darrenjlobb's Adaptronic Managed Compound Blown Derv Bus Build

[Mattcheese31]

You have really stepped up to another level completely from previous projects as i see it, I think this could perhaps be the car that makes Piggy actually try to physically bum you, for real! ..... "Joking" aside, it gets me excited looking at it fella, pressure's on but I'm sure it'll go like a charm

[hotrodjacko]

Going to echo everyone else here really but great work - it's also nice to see the technical stuff going in to the write up as well. Followed your stuff for years, although probably not as long as everyone else here!

[darrenjlobb]

Darren i would just like to say  far play to u and the work u have put in to this car/project . You are the one that makes every one  else on here  want to do there own projects :-) you have my highest respect  can't  wait to see it on the rolling road :-)

Ps so much love for this car     

Thanks chap, always great to see other people getting into the scene because of a video / thread they have seen. Thank you

You have really stepped up to another level completely from previous projects as i see it, I think this could perhaps be the car that makes Piggy actually try to physically bum you, for real! ..... "Joking" aside, it gets me excited looking at it fella, pressure's on but I'm sure it'll go like a charm

Haha cheers Matt, all the support / following is massively appreciated by myself. Felt good to have it running though!

Going to echo everyone else here really but great work - it's also nice to see the technical stuff going in to the write up as well. Followed your stuff for years, although probably not as long as everyone else here!

Cheers, Try to write up what I can remember has gone on, no doubt there are endless issues / work done that I forget about however!

[Dum-Dum]

Awesome work mate, I can't believe the attention to detail that's gone into the recent work on this. It's really phenomenal, I'd be happy if I could achieve 10% of your attention to detail.

On another note have you started blagging post offices or created a meth lab to pay for that ECU?

[darrenjlobb]

Awesome work mate, I can't believe the attention to detail that's gone into the recent work on this. It's really phenomenal, I'd be happy if I could achieve 10% of your attention to detail.

On another note have you started blagging post offices or created a meth lab to pay for that ECU?

Haha Cheers Chris, and heh, no, just the last few years of saving / selling things I don't need!!


As an update to the thread, things are going fairly well. Now really enters the time of "not alot to see" but in reality huge progress is being made on most evenings at the moment.

I am becoming much more familier with the ECU now, becoming much faster at flipping around the logic / software , which is speeding up production somewhat to say the least.

The video of it "running" was really with 0 calibration, literally just enough that the injector outputs were getting an angle and a duration, but thats about it. So the hard work has really began since that point....

I took the decision to entirely re write the logic for firstly the injection quantity, and then the injection angle logic. Have basically re-worked the logic to work with actual fuel volumes, much like EDC does, rather than just working with durations / fuel pressure. Several reasons for this, but mainly just because once you have got the logic set up in such a way that everything is working in the same unit, mm3 of fuel, it is easy to see what is going on, and also means changing injectors is just a case of entering new injector cal information, and all your previous work will still work, as the ecu will just recalculate based on the new injector cal information, and deliver the same quantity (mm3) of fuel regardless of rail pressure / injector size.

First step was basically setting up some new ECU pages, and creating the logic for it, so in essence, a fuel request map (pedal), then correction maps (MAP / temp / torque limiter etc) and then injector calibration, to convert the request mm3 into a injector duration in uS based upon rail pressure at that given time. Obviously then had to convert the idle control to also give a mm3 request into the loop, to allow the engine to idle on its own / take load and torque without stalling. 

The same has basically been done for the injection angle logic, starting from scratch, with control over both the pilot and main injection start angles in all conditions. 

Have also spent quite alot of time removing routines that arnt used but on ECU pages as stock with the e1280. Just to tidy it up, as was still alot of petrol related logic kicking around that was totally un-used now. So that has removed quite a few pages in total and seriously tidied up other pages! 

Once the car was running with the new configuration, (took a bit of fettling with the idle loop, and also learning a few things about the injector output module in Sekuku) but fairly quickly got it running, and was idling with the expected fuel quantity, so was looking good.

