I have from 1034 - 1038 and in some cases, more than one of each size.....
This doubles the number of people potentially interested in my jetting results. Now there are two people: you and Kerry.
I decided I need to make one more run with the current carburetor, with the '200' main jet replacing the '190'. I made the swap so the bike is ready but probably am not going to have a chance to make that ride until tomorrow. Meanwhile, though, I modified the 2nd 1036 to turn it into a genuine 4-stroke body.
The first photograph shows the drift ready to push the 2-stroke spray tube out of the body. The OD of the tip of the drift is 0.246" in order to slip into the 1/4" ID of the spray tube, and the OD of the main part of the drift is 0.309" to fit through the 5/16" ID hole in the body.
With the old spray tube out of the way, but before pressing the replacement into place, I then drilled the restriction out of the compensating air passage using a #10 bit (0.1935") as shown in the second photograph. After doing this and cleaning up the swarf I pressed the 4-stroke spray tube into place.
Although the carburetor appeared to be very clean when I got it, taking no chances I inspected the pilot jet passages. It's a good thing I did because the chamber under the two pilot holes held quite a bit of some unidentified fuzzy material as shown in the third photograph. I blew it out from the front of the carburetor through the hole for the adjusting screw. I then dropped #80 drill bits into the pilot holes to be sure they weren't blocked, as shown in the fourth photograph.
I have a ring light on my garage stereomicroscope which is great for most things, but not for illuminating deep holes. By manipulating a flashlight I could see into the cavity under the pilot holes where I didn't see any residue, but I couldn't hold the flashlight and take a photo at the same time so you'll have to use your imagination. To be certain everything was clean and unblocked, while holding the carburetor at an angle I filled the passages with acetone until it came out through the two pilot holes.
To minimize variables, after doing my next/final(?) run with the current carburetor I'll swap its complete float bowl, needle, needle jet, and needle jet holder, leaving the replacement #3 slide, main jet and pilot jet (and spray tube) as the only components that will be different. The #3 slide is a nice slip fit so this second body, like the first, shows no sign of wear.
While 1 of my 1000 series carbs graces a Gold Star, and your work will likely encourage me to improve the 1038. You have already convinced me I am too rich on the slide, but I already knew that, but it ran good and I was a bit too lazy to improve it.
But, the bulk of my 1000 series carbs will remain 2 stroke carbs for 2 strokes. I have stupidly got involved with flat track again. Which in turn leads into getting a couple of vintage Bultaco flat track bikes running again. And I have a belief that jetted right, the Concentric is a bit easier on a high strung 2 stroke on a 1/2 mile. IME, the trick involves the pilot circuit.
Then there is the collection of large 2000 series carbs too.....
Never underestimate the human ability to elevate stupid to a whole new level!.
This doubles the number of people potentially interested in my jetting results. Now there are two people: you and Kerry.
You do yourself a grand injustice good sir. There are lots of people interested, however for most it can only be limited to interest, the 932 was over kill on the M20 so I can not see me putting a 1000 on it till I go batty & decide to attempt a land speed record. Can't do it right now cause lake Eyre is currently full of water.
I made my absolutely last AFR measurement on the 2-stroke body 1036 today (unless I change my mind and make more...), with the results attached.
The blue areas, determined using the Mikuni Main Jet Tuning Calculator, denote the limits that the peaks and valleys of the dashed blue curve will stay between if riding in the 50-deg. range between 48 oF and 98 oF. No human should ever ride outside that band of temperatures (or even near the borders of that band).
Given these results, and the fact the bike ran every well today with the '200', if I end up "permanently" installing the 2-stroke version of the 1036 body it will be with the '200' jet in place because the health of the engine will be better if I err on the side of too rich rather than too lean. Of course, I certainly know I should change the main jet to compensate for the temperature and altitude. But, I'm too lazy to do that. I carried a range of jets with me on the Cannonball, but the Ariel crossed the Rockies with the same main jet as it had at the sea-level start and finish. If I were racing it would be different, but for civilian use 'perfect' is the enemy of 'good enough.'
After finishing the jetting run and analyzing this very last set of AFR measurements I'll ever make on the 2-stroke 1036 body (unless I make more...), I swapped the parts discussed in my previous post into the "4-stroke" 1036 and installed the slide, pilot jet, and main jet that AMAL lists for the 1036 used on 1970 Velocette Thruxtons. The bike started immediately but the idle was much too high no matter what I did to the idle or mixture screws. I got the idle as low as I could and then made a lap of the driveway. This doesn't count as much of a test, but I felt no hesitation on acceleration so the cutaway seems correct.
