We all know this stuff for the most part, but the whole point of posting on a forum is to talk it over. Ya know, the whole point of a forum.
If he didn't want any of us to put in our constructive 2 cents, I'm sure he wouldn't have posted. I'm sure he knew the steps were a bit more wire to the pump and bigger injectors to drop the base pressure, but it helps to have others reaffirm plans at times, and just maybe you get a little nugget of useful info you didn't have before.
It is amazing what a bit of current draw can do to the volts at the other end of the cable....well, perhaps not amazing, but surprising when you think you've done the right thing.
I have about 5 foot of cable running to my pumps (AEM380lph main + Walbro 255 intank) and was annoyed to see I had 0.5V drop at the pump. The cable is pretty chunky and the earth on the main pump is about 1 foot long.
The V+ is wired to the battery in the trunk and goes through a fuse and relay which is triggered on by the ECU. Still, the AEM380 is performing about 10% below spec when I tested it and that isn't too bad considering this was tested by measuring outflow from the regulator set to 60psi (Pump>6AN lines>filter>Rail>6AN lines>regulator). Testing done with the engine off but the battery hooked to a variable voltage 40amp bench supply.
I tried the PWM control logic on the ECU but it wasn't successful and the pumps made all sorts of unhappy noises. Not sure if it was the SSD (40amp SSD relay), or both pumps being connected to it, or what? I haven't got around to connecting up the oscilloscope to see what's up yet. Flyback was being damped by a large rectifier diode, but perhaps the flyback from two pumps connected to the same SSD just caused some funked up shxt. Never got around to testing that either.
PWM control of the SPAL 16" fan using the same type of SSD relay has been flawless! I love have the fan providing at least SOME cooling at anything over 85degC, and the coolant seems to maintain a very constant 90degC. Much better than a 95degC on and 88degC off type strategy - IMO.
As an aside, rather than install twin pumps, I have been looking at the voltage buck-boost modules that are around which can boost the supply voltage up to provide 15-16V to the pumps when under boost. But I'm thinking that perhaps KISS applies here. Or at least I do go down that road, I better get that fuel pressure sensor installed and set up those fail-safe tables!
Ps, kick ass result with the BP from 30 to 40 turbine!
Last edited by Nutzy; 12-08-2017 at 09:08 PM.
1989 E30 - M50B28 Turbo - ZF 8 Speed
Sometimes I'm so focused on "high level" strategies, I miss the basics.
I have in general, found it useful to "spell out" what I'm doing as a method of making sense of what I'm seeing. However, many readers of this thread are parachuting in on page 44, and are completely unaware of my previous efforts, experience, or abilities.
Now that I have a minute to reply...
In 2014 was looking to run 4 channels of PWM:
- Radiator Fan
- Fuel Pump
- Nitrous solenoid
- Turbo Scavenge Pump
All of those channels were wired into the engine harnesses with a common ground, shared between other devices. The Racegrade SSDs would then provide a 12V+ PWM signal to each. I should also note that the initial spec from Racegrade was 40 Amps for the SSD, so I assumed that each channel was capable of 40 amps max.
When I started trying to pulse with the SSDs, I got really strange results which I originally thought were issues with the CAN trigger to the PDM. (Providing 12V+ to the SSD channels) The PDM was run via CAN, and the SSDs were controlled by the ECU with PWM outputs. After many issues, I came to find out that there was a revised set of instructions for the SSDs where they couldn't pulse 12V+, despite having this printed on the side of each device, and the current limit was 10A continuous per channel.
So here I was, with the ability to switch over the SSDs to Ground PWM, but the harnesses were built to do 12V+. I ended up wiring everything directly off of the Motec PDM, where they were either switched on or off.
This adventure ended up costing me well over $2k in wasted parts and materials. (SSDs, wiring supplies, and fixing the engine that was damaged because the SSD intermittently turned the fuel pump off under 100% load)
Nutzy,
What SSD are you using?
Are you doing 12V+ PWM, or Ground PWM?
I have found that some SSDs have a frequency limit that leaves them suitable only for switched on or off. For fuel pumps, they might like to be pulsed upwards of 8000 Hz....
You might need some capacitance/inductance (I'm not a EE!!!) on the pump side of the solid state relay to prevent a really noisy signal getting to it. That would allow you to still control the voltage level to the pump (if that's what you're going for), but won't be slamming it with 0V to 14V on and off really quickly.
With a 9.0:1 S52 (aftermarket rods and pistons), S52 cams, M50 intake, open boostlogic manifold and GTX4088R, 3.5 inch exhaust, 60 lb injectors tune on stock ECU, 51 psi base pressure, DW300 pump with stock fuel lines and KB voltage booster, I made 677 rwhp at 24 psi SAE Dynojet at 6500 rpm.
The SSR is actually a 100 amp unit - https://www.jaycar.com.au/medias/sys...aSheetMain.pdf
According to its specs, it would appear that 300hz is its maximum frequency, but then it may not be suited to this task, period.
The ECU is grounding out one side of the SSR which is optically isolated internally. The pump load is ground side, so 'ground PWM'. An EE friend of mine suggested that this SSR is not suited to the task I am using it for and drew up a little IRnnnn FET circuit for me to construct instead.
I'm thinking I might try out a circuit using a FET instead, heck, there are ton of these pre-made little PWM FET boards on ebay for less than $5. In any case, these types of FET circuits will have zero issue running 8K PWM as you suggest.