Next wanted to see how far out the injector cal was, so Ruan sorted me out with the stock values from a 110HDi with respect to the requested fuel and torque limiter values)...idea being in theory, if my injector cal is correct, the car should perform a full power run, and return pretty much the same results as a stock HDI, as it would be seeing the same mm3 request values... So banged in all the data, ran it up on the dyno, and made a peak of 106hp, pretty much bang on what it should have. And made a surprisingly good dyno graph at the same time! So that was great news / basically confirmed the theory was correct, and the injectors were putting in pretty much bang on the fuel we were asking them to!.

Noticed during these pulls that the crude initial boost control wasnt working to well, due to the nature of the compound setup, so next step really is to sort out re writing the logic for a true compound turbo control system keeping the primary in its peak flow area at all times. 

Boring update really but making some big leaps forward in reality, just nothing to show for it! All a massive learning curve for me, and only now am I really starting to see the potential of this ECU, more I get familier with it, the more I realise it is capable of!

Cheers!

[NaughtonGAA]

Can only imagine what sitting at a laptop making changes to hundreds of values all evening must feel like. Shall all be worth it in the end though [SMILING FACE WITH OPEN MOUTH AND COLD SWEAT]. Exciting it being this close to on the road.

Sent from my SM-N910F using Tapatalk

[Dave]

Good update Darren, can imagine although this is the part with the least pictures/changes to report, it would also be one of the most interesting/rewarding now that the hardware has been sorted, and can actually tweak/tune/play with all the infinite settings that this thing can do, and see instant real-life results from it - this surely must be the most synonymous process associated with the old tweak with the 13mm? proceeded by a 'crikey she shifts on boost now'

[bashbarnard]

I know there are no pictures etc but this is one of the parts that interests me greatly. Its good that you know your setup is pretty much bang on running the stock figures based on ruans figures. Currently trying to learn edc15 at the mo and i find it difficult if im honest. Let alone a whole new interface of things. Ive always wondered since you started this how "long" this bit was gonna take but you seem to be making cracking progress keep up the good work


Sent from my iPhone using Tapatalk

[zx_volcane]

109bhp and rising!

[MY95]

Amazing build as said before, what power figures are you aiming for with this? Hopefully see this beast at FCS/Pugfest this year

[Dum-Dum]

Haha Cheers Chris .... selling things I don't need!!

.....

Next wanted to see how far out the injector cal was, so Ruan sorted me out with the stock values from a 110HDi with respect to the requested fuel and torque limiter values)...idea being in theory, if my injector cal is correct, the car should perform a full power run, and return pretty much the same results as a stock HDI, as it would be seeing the same mm3 request values... So banged in all the data, ran it up on the dyno, and made a peak of 106hp, pretty much bang on what it should have. And made a surprisingly good dyno graph at the same time! So that was great news / basically confirmed the theory was correct, and the injectors were putting in pretty much bang on the fuel we were asking them to!.

.....

Boring update really but making some big leaps forward in reality, just nothing to show for it! All a massive learning curve for me, and only now am I really starting to see the potential of this ECU, more I get familier with it, the more I realise it is capable of!

Yeah I'm at that stage, cleared out my box of turbos, binned what was scrap and put everything else on here and ebay along with a load of other parts.

Peter (TB205) who used to do my mapping was able to do that, he could calculate the MM3 and through knowing the efficiency of the engine could say almost exactly what a car would make on the dyno on a particular map to prove his calculations.

I think we've always known the capabilities of ECUs, it's just the cost and the learning that's stopping people from doing it. Compared to turning a few screws on Bosch pump it's alot of effort for what (for most people) is just a few % gain. So impressed as having played with megasquirt (hugely simpler) and had trouble getting my head round it I can only imagine what your learning. You must be sleeping well after sessions on this LOL.

Are we going to see some of these dyno pulls as no doubt your doing back to back comparrisons of little changes to see how stuff works and us saddos are interested in that sort of thing.

[deggerz]

Looking forward to a dyno video , be nice to hear what this sounds like under load


Sent from my iPhone using Tapatalk

[Piggy]

Not aboring update imo. Maybe no pics for the beano mag lovers but the steps forward are huge.
You've confirmed it all works to start with! Thats a big relief hurdle right there

[darrenjlobb]

Another end of week update!

