Many years ago, when my SUs were worn out, in a search for more power for my 'B I decided to try fitting a 45DCOE weber.
I've just recently completed a minor rebuild of my SUs (oversize throttle spindles, new jets and floats) and have refitted them in place of the Weber. This gave me the opportunity to compare the SUs and Weber back to back, and I wanted to capture my thoughts before they faded into the past.
The car is a 1970 MGB roadster, well worn, street use only. Driven daily to work unless there's snow and/or fresh salt on the road. 1800cc engine, 030 overbore, standard grind cam, Crane electronic ignition. Hardened valve seats. The head and seats were mildly cleaned up by me (not ported). 185x70 radial tires on 14" aftermarket wheels. No overdrive. Stock manifold and exhaust. I've tried an Ansa exhaust and header, made the B sound like a truck with no noticeable performance improvement.
The recent increase in fuel costs motivated the switch, and is the primary difference between the SUs and Weber. With the Weber I was getting 19 to 21 MPG (imperial - 4.54L/gal), depending on the time of year and my driving (highway vs the short drive to work). This is typically 190 miles on a full tank, followed by an approx 42-44 liter fill. With the SUs I get 22 to 26 MPG in the same conditions.
With the exception of the idle jets, air horns, PCV valve and float height, this was the out of the box settings.
With the vacuum advance tube connected to the weber manifold (the distributor VA can is the type designed for the SU carb take off), the idle is a little lumpy - probably too much advance at idle. Without the tube the idle is rock steady, set to around 900rpm.
There's a definite difference in the intake sound, it's more noticeable and more aggressive vs the tinnier sound from the SU cans.
The vacuum advance seems to yield about 5% better fuel consumption, but this is a very subtle effect for most of my driving which is not long distance. Part throttle acceleration is definitely better. Without the advance, more static advance is required for the same effect, but this resulted in pinging and/or run-on. With the vacuum advance and the milder static advance setting (10º BTDC) the car never runs on.
Throttle response is very crisp, the RPMs return almost instantly to 900rpm with my foot off the accelerator and there's a very good engine braking effect.
Power delivery is smooth and strong from 1800rpm up to 6000rpm, with no signs of drop off.
There is a slight stumble when accelerating from idle with part throttle, and a definite bog with a WOT acceleration from the lower RPMs. Part throttle yields the best acceleration until the RPMs increase to 3500 or so (this keeps the carb on the progression jets until the intake air is moving fast enough to pick up fuel from the primaries).
Torque seems to be higher than with the SUs, it's easier to "chirp" the tires when starting off, and there's certainly more of a kick from 70-80mph when I open the throttle. There's also a nice, low, throaty roar from the intake when I do this. No problem getting above 100MPH.
Starting cold is easy, one or two pumps of the accelerator and the car starts first crank after sitting overnight, even in winter. The enrichment circuit is difficult to use, and unless just right the car bucks, hesitates and sneezes until it warms up completely. There is no facility for a higher idle speed during warmup. It is a racing carb after all...
Starting from hot, after a short time, can sometimes be a little difficult. If I forget it's hot and pump the accelerator, it will take a few cranks to clean out the raw fuel that's been dumped into the cylinders. As soon as the car catches, cracking open the throttle will immediately clear it out and all is well. While it struggles (compared to a cold start), the car has never failed to start even when hot. After a longer heat soak (say 1 hour), there's no problem at all with starting.
It's a lot of work. I also found that the book I had wasn't entirely correct in it's theory and set-up, and most of the 'net help I found applied to the DCOE-9. Out of the box, the 50F9 idle jet was totally inadequate. The car would surge and sneeze at part throttle and low RPMs, esp when cold, and pop and spit on deceleration at all times. This was finally traced to two problems (after spending a lot of time trying to chase a non-existent vacuum and exhaust leak). The first was that the float height of 12.5mm stock does not work for the B. The cross over from progression to primary jets resulted in a wide rpm lean condition. I verified this by hooking up a vacuum gauge, and slowly opening the throttles with the idle screw. The gauge will drop at certain RPMs and the engine would stumble. Turning the idle mixture screws can be used to figure out if it was lean or rich (it was lean).
I tried various idle jets to attempt to remove the stumble, the richest being 65F9. The easiest way was to use the same vacuum gauge technique. The 65F9s worked the best, but I needed to reduce the float height to 9mm, which removed the lean stumble and the deceleration pop at the same time. However I was never comfortable that this lower float height ensured raw fuel wasn't dripping into the intake manifold (see hot starting). I know that 6mm was definitely too low (though it worked even better, eliminating the popping even with the 50F9 idle jets). For best running, the idle mixture screws were set to the rich side, rather than the recommended "lean is best" setting. It's possible 65F8s would be even better, but I don't have them (they're cheap, but shipping and duty adds up...). It's also possible a smaller secondary venturi would work better, but they're more expensive.
