N13 Turbo - Round 2... Fight!

Converting To Delco ECU From Carby Or Other Injection Systems
User avatar
charlay86
Posts: 584
Joined: Thu Sep 17, 2009 2:00 pm
cars: VT S1 SS (L67)
Location: Perth, WA

Re: N13 Turbo - Round 2... Fight!

Post by charlay86 »

BennVenn wrote: Can PWM be done in 12p to control a boost solenoid?
Yes via the flex outputs, you need to use the flash tool to set it up.
User avatar
Holden202T
Posts: 10311
Joined: Sat Feb 28, 2009 9:05 pm
Location: Tenambit, NSW
Contact:

Re: N13 Turbo - Round 2... Fight!

Post by Holden202T »

I’ll post more info later when I’m at a computer, I have made a pcb to do exactly this, uses a flex table to produce a pwm into an Arduino along with a map sensor and then a pid controller to control a high current output to drive the mac valve
BennVenn
Posts: 475
Joined: Fri Mar 04, 2016 10:35 am
cars: R33 GTST, '60 Vw Bug, Express (4G63T), GW X200
Location: Windellama, NSW
Contact:

Re: N13 Turbo - Round 2... Fight!

Post by BennVenn »

Gareth wrote:Sure, say $100 plus postage suit?
Sounds good, Messaged you.
Holden202T wrote:I’ll post more info later when I’m at a computer, I have made a pcb to do exactly this, uses a flex table to produce a pwm into an Arduino along with a map sensor and then a pid controller to control a high current output to drive the mac valve
That's a pretty cool way of doing it, so you can pass boost values in that table instead of percentage and the arduino will do the rest!
User avatar
Holden202T
Posts: 10311
Joined: Sat Feb 28, 2009 9:05 pm
Location: Tenambit, NSW
Contact:

Re: N13 Turbo - Round 2... Fight!

Post by Holden202T »

ok so now the long story!

probably like everyone, i've always wanted to do proper boost control with the delco, but sadly it just doesnt have the coding to do so.

So totally un-related to this, when building the FX i got myself a vacuum pump from a Holden, and plumbed that into my brake booster, all worked well except the pump running constantly got hot pretty quickly, so i decided to make a small circuit with an arduino, map sensor and relay control to basically read the map value on the brake booster, and once it dropped to a certain level it would turn the relay on, and then once it got back to high vacuum turn it off.

before i actually made the board, i started thinking of what else it could do, it quickly became obvious, with the right additions it could easily be a boost controller, so i added an on board map sensor, an external MAT & MAP input and a dual channel high current driver which can run an injector / boost solenoid etc. (all circuits courtesy of speeduino schematics)
IMG_0193.jpg
IMG_7499.JPG
It worked really well for the brake booster control, and then i got to thinking if you could read the delco flex table output into an arduino then you could do what you need for boost control.

it required a few hardware mods to make it work, but otherwise all good.
IMG_0192.jpg
with a bit of help from Jayme we got a working input into the arduino, and i tapped into my 3 bar map sensor that the delco uses, and then i was pretty much set for coding it up. I found a PID controller arduino library, and that was pretty much all i needed to make it all work.

So i changed the calculation in the flex table to display 0-30 (PSI) instead of the 0-100% duty cycle it shows, and then basically put in your desired boost, which then sends that value to the arduino.

as per below here you can see the raw map sensor value, calculated map value, flex input raw value, then converted to a desired boost value, and the the PWM out is the value sent to the high current output to control the mac valve.
2022-07-05_11-07-37.jpg
at this stage that is basically all i have done with it, as i do not have a registered car with delco and turbo to test it with. I am expecting the PID values will need some adjustment but the theory is at least proved.

im happy to work with anyone who is in a position to test this :)
BennVenn
Posts: 475
Joined: Fri Mar 04, 2016 10:35 am
cars: R33 GTST, '60 Vw Bug, Express (4G63T), GW X200
Location: Windellama, NSW
Contact:

Re: N13 Turbo - Round 2... Fight!

Post by BennVenn »

Very nice work! I'm wondering if you could simplify it further by instead just connecting to the serial comms from the delco and retrieving RPM/MAP/Desired Boost as it's all available over the stream. Would save a map sensor and any difference in the ADC between the delco and the arduino. Also conversion back from duty to requested boost would be more accurate.

I'm working on a 3.5" IPS panel that can read the stream and display it in gauge/digital format. Planning on mounting it in my dash cluster where the temp gauge is. I've got the IPS panel here running on a dedicated PCB with FPGA/video RAM etc. Just need to look into getting the delco to stream data upon request.