Got the car all strapped down on the dyno ready for initial testing / setup.








Again its a bit of a boring update with little photos, but have now got the fuel control fairly dialed, and also got the pilot injection sorted, although not fully mapped yet, but working / able to be.

The main thing I have been sorting is the boost control for the compound setup. First step was to create two ECU pages, to contain the PID loops for each turbo, so one for the WG turbo and one for the VNT, in basic terms, these pages basically take a target MAP (pressure) input, a feedback pressure (map pressure input) and output a correction duty cycle to open / close wastegate / move vanes in turbo to correct the pressure towards the target.  The WG turbo is just using an external WG with a 3 port MAC valve, so this is able to be controlled directly off the back of the PID loop into one of the AUX outputs, setup for PWM. The VNT however is being controlled with a Hella Electronic actuator, this has its own ecu page for the "actuator control", which essentially is a small PID loop of its own, using the duty cycle output from the turbos PID control page as an input, then a 0-5v feedback on an analog input channel for feedback of vane position (the direct control hella units have a 0-5v feedback output) and then output the final output to a pair of AUX outputs to drive the high side of the servo via a DBW card (bunch of mosfets to cope with higher current).
So essentially with the above, each turbo has control, and just needs to be fed a target. So for that I created another ECU page, for the "Boost Control" which essentially contains the tables / maps, and some logic. It basically looks at MAP1 and MAP2, and calculates the current pressure ratio, and then corrects the output to hold the VNT's P/R in the current request area, while also holding the overall map pressure at the requested map, by letting the WG turbo's output raise / fall as required. So essentially this entire thing is controlled with a 3d map (RPM vs TPS) for target MAP pressure across the engines range, and also a 2D map for the target VNT P/R across the engines range. 

This took quite alot of setting up / getting right on the dyno, had to add a base duty map for the WG turbo as was finding the PID controller was getting a bit wound up at higher boost levels and oscillating a bit. Not yet found the need for the VNT, I think mainly just due to the lightening fast responce with the electronic actuator...

Also swapped out the Zeitronix 5 bar map sensor for an AEM unit, the sensor has been being used on the dyno for logging for last coupe years, and think has taken a knock at some point as always reads 4-5 psi out, so not idea, and had to have a summer in the logic to "correct" it, so that is now reading correctly.

Next project was making a stock dash tacho work off the ecu, wanted to use a GTI6 dash as the DT dash only reads up to 6k, and has always annoyed me having the redline at 5k?!

So took apart a couple GTI6 dashs and made up a bit of a hybrid, and managed to adjust the milage to match my original set, not that it really matters on a car like this, but just wanted to keep it correct! Found the wire on the back of the dash for the tacho input, and setup and aux out on the ecu, connected it up, and then setup the output to be a PWM output, relative to the current RPM, simple multiplyer in the logic sorted this, and hey presto the tacho was working!










Other than that not much to report really, have been slowly working on the tune. Pretty much at the limits of my current injectors now for the power delivery I am after, so now starting to sort out a much bigger set to replace them with. Have been trying to tame the power curve to a nice linear petrol like output, rather than the typical huge surge of mid range resulting in lots of tyre smoke / snapped shafts and clutchs. Car is currently sat at around 205hp @5krpm, with durations pretty much at there limit without making the EOI problematically late.

Safe to say I am learning alot in this process, very quickly getting the hang of the ECU / software now, and really starting to make some good progress. Also great to finally be able to use the dyno for what I bought it for, tuning this car! And it really is prooving to be the most useful investment I have ever made with respect to this car!  The work I have been doing at the moment with the turbo control etc would be simply impossible on the road.

The latest video probably shows more than I can explain in here.

Thanks for reading

https://www.youtube.com/watch?v=6ZmGIbnWPl0

[tigerstyle]

Sounds ace! (in the vids)

[Piggy]

More fuel...
More boost...
Have a brew.
Repeat.