I then removed the carburetor, tore it down, and installed a #25 pilot jet in place of the recommended #30. This time I dropped a slide in the other 1036, noted how far above the bore that the idle screw lifted it, and set approximately the same height on the "new" 1036 before reinstalling it. However, the smaller jet seemed to make no difference. In fact, I closed the mixture screw all the way and it still idled too fast. I removed the air cleaner to check if something was holding the slide from dropping all the way, but it was fine.
I tried to stay hydrated while working and riding today, but by the time the second pilot jet hadn't worked the combination of 100 oF temperature and gasoline fumes had given me a headache of 10 on the Richter scale so I called it quits for the day. I doubt if I'll be thinking clearly for some hours so I would greatly appreciate suggestions to help me locate the cause of a too-high idle.
p.s. In case it's relevant, the "new" 1036 has a small hole in the part of the casting where the tickler would be located if it were a RH version while the "old" one doesn't. Gasoline spurts from that hole when the carburetor is tickled (which is why I noticed it), but dribbles from the tickler on the "old" 1036.
p.p.s.. As a matter of possible interest, the third image is a ~30 second plot of the AFR after I returned from my last run with the 2-stroke body and let it idle without touching the throttle. It idled quite smoothly and regularly during those 30 seconds, but note how much the AFR wandered around.
Last edited by Magnetoman; 06/22/192:14 am. Reason: p.s. and p.p.s.
What a difference a not-headache makes. Re-examining the carburetor this morning I quickly found the throttle cable is taut and so it was holding the slide above the throttle stop. When I looked into the bore at the end of the day yesterday it only appeared to be going all the way down. This explains both the high idle and the way it responded to the mixture screw.
My problem now is that the #3 slide is an aftermarket brass one that, although it fits very well, possibly has the platform for the cable (and needle) a bit too low since after shortening the adjuster as far as it will go it still left the inner cable with tension on it. Another possibility is I didn't get the ferrule fully in the countersunk slot for it when I changed the slide (or the countersink isn't deep enough). No matter what, I'll remove the slide and make careful measurements of it vs. a genuine AMAL one to also be sure having the needle on the #1 slot drops it into the needle jet by the same amount as if it were in an actual AMAL slide. However, shortly we leave for a birthday party for my younger granddaughter so further carburetor work is on hold until tomorrow.
After performing lifeguard duty for a dozen screaming kids for two hours in the 98 oF sun we turned that part of the birthday chaos over to the parents and made our escape. Although it was too late to do much work on the bike, I needed to know what was causing the problem.
Originally Posted by Magnetoman
(or the countersink isn't deep enough
Bingo! Give that man a prize!
As the first composite shows, the soldered fitting on the throttle cable projects below the bottom of the brass aftermarket slide, but is well countersunk in the AMAL. Measurement showed the AMAL countersink is fully 0.168" deeper, explaining why the aftermarket slide didn't touch bottom yesterday. Or why the cable was still taut today even with the adjuster at the end of its range.
Although the countersink was made incorrectly, the second composite shows that the needle projects beneath the bottom surface by the correct amount to within a few thou., and the third composite shows the brass slide is only 0.016" taller than the Amals (since the needle projects the correct amount, the bottom is the correct height, which means the excess 0.016" needs to be trimmed from the top). At worst a too-tall slide would project slightly into the airstream at full throttle, although it doesn't appear the 0.016" is enough to cause this. However, since it's easy enough to do, and since I'll be drilling the countersink deeper on the mill anyway, I'll take this much off on the lathe tomorrow.
Related to AMAL work, the 3½ and 5 slides from England arrived today. For future tuning work I now have a full set of 1000-Series slides from 2 through 5 by halves, missing only 4½.
The heights of four AMAL slides all were less than that of the aftermarket slide so I skimmed 0.012" off the top to make it the same length as the tallest of the Amals, as shown in the first photograph.
AMAL seems to have used a #29 (0.132") drill bit to make holes for the cable in the slide, but the holes in the brass slide were slightly larger at 0.138" so a 3.5 mm drill bit probably was used. Drill bits don't like to track straight in holes with slots on one side so I used a miniature 9/64" end mill (0.141"). The depths of the countersunk holes for the throttle cable were 0.302" in three of the slides and 0.294" in one of them so I increased the depth in the brass slide to 0.300" as shown in the second photograph.
