MGB - Finding Vacuum Leak

Following my last post about Oliver (the 1978 MGB) when I got his timing square, I knew I needed to get his carb tuned. Today's post covers those efforts. This continued in parallel with the body work on Zed (until the weather turned, when Zed work stopped). These spring days, when the sun doesn't go down until 9PM creates LOTS of play-on-cars time. Love it.

A-typical Carb

We start with the most basic learning: what kind of carb is this? When I bought this car, the prior owner had been executing some improvements. For the most part, I can't complain about any of them... except the hack-job on the exhaust. Honestly, he could have been just making do until he had the cabbage to do what I did. Anyway, one of the improvements was to eliminate the stock dual Skinner Union (SU) carbs and swap in a side-draft carb. This isn't your standard Weber DCOE, though. Instead, the prior owner went with a SK Racing carb. The SK Racing (now known as OER Racing) side draft carb was (according to Dog283) one of the best engineered side draft carbs built, just as carburetors were falling out of favor for the new computer-assisted fuel injection systems. Dog283 continued "(the SKRacing side draft) combine(s) the best of the the Mikuni Solex PHH, the Weber DCOE and the Dellorto DHLA". That's a fairly strong statement. Unfortunately, the implementation on Oliver had fallen out of tune, so our MGB was not getting the lofty results this statement describes.

We start with looking for, and ultimately finding, a shop manual for the SK Racing side draft carb. If you have one of these carbs and need a copy of this manual, I'll happily share it. Near the back of it are the set-up steps detailed in the image of the page on the right. Basically, set the idle speed screw so it is just barely touching the tang, thread the idle adjustment screws all the way in until they just barely seat and then rotate them back out 1 full turn. Start the engine and get it to normal operating temperature. Then, tune the adjustment screws first until the engine runs smoothly, then set the idle speed. I did that, but I still got sporadic backfires when I rev'd the engine. I was starting to think that all that ignition work was for nothing.

Exhaust Leak Checks

Before I started investing time in the carb, I figured it would be a good idea to see if there were any leaks in my exhaust which could account for the backfiring. There were. To test, I started by cleaning out my shop-vac (washed out with a hose) and grabbing a squirt bottle of soapy water. I thrust the exhaust end of my now-clean shop-vac up the tail pipe and turned it on. This created backpressure through the exhaust. The tail pipe was not completely blocked so lots of air rushed right back out, but there was enough pressure to execute my test. I started up front, in the engine compartment and shot soapy water on the mate-point between the header and the head. No bubbles. I then got under the car and hit every joint, and found some bubbles. I concluded that I created leaks when I added in the catalytic converter. I turned off the shop-vac, loosened the joints, slid the pipes apart and applied some copper exhaust gasket maker and then re-connected the pipes. I nutted them back down and re-checked with the shop-vac. Things looked fixed (no bubbles at all but one joint where there were very very few), so I moved on to tuning, expecting my backfire issues had been identified and resolved.

Gunson ColorTune

Back when I was doing the other work on Oliver, I kind of expected difficulty with getting the carb to adjust. So, before I finished getting the ignition installed, I ordered a Gunson ColorTune. These things are pretty neat, but not terribly useful, I suspect, for the computer-controlled fuel injected systems. The kit includes a looking-glass spark plug, a lead, a tube with a mirror and a brush. I only needed the glass spark plug and the lead. The looking-glass has a solid-center where the spark is produced, with a ring of glass around it. The glass ring is then encircled with threaded metal so it can be threaded into an engine. The tube with a mirror is for spark plug holes that are hard to see. To use the kit, you remove one spark plug and thread the looking-glass plug in it's place. You connect the spark plug lead to the end of the lead from the kit, which is threaded onto the glass plug. Then, you start the engine and look at the color of the combustion in the chamber through the glass. If the color is orange or red, your mixture is too rich. You want the color to be "Bunson blue" = the color of a Bunson burner, from, like science class. The images on the right, here, help lead you to a good tuning. Notice that light blue or white-ish is not on here. If you are in the white-zone (Airplane, the movie reference here), your tune is too lean.

I was unable to get cylinders 1 and 2 anything better than a very light blue even after adjusting the idle mixture out past 1-1/2 turns. Cylinders 3/4 did not do much better. Based on the manual, I believe the idle jets may be too small, so I ordered a pair of 50F9 jets ($7US each). If I am correct, the current idle jets are 45's and this change will allow idle-mixture within 1-1/2 turns of fully seated to sit in the Bunson-blue color.

Still, I kept going, wanting to get the tune as good as I could. While looking through the glass I increased the fuel mixture incrementally, but the misses and backfires persisted. I could see the backfiring through the glass in cylinder 1, and concluded there had to be a leak somewhere. I also figured that backfires within a cylinder that has a glass plug was probably not a good idea. I had already looked on the exhaust side for leaks, so that left the intake.

Intake Vacuum Leak Checks

There are not many places where a leak could appear in the intake: the mating point at the head, the mating point with the carb, the carb itself and the brake booster. With the engine running, I carefully sprayed some WD-40 onto the various mate points. If there had been a vacuum leak, the vapor would have been drawn into the engine and the RPM would have bumped in response. Spraying something flammable onto a hot engine is dangerous. I strongly urge you to keep a fire extinguisher handy if you do this. In my case, the engine did not change and nothing caught fire.

So, I shut everything off and considered the brake booster. I removed the hose from the check valve, which is threaded into the intake manifold heading to cylinders 1 and 2. I checked the vacuum of the brake booster through that hose with the MityVac. It would not hold vacuum. I removed the check valve and cleaned the valve and the intake manifold where they met. Then, I put some copper gasket maker on the threads of the check valve and threaded it back in. I figured if that was the leak, I just solved it. I jabbed the hose back onto the check valve and hose-clamped it tight. Then, I eliminated the brake booster by threading a bolt into the end of the hose which had previously been attached to the booster, and started the engine. Oliver ran great! I could rev him up and down without any backfires demonstrating that the brake booster was my vacuum leak.

check valve in foreground,
intake in background

I wanted to prove it for sure, so I hooked the booster hose back up to the brake booster and started the engine again. I rev'd the engine up and down... the back fired returned. I pinched the vacuum hose with a pair of pliers and I could hear the engine RPM's settle. Neat. Rev rev rev... no back fires. Remove the pliers, RPM increases... rev, rev, rev... backfires.

Fortunately, these brake boosters are being manufactured now. I didn't realize these were not available until recently, so as much as I would like to self-blame for not replacing the booster when I replaced the brake and clutch master cylinders, it wasn't available then and they are fairly expensive now (~$200US). As of today, they are not available through Moss yet. I had to go through another vendor (EnglishParts.com, they're lovely). I concluded that any carb tuning I had done would need to be redone once the brake booster swap was completed. 

That's it for today. I will install and post about the brake booster when it arrives from the MidWest. Thanks, as always, for following along-