A while back we were discussing the possible causes of my MGB’s rough idle and many of you suggested that the carburetors could be the culprit. I wanted to blame it on the valve train, but before doing anything drastic, I decided to try out your suggestions instead. If you’ve read much online about SU carburetors, then you’ve probably noticed that a lot of people complain about them. I’m not sure why though. SUs are simple, well-built, and easy to work on. Sure, they can wear out and become impossible to tune, but so can any other carb. As such, it’s a good idea to address any other potential problems before fiddling with the carburetors. I gathered together my set of standard open end wrenches, screw driver, carb cleaner, and a couple of rags and got down to business!
To get a good idea of what was going on inside, I needed to remove the air filters and that shinny dome (11) sitting on top of each carb. After removing three screws and the black oil cap (15) on top, I was able to carefully slide it off. Once out of the way in a safe spot, I removed the springs (14) and dashpots (12) as well. There is a long needle (19) attached to the bottom of the dashpot, so care should be taken when pulling it out as to to not damage anything. A few squirts of carb cleaner and a wipe with the rag removed any grime present.
I noticed that one of the needles was set a little higher than the other so I loosened the screw that held it in place and realigned the needle before tightening it back down. Things looked pretty good in there, but I decided to remove the lids on the fuel bowls (31) too just to make sure there wasn’t any crud in there clogging up the fuel delivery. Another rag was used to sop up the fuel and then I inserted the carb cleaner tube into the little hole where the needle sits (26). After covering the fuel bowl with a rag, I gave the can a few squirts to blow everything out of the seat and into the bowl. There wasn’t any visible sediment present, so I reassembled everything after giving it all a good cleaning.
I went ahead and replaced two vacuum hoses, the fuel filter, and some fuel line while I was in there too. With everything put back together, I turned my attention to setting the idle and fuel mixture. Someone must have tried to compensate for that high needle by enriching one side more than the other. So, I started over by turning both mixture nuts (3) all the way in and then turning them back out three times. I also turned the idle screws about one turn in. The engine started right up with those settings and already I could hear an improvement.
I set the idle (10) and synchronized each carb using the old air hose method. It may not be the most scientific way of doing things, but by using the little lifting pins (5) on the sides of the carbs, I was able to find the right combination of fuel and air. If you push the pin in and the engine stumbles, then you know it’s too rich. If you push it in and the rpms go up and stay up, it’s too rich. I found what seemed like the sweet spot right in between and then took her out for a test drive. Well, the work was definitely worth it! The engine runs much smoother and and seem to have more power throughout the whole rev range. It’s still too noisy and rough at idle for my liking though. Perhaps I’m expecting too much from this old girl?
I did notice something odd when everything was pulled apart. Can you see anything out of place? The bracket that should be holding one of the downpipes to the exhaust manifold is just flopping around down there. A new exhaust system was installed by the previous owner and they must have have forgotten to tightened everything down. Nothing that a couple of fine thread 5/16″ nuts can’t! After tightening everything up from underneath, I turned the key to see if that made any difference. The rattle that was sometimes heard on startup was now gone, but it didn’t really help with the rough idle.
It’s always a good idea to do tests before replacing parts or taking things apart. So, after doing the opposite for a while, I went to the parts store and picked up a vacuum gauge. I had never used one before, but we needed a pump for sucking brake fluid out anyway and this tool had both. There’s an access port right on the intake manifold so I hooked up the gauge and started the engine. I have posted a video of the results above. Wow, that needle is bouncing all over the place! After looking through the manual that came with the gauge, I’m guessing it could be weak valve springs or a leak somewhere on the head? Any experienced vacuum gauge readers here who can help diagnosis this?
Good job so far. I would suggest adjusting the valves cold to .015 intake and exhaust. The valve train sounds awfully noisy. Make sure the rocker shaft is getting plenty of oil. When you remove the valve cover, oil should be pooled in spots around the head. After that get a can of WD40 and spray around the intake/exhaust gasket to check for leaks there.