With the deeper countersink the cable adjuster allowed a little slack so I installed the carburetor, started the bike, and now the throttle stop screw worked as did the mixture screw. I had changed back to the #30 pilot jet and the mixture seems best with the screw 1½ turns out.
If I had been thinking more clearly prior to my first jetting run today I would have realized the recommendations for the Thruxton probably assumed a temperature closer to 68 oF than to 98 oF which according to the Mikuni calculator means I should have started out with a 310 jet rather than a 320. to have the equivalent of Thruxton jetting. However, that wasn't my biggest problem today.
I had carefully put two larger jets, and two smaller, than the 320 in a plastic bag, and just as carefully left that bag on the workbench. I did, however, remember to pack the tools needed to drain the float bowl and change the main jet. So, after the bar display on the Innovate showed an AFR of ~10 on a full throttle run I stopped on the side of the road, drained the fuel, removed the 320, and only then discovered I didn't have the bag of jets in my backpack. So, I put the 320 back in the carburetor and continued on while concentrating on lower throttle settings where the main jet doesn't have much effect.
Unfortunately, the missing bag of jets wasn't my only problem. I use a 12V, 2.3 A-hr sealed battery to power the Bosch sensor, which draws 1 A. I deliberately bought a small battery that is good for somewhat over two 45-min. jetting runs to make it easiest to attach to various places on a bike. The low capacity is fine because the Innovate only records for 44 min. and it takes at least an hour to download and analyze a set of jetting data and make the necessary changes to the carburetor. That allows ample time to charge the battery back to full capacity using a 1.2 A Battery Tender. That is, if I remember to attach the charger. Which I hadn't. So, a few minutes after not changing to a smaller jet on the side of the road, the A/F meter showed the battery was too weak to continue powering the sensor.
Back at home I analyzed the ~20 min. of data that had been captured before the battery died and determined that if the 320 were correctly marked than a correctly-marked 270 would give 12.1:1 at full throttle. So, I swapped to one marked 270 and on the next run it gave 11.5:1 at full throttle, which is within what I've come to expect as the variation in the marked sizes of these jet. Although still a bit too rich at full throttle the data shows the mixture is too lean in the range controlled by the cutaway.
The easy part tomorrow will be to switch to a main jet marked 260, but it will take slightly longer to change the #3 slide to a #2.5. Fingers crossed that one more run will have the jetting for the 4-stroke body nailed down
Even after many hours spent at the flow bench measuring a half-dozen carburetors having different spray tubes and air jets, and more than a dozen 30-min. jetting runs resulting in 150 pages of printouts that determine the AFR vs. throttle position with different main jets, slide cutaways and needle positions, I can't claim I know everything there is to know about Amals. But, I now have a much better understanding of what I don't know than I had a few months ago. I now know that I don't know things that a few months ago I didn't know there were to know.
One thing all the AFR data I've collected tells me is that anyone who says he can correctly jet a carburetor by looking at the spark plug has psychic abilities I don't posses. The complexity of the interaction between slide, jets, and needle position, and the fact the bike "feels" like it is running well for significantly different AFR values, means that no matter even if reading the spark plug gave a perfect value for the main jet,[*] that's only a single data point on a curve that undulates up and down in regions where that one data point has essentially no effect
[*]That's assuming the full-throttle plug-chop that was used to give a perfect reading the plug was done only after the bike had reached near maximum speed. If the plug-chop is done even after reaching 75 mph the AFR would have been still dropping so the plug reading would have given an erroneously lean result for the mixture.
Over the past weeks I've determined that my Gold Star starts, idles, accelerates, and runs well for AFRs that can be anywhere between 10:0:1 to 15:0:1 for a given throttle setting. Even if the main jet were perfect, only if the seat of my pants were calibrated a lot better than it is could I hope to say based on how it "feels" whether the AFR at, say, 1/4 throttle were 15:1 or 11:1. As an example, at the end of today's run I let the bike idle for 40 sec. without touching the throttle. The seat of my pants told me it idled smoothly the entire time. However, the attached graph shows the AFR varied between 11.3 and 14.4 during that time.
I left the #3 cutaway in for the moment, but today's jetting run was a disappointment in that the '250' main jet I put in this morning turned out to be slightly leaner at full throttle than the '260' it replaced (despite which, the seat of my pants told me it ran great with both main jets.). Aside from the annoyance of AMAL jets "never" flowing what their markings say they will, the issue I want to address is it goes very rich (~10.5) over a narrow range of very small throttle settings, but then goes through a lean-ish (~14.0) peak at ~1/4 throttle. The half-size smaller cutaway I had planned to install today was to address the lean-ish peak, but it would make that too-rich valley even richer.
