Project: 1939 Jaguar SS100 Replica Classic Roadsters, LTD. "Duke"

[Bruce Nunnally]

Idle was loping and not running evenly. Pulled plugs 1 and 3 and both were running rich, and black smoke from tailpipe.

Started over on carb idle mix screw adjustment -- all the way in, then turn out 2 turns, then adjust with engine running.

Turn tighter until engine runs better, then continue until rougher, then back off through better again,

until engine runs rougher, then back to better.

Now engine is running much smoother. Adjusted the idle up a bit. Ran the car around the block, and around the

neighborhood and it is well behaved and drives well.

Edwards Mechanical still otw later today to work on the front suspension. Then we're back on the program of

driving it more and doing more checkout to get ready for a 21 mile trip to Truett Worrall spring shop.

[Bruce Nunnally]

Today I reset the idle mix and idle speed and idle mix

again after a timing and voltage reset yesterday.

I stretched the throttle spring to force the

Linkage home more consistently.

The idle at temp is 1100 in neutral or park but

falls to 800-900 in drive (due to poor auto

clutching?)

The roadster is driving and idling okay.

Near term To do list: (Tues)

Stabilizer bolt repair, bushing install, bar install

Strut arm bushing replacement

MSD install trim

Another day issues in this phase: (susp & safety)

Rear suspension - trip to Truett Worall, alignment

Fuel sender identification, repair and replacement

Oil pressure gauge reading varies depending on gear

Rear Drum brake shoes and cyl and spring refresh

(parts in hand except new drums)

Brake master cyl replacement and fluid refill

(parts in hand except fluid)

Coolant sys drain and refill

Backup lights not lighting

[Bruce Nunnally]

I added some bungie-cord tie-downs for the top in the Roadster to keep it from parachuting. This tested well on my

drive, and the top stays out of the rear-view mirror and out of the action at 60+ mph.

Today I took the Roadster 8 miles north to a flat road to try to get a performance baseline. The old hardware worked fine,

and the Roadster got there and back alright. Good news, as I am trying to build confidence for a 26 mile run to the Spring

Shop. The test gear did not, and I ended up with 2 useless runs.

Not sure yet what to do differently. I am using PocketDyno+ on the Ipad1 due to no aux power hookup in the Roadster.

I have found that one needs to calibrate the accelerometer in the ipad on each use for good results so I did that ahead

of leaving home.

The first run I tried leaving the shifter in low and holding the Roadster until 'redline' at 6000 RPM.

This resulted in a 0-30 mph result with a max recorded speed of 39.58. I think 1st gear has a max actual speed of

47 mph.

PocketDyno+ Run 1:

0-10 0.61

0-20 3.02

0-30 8.68

Max hp was figured at 72 WHP (53 kW) or 72/.8= 90 crank HP.

The second run I shifted at 5500 rpm or so to 2 and the Roadster shifted smoothly into 2nd and continued to accelerate

through 60 mph. This method worked well from a seat-of-the-pants evaluation.

PocketDyno+ Run2:

0-10 0.52

0-20 1.33

0-30 2.23

0-40 3.43

Max velocity was recorded as 43 mph, max power as 114 whp (85 kW).

The shape of the curves is perhaps right, but the values are unreliable.

We are hoping for something more than the stock 75 hp since the Roadster has less exhaust backpressure

and more carb, but I am literally running these tests to try to get a baseline, and the results are not yet

consistent enough to use as a baseline. More testing needed I suppose.

.

[Cadillac Jim]

Even for very loose calibration, those numbers are very encouraging. They basically say that you are getting pretty much what the 3T-C was designed for, and probably more, as you are hoping with the carb and exhaust. That's great, particularly after the ignition breakdown and having to kludge the transmission shifting.

[Bruce Nunnally]

I put on my car coat and had a nice morning drive in the Roadster. A bit overcast here but didn't look like rain.

I wanted to try an alternate approach to data collection. One upgrade from the ipad1 I have to the ipad2 or iphone4 was an improved accelerometer. The newer models gained 3 axis accelerometers. I use an iphone 4 as a daily phone (one my Wife discarded) so easy enough to install the pocketdyno+ there and use it for data capture.

Same process-- entered the car data, calibrated the accelerometer, then did some runs on a flat level road to capture data.

Unfortunately, same result -- no usable data.