Pete
1989 E30 - M50B28 Turbo - ZF 8 Speed
Great info, thanks!
I'm hoping to resolve a few remaining issues with the car this week, then a dyno day will follow. (Dynojet)
Please keep sharing your progress; I think a lot of people would be interested in this as flow requirements continue to climb.
I still have the Ground PWM side of two Racegrade SSDs to play with....I will share how that goes once I revise the wiring to suit.
Yup i looked at the pretty pictures flying to Wisconsin and never read much until recently
Nutzy please share the Ebay board. I am going to be adding a SSR very soon for fan control, however i have not found a relay that looks like it will work well.
A external FET drive would be sweet to add. Send me a link and ill add it to the ecu in the next 2 weeks.
Last edited by euro2fast4u; 12-11-2017 at 10:52 AM.
Capital Driving Club Car # 102
How to turbo your car:
Step one. Install ecu and learn to tune and or have it tuned.
Step two. Install injectors and retune.
Step three. Install turbo parts and bits. get it running with out leaks. DO NOT DRIVE IT. Idling should be ok
Step four. Retune car
Step five enjoy
The ID 1050x's are in, and the intake manifold wiring harness is revised for the new connectors.
Because of the shape of the M54 intake manifold....I had to mount the injectors sideways....with the connectors facing the front/back of the car. I suspect there is a spacer to make them fit differently, but I don't mind.
The wiring changes on the back of the car are done.....now I just have to revise the front of the car to match....then I'll do some logging of voltage vs pressure to see how the pump works with the "old" wiring, vs the new wiring. (2nd 12V+ run to the pump with a DTP connector so I can run it connected vs disconnected) I think I also might move the injectors off of the A1000 fuel pump circuit, but I haven't figured out where to put them yet. (Juggling channels on the PDM)
- - - Updated - - -
Oh...and it snowed....so no test drives for a little bit....
Got the rest of the car wired up, and started doing testing.
FPR set to 43 PSI static with engine off. 40A Power supply connected, (It's actually installed in the car) showing a constant 13.0 volts on the dash.
Pump -> GND = 10 awg (3 feet)
Pump -> PDM = 12 awg (20 feet)
11.8 Volts
Adding another leg of 14 awg (20 feet)
12.5 Volts
If I were to redo the entire system again, I'd probably run 8 awg.
Most of my time this morning was spent working on the fuel film model in the M1. The ID injectors have a different spray pattern from the Bosch 630cc injectors, and I've tilted them slightly to be pointed more towards the valves.
I didn't expect this to be such a big difference, but I had to re-tune a LOT of the fuel film model. Eventually I had the tip-in at idle so it wouldn't stumbled, and I moved to doing load testing.
Now....this is where things got tricky.
With the car on the lift, and the transmission in 5th gear, I would apply the brakes slightly while giving it throttle. With a bit of practice, I could hold the car at an RPM point, then ease off the brake a bit with constant throttle to allow the RPMs to rise. After a "pull", I would keep the car in gear for a few minutes at 3000 rpm to let the brakes cool down a bit. Eventually, I worked up to doing full throttle pulls from 3000 to 7600 RPM...although to be fair.....once there was 15 PSI of manifold pressure....the RPMs went up faster than I could modulate the throttle. Yes...my brakes are taking a beating....but this is a hell of a lot safer than being out on the street. (In the snow) I think I might even look at getting some leaf blower fans to place behind each wheel to help with cooling....this method is working pretty well so far.
I got a lot of VE tuning done using the above method. Some of the load points were pretty much what I expected....but I continue to be amazed at the VE from 4500 up to 7500 RPM.
5000 RPM @ 100kpa, the VE is 112%.
6000 RPM @ 200kpa, the VE is 126%.
7000 RPM @ 200kpa, the VE is 120%.
In regards to slip% for tires, my car did the best at 18% with the bias plies and 12% with the radials. Both scenarios the car would leave marks on the pavement but you couldnt really tell it was spinning.
328i Sedan Twin s366's, 6.0LS, TH400, MS3 Ultimate
9.20 at 150 on 22psi
We do the brake loading in the semi trucks to get various speed/load points a little outside what is normally available. Works well, especially up grades (which you clearly can't do in the garage.) The mechanics hate us though
Just came back from a test session on a long uphill back road. (No houses, no cars, no intersections)
While testing 70% throttle position....I scared myself and called it quits. (The back end broke loose at about 100kph in 3rd gear at 5000 RPM)
The goal was to build a 10 PSI, 12.5 PSI, and 15 PSI boost strategy....but I was having problems with 10 Psi. I did get a LOT of lower load data though to refine the VE table.
On the way back from the test location, I continued to refine the VANOS, DBW throttle position, and WG strategy to improve spool....and boy did I do that!!!!
In 3rd gear, I'm now getting 10 PSI at 3460 RPM.
Test from 2000 rpm steady level ground in 4th and tell us when you hit 10psi. Should be better response due to gear load. With an open manifold, I was hitting 10 psi in 4th with GTX4088R and non optimized tune from 2009 at about 3750. A twinscroll would hopefully beat that by 500 rpm and an optimized tune improving spool a few hundred rpm more.
Id like to add that i had my first tuning expierence with a Motec m48 yesterday and holy crap is it old. DOS based, game boy looking interface. Crazy.
328i Sedan Twin s366's, 6.0LS, TH400, MS3 Ultimate
9.20 at 150 on 22psi
Pectel was like that up until about 2011.....DOS GUI.
Bookmarks