I am making the assumption that you have done a compression test and and checked timing. Compression needs to be within 10% or so on each cylinder. New plugs, points and condenser and wires? These things are easy to play with and given proper care will bring years of fun.GC
Seconded, the valves should be a lot quieter, re-set the clearances and test again, that wobble on the gauge could be nothing more that valve adjustment being off
Could too much valve lash really cause that kind of action on the vacuum gauge? I sure hope that is all it is!
Thanks for the suggestions George. I should have mentioned that I did do a full tune up and compression test before I started messing with the carbs. The natural inclination is to start turning those idle screws when things start running rough, but I held back until we had addressed the most likely culprits first. Anyway, here are the results of the compression test:
#1 – 125
#2 – 130
#3 – 120
#4 – 125
Those numbers are a little low, but the variation between them doesn’t seem too bad?
I wouldn’t worry about those compression numbers. They are all within 10% and remember that is a 8.8:1 compression engine. Those numbers can be a result of valve lash, but look for a vacuum leak. I’m a long way from the Pacific NW but that is where I’d go next.
Jesse, I’m with George—I think 125 PSI +-5 PSI is fine, and definitely not too low for good running.
Below is a link to discussion about a newly rebuilt MGB engine set up for supercharging that had 175 PSI. Aside from that being too high for a supercharged engine IMHO, the discussion is interesting because other commenters also think 175 is high for an MGB engine.
http://www.mgexp.com/phorum/read.php?1,2047371
Different make/model engines can vary a lot on compression pressures and all still run well. Last week I did a compression test on my Alfa GTV6 engine and pressures were all around 170 PSI. Other good running performance engines I have tested had pressures that were significantly lower, but even.
Often on these MG engines a bunch of little things add up over the years to cause rough running, and it’s just a matter of chasing those little things down and making them right, so at the end the engine ends up running about like it should, given its age & condition. And in my experience 4 cylinder engines will never run nearly as smoothly as an I-6 or V8, no matter how perfect the tune or condition.
If pressures are around 125 I think any 4-cycle engine should run well providing everything else is good. And the fact that your pressures are +- 5 PSI is very positive.
Hi Jesse, Sorry can’t be of much help but I do know these twin carbs can be a real pest to tune up. My father, who is now deceased, was a whiz when it came to cars.A cousin of mine had a triumph vitesse over in England and had been to several garages trying to get it running smooth and get the power it should have. Finally he brought it to my father and after ten minutes of playing with the carbs he had it running smooth and with tons of power. Unfortunately I am not as gifted as my father. So I think you may still have to play with the carbs. My father done it just by listening to the engine and no machines.
Got a UniSyn, Jesse? A magic little device, to say the least! Don’t know if they’re even available any more; haven’t needed one for years, alas, but would not touch a SU carb without one.
Also, the needles can bend easily, which can cause a rough idle. Ditto sticky dashpots. Clean and polish (with extra-fine emery paper) moving surfaces. But I’m sure you know all this stuff already!
My first guess for idle problems on Brit cars with SUs is always something off-kilter with the carbs. Then ignition issues and vacuum leaks.
I did have a Uni-Syn here somewhere… I’ll have to dig around in garage again and see if I can find it. Looks like Edelbrock still makes them.
” Could too much valve lash really cause that kind of action on the vacuum gauge? ”
Well , of course… and one valve too tight can kill the vacuum and drive the needle crazy
Well, let’s hope that is all it is. I can see why a tight valves can cause problems (like burnt valves), but I didn’t realize that having them a little loose could cause the vacuum to jump around like that. I had better check the bottoms of those rockers!
Too many years ago when I was operating and racing Austin Healeys I used a doctor’s stethascope for tuning and balancing SU’s, a particularly tricky task on three carb Austin Healeys.
Vacuum leaks on the intake manifold are common, but another common issue is worn throttle shaft (43), uh, bushings? Anyway, the carb body wears where the shaft goes through the carb and you get a vacuum leak there. If there’s any change when you spray WD 40 on it at idle, there’s a problem.
I do like the idea of using WD 40 more than carb cleaner. Some guys even say to use propane.
Butane or Acetylene work fine.