Next I am planning to install an aux power outlet in the Roadster to allow use of my gps based performancebox.

This will also allow me to check the Roadster speedo in real time.

On the bright side, I had a nice drive. The engine is running well, but it has more vibration at idle than I would prefer, so we'll have to look into that. Max speed was 55-60 mph for the trip today. Overall the Roadster has returned to solid reliability so far, and starts and runs well.

[Bruce Nunnally]

Nice; returned several spring accessories we didn't end up needing / using and that paid for the items I wanted to install a new power outlet with fuse and appropriate wiring gauge, etc. for the Roadster.

Also I splurged and a 2000 lb scale is coming from scaleman so we can agree on exactly how much the Roadster weighs. Weigh each corner, add the 4 together, voila.

[Texas Jim]

A few miles N of you on 75.. at exit 48... there is a Flying J truck stop...

About 6 or 7 bucks and they will weigh the car...

[Bruce Nunnally]

This morning I did a temporary install on the auxiliary power outlet just to try things out. I spliced additional wire and a inline fuse to

the aux power ponytail. Then I added some alligator clips I had from the kids' science project leftovers, which let me grab power

and ground connections temporarily. I just left the outlet in the passenger floorboard, and plugged in the PerformanceBox.

I would like to point out how flexible the suction cup mount on the PerformanceBox is -- nice simple piece of engineering there.

This let me use the gps-based PerformanceBox for testing on this morning's runs.

Run num Speed(mph) Time(s) Dist(f) Graph run Description

2 0.00 0 0.00 Speed Start
2 10.00 1.22 8.55 Speed Line
2 20.00 2.71 42.07 Speed Line
2 30.00 4.64 42.07 Speed Line
2 40.00 6.77 222.98 Speed Line
2 50.00 9.67 415.39 Speed Line
2 60.00 14.02 769.02 Speed End

3 0.00 0 0.00 Speed Start
3 10.00 1.11 8.83 Speed Line
3 20.00 2.68 44.14 Speed Line
3 30.00 4.73 119.61 Speed Line
3 40.00 6.97 234.93 Speed Line
3 50.00 9.99 435.60 Speed Line
3 60.00 14.43 796.89 Speed End

4 0.00 0 0.00 Speed Start
4 10.00 1.13 8.59 Speed Line
4 20.00 2.66 42.95 Speed Line
4 30.00 4.61 114.88 Speed Line
4 40.00 6.79 226.96 Speed Line
4 50.00 9.67 418.58 Speed Line
4 60.00 14.04 772.48 Speed End

The results were surprising -- 0-60 mph without roll-out (which would be 0.3-0.4 sec less in this case) in around 14.04-14.4 seconds.

The two lower numbers were in the same direction, so say 14.2 sec is fair.

This is much better than expected, and suggests the Roadster is lighter than thought. Weight, performance, and power are related in measurement.

So if the Roadster weighs 2,800 lbs, it would need 118 hp / 118 lb-ft of torque for this type of performance.

If the Roadster has only 100 hp / 100 lb-ft, it must weigh around 2,300 lbs for this type of performance.

If the Roadster only has 90 hp / 90 lb-ft, it must weigh around 2,100 lbs and so on.

[Bruce Nunnally]

What do I think the 3TC in the Roadster puts out in its current state?

In the 79-81 Corolla the engine was rated at 75hp @ 5000rpm and 95 ft lb of torque @ 2600rpm.

+5 for timing adjust

+unknown for Weber 32/36 carb (15-25??)

+5 for opening the exhaust (free-flowing)

So it is possible that the current state 3TC might put out 100-110 hp. So it could be at 110 hp / 115 lb ft at this point?

If that is the case, 110 hp / 115 lb-ft of torque, then a 2,600 lb Roadster would have a 0-60 time of 14.2 sec.

We'll know more once we get the weight of the Roadster. I THINK I know the gearing, so the weight will get us in the ballpark.

In test by the Autocar magazine in 1937 the 2.5-litre (20 RAC hp rating) car was found, with the windscreen lowered, to have a maximum speed of 95 mph (153 km/h) and a 0–60 mph (97 km/h) time of 13.5 seconds. With the 3.5-litre (25 RAC hp rating) the top speed reached the magic 100 mph (160 km/h) with a best of 101 mph (163 km/h) over the quarter mile and the 0–60 mph (97 km/h) coming down to 10.4 seconds.

