All the way along, speedway racers have used mechanically run fuel injection (PI) pumps (the pump was actually belt driven off the front pulley) which then pressure fed alcohol to the injectors. The mixture settings were a 'black art', changed by the fitting of larger or smaller pills. This system worked really well compared to carburettors for methanol guzzling flat out racing type applications but was as subtle as a fart in the state library when it came to daily driver applications.
Volkswagen has used (on some models more successfully than others) varying Bosch fuel injection systems
The early TLE T3 Volkswagens and 76 to 82 two litre air cool’ds ran the Bosch 'D' Jetronic and 'e' Jetronic respectively. The injectors tended to wear relatively quickly and the system ran lean causing overheating, premature engine wear and burnt valves. More than half of the models fitted with this model PI would now have been converted to carbies
The 1900 and 2.1 waterboxers ran Digijet and Digifant systems respectively; GTI Golfs—L Jetronic; VR6 Golfs— Motronic; Audis varied from KE Jetronic to L Jetronic systems.
As can be seen there were many and varied systems manufactured mainly by Bosch. However, they did have mixture trims but could not be drastically 'retuned' to take into account larger capacities, long duration cams etc and that means all the nice hot up work that we VW loonies do to our motors to get them to perform and goooooo!
The PI system that I am writing about here is the Haltech F3 Programmable Fuel Injection (HPPI). This system uses a very sophisticated programme that is accessed through any IBM compatible 386 or 486 laptop computer.
Don't think for a moment it is a job that you can do with a few mates on Saturday arvo' with a case of beer and a few tools. It definitely is not, but by the same token if you take you time, plan it out and get some correct advice from someone who has fitted one, the job is not very daunting. This is especially so now that PPI has been around for about six or so years and all the ancillary parts are pretty well available off the shelf or can be sourced second hand from a wrecker of late model cars.
The great advantage of PPI is the varying applications that it can be matched to, such as; Multiple throttle body direct port injection, single throttle body direct port injection, single throttle body single injector injection, single throttle body direct port turbo intercooled, multiple throttle body direct port supercharged, single throttle body direct port staged (second set of injectors phased in at say 1.0. bar boost ), turbo intercooled, single or multiple throttle body direct port, staged injector supercharged. The last two can be set up so that the second set of injectors can be programmed to come on at any rev point or pre-determined boost pressure.
Add to this the newer systems which allow for total engine management i.e.; timing function as well, you could very easily get that retard on your dizzy when the boost comes in, to stop all that pinging and detonation on a boosted engine system. Basically they can do any configuration engine! Full stop!
Once upon a time there were two VW psychopaths... In 1988 Joe
Smith and myself were racing the 'Bug out off Hell" race car under the Rogwin
Motorsport banner in the NSW Hill Climb Championship. As the racing fever got hotter and
hotter we kept upping the ante to try and catch the three litre Porsche of David Withers
(eternal optimism had us entered in the 2-3 litre Road Registered Class in a 1904 cc 'dub
(74 x 90.5). ( Funny, the 2161 cc (84 x 90.5) motor never did eventuate that year). We had
started off with 45 Dell's and standard valve ported Berg heads and ended up with a set of
48 IDA's (a friend had bought them back from the States and Henry Spicak had set them up
for us) and a set of Mark Walker 40 x 37.5 ported heads. The passion for speed was still
there as David ended up taking out the championship by seven points; Joe and myself
finishing third and second respectively. We must have made an impact coz' Porsches were no
longer allowed to run the Road Registered class only as Marque Sports Cars for 1989.
It was during the later half of'88 that a friend of a friend (Wayne Glasser) had started taking a fair bit of interest in our racing activities and was calling around a couple of days a week hassling us to try one of the new Haltech systems. I cannot remember what I had sold but the money from its sale went into the purchase of Haltech F3 unit number 25.This was early days for PPI and I hadn't yet come to grips with computers so the unit just sat there for a couple of months.
Besides 1988 being our Bicentenary Year it was also a pretty big year for us with all the racing and the prospect of strutting our stuff in the hot-mix lap dash battle at Valla Park in August. As it turned out, Jeff and Joe always headed up to Valla early so we could have a bit of a rest before the VW invasion took place.
Donna Pell had told us that we would have to meet Gene and Dee Berg at Nambucca Railway Station. We were the two free souls up here who had nothing else better to do. For the previous years all we had heard was "Berg this, Berg that" from Richard so we joked about being like Moses and having to go to the station and receive the Ten VW Commandments (For those of you who don't know Richard, he has the ability to talk a Tom Thumb bunger up to a Hiroshima H bomb blast. There's no disrespect there Richard but you had us feeling pretty awed out.)
