The walbro-type carburetter is the commonly used type with model gas engines.
I believe there is a major problem with these carburetters that nobody realy
seems to master.
The definition of the problem is:
A good setup on the ground, full power ok, idle ok, transistions ok...on the ground.
Once in the air the setup changes
Those who haven't experienced this problem probably will say
I'm nagging and the rest will know what I'm talking about.
The problem is the apparent sensitivity for airpressure variations and /or
turbulence in flight. Some set-ups won't, others will, and from those who
do, some succeed in resolving it totally or partly by experimenting with all kinds of tubes and scoops
and others just keep struggling, I think that's about the size of it for the moment
I had a ZDZ50 in a CAP232 that changed RPM in every other flight position. All kinds
of changes under the cowl, tubing and enrichting and leaning of the mixture minimised
it at a certain moment but it never went completely away and after all these changes I cannot put my finger on what
exactly did the trick.
I have a DL50 in an EXTRA260 which virtually doesn't has the problem.
I have a GF45I in a ZLIN526 with the carb sticking out of the cowl which runs
very well on the ground only if I use an airfilter or a scoop but always
backs off in flight (cfr Rusty Tumbles in the GR45I thread).
I know guys with DA50 which do it in some models and don't in others.
I think this is something that is related with the system used with this carburetters
and that's it. Next step for gasser industry should be to deal with this by developing
a new kind of carburetter that is airworthy. Nice theory but probably industrial rubbish
cause someone will copy and offer it at a lower price.
So it's up to the hobby scientist I guess.
I was thinking about electronic mixture control using a few sensors and controllable needles as it
is(was?) used by OS in their high end 140, I don't know if it was a success though I don't see it listed
in their catalog today, maybe a sign.
It's a problem we don't know of in the nitro branche, why? Is it because they use different type's of carbs
or because the nitro/air mixture is totally different (1/6 versus 1/15 for gas) and therefor not so
sensitive for variations.
Anyway, it would be nice if someone could come up with at least some guide line to minimise this effect
so we don't have to unscrew the cowl a dozen times to try another setup.

The problem is known and recognized.
I addressed it in my forum at http://www.prme.nl/forum/viewtopic.php?t=205 , and on several occasions in threads here and on FG.
Just provide a stable reference pressure to the carb regulating membrane and be done with it.

"referrence pressure" is not in many users vocabulary.
Nor is "vacuum signal"
The fixes suggested for getting the pressure constant (across the topside of the regulator diaphragm) are endless and many are worthless.
The carbs on glo engines also suffer from air pressure problems .
The use of exhaust pressure to the tank can fix most of these.
Glo engines because of the fuel used ,typically can tolerate much more varience in fuel to air ratios.
My ZDZs and Evolutions etc., don't have fuel mix problems as noted -because I always center the tanks on the carb .
This keeps the fuel head at a more constant level
Also my cowlings are always setup to provide maximum air extraction at the rear or bottom of the cowl. this helps stabilize interior pressures and keeps the engines cool to boot.
For what it's worth , a large plenum attached to the top side of a Walbro type carb, will stabilize any pressure problem. a balloon on a piece of tubing attached to the diaphragm vent can do this. OR a line from the vent to a box at the carb inlet, etc., can also work

...don't forget the condoms, and empty film cannisters that have also been used

Isn't it time to change that?
The reference pressure is the equivalent of the float height in a standard carb, or fuel pressure in injection systems. Change it, and the mixture goes off bigtime.
If people ask for solutions to their problem, they may not like or understandthe answers they get. Can't help that.

Somwtimes ya just have to step up to the historically recognized vernacular.

BTW, the solution to all of the carb reference pressure issues is to dump the carb. Injection does a beautiful job of constant everything. Unfortunately there will never be a, effective, miniaturized version for small gassers that modelers will want to afford. The only thing that even comes close to what a modeler might someday pay for is much too large and heavy for models less than 60% scale.