So that is still our goal -- to get down from 0-60 in 14.2 sec to 0-60 in 10 seconds, which will need 155 hp with the current transmission

or around 140 hp with a T50 manual transmission.

[Bruce Nunnally]

The Siltec WS-2000 one-wheel scale arrived from Scaleman today. With one scale, jack up each corner, let it down on the scale,

record the reading, jack it up, remove scale, let it down. Rinse, repeat.

First weigh-in for the Roadster:

RF: 560 lb

LF: 626 lb

RR: 586 lb

LR: 552 lb

Total Front: 1186 = 51%

Total Rear: 1138 = 49%

Total Weight: 2324 lb

This is less than the 2,800 lb on the paperwork or the weight I was guessing of 2,600 lb.

I will repeat the weighing again next opportunity; I forgot to 'bounce' each corner after putting on the

scale so could be +_ 10 lbs each. I wanted to stop and rethink all my other assumptions based

on the approximate weight of 2300-2350 lbs.

The good news here is the car is relatively light, and that the front to rear balance is close which is good for handling.

it also means the Roadster weigh less than the original ~2,600 lb SS100 (not quoted if that is 2.5L or 3.5L)

One Cartest simulation tiral-and-error fit would be 97 hp at 5000 rpm and 110 ft lb of torque at 2,000 rpm with this gearing

and weight would yield similar performance numbers to my observed numbers above (and a top speed of 97 mph).

[Cadillac Jim]

One thing to watch is that the brake is not engaged on the wheel on the scale.

[Bruce Nunnally]

Good I will watch that next test. Car was in park no parking brake on.

Today's embarrassing news is I just realized that the speedo/tach relationship I am seeing does not match

the final drive I was thinking the car has. On re-examing the sources this era Corolla had a

41/14 = 2.929 final drive not the 4.1 I was figuring. The tach in the car in 1gear mirrors the speedo

2000 rpm ~ 20 mph, 3000 rpm ~ 30 mph etc. with tires 205/75-14 or 26.1" diam that suggests

the 2.929 also. The other likely choice would be a 3.08 which was a Pinto final drive.

I see a new speedo that I think was put in with the 3tc engine and the differential visually

matches a toyota.

If toyota 2.929 1gear 2.452: 64.87@6000 rpm

If pinto 3.98 1 gear 2.452 61.69@6000 rpm

So next I will get a rider or a video and record.

Also means should hit 60 in 1 gear for 0-60 testing instead of shift arnd 5500 as I did

[Cadillac Jim]

A drive wheel with the car in park may have weight transfer too.

A low gear set up for 10 mph per 1000 RPM is great for a high-performance setup. I would leave the gearing, whatever it is.

[Bruce Nunnally]

Next on the engine tuning I need to get my wideband O2 installed and see where the air fuel ratio is.

I think default main jets for weber 32/36 carb are 137/140 and my 3tc needs only 130/130 or similar.

Likewise idle 60/50 default and needs 45/50. So I expect to find we are very rich.

I kludged a virtual dyno chart by calculating rpm via speed and gearing into the output from

the performancebox. That suggests hp is peaking at 4000 rpm and holding to 5500 rpm

instead of peaking at 5000 rpm (Corolla) or 6000 rpm redline (more carb, exhaust).

The hot ticket is sidedraft weber 40DCOE carbs instead so perhaps Instead of further tuning 32/36 I

should just refit for DCOE's? ($1400)

Hopefully spring shop trip for diagosis/evaluation Wed Sept 11

[Cadillac Jim]

Be careful with the DCOE switch. I'm sure that there are a lot of varieties of DCOE, but the famous ones have no choke and little support for low-speed or idle operation. They are basically velocity stacks with analog fuel injection.

You will need to start with making sure that the venturi size (32 mm primary, 36 mm secondary on your current carb, which works like half of an American four-barrel). I recall an article linked form this Topic about a tuner that cleaned up a previous tuner's work on a Cosworth Twin-Cam that included a throttle plate to reduce the capacity of the carbs that the previous tuner had put on; the guy that wrote the article removed both the throttle plate and the carbs and put the right carbs on the engine.

As I recall from distant memory, tuning the racing DCOEs requires adjusting the length of the velocity stack (bolt-on variations), jets, and low-speed linkages for the RPM range that the engine benefits most from the intake resonance, and to get the mixture right. The mixture will vary when the resonance takes hold, so getting the jets right is an art.

