Board Logo
Go To Bottom

Printable Version  
[ Total Views: 135480 | Total Replies: 190 | Thread Id: 27098 ]
 Pages:  1  ..  5  6  7
Author: Subject: Aftermarket EFI - cont'd
Membertassupervee
A.k.a.: Knob Jockey
Custom Title Time!
That really hurts doesnt it!
*****


Avatar


Posts: 1171
Threads: 25
Registered: September 21st, 2003
Member Is Offline

Location: Huon Valley SouthernTasmania
Theme: UltimaBB Psyche Blue
Mood: Moody!

posted on February 16th, 2014 at 11:31 AM



Hey Ben
Nice to see a totally home grown setup.
Single point (TB) injection is always a nice thought, especially on an x-flow engine with short inlet runners but on a Kraut boxer, it becomes more or less an "electric" carby with all the annoying issues that beset the carb setup such as fuel condensation and separation in the pipework blah.
Im sure this is what has prompted you to go to multi-point, straight onto the back of the valve injection.

As a matter of interest, have you grouped your injectors in banks, sequential or batch fired them?

I see you have cast your own TP manifolds which is a bloody nice step! A simple pair of cable pulled 40mm TB's on top of those and your grinning very widely. No need for funky and expensive linkage setups designed for clunky old carbs.

I love this kind of imagination and initiative. Well done M8.

E




Im not a complete idiot, quite a few parts are missing....
MemberBennVenn
Casual Dubber
*


Avatar


Posts: 26
Threads: 1
Registered: April 13th, 2012
Member Is Offline

Location: Windellama, Canberra
Theme: UltimaBB Pro Blue ( Default )

posted on May 9th, 2014 at 10:03 PM



Hey tassupervee,

I'll be going with full sequential injection. I'm re-designing the whole setup, yet again. This time i'll be using a RB20 Cam Angle Sensor which provides both 1degree pulses and 90degree coded position markers. Having this kind of precision, fed into an FPGA for direct hardware control of Dwell, Ignition timing, Injection timing as well as injection pulse delivery will off load all the timing restrains from the micro controller, giving it plenty of time for accurate MAP based fueling, self learning, etc. Having an FPGA control all the hardware gives me as much IO & PWM channels as i want, will allow a complete re-flash of the MCU while the motor is running, have lots of nice features like precise control over when the injection will occur and not have a trigger wheel and VR sensor to fiddle with when things start to go wrong. Reliability is priority with this build.

I'll also be removing the doghouse cooler and fan and going with a custom electric cooling. I've read this simply won't work, as it doesn't have the 'pressure' to cool the cylinders and head effectively. I think it will, if it can pull the rated flow through a radiator core with 1mm gap between fins, it'll push through a VW head with 5mm gapped fins. I'll be fitting a large after market oil cooler through the firewall and mount it in an enclosure behind the rear seats, dumping the hot air under the car or inside the cabin in winter. And as the oil cooling has increased, a secondary pump will provide bulk oil to the heads for additional cooling, as well as feed the turbo when it gets fitted. I've seen some nice variable geometry turbo's coming out of china recently...

I'll add a new page with this new build shortly to my website. I'd love to hear from anyone who has fitted a cooling system other than the standard!
Membertassupervee
A.k.a.: Knob Jockey
Custom Title Time!
That really hurts doesnt it!
*****


Avatar


Posts: 1171
Threads: 25
Registered: September 21st, 2003
Member Is Offline

Location: Huon Valley SouthernTasmania
Theme: UltimaBB Psyche Blue
Mood: Moody!

posted on May 10th, 2014 at 09:15 PM



RB20 is a 6 cylinder so trigger slots are at 120 degrees (crank) So not too much use to you.
Im guessing either SR20 or CA18 will be what you need!! Thing is by using tha kind of digital CAS, and account for crank to cam and cam to dissie gear trains, you introduce timing scatter issues associted with cam driven CAS's which kind of cancel out the precision your after by using such a high count rate on the synch timer to begin with.
Nissan is pretty peculiar with using such a high resolution synch timer and the rest of the automotive world seems perfectly happy with far lower timer counts.

After all, its only a reference timer. System triggering is still a function of the 4 trigger slots and their accuracy along with scatter from gear trains!!!! Do you think you might be over thinking this a little??

