a restriction, but this might be the only way to achieve the target torque without changing hardware, which for autocross is the key factor.

FYI, the plot above of my boost duty sweeps were with the IWG housing on a 0.64 T25 EFR 6758 with the kraken low mount and a 7psi single port turbosmart actuator which should flow way less than the straight gate but was still able to keep close to spring.

 

Everybody keeps talking like I have a boost control problem. I don't. It hits boost target and is stable. The problem is it takes too much rpm to get into boost threshold compared to my old turbo. The wicked jump of going from 200 to 400 ft-lbs of torque in less than 1000rpm around the 4000 rpm mark is comical to drive through but not helpful to make an autocross car faster. Then driving within the boost rpm threshold it will immediately zip to boost target even at part throttle rendering allot of my throttle travel useless for control. Some TPS based boost control might fix the latter part but I'm not going to improve that boost threshold problem with boost control strategies. Smaller than the .72 AR turbine housing might help but the only option smaller is the .49 internally gated T25 housing which I think is too small for my purposes it would probly work great on a 1.8l less extreme build. .

For autocross in a high hp car Ideally you want to be able to go between 30mph and mid 70's in one gear and have torque and response everywhere. I'm thinking the car produces better lap times with my old turbo I'm pretty sure. A GTX3071 with a .61 AR turbine housing looks like the turbine would match my .64 GT2971 and still be more efficient. I'm not sure on the compressor wheel though the Gen2 map looks really broad and wide with more power potential but the Gen1 looks like it might be more efficient in the range I need and closer to my past GT compressor wheel.

 

You do have a boost control problem because you are unable to limit the boost to anything close to your wastegate spring rate.

Unless of course you are making 400lbft at 6psi, and if so, more "power" to you.

Honestly, the screen shot of the log above looks very suspicious, like you have it hooked up incorrectly or a configuration error.

 

Bob, your description just reminds me of the 90's Lotus F1 cars like the 98T. I'm talking about sim racing, so obviously not 1:1, but it's just wait, wait, wait, oh my god here we go. Like you said, entertaining, but you end up spending more time focusing on controlling the car than on driving good lines.

This thread has been interesting to follow. It would be a bummer to have to jump back to the old setup, but if it makes the car better than I suppose you might as well.

 

I'm not trying to run it at 6psi.

 

Bob, with all respect and I know you are an experienced guy, but you DO want to be able to run 6psi at any rpm you want, right? That's the whole point. If you're able to do just that, you have the control you're after.
Welding the gate directly to the turbine housing might be a fix, but you need a good fabricator. And you need the space in the engine bay.
I can't believe a new Link isn't able to control boost like an ms3 can with pretty basic strategies.

 

I have a feeling the retarded timing from traction control is creating more boost.

 

You can look at the boost curves on my dyno plot. it comes up to target and stays flat. similar curve for all my preset boost levels when I hold constant throttle.