The DCOEs were replaced in racing by Hillborn injection in the 1950's and 1960's, and you can get the equivalent only better from MSD, FAST, Professional Products, Holley, Edelbrock, Accel, and others from Summit Racing and other outlets, although I would go through Toyota communities and the manufacturer to pick a system. Many of them bolt onto a carburetor mount and are throttle body injection, but you would want port injection with wideband oxygen sensors and open-loop PCM operation for performance, reliability, emissions (smell), starting, and street driveability.

I think you are going to end up with DFI eventually, so you can save the time and money, and perhaps have some warm weather driving time this year if you skip a step. The DFI system may very well be transplantable, except for possibly the intake manifold bolt pattern, if you switch to one of the DOHC Toyota engines later.

If you get street DCOEs, with diffusers on the velocity stacks (used in racing to broaden the resonance peak) and an air cleaner, low speed passages and a vacuum/ventury piston with needle valve and such like your existing Weber carb, you probably won't have anything more than what you currently have except for a damped velocity stack. I would look at the existing intake manifold, and perhaps look into its supplier, to see if there are intake resonances designed into it, as there are in most DFI intake manifolds and many carbureted manifolds (see Ram Air intake systems, the "snakes nest" look of most DFI intake manifolds, etc). Even the lowly Holly high rise aluminum intake manifolds for SBC bowties have equal-length runners tuned to 5500-6500 RPM with a four-barrel Holley mounted on it, and I could feel another resonance around 7,000 RPM in my 1964 Chevrolet tricked-out 327.

[Bruce Nunnally]

Yes the DCOE sidedrafts require more tuning and care, but oh the lure of one venturi per cylinder.

I agree that I should tune what is there as best can be at each step then decide on the next course.

On the engine front next is get the wideband installed so I can see what the engine is really

doing. I do think the main/sec jets are too large.

Of course the spring shop trip and fixing rear suspension a priority.

Not sure if I will keep it carb for historical interest or figure out an injection setup.

A tunable injection system with ignition timing control would give a lot more info for

characterization and tuning.

[Cadillac Jim]

Equal length runners == one Venturi per cylinder.

My feeling is that we have uncovered a 2300 pound roadster with vast sports car performance potential as well as amazing visual appeal. A DFI won't detract any more from the visual appeal than the MSD ignition or modern fan control module, and will just add to the stand-back-and-admire quality that is emerging under the hood.

The advantages of closed loop port DFI, particularly with the broadband oxygen sensors, include but are not limited to optimum performance, optimum fuel economy, lower unburnt hydrocarbons in the blow-by and thus better oil life and quality, lower emissions, cleaner combustion and lower maintenance, better exhaust system life and performance, and portability of most components with upgrades in engine hardware. Your MSD ignition is designed to be compatible with most ECMs, OEM and aftermarket. And, with a good ECM that accepts MAP, MAF, ECT and other such sensors, your spark advance map problems are solved.

[Bruce Nunnally]

I received several main jets and an idle jet I ordered. This reference suggests a mild 3TC needs a main/aux jets of 130/125. The default Weber 32/36 carb comes

with 140/140 jets. The main jet is responsible along with the air correction jets for the upper RPM performance, although it can influence across the RPM band.

My test results suggest the engine is pulling okay to ~3000 RPM, which would be on the 'idle' circuit, then flat from 4000-6000 rpm.

What I am hoping is this is because with the 140/140 jets my 3TC was very rich. I need to get my wideband o2 over to the muffler shop and get an o2 bung

installed to tell exactly where we are.

Meanwhile, I have updated my main/aux jets to 130/125. I left the idle jet alone. I re-tuned the idle after updating the main jets.

The test drive went fine and the car seemed to rev okay -- not much seat of the pants difference. I hope to get new test

results on Sunday morning early, which would tell more. What I am hoping to see is that the hp continues to go up to redline.

This is a BEFORE Virtual Dyno plot from Performancebox data. Performancebox doesn't capture RPM so I have calculated it from the speed values.

Results are suspect, but the shape of the curve is of interest. If the car is peaking at 4000 rpm instead of 5000 rpm then there is a problem

with the main fueling circuit -- perhaps running too rich since it is jetted too rich -- and hopefully swapping to the smaller jets will correct.