Re: adjusting the valves – you need to check to see if the valve stems have worn round holes into the rockers – common on these cars – if the rockers have holes worn into them by the valve stems it will be close to impossible to get the correct valve lash setting – what happens is that the feeler gauge will bridge the hole giving a false setting and after removing the feeler gauge the rocker will once again settle into the wear hole – giving you a much wider gap than you had set it for with the feeler gauge resulting in a very noisy valve train…….take the rocker shaft off and check the surface of the rockers – if there are wear holes in them they need to be resurfaced or replaced……cheers
I think you may be onto something here Richard! I haven’t checked the rockers yet, but when I last adjusted the valves one seemed a little tight. The rest were right on or a little loose, but afterwards they almost seemed noisier. Wear on the rockers could explain how that could happen.
If rough idle gets better as it warms up or if you can’t get the idle to set low enough:
Check the intake and gaskets. Carb cleaner sprayed or Wd 40 while it is running should locate the area if there is one. Even if the intake bolts are tight
against the intake, the bolts (Studs) can bottom out on the block and be tight,
but not hold the intake tight. Check Carefully at the block and the intake mounting ears to be sure they are not bent in.
The idle actually gets worse as it warms up. Especially when the ambient air nears triple digit numbers. In fact, when it’s really hot sometimes I have to use pull the choke out a bit to get the fast idle to kick in so it will stay running.
Jesse,
Remember it’s probably the something simple and cheap.
I wouldn’t remove the shaft until you are ready to change the head gasket and lap the valves. Is this an air port head?
Hopefully you’re right George! The head does have the air injection holes. I’m not sure if harden valve seats have been installed or not, so your suggestion of lapping the valves seems like a good one.
Check valve lash and leaks as said, but that is the classic vacuum gauge indication for leaking valves. Valve(s) could be burned or just too much carbon causing one or more to not seal properly on closing. Confirm and find which valve with a leakdown test (listen at exhaust, intake, etc.). You could try ‘staking’ the leaking valve, but that hasn’t worked very often for me.
You’re quite right about SUs being simple and very reliable carbs. If you decide to clean the dashpot and cover, don’t use anything abrasive, just clean with Nevr Dull or some non-abrasive cleaner. The dashpot and cover are a matched pair to keep just the right amount of sealing for proper dashpot action. Cleaning either with abrasives upsets that seal.
If it turns out to be a leaking valve, just run her anyhow and enjoy the rest of the summer. Those old B engines are the equivalent of a Chevy straight six–completely understressed, they’ll just keep on soldering along. Then get into it this winter.
I’ve never done a leak down test, but it doesn’t sound complicated.
I was careful while cleaning up the carbs and made sure to not mix up the dashpots or domes. Cleaning and adjusting them did help, but obviously that wasn’t the solution.
Good advice up above. Agreed on checking throttle shaft leakage. I have had some success “peening” the carb bodies and using a mild abrasive solution (toothpaste/polish) to re-lap the shaft into place. Before you folks complain, I was a college student on ZERO budget, and yes, it does wear back because you don’t have the entire thickness of carb body to make a seal. Jesse, if you need a UniSyn I can mail you one…let me know.
Thanks Jamie. We had one kicking around somewhere here… Hopefully I didn’t ship it off with the Singer on accident!
George is right on. The carbs can be adjusted to perfection, but if there are engine problems the engine won’t run smoothly so matter what you do.
Uneven compression pressures will make any engine run rough, especially a 4-cylinder engine. Unless you KNOW the compression pressures of your engine, job 1 is a compression test.
A valvetrain with uneven clearances will also run rough because different cylinders will be getting slightly different amounts of gas/air in, or different amounts of exhaust out, or both.
One poor plug or plug wire will make the engine run rough. Again, all of these things will make a 4-cylinder run rougher than if there’s one bad or ‘different’ cylinder on a 6 or 8 cylinder engine
Yes you can still get a Uni-Syn. You sometimes see them for sale on Ebay. Probably other sources also. A UniSyn is useful for equalizing the air flow between carbs at idle, but there are things you need to do to tell whether the mixture is right first.
The best quick way to do that is to lift each dashpot up slightly while the engine is at idle. About 1/16th to 1/8th inch is good. Engine RPMs should rise slightly, then drop a little. If RPMs speed up a lot and stay there, lean out that carb. If RPMs drop immediately and the engine wants to stall, make the mixture richer. Best to block off the other carb while doing this so you have a better read on the carb you are testing, but I did not find this essential. Do this for each carb. You can use the little lift pin that many SU carbs have, or just use a long thin screwdriver to lift the dashpot.
