Slow Car Fast
July 24, 2011 - It wasn't all fireballs and oversteer yesterday.
Janel spent most of the day running around town in the little car, and seems to have quite enjoyed it. The car still has a tendency to pop out of reverse (and will continue to do so until I pull the transmission to fix it) so she planned her stops as pull-through only.
Until we were driving home from the track. I was leading the way in the Targa Miata, enjoying the view of the little tiny black car in the rear view mirror when she suddenly pulled over the side of the road about a half mile from the house. The mysterious fuel problem had returned.
Checking the fuel pressure revealed the problem - 20 psi. It's supposed to be about three times that. Between the inoperative fuel pressure gauge and the spill earlier today, I (once again) decided the best thing to do was to fill up the tank. And this time, it worked. Instant 60 psi pressure and the car came to life after a couple of tries. We took it home, then ran around town some more with the car at night without any problems.
I'd really like to know more about what's inside that fuel tank. No baffling, that much is clear. But where is the pickup? Janel described the car as losing power after a right turn, which would have sloshed fuel away from the side of the tank that has the pickup. And it just didn't come back all the way. If the extra fuel in the tank hadn't fixed the problem, I was going to see if I could bleed off any air bubbles in the system. But that may have to wait until the car fails again.
As a possibly related problem, the car sometimes doesn't start quite right. It'll pop once and then just crank. Cycling the key brings it back to life. I really want to know what's in that tank, as everything else in the system is something I'd consider to be a known quantity.
The car really does look cool out on the streets, especially from a distance. It gets a huge amount of attention. And Janel quite enjoys the effect of the accelerator. A hot-rodded vintage pickup passed us in traffic, and I thought Janel was going to rear-end it when things cleared. She was quite disappointed that his big fat tires were mostly for cruising up and down the strip.
The picture has nothing to do with driving around town, it just shows the car on track and not on fire.
entry 633 - tags: fuel, testing
December 17, 2011 - MG for sale!
No no no, not the V8 one. The parts car. I'm thinking I'm probably done with it, and it needs to move out to clear some room in the driveway.
It's a 1969 with a straight body, no title problems and no engine or transmission. A perfect candidate for an engine swap, actually. There's a small bit of rust on the sills and the paint appears to have been applied with a brush, but the floors look clean and it's a southwestern car as far as I know. A few of the parts have been removed, but it's mostly complete. I do have a new set of taillights of the 1972 style ready to be installed as well as a few bushings and brake lines intended for the MG suspension. If you're interested, drop me a line.
The black MG has been resting for some time, mostly due to the Targa Newfoundland and the chaos that surrounds it. If you read about this car but not the Targa, you might want to see how that went. After the fireball experience on track and yet another stranding episode as the car decided to stop feeding fuel to the engine, I parked the MG for a while. The plan is to build a better fuel tank and fuel system. The current tank has a known pickup problem as well as some major dents in the bottom. It's also close enough to the exhaust that the sensor for the fuel level shorts out once in a while. So that's the big plan for the car.
entry 634 - tags: for sale
May 22, 2012 - The car has been sitting for quite a while.
There have been a few reasons, but the biggest ones center around the fuel tank. Janel wasn't a big fan of the whole fireball escapade on the track for some reason. The fact that the car also stranded her on the side of the road on the way home didn't help.
Of course, we also took some time to run the Targa Newfoundland.
But it's time to get back at it. I really wanted to get a look inside this tank. So I yanked it out, filled it with water and used a Sawzall to cut the top off. The water was to force out any gasoline vapors, which could have caused extra excitement. It worked.
entry 635 - tags: fuel tank
May 22, 2012 - There was a big surprise waiting inside.
A BIG surprise. This giant canister. That's the top of the tank, upside down. The line running under the tank is the one that was originally attached to the charcoal canister. There's an itsy bitsy hole in the top of this canister, probably to prevent heat expansion from forcing fuel into the canister.
Great. But there are two problems with this. One was a big aftermarket dent in the bottom of my fuel tank. It kinda looked like someone had tried to jack the car up via the tank. This was perfectly placed to close off the hole in the bottom of the canister. You can see where it's been pressed up against the bottom of the tank.
The second problem is that I was using that charcoal canister fitting as my fuel return. I'm amazed any fuel got returned into the tank at all, and that's probably where my oddball fuel starvation problems came from.
This was quite a surprise.
entry 636 - tags: fuel tank
May 22, 2012 - Here's a view on how the giant canister was placed in the tank.
You can see a couple of dents in this shot as well as the one baffle that was in place. It's to keep fuel from sloshing around and upsetting the fuel gauge, I assume.
But here's a neat trick. The fuel filler dumps into that right side section of the tank. And there's only one hole joining it to the rest of the tank. That hole doesn't go all the way to the bottom. So once you're down to a gallon or two, fuel doesn't transfer across. The fuel pickup is on the other side. So the pickup will continue to suck fuel out of the left side while the right continues to show fuel in the tank.