Ideal would be to see it slope up to 80-90 whp at 5000 rpm

[Bruce Nunnally]

Hmmm. Ran some tests this morning with the 130/125 main/aux jets, and not the hoped for results.

New test results:

Run num Speed(mph) Time(s) Dist(f) Graph run Description

1 0.00 0 0.00 Speed Start
1 10.00 1.09 8.46 Speed Line
1 20.00 2.63 42.99 Speed Line
1 30.00 4.55 113.70 Speed Line
1 40.00 6.68 223.35 Speed Line
1 50.00 9.50 410.64 Speed Line
1 60.00 13.69 748.99 Speed End

3 0.00 0 0.00 Speed Start
3 10.00 1.14 8.64 Speed Line
3 20.00 2.69 43.54 Speed Line
3 30.00 4.64 115.19 Speed Line
3 40.00 6.77 225.22 Speed Line
3 50.00 9.61 413.44 Speed Line
3 60.00 13.84 755.57 Speed End

Previous:

Run num Speed(mph) Time(s) Dist(f) Graph run Description

2 0.00 0 0.00 Speed Start
2 10.00 1.22 8.55 Speed Line
2 20.00 2.71 42.07 Speed Line
2 30.00 4.64 42.07 Speed Line
2 40.00 6.77 222.98 Speed Line
2 50.00 9.67 415.39 Speed Line
2 60.00 14.02 769.02 Speed End

So that is all well and good -- previous was 14.04 or 14.06 in this direction, so less is better.

I was surprised to see that 6000 RPM indicated on the tach was not 60 mph tested -- I blew a couple of

runs by stopping after I hit 6000 RPM, and then got these by shifting at 6000 rpm. So I still don't

have the correct final drive ratio, or gear ratio one, or perhaps the 'swing' of the tach is delayed?

So, hmm. These look almost identical to the previous 2 runs with the 140/140 jets. Bill Jordan points out --

that is such a small change you are likely not to see any difference at all unless you are in cold weather or altitude. That might be a 0.0040in change in most manufactures jet size. 

The engine is still hitting a flat spot over 4000 rpm and not revving making more power up from there. I really need to get my wideband o2

hooked up so we can see what is really going on with the carb tuning.

[Texas Jim]

Just a thought.............

Does that carb have a POWER ENRICHMENT circuit ??

If it does... it does not appear to be working, and that could account for the lack of power above 4000 RPM...

The power curve should not fall off like that...

[Bruce Nunnally]

First, let me say that I dont' know that much about the carb (or the car) so I am open to suggestions and you may be on the right one,

but here are my thoughts. In the big picture, this carb/setup should be plenty from mild 1.5L to tuned 2.xL engines so we should NOT

be running out of anything for the 1.8L lol.

The Weber 32/36 does have a power valve, but it is part of the progression circuit responsible for the transition between the low-RPM idle

circuit (0-2500 to 3000 RPM) and the high RPM power circuit. The power valve opens on low manifold vacuum and adds more

fuel to the 'idle circuit' as the engine revs and transitions to the power circuit. If the engine is not making vacuum the power

valve would not open. You are right though that fueling at high rpm WOT then consists of idle & aux idle, power valve, and main & aux jets.

So it could be part of the issue.

This is the power valve on my 'old' carb -- it is on the lid of the carb, and hangs down into the bowl. Those are the carb floats on either

side of the post in the center which is part of the power valve. I surmise that the center post has a long spring that holds it shut against manifold vacuum.

Other possibilities: To make power we need air, fuel, and spark.

Spark: new ignition, new coil, new wires, new plugs -- we seem to have a trend here. Could try a .05 larger plug gap with the MSD but ok here.

Rechecked the MSD rev limiter, which was not set -- and it would have kept the car from literally revving past 4000 if that was the issue.

I reset it to 7000 rpm.

Fuel: Previous owner replaced mechanical fuel pump, but used one for a different engine. This could be a macro-issue if the fuel pump can

only make enough volume for 4000 rpm, but seems unlikely. Pump on the car is a mechanical pump. I have an electric pump on hand

ready to install, so perhaps put that in.

Air: New air filter, new carb. Could have some blockage in the intake?

The DGEV 32/36 is a progressive carb -- both 'barrels' are open at wide open throttle, so that should be the case throughout

the run. At WOT the linkage appears to correctly show both opening with the engine off.