To lean the mixture, raise the needle seat. To richen the mixture, lower the needle seat. It’s also possible that the carb(s) have the wrong needle(s) for your driving conditions.
All this assumes that the carbs are clean, the dashpot goes up and down smoothly with zero binding, and that the oil in the carbs isn’t too thick. Some people use 3-in-One oil, but I used to use SAE 10W30. One thing that can make a dashpot bind in an uncentered needle. Also assumes no air leaks at the throttle shaft. Rebush them if necessary.
How did I learn this stuff? Yes, a first-year Sprite, then an MGA, but the real reason was the tri-carb Healey I used to own. When those 3 carbs were spot-on, it ran real strong. When they were off, you hated them. I liked them to be spot-on.
Thanks Dolphin! I listed the results of the compression test above, but I have a feeling that the valves are my problem. I’m going to check the rocker arms for wear as suggested above, but I’m guessing that if they are worn, the head should probably come off and get a full valve job. I bet the Healey was fun to tune!!!
Jesse, I looked at your compression numbers above and they look fine to me. I added another comment above.
Hi Dolphin,
Any chance getting a good write up or source from you on full procedure for syncing 3 SUs? I’ve looked online and never found a really clear one. I’ve got a Jag XK 150S (3 HD8s) and a Mark IX, and the Mark is smoother and pulls more consistently over the range despite only having two SIs.
what RayT said……any SU tuner should have one…..even if it on a Morris Minor – like mine….
Doesn’t anyone know how to tune SU carbs with a coat hanger?
By the look of the vacuum gauge readings you have, at least one, or more cylinders is very low in contributing to engine performance. I’d recommend a compression and cylinder leakdown test, as a baseline. I’m also concerned about the valvetrain noise? Both of these symptoms reek of engine mechanical problems. I’d recommend doing some basic engine health checkups to see what the problem is. A vacuum gauge never lies!
Yep, the compression test results can be found above. I replaced the wires, plugs, and even the worn out distributor before getting this far, but have yet to do a leak down test.
That is one noisy valve train. Hopefully you can adjust the valves and quiet it down.
Like Dolphiin, my first SU’s were on the Bugeye I bought new in ’59… Dolphin laid out the correct pattern for diagnose and service. You should follow everything he says, if you skip a step, you will be trying to compensate by incorrectly setiing something else.
When adjusting the mixture on one carb, I don’t block off the other one, I just unscrew the idle screw so the throttle plate is closed.
Actually, you should loosen the linkage so one throttle plate does not move the other one, until you have each carb adjusted correctly, THEN set both idle screws equally and tighten the linkage..
I don’t use a Unisyn, I use a setup with liquid ‘J’ tubes, so I can see both ( or 3) carbs airflow at the same time, , but use whatever works for you, to balance the throttle plates. I like the ‘J’ tubes, because I prefer to also check the balance at 2000-2500 ( steady) rpm’s
Nobody mentioned the ignition timing, there is not much sense trying to tune an engine if the ignition at peak advance is not assured with a timing light. Check for about 30 degrees advance at 2500-3000 rpm
As far as the vacuum gauge, we use it every time the ‘hood is open’ for service, because it is the ONLY way to see what is going on INSIDE that lump. Don’t ‘blip’ the throttle, record what the needle is doing at 900-1000 (STEADY) rpm’s…and at a STEADY 2000 or so rpm’s. BOTH readings should have a needle that is not fluctuating …. While you accelerate from 900 to 2000, the vacuum reading should drop, and recover when the rpm is steady at 2000. . Your video shows there are problems, but because of the way you managed the throttle variations, and with limited knowlege of what you have done, there is no way to nail anything down. One bad spark plug, one bad valve, various intake leaks, I could go on an on…
The important thing to know, is that the motor MUST be able to run at a steady rpm…if the rpm’s vary, the gauge will be of limited value. ( Intake leaks are the biggest ( but not the only) cause of floating rpm’s, assuming the carbs are clean )
Lots of good info Dave! I did disconnect the linkage while adjusting the carbs, but the rough idle made the job a challenge. I did set the timing a while back too, but haven’t checked it again, so that will go back on the todo list. I should have let the camera film a little longer with the vacuum gauge hooked up. I’ll try to do another one at both RPMs. Thanks!
