Something I haven’t mentioned on here yet is the fact Schnellwagon didn’t even make it a month before breaking down. Thankfully it was simple to resolve.
I took the car to work a couple of weeks ago, and everything was going fine and dandy until about a block from home, and BAM, as soon as I came on boost the engine falls flat on its face, cuts power and drops a cylinder. I carefully nursed the car home, and although it was running badly, I noticed I didn’t have a Check Engine Light.
I shut the engine off, restarted, and everything was fine again. The engine was running fine, so I took it for another quick drive. Once again, as soon as I came on boost, the engine cuts power and drops a cylinder. Something wasn’t happy.
29DC – DME: Cylinder Injection Cutout and a shadow code for 2AAF – DME: Fuel Pump, Plausibility. Googling around, and these seem to indicate either the High (HPFP) or Low (LPFP) Pressure Fuel Pumps causing issues. The previous owner of the car claimed the HPFP was replaced but the workshop wouldn’t release the receipts to him (due to privacy) and it does look quite new, but the previous owner also replaced the LPFP with an aftermarket AEM pump so he could run ethanol fuel (Ethanol blends need more fuel than standard petrol so needs a higher output fuel pump, but ethanol makes more power than regular petrol so people like to use it on high power engines).
I did some data logging using MHD, and it became crystal clear what the issue was
The scale is a bit whack because of the huge difference the High Pressure fuel makes, going up to over 2000psi. The pink line is the output from the HPFP. It takes a MASSIVE dip from almost 2400psi, down to 304psi when under boost. Not good at all.
As boost (orange line) goes up, fuel pressure (blue line) goes down until it hits about 27psi (the LPFP should be pushing much more than that, although not as much as the HPFP). Even at idle or with the engine off, I’m only seeing 47 odd PSI.
So what does that tell me? Well, it indicates that the HPFP is OK, but what’s happening is that the LPFP is starving the HPFP under load/boost. The HPFP needs more fuel to keep up, and the LPFP cant supply it, it goes lean, and the engine goes into limp mode to save itself. This shouldn’t be happening with a reasonably new high output AEM pump.
It was time to whip the pump out and see what was happening. Thankfully the previous owner retained the old fuel pump, so my plan was to remove the AEM and fit the original pump as I wont be running Ethanol as its not readily available locally, and already make enough power.
Access to the pump is very easy. It’s under the back seat, and the base cushion is only held in with a clip at the front on either side of the car. I placed a hand under the corner of the cushion, and lifted sharply to disengage the clip. With both clips disengaged, the base can be carefully removed from the car.
These hoses have a clip on the end that needs to be pressed to disengage it. The blue bit on the hose needs to be pushed inwards to disengage it, whilst pulling up on the hose. I used the pick to press in the blue clip.
The fuel pump is secured with a large steel ring which needs to be removed. There are tools to remove it, but not having them, I (and the previous owner) used a large flat blade and a hammer to tap it around and remove it.
With the ring removed, the pump can be removed. I will recommend though, that you don’t do this job with a FULL tank like I did. It was bloody messy and smelly.
So yes the pump was indeed modified. This is a fairly common modification, but none of this was off the shelf, all custom made by the previous owner.
These run an unusual setup compared to typical 90s JDM cars that I’m used to (which just dangle the pump from the top of the pump, straight into the tank). The pump sits inside a “bucket” which acts like a surge tank, keeping the pump submerged in fuel. The metal pipe in the above photo is the output from the pump. This is also a T-piece, which also goes back down into the bucket and feeds fuel to the Venturi in the bottom of the bucket. The Venturi uses fuel from the pump, to suck more fuel into the bucket via a one-way flap in the bottom of the bucket, so even if the tank level is low, the bucket is always full of fuel.
The first issue I encountered, was that the hose that connects from the pump to the Venturi was disconnected at the Venturi. The hose had gone soft and was starting to perish. Obviously this bit of hose wasn’t suitable for submerging in fuel.
In theory if this isn’t connected to the Venturi, a large amount of fuel from the output of the pump is just being pumped straight back into the bucket. This still shouldn’t be a huge issue though; this pump should be able to output enough fuel to keep the HPFP supplied, and with the full tank of gas, filling the bucket up isn’t an issue.
The pump had slipped in its zip tie mount, and the inlet was sitting on the bottom of the bucket. This restricted the inlet of the pump to only a small section of the filter, which also happened to be blocked with gunk. Considering how clean the tank is inside, I don’t know where this gunk came from.
Now it was a case of installing the standard pump and putting it all back in again.
To fit the AEM pump the previous owner modified a few things. One of them was the Venturi, by removing the clip that holds it to the standard pump. This clip also holds the filter to the pump, so both are now just attached to the pump by friction. The Venturi is the weird thing with tubes on it, in the top of the photo.
I later found out that the zip ties do not fit into the tank when the head is on the outside of the bucket like that, so I had to remove them all and put all the zip ties back in with the heads inside the bucket
Once everything was all back together the pump went back into the tank, and it was time to see if I was right, or if the HPFP was actually the fault.
I logged the pressures, and as soon as the pump primed I immediately saw 89psi. Almost twice what the AEM pump was putting out, and that’s even before the engine was started.
A quick drive showed that the car was running better already, and a hesitation when cold was gone now. Once warm, it was time to turn on the logging and hit some boost.
Sure enough, it ran and boosted like a dream. No cutting out, and no misfire.
At the same level of boost, there was no longer a dip in the HPFP output, and I had over 2600psi instead of the 304psi previously. The HPFP was happy.
The HPFP was no longer being starved of fuel, and everything was happy again. Both of the codes have not returned since being cleared.
There was one little catch. I didn’t like that everything was held together with zip ties, and I didn’t know if my old zip ties were going to last in fuel, so I needed to get a standard replacement and swap it in.
I pulled the bodged pump out today, and swapped the replacement in. I made sure the tank was almost empty this time, and it made for a much cleaner, and nicer job. Unfortunately I broke the outlet off the bodged standard pump when trying to remove the hose, so that’s only fit for the bin now. That hose is on tight, and there isn’t a lot of room to press the clip, and pull the hose up.
I still have awesome fuel pressure on both high and low rails, and everything is running as it should. I’m bloody happy it wasn’t the HPFP at fault, they aren’t cheap.