For simplicity, why not just stick with a simple digital 60-2 or 120-2 direct off the crank. Or if you want to be real pedantic, use the starter ring gear as a synch timer (assuming the tooth count divides correctly) and a single crank trigger slot on the front pulley.

Im using a dissie out of an E15ET in mine with the timing wheel flipped due to the opposite rotation direction of the Kraut dissie. I dont really care for absolute precision!!!! Function over form! Actually, the old EMS system im running does not have provisions for sequential triggering so i have to use a dissie anyway but it keeps everything neat and tidy!!!!!
Otherwise a simple 60-2 wheel and a 2 bob Ebay Hall trigger would be more than adequate for a zillion rpm and HP.

You can't really compare the air and water cooling systems in such a simplified way with regards to an electric fan.
Heat transfer from metal to water and back again is far in excess of what metal to air can achieve and the radiating area and volume (heat sinking) of a typical radiator/watercooling system is such that even a small electric fan can move sufficient air keep a system cool.
The air cooled setup requires CFM, and a fair bit of it to transfer heat off the heads. I get the addition of the oil cooling setup but...external, enclosed coolers, fans, an oil circulation pumps, oil squirters in the heads pipes and hoses..... shit man, its getting awfully scientific just to say "I did it" (with electric only). Its really kind of a mod purely for mods sake which is largely pointless.
The Porche style shrouds with a belt driven fan are perfectly adequate for the job. Simple and effective and look cool!!! (No pun)!

Just me tho!!

E




Im not a complete idiot, quite a few parts are missing....
MemberCamo
A.k.a.: Kev
Custom Title Time!
*****


Avatar


Posts: 1936
Threads: 209
Registered: October 29th, 2007
Member Is Offline

Location: Sydney - Liverpool area
Theme: UltimaBB Pro Blue

posted on May 11th, 2014 at 08:08 AM



Sorry if off track here as I think you are trying to build this stuff yourself, but what about this setup. States this is for sequential injection.

http://www.cbperformance.com/ProductDetails.asp?ProductCode=7074

and a little bit of a writeup here

http://www.cbperformance.com/Articles.asp?ID=252

Could make life easier :lol:

Kev




I thought rebuilding a VW would be cheap - Shit I was wrong, just don't let the wife know !!!!

Click here for my (slowy) modded 1961 Beetle build.
Membertassupervee
A.k.a.: Knob Jockey
Custom Title Time!
That really hurts doesnt it!
*****


Avatar


Posts: 1171
Threads: 25
Registered: September 21st, 2003
Member Is Offline

Location: Huon Valley SouthernTasmania
Theme: UltimaBB Psyche Blue
Mood: Moody!

posted on May 11th, 2014 at 09:22 AM



Kev
Hes (Ben) going a totally home grown kit using his own ecu and coding.
There is a legion of aftermarket goodies around to make what is as close as can be to a plug and play efi setup. I get where Ben is coming from. Hes making it a project, rather than just a bolt on.

Thats a nice bit of clbber there but it still can suffer from what i have mentioned above if REALLY chasing the ultimate in system resolution. And high resolution is what b
Ben is shooting for by specifying the goodies he has.
Somthing that i dont think is really worthwhile when relying on a distributer drive for accuracy. Crank triggers FTW!!!!

Cheers
Evan




Im not a complete idiot, quite a few parts are missing....
MemberBennVenn
Casual Dubber
*


Avatar


Posts: 26
Threads: 1
Registered: April 13th, 2012
Member Is Offline

Location: Windellama, Canberra
Theme: UltimaBB Pro Blue ( Default )

posted on May 11th, 2014 at 10:08 AM



Ahh yes it is a CA18DET CAS. Which strangely enough looks exactly like my RB25 CAS... Either way it wouldn't matter as the trigger slots are just there to keep the 720 degree counter in sync. There is no actual triggering done off those slots.

Over thinking? of course! Its a hobby/project and i'm more interested in the technical challenge than what is 'convenient'. Speaking of convenient, CBperf has some cool toys, but $100 for what is essentially a 009 distributor with one cam lobe??? I was originally going to add a hall sensor set-up to the fuel pump push-rod. It is cam driven too and that will satisfy the sequential setup with my Trigger wheel. And while my trigger wheel setup worked great, there would always be a point where accuracy would drop as the RPM increases. As the RPM goes up, the micro controllers pulse-width count decreases, along with its resolution. This could be solved with using a 36/48bit resolution counter running at a ridiculous clock speed, but then the complexity and delays start to introduce High RPM errors too...