.
All good advice above, but for any car you should just do a basic checklist, and never assume the DPO (Dreaded Previous Owner) did anything right. Start with leakdown test AND compression test (Research the PROPER methods). then do a second comp test with 50cc oil, this will tell you rings or valves. Then check initial timing, and advance, then the condition of cap, rotor, wires, plugs. If in doubt, replace all of them, and start a journal and record part #s, and dates and other notes so you have a baseline. Valve adjustment is a must. (Cyl head rebuilds are a piece of cake so dont fret) Modern fuel is S**T and will clog any orifice, dont rule out the tank, pump and lines as well as filter. SeaFoam or Chevron Techron additives at high concentrations are an excellent idea regardless. Those carbs once properly setup will purr like a kitten. Motorcycle guys understand carb sync, and nothing beats tuning with a vacuum gauge for both timing and carb setup, no room here for all the details. My wifes family have 50 years experience with MGs and everyone one of them has at least one. Some have several. They are 99% purists however they deviate on 3 things. Add a Delco remy alternator with internal regulator. Add a REAL fuel pump and regulator. Stock are pieces of garbage, 3rd, if you want the details email me off list, But they used to convert/manuf Ignition kits for MGs and several well known shops sold them. There are certain Datsun/Nissan Electronic ignitions that are drop in. (A series engines) and all you have to do is swap the drive gear. thats IT! they drop right in. Scavenge off a Datsun/nissan in the wrecking yard and make sure you get the brain box as well,.EASY to wire up. The gear is held in with a roll pin, swap in the MG one and Bobs yer uncle laddie. If you need more help, ping me off list,. My FIL is a wealth of knowledge on all MGs, and a member of the Salem Oregon MG Club. Im a Triumph guy myself, Both car and Bikes, Keep in mind the old adage of British Motorcycles,,, If you think a problem is fuel related it is probably electrical, If you think its electrical, its probably fuel related.
Lots of good advice and tips here Doug. I’m going to write down everyone’s ideas and make a new plan of attack. Tackling projects like this sure is a lot easier with a bunch of friends helping you out. Thanks guys!
I believe, david, above has it. The throttle shafts were worn on my MGB, and it never idled right, later on. ( the carbs had over 100K miles on them). I see all the air pump hardware has been removed, so that isn’t the problem. ( remember those annoying backfires on deceleration?) I think the carb housing wears as well, and bushings and new shafts are the fix.
I’ll be sure to check for leaks before doing anything else.
@Dolphin`… regarding supercharging, which is an art, not a science, because there are SO MANY variables in building an engine … 175 psi is not considered high , the Lampredi motors we build pull at least 210 psi naturally aspirated, and the Volvo B18’s at least 200 psi. ………. Blowers like the Judson only put out 50% boost , and design compression ratio is more important than psi. A tight motor will have ‘high’ psi, even tho’ the compression ratio is not ‘ very’ high .
When I got the Judson for my Bugeye in 1960, I bolted it on and and pressed the loud pedal.
Dave, I can imagine that a Lampredi V12 will have high compression pressures, but those were designed for competition, with some used in the highest of the high end street Ferraris of the day, so I would expect high compression pressures with those.
My comment was based on wanting decent reliability with a street engine, and that supercharging a street engine that already has high compression pressures is going to reduce reliability, especially if the owner uses the power a lot. Some of the commenters in the thread that I linked to seem to have the same concern about reliability.
Along those lines I seem to remember that some installations (maybe turbochargers) had water injection to keep things cool and reduce chances of the engine blowing up.
I also seem to remember the Judson superchargers put out fairly low pressures, maybe about 6 PSI, and that they tended to be marketed for engines with low CRs like VW, and so could work well.
