#anyone running a standalone? (haltech, ms3, ect.)

1 messages · Page 1 of 1 (latest)

torn flame
#

I was wondering what kind of standalone u guys are running and how difficult wiring was for the option you went with.

hidden eagle
#

Many aftermarket options come with adapters or straight up plugin options. I'd advise against MS3 unless you can find it cheap, it has been pretty much abandoned and the price has risen up there with better options making it poor choice. I'm personally likely going with EMU Black. Versatile, but affordable. Also there are PnP adapters available for it.

lost spear
#

Why though?

#

Carberry open source works great

torn flame
#

Some context, Im planning to swap an turbo EZ30R into a 96 Impreza with the only real reason being bc I want to. I could merge ecu connectors and what not but alot would need to be replaced due to CANBUS of the EZ30R ecu. My only real options would be to patch the ez30r rom with boost control and probably disable some features (a ton of time/ I dont think its been attempted) or just do a standalone. Ive worked with low level programming and tuning quite a bit so I thought it would be fun to try and implement a diy type standalone since Ill have quite a bit more time and money in the near future. I agree that MS3 is pretty dated, Speeduino is another option developing fairly quickly, it would most likely require a ton of custom code and PCB design

lost spear
torn flame
#

Went down the Speeduino rabbit hole today. Really interesting project. The current firmware supports the Subaru H6 36-2-2-2. The only real issue is that the current, official PCB only supports 4 ignition and 4 injector outputs. The manual says it does support 6 cylinder engines but I would need to do a wasted spark and semi-sequential fuel set up. A little less than ideal, but from my knowledge the gains/losses from a set up would be pretty negligible. There are some PCBs out there for microcontrollers such as the teensy 3.5 and STM32 which support up to 8 injector and spark outputs, my only concern is that it appears they have not been tested as thorough

unreal acorn
#

you want 'Six Swapped Subarus' on facebook or discord for more information. also hi... I have swapped a 30R into a '96 Impreza... and currently daily a 30R swapped hawkeye.

#

All the wiring options are roughly the same level of difficulty for that swap. I'd rate it a 7/10. Link and Haltech are the two brands that support the 30R the best, with Link being a clear frontrunner.

#

in terms of Microcontrollers, I can only confirm that the Link is Freescale based running at 180MHz. The add-on modules I develop are STM32 based running at between 80 and 230MHz.

torn flame
# unreal acorn Dual VVT (aka Subaru) is not supported by the Speeduino yet. SpeedyEFI are worki...

https://wiki.speeduino.com/en/configuration/VVT
I thought the current v0.4 board supported up to 2? I noticed it says that it can only do closed loop control for one camshaft but open loop for both. I believe Subaru uses a closed loop so I will need to research how that could work. I'm open to other platforms ig I'm just really interested to see if I could develop a PNP for this specific swap.

I'll definitely have to check out the groups for more information. Concerning you swaps, have you been able to use the stock 5 speeds or had to swap for the H6? I'm also interested to learn more about the modules you've developed with STM32 controllers

Speeduino Manual
VVT
hidden eagle
#

Subaru uses closed loop as VVT. Unless you're up to building a working one for your application it's not going to happen. You need to monitor both side intake cam angles and adjust them separately.

torn flame
# hidden eagle Subaru uses closed loop as VVT. Unless you're up to building a working one for y...

Unless I grossly misunderstand VVT, its simply a PWM controller similar to Subarus wastegate control? Fully Closed being 0% duty cycle while 100% being fully open or something comparable. Closed loop should just be a PID controller to add or remove PWM duty cycle to reach the goal position. IF a proper scaling of the VVT controller angle to duty cycle could be derived I assume that an open loop solution could work very well unless angle is also affected by fluctuating oil pressures.

Again either way a good amount of effort would be required either way but I think that's just Speeduino as a whole. I was already planning an FPGA based DBW controller so I could also look into VVT control to see what the functional requirements would be

#

I am interested in developing hardware and software solutions for this project, the list keeps getting bigger on what I will need to develop vrs already supported features. Something like a link standalone would definitely be alot easier

unreal acorn
# torn flame Unless I grossly misunderstand VVT, its simply a PWM controller similar to Subar...

You're taking what is already a massive job, and trying to do it on hardcore permadeath mode. FPGA based DBW controller? Come on... That's using a sledgehammer to hang photos.

Also the VVT actuators are hydraulic spool valves. The PWM percentage basically represents:
0-~50%: retard
~50%: hold
50-100%: advance.

The actual response rate and hold duty cycle varies depending on the wear level of the engine, the specific actuator, and the bleedback rate of the pulley.

They MUST have two separate PID functions.

unreal acorn
torn flame
# unreal acorn You're taking what is already a massive job, and trying to do it on hardcore per...

AH ok that makes sense, I just finished reading through a patent for VVT and had a similar feedback approach. I assume inputs would include oil pressure, crank speed, and spool angle. Good news is I just messaged one of the Speeduino developers and one of the new releases includes dual VVT with separate PID controllers allowing for an easy solution of that issue (granted I can get the code working properly). as far as the FPGA DBW controller, I need that 100ns throttle response 🤑 . I understand its highly unnecessary, I enjoy FPGA development so it was more of fairly simple problem that requires a very stable solution I though would be interesting to implement with an FPGA. probably could just use a high current output from the Speeduino if I go that route.

unreal acorn
#

Also Isaac is not an unknown quantity... That release is at least a year away.

#

Go with what's known and well understood, or expect failure. This swap has a 70-90% rate of failure.

#

Largely due to people either not realising they'd bitten off way more than they could chew, or budget blowout.

#

When you're working on your own project you have a problem: you're the financier, the project manager, the lead engineer, and the end customer. Four roles that should literally never be the same person.