I'll give this some more thought before deciding what to change next. Unfortunately for me, the forecast shows a steady climb to a high of 107 oF by Friday, where it will plateau for a few days, and the long-term forecast shows it no lower than 101 oF for the next two weeks. Those of you who have never experienced riding in desert air above 100 oF can simulate the feeling by holding a hair dryer on 'high' a few inches from your face. Once your body is perspiring as fast as it can, going faster only makes it worse since your internal cooling mechanism is already maxed out.
I have to say that, with the advantage of time zones (I'm 12 hours ahead of the US West Coast), this thread is my first port of call when I get up in the mornings!
I've always been keen on getting carburetion right and have long believed that the seat-of-the-pants assessment is little more than a go/no-go gauge, in terms of gross assessment of direction of change and the owner of the pants is unlikely to notice less than 10% differences between setups, whien using the 'concrete dyno'.
I had an instance a few years ago where I was running a 4 cylinder 750cc two stroke road-racer that I knew made 120hp at the rear wheel. For originality, when the rings needed changing (at 250 miles!) I reverted to that bike's original 700cc cylinder setup, which was supposed to make 90hp, ex-factory. Keeping the same ignition, exhausts (aftermarket) and carbs (rich), with some subtle piston mods, the bike made 118hp on the same dyno, though with much better mid-range power than the full 750cc setup.
Riding the bike (despite it being no slower, pulling hard through 170mph), the rider perceived the bike to be SLOWER in acceleration...all because the torque and HP curves were a better shape and had less 'step' onto the power, making the acceleration seem less aggressive. We knew from the dyno what the real story was. Actually, the bike would have been faster around a given road circuit as the modified 700 than as a full 750, all due to the piston mod-inducing mid-range go, despite all the 'wisdom' saying that the later 750 version cylinders made more power (certainly true if apples are compared with apples; the mods to the pistons in the 700 cylinders have ben incorporated on the full 750 pistons to get the 'full effect'...I should know how much in a few months!..I'm betting on 130+ hp..low numbers these days but exciting enough in a bike that weighs less than a GS)
The point of my tale is that perception is not foolproof and can be misleading, though, with effort, good carb setup can be achieved (perhaps with a stopwatch). However, the knowledge and effort of definitive data like MM produces is the gold standard. Solutions to problems where most of us didn't even appreciate the problem!
Last edited by Kerry W; 06/25/1912:18 pm.
No generalisation is wholly true, not even this one. Oliver Wendell Holmes
Reminds me of a story about Nicky Hayden when he went for the first couple of laps on the RC51. When asked what he thought, he mentioned it could use a bit more power. Then they showed him the lap times.
Also reminds me of the 1/4 mile times of a Yamaha RD 400 and a Honda 400 Hawk.
The AFR graph at idling is interesting to me, and I assume that you find some interest in it as well, as you took the trouble to post it. Thanks. I enjoy speculating. It is a small sample to draw much in the way of inferences from, but there is a suggestion of a fundamental variation with a period of ~20 sec, with a higher frequency variation imposed upon it.
At such a minimal fuel demand, I suggest that the float mechanism reveals its resultant hysteresis on the fuel level. The fuel demand is in (relatively) infrequent pulses, which may be followed by a return "after-pulse" back into the carb. The float needle is likely to be opening and closing under these influences (as contrasted with the needle hovering at varying lift from the seat at higher fuel demand). Hence the hysteresis in fuel level, as the float mechanism can only act after the event and likely allow a bit more than necessary to pass the needle before it shuts.
This hysteresis in fuel level will be reflected in the AFR, and perhaps superimposed will be the individual draws of the induction strokes (and brief pressurisation of the bowl by return pulses, if any). The latter ~20 sec of your plot does fit over the first half, in a loose fashion. The general trend seems evident. Assuming a 1500 rpm idle for illustration, those 20 sec have 250 induction cycles. The graph is of insufficient resolution to observe on that scale, what impact if any the induction has upon fuel level, and then its impact upon successive induction mixtures. I doubt that this is of more than academic interest, as the float needle will be acting far less hysterically when larger flows are going through it. The valve will then be open all of the time, just varying in opening in response to the amount of fuel level drop (and consequent float drop).