We picked up Gene and Dee and had a great night playing carpet bowls in at the Valla Resort and talking big motors. We talked more after the practice session on the Friday and had soon ordered a set of 44 x 37.5 heads, 84 wedgemated crank and flywheel, Carrillos etc. Over the weekend I had told Gene about this new high tech HPPI system that we had bought. He was excited about the concept and agreed that if it did do what was claimed it would be a world first—he couldn't come to grips with the programme that would be needed to analyse all the different inputs from the sensors and then trigger the injectors.
As it turned out Gene had been playing around with both turbos and supercharging and mechanical fuel injection—the same old problem—lean out at elevated boost pressures. And, yes he did have a couple of what he termed 'crude throttle bodies' that would fit onto a 48 IDA manifold. Gene had always talked about his pile of discarded VW parts that hadn't come up to required standard—well it wasn't until I was over there two years later that I got to look at the pile—he must have just about tested every VW hot up part that was ever made, as well as the prototypes of his own that were never made! " Yeah, Sure ' Gene said," you can have them." The first parts on the shopping list found. You bloody beauty! Our friend Wayne Glasser took the Scat track inlet manifolds and managed to graft on a patch of alloy with his TIC, so we had a platform on which to mount a set of Camira fuel injector holders. We put the whole set up in the mill, bored a hole and there was our first set of direct port injection manifolds. Now it was time to fit up the injectors. "Well what size are you going to use"? He asked.
"How much horsepower are you going to make?" asked Wayne. " The old 1904 with 40/37.5 heads, K8 cam, 45 dual Dell's and 1-5/8 merged exhaust put out 132 HP on the Maztech dyno ' I replied. So this 2161 with 50mm throttle bodies, FK87 cam,1-3/4" race merged exhaust and so on would have to run at least 200hp." The calculations were done and a set of Bosch 0280 1500 34-060 injectors were fitted up to the now modified manifolds.
Due to the non-availability of such simple parts as fuel injector rails we had to put on the thinking caps to come up with simple solutions. The fuel rails were made using brass 'T' pieces inter connected with small lengths of PI hose from Tooleys. Not the prettiest fuel rail but it sure worked. The list went on, a real Hitchhikers Guide to the Retrofitting of PPI onto a Bug.
The throttle bodies that Gene had sent were void of any velocity stack or ram tube so they too had to be made.
It was now time to fit the whole lot up on the motor in the car. The fuel rails stuck out as far as a stallion would approaching a mare. So we pulled everything out after marking out the areas to be cut. A combination of jigsaw and die grinder seemed to work best and after three attempts and three hours everything seemed to fall into place. The only other problem was that so much dirt and rocks could get into the engine compartment so we had to devise a seal to go over the rail. Six self tappers and a hunk of inner tube were all that were needed. A phone call to Finer Filters had the air filter made up and dispatched in two days.
Now that all the engine compartment manifolding and rails were in place it was time to run the tank fuel line from the tank to the pressure regulator and return line. We opted to run the fuel injection hose tie strapped to the outer pan rails down each side. The fuel pump and filter were mounted on the front of the pan opposite the master cylinder. A larger fuel tank outlet (7mm) was made from a barbed fitting to match up with the PI hose. The return line was routed so it dumped into the petrol tank filler neck—only two late nights for this one!
We had been advised by the powers that be (Mark Boxsell from EFI Technologies ) to fit a fuel pressure gauge to help with the initial Haltech set up, so a 0-500 kpa VDO gauge was purchased and fitted on the interior of the rear firewall.
Next was the CPU (Central Processing Unit) which was mounted on a one millimetre thick aluminium sheet, acting as the drivers side rear trim. Rubber mountings were used to isolate the CPU from our race suspension. A mini electrical board was also fitted with fuses and push on terminals to facilitate the set up.
The fuel pump relays and associated wiring were run and hey presto! The only thing remaining was to run the loom through the firewall into the engine bay and connect the various sensors. |
Who were we trying to kid? Still more to do.
As the Haltech CPU was set up for MAP (manifold absolute pressure) as opposed to throttle position we drilled and tapped each inlet manifold runner about two inches under the butterfly shaft and fitted up four brass fittings that would take ordinary braided vacuum line. These hoses were then connected up to a fabricated metal vacuum box that would average out the vacuum pulses so a static single vac' signal could be fed into the MAP input of the CPU.
The mixture trims wiring was temporarily run from the CPU along the centre main loop of the roll bar using tie straps. This trim would then be used to do the initial set up of the fuel curves. When I say initial, the big plan was to fit in a Celica fuel curve, do the start up tuning by ear running through the rev ranges in neutral, 'til we had a clean running engine that was at least able to drive out of the workshop and onto the road. Once on the road with an oxygen sensor enema we could then actually get the motor tuned using the mobile exhaust gas analyser (EGA) in conjunction with the + or—10% mixture trim and laptop computer. At a later date a dyno-session was to be organised to maximise the fuel curves for performance and power.
