J-series Acura/Honda

[Peter]

Or has the setup changed now?

No, that's the current setup. I just thought the secondary turbo would change the temperature enough that we would want it after it.
I got the BRV fixed today! It shows about 12.2 with the engine off, 10 cranking the engine over, and 13.6-14.2 with the engine running. Seems pretty reasonable to me. I still need to configure it. I was just going to wait until I go to my parents house this weekend, because my dad has a nice multi-meter.

Another thing that is really good is naming/dating the files well. It's good for you too. I have hundreds or thousands of log files here, it's a bit of a nightmare sometimes

lol I'll try and start naming them like I'm going to have to find them on your computer.

[Fred]

wait, one of your turbos is after the intercooler??? if so, you're right, but can you fix that, or why have you done that, i feel lost! :-/

[Peter]

Well I don't know how good of an idea it is, but this is where it came from.

It takes less work to get to a higher pressure if you cool the air in between compressions. They also have a regenerator in this diagram. It heats the intake air so that not as much heat gets wasted into the environment. I considered trying to heat the air after the second compression, but I figured it might cause predetonation so I decided not to do that. The idea was to raise the efficiency when I'm not hot rodding it around. Sure I could get a denser charge of air if I did two compressions, and then sent it through the intercooler, but it already has more than enough power to get me in plenty of trouble as it is.

[Fred]

OK! Interesting, though like you, I hate the idea of hot air going in, not because it's less dense, but because it's prone to detonation, hence my concerns about not intercooling after the second turbo.

Have you considered water injection? You can get away with a lot if you do that.

Compound setups are generally only used for high boost levels. What sort of boost are you running?

I like that it's unique, whether it's an awesome idea or not. I want to do one of these setups, they're in use on diesel fords and BMWs, and are a beautiful design, perhaps you could beat me to it:









They're aimed at low down torque and efficient high end operation in the same package, hence the attraction, turn your 2 litre into a 4 litre type stuff.

There should be four (off topic) images there, if there are only three, let me know and I'll fix it.

Sorry for the distraction!

Fred.

[Peter]

Have you considered water injection? You can get away with a lot if you do that.

Kind of, I think it would probably be more trouble than it's worth. Mostly because we have freezing temperatures 6 to 9 months out of the year, usually. This is kinda why I would like the knock sensor on the engine hooked up. That way if it starts predetonating during the summer I'll know it's time to reroute the intake plumbing. Right now it's -7C outside, so I can get away with it.

What sort of boost are you running?

No idea, I could barely spool the turbos up with the OEM setup, and I have the secondary wastegate opening up as soon as it can right now. The main goal was boost earlier in the RPM range, not so much tons of boost on top end. I'd like to point out the regenerator in your third picture. Essentially regeneration by mixing hot exhaust gas with cold intake air(EGR valve). I'm doing pretty much the same thing as the first picture without any intake air valves.

I want to do one of these setups, they're in use on diesel fords and BMWs, and are a beautiful design, perhaps you could beat me to it:

I'll keep my eyes open for a cheap 4 banger to try a complex boost off idle setup for it. I have a little motorcycle turbo for a project that never went anywhere. I think it was suppose to be able to run a .75L engine to 8grand. A simple flow rate calculation would tell us where we need to cut it out of the loop.

Sorry for the distraction!

That's alright it kind of applies. I did put up a post in user rides section with some more pictures, and Peter logic on why I did things the way I did. viewtopic.php?f=3&t=1563

IAT sensor, the weatherman says it's 19F(-7C) outside this morning.
9.0 kOhms at -7C
2.9-3.0 kOhms inside my house that's about 20C
190 Ohms at the boiling temperature of water at 7165ft(wikipedia) so about 92C according to the property tables in my Thermodynamics book.

CHT sensor I didn't take it out of the car so I only have the one reading
8.9 kOhms at -7C
The thermostat is suppose to open at 185F so I'll check to see what the resistance is at that temperature later today.
I might boil it after that if I'm feeling ambitious.