Lots to think about... Maybe i should just use my working trigger wheel set-up. Like you say, good for a zillion rpm.

As for the cooling, I really like the look of the VW engine. At the moment all you see is fan shroud. And while it does what it does, very well, it is very inefficient. My RB25's fan clutch is stuck at the moment. It makes an enormous difference in power output of the motor. Just off idle, there is no boost, the engine is working hard against the exhaust back pressure trying to spool the turbo, and engine power is more or less linear with RPM. Having the clutch stuck sucks valuable power from the motor when it needs the HP the most. It is the same for the VW. Ever started and revved your VW without its alternator/generator belt on? A completely different engine! That is what i'm trying to achieve with the electric fan setup. If supplementing the cooling with oil squirter's will do the job, it seems like very little outlay for big gains. As you said, Metal-Liquid heat transfer is far more effective than Metal-Air, and 80% of the heat is generated in the heads, i cant see why forced oil cooling wouldn't be more effective and efficient then forced air.

With the end goal of running ~12psi boost, i need as much cooling as possible where it counts, as well as a supplementary oil system to feed the turbo after the engine as been switched off.

Its good to talk with someone as enthusiastic as myself with this sort of stuff. Do you have any pics of your build?
Membertassupervee
A.k.a.: Knob Jockey
Custom Title Time!
That really hurts doesnt it!
*****


Avatar


Posts: 1171
Threads: 25
Registered: September 21st, 2003
Member Is Offline

Location: Huon Valley SouthernTasmania
Theme: UltimaBB Psyche Blue
Mood: Moody!

posted on May 11th, 2014 at 01:57 PM



The RB/CA/VG30de/t/tt cas's all look pretty much the same externally.
No denying the power loss associated with an engine driven fan but im still predicting trouble keeping temps in order running just an electric fan.
Hah, in our Formula Vees, we run the belt as slack as possibl just so the belt stays on but slips as much as possible!
The oil squirting might certainly help but will there be sufficient area of ally inside the rocker box to make a lot of difference??
I reckon machine down the cooling fins somewhat, box it up and pass oil thru there as the cooling medium
Turbo oiling after shutdown? Might have been relevant in the 80's but times have changed and turbo tech improved and after shutdown turbo oiling is pointless really. No reason not to but again added complexity.......

Sry M8 I dont have any pix. Its an ongoing project that stalled a few year back. Its a ghetto setup using hastily modified, T3 injector housings. They use a typical Type 5 Bosch hosetail injector and Im using a set of 240cc injectors right off an E15et.
The dissie is stock E15et other than I drilled the shaft and added the drive foot off the stock VW dissie.
An old T02 off the E15et remotely mounted on a bracket on top of the gearbox and a T-piece off the oil pressure sender for oiling. Drain back to the rocker cover.
Its just a shits and giggles thing.Ill get back to it eventually.

Evan




Im not a complete idiot, quite a few parts are missing....
MemberBennVenn
Casual Dubber
*


Avatar


Posts: 26
Threads: 1
Registered: April 13th, 2012
Member Is Offline

Location: Windellama, Canberra
Theme: UltimaBB Pro Blue ( Default )

posted on May 11th, 2014 at 02:18 PM



A guy I went to school with races Formula Vee, wonder if you know him. Adam Brook? I know my RB25DET has squirters under the pistons, and they have as little surface area as you can get (flat surface). Strange you mention it, I was looking at the heads yesterday thinking about welding them up and running oil through them. Also looking at how little the cooling fins on the cylinders are, compared to the heads. Definitely something to look into.

As for the oiling after shutdown, modern cars have water cooled turbos, and are positioned in such a way that a thermal syphon kind of pump keeps water circulating. A VW without a water system doesn't have that luxury. My 2014 Turbo diesel SUV has a secondary water pump for running after shutdown if the turbo is too hot. I believe Mini a few years ago recalled a heap of cars due to the shutdown coolant pump failing and causing engine fires after walking away from the car. Oil is still oil, and oil at 400degrees will coke regardless.