At the end of the day, I think there will always be a delay between demand (say opening throttle from half to full) and the fuel level dropping enough to open the valve sufficiently to equal flow demand. Carburettors will always be playing catchup in fuel level, fortunately it isn't too critical in most situations.
...this thread is my first port of call when I get up in the mornings! ...Solutions to problems where most of us didn't even appreciate the problem!
Thanks very much for your comments. This thread is a result of someone with a long career in experimental physics and instrumentation who is (mis)applying that background to the study of obsolete... er, I mean, to British motorcycles. Also, not sponsoring a race team or selling anything means I don't have any motive to keep my findings secret.
If I only wanted my Gold Star to run well I would be done by now. Unfortunately, I also want to understand why, and that takes instrumentation and time.
Originally Posted by Kerry W
Riding the bike (despite it being no slower, pulling hard through 170mph), the rider perceived the bike to be SLOWER in acceleration...all because the torque and HP curves were a better shape and had less 'step' onto the power, making the acceleration seem less aggressive. We knew from the dyno what the real story was.
In principle, having AFR=12.5 at all throttle positions would result in maximum h.p. at each setting. Thus far I've concentrated on collecting AFR data at fixed throttle settings, not throttle response data, but the attached graphs illustrate an important difference between fixed settings (and dyno data) and performance on the road.
The first graph shows AFR in purple and throttle position in red for a 13-sec. interval yesterday. Full throttle is 1.5 V so the ~0.2 V peak at 9:25 is ~1/8 throttle. The second graph just expands the time scale. Without describing all the data on this graph in detail, one segment of it shows I started rolling on the throttle at 9:24.4 and had it fully at the 1/8 value by 9:24.9, i.e. 0.5 sec. later. However, the AFR data shows there was a delay of ~0.5 sec. between changing the throttle and response of the AFR. A half-second doesn't sound like much but it's a lifetime on the racetrack. And, on the street, if my engine had hesitated for 0.5 sec. instead of responding smoothly every time I changed the throttle it would have made for an awful ride.
I didn't have the tach hooked up for yesterday's session but I'll guess from the throttle setting that the engine was at ~2500 rpm, which means it fired ~10 times in that half-second. The response time of the sensor is faster than 0.1 sec. so, taken together, this means the carburetor is to blame for the delay in response of the AFR to the change in throttle. This "transient response" behavior of a carburetor is just as important as having the correct AFR at steady-state fixed throttle settings.
Anyway, in case anyone has followed along this far in the discussion, a dyno session could determine the AFR under load at 1/8 throttle and 3/16 throttle. However, what also is important on the road is how the engine responds when the throttle changes between those two settings, not just when going up a steep hill (roughly equivalent to a dyno measurement), but also how it responds on a level road as well as going downhill, each of which will have different AFR values at the same steady-state throttle setting. Also important for good behavior on the road is how it responds if a change is made in 0.1 sec. or 1 sec. Although not too much time at a racetrack is spent going downhill at 1/8 throttle, the same considerations apply at larger throttle settings.
To a large extent the response time of a given carburetor is fixed by the design of that carburetor. However, although I have a GP, 1038 Concentric, and 38 mm Mikuni on the shelf, for reasons I'll skip for now I've arbitrarily decided on a 1036 so I have to deal with its fundamental limitations instead of those of other carburetors. But, what my ~5+ hours worth of measurements on the 1036 indicate is tradeoffs can be made between AFR in a given range and throttle response in that range. I'm sure I'll have more to write about this in some future post when I turn my attention to transient response measurements.
By the way, I'm not saying there's anything wrong with dyno measurements, just that conditions on the road aren't the same as in a dyno cell. On the other hand, an advantage of dyno measurements is you don't have to worry if a sheriff's car might be around the next bend.
Re Nicky Hayden, I may have this wrong, but I think Jeff Smith had a similar experience switching from the 500 Gold Star to the smaller lighter 440cc Victor. He got off his test laps on the smaller bike and said, "No effing good, give me back my Gold Star'. Then they showed him the lap times. End of an era.
1949 BSA ZB34 'Bitsa' 1959 BSA DBD34 Catalina 1973 Norton Commando 850 R 1974 Norton Commando 850 R (I know, one too many) 1975 Honda TL250 Trials, a new addition to the family 1998 Montesa Cota 315 HRC 2004 Ducati M1000ie
It was going to be over 100 oF today, but despite my better judgement.. OK, scratch that, I'm working on a 60-year old motorcycle rather than doing anything even remotely useful so clearly 'judgement' is lacking.