I believe the earlier (non-spanish made?) DCOEs had different characteristics. For example, the 1.5 max turns for the idle mixture on the DCOE-9's is too little for the DCOE-152s, which need something more like 3 turns (I believe the taper on the mixture screws was steeper on the 9s).
I also experimented with a crossover pipe in the manifold into which I plumbed the PCV valve. This seemed to smooth out the idle somewhat further, but that effect was subtle and very subjective.
It was important to soft mount the carb, if I had it too tight it would pop on underrun, and I suspect run lean on WOT and high speed cruising. I found the thackery washers and soft washers inadequate (both disintegrated after a short time). Instead I used short lengths of heater hose, which worked well and lasted for many, many years.
BTW, I tried different main venturi (36mm) with of course different mains and air correction jets. I tried F16 emulsion tubes, running with and without air horns, different air cleaners. None of these had any discernable effect on the progression to main cross over mixture. The 36s did run well at higher RPMs, but 33mm seems closer to the needs of a stock engine, giving higher air speed and torque at lower RPMs (IIRC - it's been a while).
The linkage that came with the weber was hopeless. Good ones are expensive so I made my own by using some brass nuts drilled to hold a return spring between the manifold cable mount and the cable attachment on the carb. It works fine, but I never did like the looks of it.
Vacuum advance was taken from the weber ports used to check the balance between the throats. This is the screw in plugs just downstream of the throttle plates on the top. I used brass barbed tube fittings tapped to fit the ports, joined with a plastic Y just above the carb and sent to the distributor.
The stock throttle cable can be used, adjusting the pedal travel with the bolt beside the pedal box. I drilled and tapped the manifold boss that takes the end of the cable sleeve so that I could lock it in place with a small screw.
The choke cable doesn't reach the Weber enrichment lever, which is on the front side of the carb. I used a universal replacement that fits just fine and doesn't look too out of place on the dash.
It was more expensive than the SU rebuild esp after factoring in the additional jets and other bits. A good challenge and lots of fun to fiddle with - a great learning experience. I know more now than I ever thought I would - too bad carbs are obsolete :).
Very smooth acceleration from any speed, including right from idle. No bucking or surge when negotiating speed bumps or stop and go traffic in any gear.
Slower return to idle, very little initial engine braking (relative to the weber). The car will run-on if I turn off the ignition before waiting a few seconds for the idle to return to normal.
Quieter, tinny, hollow sound from the intake. Not bothersome, but not as nice as the weber. Makes the MG engine very recognizable, I can always tell the sound of one running on SUs from quite far away :).
Not as much torque as the weber in the 2000-3000 rpm range, it's harder to chirp the tires. Runs out of breath above 5500 RPM, there seems to be little point holding the upshift much past 5000. Doesn't have the same kick going WOT from 70-80MPH, and I suspect it would struggle to get over 100MPH (relative to the weber). Doesn't have the same pull through a corner as the weber.
Starting is easy. A little choke in this fall weather gives a slight enrichment and just the right throttle, the engine usually starts on the first or second puff of the first crank after sitting overnight. When restarting later in the day to return from work no choke is needed, it starts right up with just a touch of throttle. No issues at all with a hot start.
Anecdotal, and the memory is starting to fade already, but I'm pretty sure from a standing start I couldn't get my foot all the way down with the weber before it was time to upshift. With the SUs I have more time at WOT in first. It's possible this is because I can engage the clutch and open the throttle at a lower RPM with the SUs (see earlier about the Weber bogging at low RPMs, WOT), but the weber did feel faster off the mark.
Easy, even the linkage is a joy. Completely stock, including the air filters. Poppet valves still in place on the throttle plates. Emissions equipment removed, except for the PCV tubes (metered bleed into carb throats). Vacuum advance (carb take-off) in place. Balanced with a unisys, tuned against idle speed and using the lifting pins, with a final check for good and matching plug colours. Holds the tune very well (so far at least).
I've got a carbon monoxide detector installed in my garage. Oddly enough, when fiddling with the DCOE idle, it seldom went off. When tuning the SUs or leaving the car to idle even for a short time, the detector almost always goes off (the garage door is open of course).
The weber is more fun than the SUs. With more torque and better engine braking it's easier to "steer with the throttle" around corners, and it gets better acceleration at speed. The SUs make the 'B a milder, easier and more economical car. If fuel prices every return to previous levels (ha!) I'll reconsider using the weber. For now I'll stick to the SUs, at least until they need another rebuild.
Reminder to self: http://www.aptfast.com/Main_Index_Page.htm for a 3.5 aux venturi, $20.25USD each. This might improve the transition problem from progression to mains and allow the use of a leaner idle jet and better fuel economy.My home page is Mike's 250lb Locost Site.