Even though we were totally and utterly stuffed by this time adrenaline soon overtook tiredness as the motor fired up first go (after getting the initial fuel pressure by turning the ignition key on and off end listening to the fuel pump purging all the air from the system. To start off with you couldn't really rev the motor as there were load points at which the mixture was wrong (it may have been correct for the Celica but not this 2161 boxer with balls.) It was crying out for more fuel! With Wayne Glasser and Mark Boxsell hovering over the EGA and laptop it was hard for us two mortals to get even a look in as to what was being fiddled with.
Time out for the two Js while the two EFI crazies waved their magic mixture trim over the CPU. Sitting outside the workshop we were numbed by the feeling of accomplishment, that yes this had been the first HPPI fitted to a VW, our ears were ringing to the sound of the 1-7/8" merged with a Supertrapp, the whine of the straight gears and our noses were blessed with the aroma of BP100 wafting out from within the bowels of Rogwin.
Before we actually had time to get the caffeine buzz Wayne and Mark had gone through the revs up to 6000 so there were no apparent "no load" dead spots only the raw bark of a boxer on heat! It had only taken 15 minutes. It was now time for the first test flight.
The oxygen sensor wiring was taped to the bumper, whale tail, roof gutter and then into the cabin. Mark adjusted the seat belt tightly (perhaps he knew how fast we were actually going to be going or maybe had seen through my hell-bent eyes).
The same process as before was going to be used go though the various rev ranges in 1000 rpm intervals, get all the load points correct using the mixture trim to change the mixture which was monitored on the EGA. Piece of Piss.
The only major problem encountered was that it was completely useless using the first two gears as they disappeared into the next before you had time to make any adjustments! Third was chosen as it gave us a little more time. As the load points were tuned the mixture trim came into its own—the motor would be lead at say 4000 revs—mixture trim 10% richer—wambo—the car took off like I just hit a nitrous button!
Half an hour later the really fine tuning was in progress, third gear disappearing as quick as the full feeling after a Chinese meal. Top gear! More hassles here, we were starting to go too bloody fast. Poor Mark was stuffed into a standard 1500 low-back seat, had to juggle the EGA on his left knee, the laptop on the other to adjust the mixture trim and hit the update button on the laptop with Jeff "Rocketman Onion" Unwin driving from 2000 to 7500 rpm in top gear around the back streets of Taren Point. I think Mark must have had a set of "tear off" undies that night, coz the stress was starting to show!
By the time 50 minutes had elapsed, we were adjusting the accelerator pump functions of the Haltech, i.e.: length of time of injection and decay of pump stroke. In five minutes there were no more flat spots. s it turned out, we never did get to put that motor on the dyno. We ran the 2161 for about four months but had heaps of trouble with oil surge- the jump from Road Registered Class type tyres to Sports Sedan slicks, stiffer suspension, sway bars and much, much more power soon turned into a total nightmare. This motor suffered a terminal blow up while representing NSW at the Morwell Interstate Hillclimb Challenge—kicking two Carrillos out the top of an ARPM case, creating a five piece camshaft, destroying two pistons and liners and digging a three mil' gouge into the crank!
I can joke now but all was not lost as the pieces ended up becoming the dreaded "Bent Carrillo Trophy" the most feared award at the annual Hillclimb presentation night. If you "win" this one, you've had the best engine blow, bar none, for the year.
In between all this trauma we did have some success which the history books proudly show but space limitations here prevent us from reporting.
Over the preceding pages I have referred to load points, an explanation is needed. If you look at a carby, it has basically three load points i.e.; idle speed and mixture, progression circuitry for transition from pilot to main circuit and the main circuit itself. The 48-IDA is a very easy carte to tune but does not exhibit much smoothness as it has no progression; basically it should be used for idling in the pits or flat out racing around; not for driving on the street for which it was never designed. The greater the number of load points the better or smoother the engine will run. What we found was that the HPPI filled in the gap under which the cam kicked in, by virtue of its 512 load points. It's like a virtual reality thing—being sucked into feeling that the motor isn't going that hard whereas it has so much more bottom end that the cam step doesn't appear to be that big at all.
I'll never forget going out to Kurnell in the wet one night with the Bug Out of Hell having 185/70/14 tyres on and being able to squeeze the throttle at 80kmh in top gear and having both tyres light up and keep spinning through to an indicated 120kays. That’s drivability at a rev range that would bog down with carbes and a similar camshaft.
In conclusion there is no doubt PPI is here, this is 1994 and we are heading towards 2000 so if you want to upgrade and update your ride so it is as smooth as all the other 1994 PI cars give PPI a really big sussout before purchasing a bigger set of carbes.