If the '260' jet in the previous full-throttle run could be trusted as having been accurately marked then I should have installed a '220' or '230' today. But if the '270' in the run before that could be trusted then the '260' would have worked yesterday rather than giving an even richer full throttle AFR than the '270'. Anyway, I got a fairly early start before it was too hot, dropped the main jet to '250' and swapped the #3 cutaway slide for a #2.5. Which brings us to the next installment of "Bet You Didn't Think You Needed to Worry About That."
The first photograph shows that when I installed the #2.5 slide in the carburetor the needle wasn't hanging straight down from it. I might have been able to shove it sideways far enough to insert into the jet but that would have left the needle constantly rubbing one side of the jet rather than rattling around and bashing all sides as AMAL intended.
The second photograph shows that the hole for the needle was drilled at an angle. How this could have happened is not explained by the measurement shown in the third photograph. With the slide held in a jig so as not to rely on the end of it being perpendicular to the sides, and the gauge zeroed at the front of the slide, the rear of the slide is higher by 0.013". However, that isn't sufficient to quantitatively explain the measured tilt of approx. 0.15" out of 2.5" = 3.43 deg. Further, the gauge found no difference in height from side-to-side while the first photograph shows there's definitely tilt in that direction. But, irrespective of how AMAL screwed it up, it needed to be unscrewed.
Drill bits tend to follow existing holes so as the fourth photograph shows I first used a 7/64" (0.109") miniature end mill to straighten the hole and then a #32 drill bit (0.116") to make it the correct diameter. The needle itself is 0.098". Note that the fixture holding the slide is sitting on parallels to ensure the hole is precisely perpendicular to the sides of the slide.
Despite the machining, the needle still had a slight forward tilt when I did a test fitting in the carburetor. This differs from bench measurements because when installed the spring presses down on the needle clip. I already had measured the face against which the needle clip rests and found it properly perpendicular to the bore, and there were no burrs on the edges to interfere with the clip seating fully, so the problem had to be caused by the clip itself.
The fifth photograph shows three clips, two the same shape with the color of copper and of steel, and a third at the upper left with a different shape. It was over 100 oF by the time I found that the third type worked perfectly so I didn't try to figure out what subtle difference in the clips was causing the problem. I installed the slide, bolted the carburetor back on, and am now inside the air-conditioned house typing this. The next jetting run will have to wait until the relative cool of tomorrow morning.
One cannot have too much tooling for a lathe or mill. I'm not sure which would have been worse, not knowing miniature end mills exist, or knowing they exist but not having a set of them when I needed one. I've found that no matter how much dust any piece of uncommon tooling collects on the shelf it always will be needed for some unusual task.
Aside from manufacturing defects, such as having the hole for the needle drilled at 3.4-deg., it would make setting up a carburetor ever so much easier if AMAL had stamped numbers on the main jets that were somewhat exact rather than crude approximations.
I made progress with today's jetting run, but there's still more work to do. For future reference, changing from a #3 to a #2.5 slide enrichens the mixture by fully 2-2.5 AFR units from just off idle up to ~0.3 throttle (e.g. dropping the AFR from ~13.5 to ~11 at 1/8 throttle). Unfortunately, the #2.5 made it much too rich from just off idle up to ~1/8 throttle, while I needed a richer mixture only starting at ~1/8 throttle and from there up to ~1/3. So, my next attempt will be back with a #3 but with the needle raised one notch. Today's results showed the present '250' main may be OK, but it might need a '240'. See the next paragraph for an explanation.
As for determining the main jet size, assuming someone could accurately decide if the mixture was too rich, too lean, or just right by reading the color of a plug, how should a plug chop be done? We know to give the engine full throttle for some length of time, then simultaneously hit the kill button and pull in the clutch. But, for how long do we need to keep the engine at full throttle before doing the chop in order to get an accurate value? Prior to starting all these jetting runs I would have thought that 15 sec. of full-throttle in 4th gear headed up a hill until reaching ~5000 rpm and ~75 mph (in a 35 mph zone...) would have been sufficient, but the A/F meter tells a different story. As the attached graph shows, the AFR was still dropping under those conditions and thus a plug chop would have been at a point where the mixture was leaner than it eventually would be if given more time to reach even higher rpm (and higher speed). For more than one reason, I suspect for determining main jet sizes that many plug chops may be only slightly more reliable than reading Tarot cards.