If you want to develop your idea, I've got a dyno at my disposal but not sure the old pulsar would be happy with much more than 10psi which would limit testing but I can put a 1psi spring in the wastegate so we'd get some testing/tuning done.
User avatar
Holden202T
Posts: 10311
Joined: Sat Feb 28, 2009 9:05 pm
Location: Tenambit, NSW
Contact:

Re: N13 Turbo - Round 2... Fight!

Post by Holden202T »

yeah thats probably a good idea, but unfortunately out of my coding ability - also for me when racing i use data logging on every run, so would need to use the serial comms for that.

as for the testing, i see no reason why it couldnt work with your low psi ... even control from 5-10 would prove it does what you want!!
BennVenn
Posts: 475
Joined: Fri Mar 04, 2016 10:35 am
cars: R33 GTST, '60 Vw Bug, Express (4G63T), GW X200
Location: Windellama, NSW
Contact:

Re: N13 Turbo - Round 2... Fight!

Post by BennVenn »

My Saab China turbo arrived with a 1bar actuator installed. It should have been 0.4bar if it was to be a factory replacement turbo.

I've since installed an aftermarket actuator but it got me thinking, would a $5 spring from bunnings reduce actuator pressure to a more stock level to keep the old pulsar happy?

https://youtu.be/TrTsoj5eIhc
User avatar
Holden202T
Posts: 10311
Joined: Sat Feb 28, 2009 9:05 pm
Location: Tenambit, NSW
Contact:

Re: N13 Turbo - Round 2... Fight!

Post by Holden202T »

at the end of the day i guess it doesnt really matter what the spring is as long as its within spec of what you need....

for example, more coils in a spring will make it have more leverage over the load applied to it, so it would be softer than one with less turns. On the same token, thicker wire diameter will be stiffer than thinner if the coils are the same length ...

but yeah if you could get a really light spring that was like 5-8psi max, then it would be a really good one to use for testing a boost controller.
BennVenn
Posts: 475
Joined: Fri Mar 04, 2016 10:35 am
cars: R33 GTST, '60 Vw Bug, Express (4G63T), GW X200
Location: Windellama, NSW
Contact:

Re: N13 Turbo - Round 2... Fight!

Post by BennVenn »

anomoly.JPG
Thoughts on this boost profile?

What I thought previously might be a bad coil or bad valve springs, turns out to be boost related (The dyno software was filtering its MAP input so much this spike/dip wasn't present...)
between runs.jpg
All boost hoses are super tight though I haven't done a boost leak test on them.

All manifold barbs are plugged except for the brake booster feed and map sensor.

Exhaust gasket is new and the manifold was ground flat on a belt sander.

Wastegate actuator was swapped from the OEM saab type to an aftermarket kinugawa actuator. This boost dip isn't present at 6psi, only when increased to 9psi or higher (stock actuator or oem). Switching boost between the two will show repeatable torque curves.

VE doesn't change when the dip occurs. AFR's don't change when the dip occurs.

I'm not sure if its a lack of exhaust flow from the engine which is lowering the boost, or if its actuator/turbo related. Could it be the head lifting between cylinders reducing combustion and exhaust energy? There is no coolant in the oil, oil in the coolant or any combustion gasses making their way into the coolant tank.

This has got me stumped... retarding the timing up to 10 degrees at the higher boost lowers torque significantly though the boost profile remains the same (moves maybe 100-200 rpm higher or lower).

It's hard to hear any boost or exhaust leaks over the dyno at this wheel speed though the induction noise does change during the dip.

Increasing to 1 bar boost pressure makes around 94kw at 3400rpm before the boost dips like the 9psi profile.

The turbo poppet moves freely when cold, the actuator arm is in a straight line to the wastegate arm. I don't think its binding.

I thought it might be the car creeping up on the rollers and off the idle rollers but the same boost issues are seen on the road ruling out the dyno.

open to any suggestions! (regarding valve springs - any aftermarket springs of the same dimensions that i've found have less seat pressure when open or closed according to their datasheets than the stock ones, by quite a margin)
User avatar
Holden202T
Posts: 10311
Joined: Sat Feb 28, 2009 9:05 pm
Location: Tenambit, NSW
Contact:

Re: N13 Turbo - Round 2... Fight!

Post by Holden202T »

thats definitely a very odd one!

its not spinning the tyres at that point ? do you have a data log of the ecu to see if the speed spikes or something ?
Post Reply