Is there enough room on top of the gearbox for a turbo? is this all in a stock beetle??

You've convinced me to do away with the CAS. Added complexity to a system i know works with just the trigger wheel...

Check out http://www.bennvenn.com/NewBrain/NewBrain.htm  I'm still writing it up but its a start...
Membertassupervee
A.k.a.: Knob Jockey
Custom Title Time!
That really hurts doesnt it!
*****


Avatar


Posts: 1171
Threads: 25
Registered: September 21st, 2003
Member Is Offline

Location: Huon Valley SouthernTasmania
Theme: UltimaBB Psyche Blue
Mood: Moody!

posted on May 11th, 2014 at 07:18 PM



Ben
Ther turbo issue just isnt a drama to begin with.
Yep lots of watercooled OEM turbos setups employ thermosyphoning but purely to prevent boiling of the coolant in the turbo core. Now think about every big truck engine. Apart from some very recent Euro engines, they universally use non watercooled turbos and diseasels run consistently higher exhaust gas temps for longer periods of time. What about every other Hi-po non watercooled turbo arrangement. They live long long time just fine!!!!
It just isnt a drama. If your really pedantic, a turbo oiler setup, pretty much a small accusump can be employed to run a bit of oil thru the turbo...... nobody uses them anymore but you can use one if you wanted.
None of my many turbo cars has ever coked up a turbo due to oil burning on the hot side and this is all your bothered about. Modenr oil tolerate the heat pretty well without burning.

Now, the turbo engine I have set up is for the Formula Vee as a shits and giggles thing just to play around with at club days and perhaps do a hillclimb or two!! All the room in the world since my gearbox has been flicked and reversed!!!! On eof the issues is finding a fast enough diff ratio to use.
Im predicting around 120Hp but the 1600 Im using is bone stock wont see much more than about 4500-5000 rpm I predict it will hit those revs very quickly on a 4.1 diff @ 450 kilos combined so i have to get on top of that. Fried tyres anybody!!!!!

Adam Brook... yes the name is familiar. Dont know him personally. Im down in southern Tassie.
Ill check your link.
Cheers
E




Im not a complete idiot, quite a few parts are missing....
MemberBennVenn
Casual Dubber
*


Avatar


Posts: 26
Threads: 1
Registered: April 13th, 2012
Member Is Offline

Location: Windellama, Canberra
Theme: UltimaBB Pro Blue ( Default )

posted on May 12th, 2014 at 08:38 PM



I guess i've just had bad luck with turbo's. This is the 4th in my R33 since owning it. The stock rear seal failed, upon inspection there was a LOT of coke all through the bearings/bearing housing. The replacement was an aftermarket KKR, This lasted around 350kms before the rear seal failing. It was replaced under warranty, with an additional oil restriction fitted in the feed line. This turbo lasted around 20thousand km's before the front oil seal failed, filling my intercooler with oil. And now im onto my fourth, 100k kms later and its still going. Only the original and the current are water cooled. The manifold and turbo can be seen glowing on the dyno, indicating well in excess of 700deg C, so i have no doubt that shutting down with a hot turbo will cause failure...

Question on fitting an external cooler, i was planning on removing the stock cooler and fitting a manifold there to send the oil elsewhere. Further reading suggests leaving the stock cooler there and getting a feed from the pump itself. Any ideas??

Is it just a matter of reversing the diff centre to give you 4 reverse gears and 1 forward?
Membertassupervee
A.k.a.: Knob Jockey
Custom Title Time!
That really hurts doesnt it!
*****


Avatar


Posts: 1171
Threads: 25
Registered: September 21st, 2003
Member Is Offline

Location: Huon Valley SouthernTasmania
Theme: UltimaBB Psyche Blue
Mood: Moody!

posted on May 15th, 2014 at 09:24 PM



I think you have just been unlucky with turbos.

With out vees, the stock cooler ports are blocked off completely and a full flow setup off the pump.

Yep just pull the diff out and flick it. Single sideplate bozes might be a challenge but ive never investigated that.

E




Im not a complete idiot, quite a few parts are missing....
 Pages:  1  ..  5  6  7


  Go To Top


Powered by GaiaBB, © 2011 The GaiaBB Group


[ Queries: 40 ] [ PHP: 19.5% - SQL: 80.5% ]