I managed today's run while it was still in the 90s, but the high will reach 106 oF. The next three days will be even worse at 107-108. In order to do a run tomorrow I'll have to get started in time to swap the slide first because already it's too hot to consider doing that today.
On a final note, AMAL's settings for the 1036 on a 1970 Velocette Thruxton include a 320 main jet, #3 cutaway, and needle on the top, #1, slot. Although I'm still not quite at the final configuration of the "4-stroker-erized" 1036 body, I'm close enough to compare the results. I'll go out on a limb to say that nearly all 1036s were made for 2-strokes, but based on my measurements that AMAL supplied Velocette with a "4-stroke" version, i.e. with flat spray tube and no restriction in the air jet passage.
Just a guess because it’s been decades since doing plug chops. I gave up when jetting a Kawasaki triple with expansion chambers and I was not so blissfully unaware that I was being chased by the local constabulary. No mirrors and couldn’t hear the siren..... anyhow... thinking in a lot of cases plug chops are pretty much pointless simply due to questionable fuel quality and varying quantities of ethanol.
To add some information about the AFR range over which the Gold Star seems to run fine, the attached graph pieces together two ~30-sec. time segments from today (i.e. #2.5 cutaway) when I was following a line of cars down a hill for over a mile at ~40 mph. Because of the speed I probably was in 4th, but because of the hill I hardly had to use any throttle to maintain my speed. Note that when the throttle position sensor was above ~0.1 V the AFR was well behaved even at a very rich ~9.5:1, but when the throttle dropped below that value (i.e. barely above idle) the AFR became squirrelly.
The behavior near times of 22:05 and 22:30 wasn't due to the engine missing, because that shows up as sharp spikes to much higher AFR values, so it probably was incomplete combustion. I could feel the engine not behaving quite right a few times when coming down that hill with the throttle barely open but, other than those times, the bike seemed to run great today and was responsive to the throttle.
The point of this is that from these and other measurements I've made, as long as the mixture is within the range ~9.5-~16 the bike feels like it's running great. Without the ability to measure AFR I wouldn't have any idea what is actually going on in the combustion chamber.
I will be very interested to see what the Innovate tells me about the jetting in the BB and Catalina, both of which feel "perfect." However, those measurement almost certainly are going to wait until temperatures drop in the fall.
Pertinent to 1036s on Thruxtons; the Australian tech editor of the theirl Velo club, in the compilation of his articles " Norm's Technicalities", says he has encountered Thruxtons that came equipped with two stroke configuration carburetors. His assessment is they work fine as long as you do not mix four stroke and two stroke parts. The basic settings do not vary between the two versions.
the Australian tech editor... assessment is they work fine as long as you do not mix four stroke and two stroke parts. The basic settings do not vary between the two versions.
My flow bench measurements show that if AMAL's recommended 320 main jet is correct when in the 4-stroke body, it would flow the same as a 390 in a 2-stroke body. Stated differently, if the 320 jet gave 13:1 at full throttle in a 4-stroke body it would give 10.7:1 in a 2-stroke body.
At 1/4 throttle, where the slide cutaway dominates, using the same #3 slide in a 2-stroke body would increase the richness by 17%, equivalent to increasing the richness from 13:1 to 11.1:1. I didn't do any measurements with the needle in the Thruxton #1 slot so I can't comment on that.
Although my measurements on the Gold Star show the bike doesn't exhibit obvious problems over a wide range of AFR, that's not the same as claiming "the basic settings do not vary" between the 2-stroke and 4-stroke versions of the body. They certainly do. By a lot. "Seeming to work fine" isn't the same as "actually working fine."
On a different note, I might have commented on it before, but the Eddie Dow TLS with Vintage Brake linings on this bike is the closest thing to having a front disk that anyone could hope for.
p.s. my younger granddaughter had an end-of-summer-camp dance recital at this morning and by the time I got home it was already 105 oF. Despite that, since I've done it enough times I could do it blindfolded (but please don't test me on that), I removed the carburetor and swapped for a #3 cutaway slide with the needle now on the 2nd slot and a '240' main jet. After lunch when it's up to 106 oF I'll do a jetting run... just kidding, no way am I riding at that temperature. Tomorrow there won't be a dance recital so I'll hope to get on the road before the snow melts.