[dr_jekyll]

Sat watching this at 2am when i couldnt sleep. Just cant get my head around so much of it. You sir are one talented mofo [VICTORY HAND]

Sent from my SM-G920F using Tapatalk

[Magenta Sunset]

that is one hell of a beast/street-sleeper.
never knew a derv could rev so high.
most of the tech is above me but it's a great build.
correct me if i'm wrong, but it sound like it's 'hunting' a bit at idle. reminds me of the old lorries from years ago.

[welshpug]

that's not high tbh, I'm sure Darren revved one of his old engines further, fairly commonplace for engines with more valves to rev higher, though the old M51 BMW inline 6 goes to 4800 and that's just a 12 valve.

its usually more a function of bore vs stroke plus rod lengths.

[unclebeast]

Id have to sell the house to get insurance for that with a M9 postcode..what a build...The John Deere Missile...wonder if they would sponsor it?

[cully]

good going Mr Lobb keep at it

well done

[Dum-Dum]

Not watched the video yet but it's awesome that in a week you've got so much licked and doubled the power output.
She's looking awesome in the photos. Can't be far from the road?

[Paul Baldwin]

Thouroughly enjoyed that write up and video ta, all impressive stuff and quite the achievment for a home-brew project!!!

That certainly looks like a very useable power output, should be a pussycat to drive where traction is limited hopefully? Guessing from the dyno graph that the rev cut is just shy of 7K so pretty high as dervs go.

[Danny Wideboy]

Read all of the last update, watched the video.
I don't understand any of it but I do know it's pretty feckin' cool.

[darrenjlobb]

Sounds ace! (in the vids)

Thanks chap,

Sat watching this at 2am when i couldnt sleep. Just cant get my head around so much of it. You sir are one talented mofo [VICTORY HAND]

Sent from my SM-G920F using Tapatalk

Thanks mate, it all makes sence when you read the entire thread I think!

that is one hell of a beast/street-sleeper.
never knew a derv could rev so high.
most of the tech is above me but it's a great build.
correct me if i'm wrong, but it sound like it's 'hunting' a bit at idle. reminds me of the old lorries from years ago.

The car isnt actually revving that hard at the moment, its hitting my fuel cut limit at 5800rpm in all the videos, infact some of the first videos its hitting the limiter at 5200rpm. On the limits with durations so the power is rolling over around 5k at the moment due to simply not being able to get the fuel in any faster.

that's not high tbh, I'm sure Darren revved one of his old engines further, fairly commonplace for engines with more valves to rev higher, though the old M51 BMW inline 6 goes to 4800 and that's just a 12 valve.

its usually more a function of bore vs stroke plus rod lengths.

To 8 valves or 16 valves, diesels will revs, short stroke obviously helps, but more importantly than anything is the turbocharger / turbochargers ability to keep moving air as its required up there, and the injectors ability to get the fuel in fast enough without the durations getting out of hand up there. More RPM's will come with time.

Id have to sell the house to get insurance for that with a M9 postcode..what a build...The John Deere Missile...wonder if they would sponsor it?

Haha, I dont think Deere would  be massively impressed with the PSA powered machine tbh!

good going Mr Lobb keep at it

well done

Thanks Paul,

Not watched the video yet but it's awesome that in a week you've got so much licked and doubled the power output.
She's looking awesome in the photos. Can't be far from the road?

Cheers Chris, to be honest still a long way from the road, I see little point in wasting money insuring it / MOT'ing it and taxing it when its still in development. After tuning on the dyno, it has made me realise that any attempt to "tune" an engine on the road is largely next to useless, esp with this sort of build. So its better off staying on the rollers for the time being where I can keep a very close eye on everything thats going on, and use the logs / dyno to make progress.

Thouroughly enjoyed that write up and video ta, all impressive stuff and quite the achievment for a home-brew project!!!

That certainly looks like a very useable power output, should be a pussycat to drive where traction is limited hopefully? Guessing from the dyno graph that the rev cut is just shy of 7K so pretty high as dervs go.

No as above its hitting my fuel cut at 5800, this will increase in time, but right now fighting injector capacity more than anything! But she will scream once again soon! And yes, all still built in an ex calve rearing shed on a farm in Cornwall!

Read all of the last update,  watched the video.
I don't understand any of it but I do know it's pretty feckin' cool.