It's like MG wanted people to run out of fuel on the side of the road.
entry 637 - tags: fuel tank
May 22, 2012 - Time for some re-engineering.
The big canister has been repurposed - more on that later. I decided to move both the fuel pickup and return to the right front corner of the tank for better exhaust clearance. I used 3/8" and 5/16" brake line to form the lines, and kept the fuel pickup in the original location. I put a Miata "sock" on the pickup to replace the small MG one. Here I'm just starting to shape the pickup line.
The dents in the bottom have been hammered flat by this point.
entry 638 - tags: fuel tank
May 22, 2012 - The lines have been shaped.
The big canister is now a big round baffle. There are some angled flaps on the bottom to make it easy for fuel to run in but a bit more challenging for it to run out. The return line also dumps into the canister to ensure it's got fuel in it. The lines take two different routes to avoid the float for the fuel gauge.
I'm not really planning on spending much time on the track with this car, but I couldn't resist making it better.
entry 639 - tags: fuel tank
May 22, 2012 - To avoid the problem of the baffle segregating the tank into have and have not sections, I drilled a few holes.
They won't allow rapid flow of fuel from one side to the other, but it will be enough to let me use the full tank capacity. Between this and removing the giant sealed off canister, I've probably doubled my tank capacity.
Once this was done, I cleaned out the tank and started welding. First, the big round baffle was tacked into place. Then the fuel lines were welded in. The original fuel pickup was welded shut. And finally, the fun job of welding the top back on again. I used a MIG welder for this which does usually leave some pinholes, but I did my best to get a good seal. I also spread some gasoline-resistant liquid metal over the seam to catch any little leaks. The seam is right at the top of the tank so it's unlikely to ever be wet on the inside, but I still don't want any leaks!
entry 640 - tags: fuel tank
May 22, 2012 - With the tank back in, it's time to sort out the spitting fuel problem.
That's what led to the fireball at the track. I probably should have put some baffling inside the filler neck to avoid any splashes, but instead I decided to borrow yet another Miata part.
The filler on a Miata has a spring-loaded flap in it. Perfect. I cut one out of a Miata filler and welded it into the MG neck, just below the cap. There are two small vents on each side that act as drains and a place for air to escape when you're filling up. They're usually attached to a separate pipe, but I decided to just let them go into the main filler neck. There's a small chance of some splashing through them, but it's going to take some real effort to make that happen. I think it'll be a decent solution, and it's invisible.
As an added bonus, the MG now looks like a modern car that drinks unleaded when you unscrew the fuel cap.
entry 641 - tags: fuel tank
June 3, 2012 - Hang on a second, the tank is supposed to be in the car.
The car was running just fine...until I had it backed out of the garage and it died. No fuel pressure. I jacked the car up there and started running diagnostics. Of course, the first thing you suspect is the thing you just did. But because it was such a pain to weld the tank back together, I started looking elsewhere.
And I found something. It appeared that the regulator had died on me. When I disconnected it, I got a good gush of fuel out of the pump. Huh. The regulator is built in to a stock Corvette fuel filter that should be good for extended use - but maybe something from the tank plugged it. So I swapped in a new one. Nope. The pump struggled to get up to 40 psi (slowly) when it should have popped up to 60.
So I plugged a new hose on to the pump inlet and dropped the other end into a bowl of fuel. Voila, instant 60 psi. I'd tried this earlier without success, but I think I had the bowl too low so the pump couldn't suck it up. So out the tank had to come.
When I cut the top off, I didn't see anything obvious. But plugging the end of the pickup and blowing in the other end revealed the problem. A small pinhole in the new line where it had been tackwelded to the interior baffle. Not much, but once the pump was pulling hard it let in air instead of fuel. A half second zap with the welder and the pickup line was nicely airtight. Argh.
Welded it all back up again, smeared liquid metal around the seam again (to cover any pinholes in the welds) and the tank went back in. I tested it with a couple of gallons in the tank and again, 60 psi. Excellent.
Then the car refused to start. It cranked and backfired, then went "clunk" and wouldn't crank anymore. I'd been working the battery pretty hard during testing, so I threw on the charger. Just to check, I stuck a test light on a hot terminal in the engine bay. Nothing. Hmm.
I checked across the battery terminals, no problem. I checked from the grounding point to the terminals, no problem. But then I checked from the bumper to the battery. Nothing. Closer inspection showed that the main grounding point (a bolt running through the sheetmetal) was a bit loose. I tightened that up and all of a sudden I had power through the car.

If you remember back, the car's had the habit of not always starting on the first attempt. It would backfire and pop, and then would fire happily if I just cycled the key. It also had a habit of occasionally stopping at the side of the road. I'm thinking the loose ground strap may have been the culprit for the first, although we'll see if that behavior continues. The latter might have been related to that heavily restricted return line inside the tank. That's certainly not a problem before. We'll see what happens now. At the moment, the car is on "close to home" probation.
entry 642 - tags: fuel, troubleshooting, wiring
June 5, 2012 - So, how does the new fuel tank work?