Last edited by Magnetoman; 06/28/197:04 pm. Reason: p.s.
the 932 was over kill on the M20 so I can not see me putting a 1000 on it
I think you underestimate how much the lack of a proper carburetor is restricting your M20's h.p. potential. Yesterday a friend gave me a box of a half-dozen small Amals, one of which is marked '41 M20. As the attached shows, with a 1036 from a Gold Star your M20 no longer would have its h.p. restricted by only breathing through the tiny main bore, it also would get additional air from the mounting holes. Since both engines are 500cc I'm sure you can find the necessary adapter from one of the usual suppliers.
Mission Accomplished. It was 101 oF when I made the jetting run this morning so the snow plows already had cleared the streets and were nowhere to be seen.
After reviewing, and revising where needed, today's and previous results, the first graph shows the best jetting I found for 2-stroke and 4-stroke AMAL 1036s after ~15 jetting runs covering ~200 miles and resulting in ~160 pages of printouts. More -- much more -- information about these carburetor bodies and jetting will be in a forthcoming multi-part thread.
Keep in mind that at a more human ~75 oF the equivalent main jets would need to be one size larger to give the same results. Or, keeping the same jets, the curves would be ~0.5 AFR leaner.
From my full-throttle flow bench measurements I would have predicted that given an accurate '200' main jet in the 2-stroke body, it would have taken an accurate '255' main jet in the 4-stroke body to give the same AFR at full throttle. Given the issues already noted with the marked sizes of AMAL main jets, the results on the graph have to count as in excellent agreement with the flow bench measurements. If the two jets were actually 193 and 248 (i.e. less than one size off from the marked numbers) the agreement would be perfect.
It's not clear to me that the "flatter" overall behavior of the 4-stroke curve gives it any sort of performance advantage. However, although both versions of the 1036 body work well on a Gold Star the 4-stroke version does have the advantage of being leaner (but still rich enough) at nearly all throttle settings so it would be less damaging to the wallet. As a very rough estimate, if most riding is done between 1/4 and 3/4 throttle the fuel savings with the 4-stroke version will be ~5%.
The second graph shows that starting from near top speed in 1st it took 15 sec. at full throttle heading up a hill before the AFR stabilized at its final value. I was going over 75 mph at the end of that run which means during that time I covered ~1/4 mile. A shorter run reaching a lower speed would have given an erroneously high AFR reading.
Not to pick on him, but this post from two years ago illustrates an important point:
Originally Posted by Belfast Bikes
DBD Clubman ... AMAL 1038 carb that is fully converted for four stroke use. The bike has a standard exhaust pipe and muffler,... a "minimalist" gauze type air filter. ... the bike runs well most of the time ... except at wide open throttle ... At that point, the RPM burbles and bobbles around 4500 to 5000 RPM, not 8-stroking and not surging, just hitting and missing until I close the throttle adequately to get things cooking again. ... current carb settings are: 35 pilot jet, 4-stroke 107 needle jet, 4-stroke (2-notch) needle in the richest position, 3.5 slide (a well-fitting, brass one), and a 4-stroke spray nozzle. I have tried main jet sizes from 330 to 430 ... all with the same result..
Even if "fully converted" really was the case, i.e. including drilling out the air jet passage, the smallest main jet he tried is 8 sizes larger than the one that my tests show would give the proper mixture in the cool air of Belfast. His smallest main jet would have resulted in an AFR at full throttle of ~9:1, along with the burbles and bobbles that come with being way too rich.
His use of a 107 needle jet along with the needle on the #3 position would have made things quite rich at lower throttle settings as well, only partially compensated by the weaker #3.5 slide. That his jetting worked at all is consistent with the fact my Gold Star seemed to run fine with AFRs anywhere in the range 10-15:1. This tolerance of incorrect mixtures also explains the fairly wide range you will find if you search for 1036 and 1038 jetting recommendations.
Speaking of Concentric carbs .. My departed friend GStar Ron (RIP) gave me the idea for a idle stop for the "smooth bore" racing carbs. Merely tapping the vent hole in the top of the slide and fitting a short #2 screw that protrudes into the inner with a locking nut on top works well after 'fiddling' adjustments for best idle speed. The end of the screw strikes the alu casting in the slide cavity. Make sense?
BTW, having a idle stop makes engine starting much easier vs trying to hold the throttle barely open.
Another issue on this carb Ron and I figured out, was the float bowl level running low especially when the fuel level in the tank was low. This condition will drive you silly especially when attempting to jet with a full throttle high speed run. We unsuccessfully tried drilling out the float bowl needle seat. Again Ron came up with good fix by ordering up the higher flowing alcohol float bowl from the UK AMAL supplier. An expensive fix, but it works.