#

DBW also uses a H bridge driver, can't just use a high current output. Motor is driven at 700Hz PWM or thereabouts. Two PID loops, one for drive open, one for drive close. Can't use the same for both as the resting position is about 2.5% open and the spring rates in both directions are different.

#

Also uses dual feedback on both the pedal and the actuator. Logic routines need to have a "fast fail" function where if ANY attribute goes out of spec, power is cut to the throttle motor and a fault condition is raised.

torn flame
torn flame
unreal acorn
#

Link or Haltech.

#

Haltech is more plug and play. Link is more flexible.

#

If you want to build your own add-on modules, Link is the only real option. Has a fully configurable CAN bus that will interact fully with the ECU functions

torn flame
#

was trying to keep the whole build under 10-15k for now but can spend more if Im going to be working on this long term

unreal acorn
#

10-15k is about mid range for the total cost of an NA 30R swap

#

Should be doable. I'm running a R180+Legacy 6MT combo. I started with an auto. Last time around I ran a standard R160+5MT setup.

#

Can usually pick up a G4X Xtreme (Xtreme for road cars, Fury for race cars as they don't preserve all functions) used for like US$1500. Or new about 2500. There's a good 20-50hrs of wiring and about 15-20 of fab work needed.

#

Most of the extra cost is either in things you can't do yourself or extras like fuel pump, radiator, fans, exhaust etc.

torn flame
#

Also what fabwork usually needs to be done? other than the exhaust I thought it was mostly drop in

unreal acorn
#

Probably 2-3k more unless you want to open the engine.

unreal acorn
#

Lots of little shit that adds up.

#

Extra cost for turbo is mostly in the turbo, the intercooler, and the raw materials.

#

When you buy a standalone you also need the appropriate wideband, in this case a Link CAN-Lambda.

torn flame
unreal acorn
#

The cooling issues mostly come from people using underpowered fans.

#

The stock cooling setup has two fans with a total power of around 720W.

#

A lot of people try to use 80-100w fans which simply don't have enough static pressure to overcome the air-dam effect of that massive timing cover. I'm using a BE/BH 3.0 rad and fans, and have no cooling issues whatsoever, even on long slow climbs.

#

The BL/BP rad is super thin, and the fans for it are also both incompatible with the drive system used by most standalones (PWM variable speed) and the stock wiring in most cars.

#

Have you done much automotive wiring? There's substantially different design rules that should be followed for good results

#

In general, the recommendation is: M22759/32 or /34 wire, and no soldering, open barrel (V or U) crimps only.

#

Deutsch or Amphenol DT/DTM or AT/ATM connectors with solid terminals are great where you need to add connections. Use the Iwiss crimp tool instead of forking out $600-1200 for either a DMC or genuine tool tho.

torn flame
#

Ive done some pretty straight forward gauge wiring and had to rewire my hatch but nothing extensive. I know a guy who does alot of first responder radio installs so im sure he has the required tools. Ill be sure to stock up on those connectors and get the correct wiring

#

I was thinking of pulling a stock engine harness to have all the OEM sensor connectors and possibly make wiring a bit easier to merge

unreal acorn
#

You will use the stock block harness yes. And you'll want a bl/bp bulkhead harness for parts.

#

BE/BH rad and fans is the best combo

lost spear
#

Btw you might wanna talk to Brian at iwire, he's got experience running 6cyls in non-h6 chassis so i'm sure he'd have the best recommendations

#

also just a really good dude/team there at iwire

torn flame
lost spear
#

and I bet he can probably make you whatever wiring harnesses you might need for your project

unreal acorn
torn flame
#

I also just read that on their website😔

#

I saw they also switched that EZ36 swap back to the stock STI powertrain

lost spear
#

aww

sacred kite
#

The 2nd gen rad+fans

unreal acorn
sacred kite
#

Would aftermarket rad be a consideration as well? I.e look for something more effective for the size

unreal acorn
#

BE/BH radiator, either very powerful slim pull fans, if you can even fit two, or equally powerful pusher fans and a shroud between the rad and the AC condenser

sacred kite
#

Push is typically better than pull right?

hidden eagle
#

With a proper shroud I'd say pull is better, because you have bigger intake area than exit area. Shrouded push is the other way around

unreal acorn
sacred kite
#

I totally forgot about the fact the rad gets direct airflow without fans, Im used to PC rads

unreal acorn
#

pull is preferable with a PC too

sacred kite
#

Not really

unreal acorn
#

just nobody does it because its not as pretty

sacred kite
#

Push/Pull is the best, but pull alone is worse than push

unreal acorn
#

on what grounds?

#

this is off topic for this conversation in any case.

sacred kite
#

I could have sworn I had some data lying around, but the only thing I can find offhand is Push vs Pushpull

covert rover
#

I am kind of late to this discussion but I run a Haltech elite 2000 on my bugeye wrx with their Plug n Play adapter. the 2000 will only work on cable throttle cars as the 2500 and nexus series are the only ones with DBW capabilities. I'm running staged injection with the radium dual injection rails and housings (two 1050x injectors per cylinder); It helps to maintain a 14.7 idle afr and also allows for ~2000cc maximum. <-- this is only possible with a standalone and is really the only reason i would switch to one. Other things that change are Idle tunability and quality, Motorsports features such as antilag, launch control, boost by gear, NLS, custom pwm output mapping, the ability to run hybrid maf/SD fueling tables, map switching on the fly, E85 without having to run an adapter box, lap timing/0-60 etc..., the ability to run several aftermarket CAN enabled dashes (powertune digital, cd7, ic7 etc...), The can run wideband AFR sensors natively, you can cut down on gauges by wiring sensors directly, etc....

unreal acorn