So far, so good. I used the car for some errands today (which somehow required sudden bursts of acceleration for no apparent reason) and it worked fine. It was hot, too - somewhere around 95F and the car managed to keep its cool.

There was one troubling moment - after stopping at an auto parts store, the car didn't want to fire, just crank. Eventually, it did roar into life. But it was an odd moment. Next time that happens, I'll check the fuel pressure. The main fuel line does run awfully close to the exhaust at one spot. Could it be heat soak from the hot exhaust in the high ambient temperatures?
entry 643
June 6, 2012 - Insulation time!
The exhaust has to run just below the main fuel feed at one point. I had it insulated with the foil-backed fiberglass insulation at right, but figured it wouldn't hurt to improve it. The rubberized...umm... stuff is the same sort that's used for protecting wiring for oxygen sensors, so it's easily able to deal with high heat. It's actually from a company that also made parts for the space shuttle, only that version had more Kevlar and looked cooler. Anyhow, it was a fairly short job pop the fuel line off the filter/regulator and slide the black insulation over top. Let's here it for quick-release fuel fittings!
The silver insulation shows no sign of heat damage, so this may not do anything. But it won't hurt.
entry 644 - tags: fuel
June 12, 2012 - When I repainted the dash a while back, I simply covered up a crack with black tape and sprayed over it.
This was intended to be a temporary thing, with a future plan of refinishing the complete dash. But then the 1969 parts car showed up with a plastic dash cap - and you know what? It looked pretty good. I think I saw one in place when I was at the Mitty as well. So I decided to take the easy route.
Unfortunately, when I unpacked the cover there was a big gouge out of one side. Nothing functional, but when a part is 100% aesthetic it's difficult to overlook a problem that's right in front of the passenger and very much not aesthetically pleasing. The box wasn't damaged and the gouge appeared to have been in place when the part was wrapped up at the factory.
So I called Moss Motors and got Kevin on the line. He offered to send me a new one - and after I told him it was damaged inside the plastic, he actually went into the warehouse and started unwrapping. He called back the next day and said they were all a bit rough, but he'd picked the best for me. It was shipped out that day and arrived shortly after.
Now that's excellent customer service. I've been on the other end of that phone call. Would our company have handled things as well? I don't think so.
The cover will likely go on this week. Janel thinks it's a big step up.
entry 645 - tags: interior, dash
June 12, 2012 - Time to work on steering.
I've been trying to track down some free play in the steering wheel. It's definitely not the tight, responsive steering it should be. It appeared to be in the rack itself. First I tried adjusting the rack to see if that would help, but no. So today was spent pulling out the old rack and slotting in a replacement. It was a bit of a challenge, one of the bolts was almost completely inaccessible with the engine in place. One custom-length socket later and it succumbed!
On a quick drive around the block, the new rack appears to be a success. We'll see how it feels on a longer drive tomorrow.
entry 646 - tags: steering
June 12, 2012 - Another steering tweak.
The steering feels overassisted - it's too light for my taste. I have a few things to learn about power steering systems in terms of the results of different pressure versus different flow, but the RX7 crowd seem to think that trimming the relief spring should drop the pressure a bit. Others claim that the only real solution is a Turn One pump for $240.
The spring cut makes logical sense to me - and if it doesn't work out and I end up going for the Turn One setup, it doesn't make any difference. So I made the cut. A fairly conservative one at that. It's not a massive change, but there is a bit more meat to the wheel. I might try one more little snip later, but first I'll drive it for a bit to get used to the new rack and assist.
entry 647 - tags: steering
January 5, 2014 - The MG's seen some action.
First, I took it to the track again. I finally got the chance to see how it handled at the limit without the big fireballs. The answer? Not that well, really. The car was oversteering badly, feeling like an arrow flying backwards. On the sweeper, it was actively trying to kill me constantly. The steering still had the wobble in it I didn't want. I tried adjusting the rear suspension pickup points (based on something I came across in Mark Donohue's "Unfair Advantage", of all things) and it helped a bit. But it was still twitchy and difficult to even drive on the interstate without the car exhibiting roll steer.
But I noticed something. The front tires were worn as if I had significant toe in. Lots and lots of it. So I checked, and sure enough. The toe was way out. I fixed that and all of a sudden the car was much better. The odd roll behavior went away and the steering dead spot disappeared. All of that chasing around, and it was just a bad alignment. Which is odd, given that it's been aligned before...
I also took the car out to a Corvette autocross, which they found fairly entertaining. Unfortunately, I simply didn't have enough tire to deal with the torque of the engine on a first-gear course and finished 4th overall if memory serves. Still, it was a fun day.
So it's mostly a driver right now. There are still some things that need to be sorted - in particular a vibration from the transmission leaning up against the tunnel under load and the bad reverse selection - but those don't stop me from driving it. That steering problem really took the fun out of it. That's been sorted!
entry 648 - tags: track, steering, handling, autocross