Thanks Danny!

---

So another little update, looks like we are out of injector pretty much now. Have been slowly increasing power over the last week, and are now at the maximum rail pressure I can run with the current rail pressure sensor allowing some control overhead, but are hitting my "duration limiter" I set in the logic to make sure there are no accidental fails! 

Although we could get away with some extra duration in the mid range where is there is more time to get the fuel in, we are sat on the limiter up in the revs where we want to be making power. Here is a screen shot of the logs from a pull, in the top graph the white line shows current RPM, the yellow line shows requested fuel "mm3" after all correction maps, and the red line shows the injector duration (actual output) at the time, you will notice at around 4krpm the duration just hits the limiter (hence the flat line), so although the request continues to rise, the car will put no extra fuel in as its already at max rail pressure, and max duration. So little option now than to pull the injectors out and see about sorting some bigger hose pipes out. For reference, bottom graph is showing rail pressure in red, and MAP1 in green.




So this evening set about dropping the car off the Dyno, and pulling the injectors out:






Will be getting these sent off this week, initial plan is to increase the needle lift on one, and stick it back in a flow bench, and literally just run it up at 1800bar and see how much extra flow we can get from the current nozzles. Im thinking although it will be a significant increase, it will likely not be enough for what I have planned, so will likely be needing another set of nozzles, but we will have to wait and see on that. Sadly I have no means of setting up Common Rail injectors myself and / or flow testing them, so have to get it done externally, which is fairly expensive, so really need to try and get it right first time if I can. Have debated building my own common rail flow bench a few times, but for what it would cost to do a decent job, I could have a few sets done, and cant really justify the expense right now. 

Have noticed a couple minor oil leaks that need to be resolved also, so while the engine is apart will be able to take the time to change the oil, sort a few little leaks / niggles out and work on some other areas that still need sorting out!

Thanks for reading along

[tigerstyle]

Did you predict you'd be 'out of injector' so soon? (it feels fairly soon into the project/testing to me?)
But a nice ceiling to hit soon, rather than say the pump not being able to deliver enough

[toseland]

in awe....

[mr_fish]

Made a note of how many miles has she racked up on the dyno?

[Matt-Rallye]

Just had a big catch up session on this thread...

Stunning progress! I really am trying to get my head around the massive amount of technical detail your throwing around there and still havent got a clue! i really take my hat off to you

So is it now a case of improve the injector flow and see whats capable on the rail pressure at 1800bar?

[Ruan]

that's not high tbh, I'm sure Darren revved one of his old engines further, fairly commonplace for engines with more valves to rev higher, though the old M51 BMW inline 6 goes to 4800 and that's just a 12 valve.

its usually more a function of bore vs stroke plus rod lengths.

Airflow through the engine is a tiny portion of the story... Valves do help move more air at high RPM, but you'll tend to always limited by the injection system rather simply by airflow - which can usually be upgraded more than enough by fitting decent turbos.

Usual limit is simply the time you have to start the injection, get the quantity of fuel you need in and finish the injection, ensuring it's all in before the compression is fading away - there's a limit to how much timing you can give it... There's also a fixed amount of time that it takes from the point of fuel going in until the first sign of combustion (ignition delay) which has to be factored in...

At 1400uS duration (yes, 0.0014 seconds...) at 4000rpm - it'll take 33 crank degrees to get the fuel in... Move that to 5000rpm and that's now taking 42 crank degrees... At 5800rpm you're up to 49 crank degrees... That's just to get the fuel in, ignoring the fact that 50 degrees ago the first of the fuel went in and is now starting to combust - the last of the fuel still has some time to go before it even begins to combust! To make that power up high you need insanely short injections and quite a bit of timing, even to ensure that the fuel is in early enough to compensate for the ignition delay and ensure the fuel is combusting at the right time.

Joy of commonrail is that you can have those huge injectors that inject a huge amount of fuel in a very short time at high pressure... However even at lower pressure - you can get an acceptable idle and midrange and ensure you're getting good spray patterns for decent combustion, but crank the pressure and you've got craploads of fuel going in a short period. Just need some bigger nozzles!