[Fred]

FreeTherm will run on windows, I don't mind adding a table to the code base just for you if you come up with one that makes the code read correctly across the range. Grab the latest version of it and give it a whirl.

[Peter]

FreeTherm will run on windows

I sure can't get it to work. Maybe I'm doing it wrong.
I open up GitBash
git clone https://github ... FreeTherm
cd FreeTherm
make FreeTherm.exe
Then it outputs some errors. I've tried it in XP, 7, and Ubuntu. The s19 files will build in XP and Ubuntu, but not in 7. Do I need a different tool-chain or something? I can't seem to find much in the way of instructions on it. I also tried to use the release with the exe already built in it, but it didn't work either.
At this point of screwing with it I think I would rather input all the values in the table individually. I just haven't figured out the conversion from the temperature to the 0-65535 range yet.

[Fred]

Are you messing with me? :-)

Try double clicking the exe, it's a windows CLI app and should work on any winblows machine with CLI libs and runtime installed.

make needs two things:

A makefile
A matching target in the make file

I think you have neither with FreeTherm :-)

Re not building on 7, can you be more specific? I've built on 7 with the tools from the website.

Inputting values directly will be a pain. The app goes through some math to come up with the curve based on the bias resistor and three points you supply. If you provide a good set of data for a curve and your bias resistor value I can do it for you. It shouldn't be this hard, though :-p :-)

Feel free to hit me up on skype, irc, gmail, fb, whatever rather than struggling alone. A quick question could save a lot of fucking around at your end! :-)

Fred.

[Peter]

Are you messing with me?

Man I wish. I think windows knows it has lost my loyalty.

This is what I get when I click on FreeTherm.exe:
FreeTherm.exe - Common Language Runtime Debugging Services
Application has generated an exception that could not be handled.
Process id=0x74c (1868), Thread id=0x1374 (4980).
Click OK to terminate the application.
Click CANCEL to debug the application.
OK Cancel

Trying to built it with GitBash and Windows Command Prompt gives me the same error.
Microsoft Windows [Version 6.1.7601]
Copyright (c) 2009 Microsoft Corporation. All rights reserved.
C:\Users\user>cd C:\Users\user\Desktop\FreeTherm
C:\Users\user\Desktop\FreeTherm>make FreeTherm.exe
mcs -target:winexe -resource:FreeTherm.glade.xml,FreeTherm.glade -resource:diy.p
hp.png,diy.php.png -pkg:gtk-sharp-2.0,glade-sharp-2.0 -out:FreeTherm.exe FreeThe
rm.cs
process_begin: CreateProcess(NULL, mcs -target:winexe -resource:FreeTherm.glade.
xml,FreeTherm.glade -resource:diy.php.png,diy.php.png -pkg:gtk-sharp-2.0,glade-s
harp-2.0 -out:FreeTherm.exe FreeTherm.cs, ...) failed.
make (e=2): The system cannot find the file specified.
make: *** [FreeTherm.exe] Error 2
C:\Users\user\Desktop\FreeTherm>

Re not building on 7, can you be more specific? I've built on 7 with the tools from the website.

Nevermind, it builds in gitbash, but it won't from the command window. It gets in some sort of a loop, and won't leave it. I've just been too lazy to type everything into gitbash until now. I'm also to lazy to go learn how to copy and paste into gitbash.

Inputting values directly will be a pain.

More of a pain than getting Windows to work properly?

Feel free to hit me up

I know, but I feel like it's going to be something stupid like dragging the boxes down to the dark area in OLV.

[Peter]

There must be something seriously wrong with the car's oem ecu. I drove it about 200miles(320km) this morning, and got 28.5mpg(45.7kmpg). Which is better than the best I ever got out of the oem ecu, and it runs better. The stickers and the wideband O2 sensor showed up today.