[Texas Jim]

First, let me say that I dont' know that much about the carb (or the car) so I am open to suggestions and you may be on the right one,

but here are my thoughts. In the big picture, this carb/setup should be plenty from mild 1.5L to tuned 2.xL engines so we should NOT

be running out of anything for the 1.8L lol.

Agreed

The Weber 32/36 does have a power valve, but it is part of the progression circuit responsible for the transition between the low-RPM idle

circuit (0-2500 to 3000 RPM) and the high RPM power circuit.

The power valve opens on low manifold vacuum and adds more fuel to the 'idle circuit' as the engine revs and transitions to the power circuit.

If the engine is not making vacuum the power valve would not open.

You are right though that fueling at high rpm WOT then consists of idle & aux idle, power valve, and main & aux jets.

So it could be part of the issue.

I don't understand the highlighted sentence...

Actually.. the first sentence is correct... But the second sentence contradicts the first one...

This is the power valve on my 'old' carb -- it is on the lid of the carb, and hangs down into the bowl. Those are the carb floats on either

side of the post in the center which is part of the power valve. I surmise that the center post has a long spring that holds it shut against manifold vacuum.

Actually vacuum holds it closed...the spring opens it when the vacuum drops.

Other possibilities: To make power we need air, fuel, and spark.

Agreed.

Spark: new ignition, new coil, new wires, new plugs -- we seem to have a trend here. Could try a .05 larger plug gap with the MSD but ok here.

Good...

Rechecked the MSD rev limiter, which was not set -- and it would have kept the car from literally revving past 4000 if that was the issue.

I reset it to 7000 rpm.

Good...

Fuel: Previous owner replaced mechanical fuel pump, but used one for a different engine. This could be a macro-issue if the fuel pump can

only make enough volume for 4000 rpm, but seems unlikely. Pump on the car is a mechanical pump. I have an electric pump on hand

ready to install, so perhaps put that in.

Probably not needed...

Air: New air filter, new carb. Could have some blockage in the intake?

Not likely...

The DGEV 32/36 is a progressive carb -- both 'barrels' are open at wide open throttle, so that should be the case throughout

the run.

At WOT the linkage appears to correctly show both opening with the engine off.

Good...

[Bruce Nunnally]

I see; the engine is supposed to make manifold vacuum at idle. This vacuum goes low when the throttle is opened.

The vacuum holds the power valve shut. The valve opens when the vacuum goes lower than the spring.

Discussion

So if I have a vacuum leak then the power valve would be open at idle.

If the power valve were failing to open, I think we would see it as a dip at mid-RPM until the main jets come on.

The counter-argument is we would see it as a dip at mid RPM that continues through high RPM but I think

once the main jets come on -- especially with the previous jets -- we would have gotten enough fuel.

[Texas Jim]

I see; the engine is supposed to make manifold vacuum at idle. This vacuum goes low when the throttle is opened.

The vacuum holds the power valve shut. The valve opens when the vacuum goes lower than the spring.

Discussion

So if I have a vacuum leak then the power valve would be open at idle.

If the power valve were failing to open, I think we would see it as a dip at mid-RPM until the main jets come on.

The counter-argument is we would see it as a dip at mid RPM that continues through high RPM but I think

once the main jets come on -- especially with the previous jets -- we would have gotten enough fuel.

Not necessarily... The carb is DESIGNED to smoothly go from the low speed jets to the high speed jets with out any noticeable DIP...

When you STAND ON IT, and vacuum goes to almost zero, the power valve opens and lets a stream of fuel about an 1/8 of an inch diameter into the throat of the card at WOT.

This supplements the high speed jet to give you proper amount of fuel at WOT.

My guess at this point is you are running really lean above 5000 RPM...

[Bruce Nunnally]

I think coincident to our discussion, but worth mentioning -- the choke butterflies were opening on wide open throttle, but were not properly

engaged with their close spring. I adjusted the lid again so that the linkage is properly moving on the close spring and opens at WOT.

I re-adjusted the idle mix. During the test drive the car drove fine, but is uneven at 3500+ rpm -- pulls when you accelerate, but

spits/coughs a bit too just running along at 3500-4000 rpm. So I agree the mixture is perhaps off at that point, or this is another symptom

of why it isn't pulling well from 3500-6000 rpm.