Here is a video form John Twist at University Motors. Lots of information that people tend to over look. Even if you are an old hand at balancing and tuning SUs this worth watching
https://www.youtube.com/watch?v=4nvGLgO6pj0
great video. i find very interesting what he said about removing/pluging the emission equipment. could that be an issue with this car. also, are you running pure gas ( pure-gas.org ) or pump gas with what % alcohol?
I’m sorry if you have already checked this. It seems that the last couple of British cars I have had that ran rough ended up being the distributer. The last sounded similar to yours and only needed a distributer shafts bushing. The worn bushing was messing up the advance.
I thought it could have been the distributor before too Rick so we replaced it a while back. It definitely helped the engine run better, but it wasn’t the cure-all.
I have some good news! Yesterday we went and purchased a dial indicator and a leak down tester from Harbor Freight. I figured that the dial indicator gauge would tell me if the rocker arms had groves in them. Well, there were a few valves that were off by 1-2 thousandths so I tightened them down and that did seem to help with some of the valve train chatter. The leak down test showed that each cylinder had leakage of about 20%. That’s not necessarily a good number, but since there wasn’t a huge variation between each cylinder, I think its safe to assume that we don’t need to tear the engine apart just yet. Next, I’m going to see if we have any vacuum leaks around the intake or throttle shafts and then I’ll tune the carbs again with the Uni-Syn. We’re getting closer!
20% leak down on a street car isn’t bad at all. I wouldn’t let that bother me.at all. Once you get the carbs balanced and vacuum leaks fixed drive it a month or two and do another leak down. you may see the numbers increase with carbon build up. Keep looking for leaks, you’ve got air getting in somewhere.
So, I synced the carbs with the Uni-Syn today and the hose method apparently works fairly well because they were close. That didn’t make any noticeable difference. I then checked for leaks and wasn’t able to find anything obvious. That doesn’t mean there aren’t any, but I don’t think there is anything bad enough to cause the rough idle. The valve train seems pretty noisey though. Could weak or even broken valve springs be our culprit?
Here’s another video of the vacuum gauge: https://youtu.be/cLYw-K-0qo0
You bet your sweet bippy a broken spring could cause a rough run. It should be easy to pot with the valve cover off. Before you do that, fire it up and pull one plug wire at a time to locate the cylinder that seems to be off. Pull the wire and if the motor is worse put it back on. Continue until you pull a wire and no change in motor sound occurs. That will tell you which cylinder to look for. Now when you remove the valve cover you will know where to look.
I was able to pull the wires on #2 without causing much stumbling. Pulling #1 didn’t cause much either though. Guess I’d take a look at the springs. Any suggestions on how to tell if they are bad?
I’ve only broken valve springs on my race motor. The inner spring is most likely to fail. When It has happened to me, it was vary obvious because of broken pieces of spring. Is there any way you can contact me off board? We are chasing our collective tail here
I apologize if this suggestion is noted in one of the many other comments, I haven’t the time to read them all, but check for a common sign of age in these carbs, worn throttle butterfly shafts and housings.
I made my living back in the old days (when sex was dirty and the air was clean) working on imported cars mostly with SU carbs. Worn butterfly shafts were quite common, and resulted in erratic idle because that is a distinct vacuum leak (high one time and low the next, etc). Frequently an owner would just add another spring to slam the carbs back to idle. Check the idle by squirting some thick oil around the shaft/body bushings at idle to see if the idle quality changes.
The repair might be as simple as a new shaft if the body isn’t too worn, or the body might need to be line bored and a bushing installed if you can find somebody who does it now days. I had a lathe set up to do it back then.
Cheers,
Keith
Keith has a good suggestion. What do the plugs look like after driving the car for a few minutes? See if you can get a clean reading. At idle the plugs will load up and I’ve never gotten a good reading from a stationary motor unless it was on a dyno. Was there oil pooled around the rocker towers and low spots on the head? Again it’s something simple
SU carburetors are the definition of simplicity………….people usually work on them until they make the car run bad………….always MAKE SURE that the IGNITION is perfect before doing any carburetor work………….AND, a major overlooked fault ( and one that is almost impossible to investigate without just replacing it ) is the INTAKE / EXHAUST manifold gasket….if it is leaking, the carbs can never be made right………..