Even fuel injection systems need a static pressure reference

It is possible to do a pretty darn good closed loop setup for SOME uses
My favorite is a tuned exhaust (for quiet operation and increasesd power)
add to this SOME type of balance correction using the exhaust pressure which increases fuel flow as the exhaust goes resonant.
"HUH?"
I have measured this many times and in all cases , I found that as the power increases the need for fuel increases sharply. Note I did not say as rpm increases tho typically the rpm does increase.
Where many modelers lose it -is by leaning out the mix till rpm is maxed in level or slightly diving flight
The engine may make more rpm -but the load is lower
as soon as the load is increased . the engine goes lean. This situation is common as fleas on a hound.
The so called pressure regs sold by sme outlets are yet another botch up
I measured the pump to pressure readings on all the early OS pumps Perry etc., and always got the same result
max pressure at lowest rpm.
Worthless
The fuel DEMAND relative to power goes up very sharply
On Walbro setups - the easy fix is to insure that the reg diaphragm does not see a decrease in vent pressure as the engine goes to full power
Also the tank should flow as easily under high demand as under low demand
I believe in a forward facing vent line and no filters in the lines.

Agreed, but a good fuel injection system has the ability to compensate for pressure references differences through software, especially when static pressure is read at a source location away from the engine. Along with true miniaturization, the computational side of a good injection system, direct or otherwise, is why modelers will never be buying it.

TOM
So are you saying the cost will be too much or the complexity?

Motorcycles with injection are actually easier to tune if you have the software and the hardware interface. I had been of the opinion that carbs were easier but when you can see a graph of what you are changing this old man changed his tune.

John

The cost for modelers will be too great for the miniaturization and complexity. People don't want to pay the higher costs of a high end engine now. They sure as heck won't want to spend an additional $1,000.00 plus (that's a low number) for a well designed injection system for our size engines. You won't want to use a copied design. Far too many places for things to go wrong, so the original designs and manufacturers would have to be the purchase sources. Other users don't have an issue with the cost.

The Walbro carb was never really designed to be placed in the environment we use, but with the above mentioned "fixes" they can be made to run very well if problems arise.
An injection system would be nice, and as TOM mentioned can and does work, complexity is not the problem, most would not be willing to pay the price.

Unlike our glow counterparts, gas engines are quite forgiving to power changes in relation to tuned pipes.
In glow engines, you use a good deal of expertise and a prayer to get the engine to transition from overly rich running to running on the pipe.
This also is the case in piston ported gas engines, though for different reasons. With reeds or rotary valve, float and constant vacuum carbs, tuning with pipes is child's play in comparison, and the carb follows the engine's fuel demand quite well IF the carb is set up right.
Since fuel injection is not (yet) an option for our small purses, we have to make the best of what we have. This requires a good understanding of what is going on inside the carb.
What I read in many places, is more like the lame leading the blind, both clutching at straws to make progress. So busy, there is no time to stop and think what is realy going on, myself included.
I have made glow fuel pumps/regulators out of walbro carbs, with exhaust pressure as reference on the membrane. Combined with a normal glow carb and tuned pipe, there hardly was a rich/lean "jump" to speak of. The carb just followed the fuel demand. So there is a lot of potential to play with without the need for fuel injection systems.
So far, The static tube reference pressure on the regulating membrane has worked for me in all cases, where other "solutions" failed. It is not a cure-all. Large pressure differences and turbulence at the carb  mouth entrance still can play havoc with mixture strength, so it is important to create a calm air zone for the carb to work in.

Very nice post.

No Good to tune for max on the ground and leave it there. On the ground, the prop is somewhat stalled and not getting a good bite, so less load. In the air, any extended upline and the motor is running lean.

You are correct about the prop being partially stalled on the ground in a static situation, but that means more drag on the prop and load on the engine, not less. The engine unloads in the air as the airplane approaches the prop pitch speed. (If you don't believe me try some data linking and watch your rpms in flight.) A climb will then load the engine, and on an non-pumped fuel setup, lean the mixture.

AV8TOR

Fuel injection is one possibility that's perhaps too far away, but what about electronic in flight mixture control using exhaust temp
sensors and rpm readings the way OS a few years ago applied in it's 140.
Or isn't it possible for manufacturers to provide for those "solutions" in the form of a box/airfilter to screw on the carb in stead of leaving it to the mercy of the modellers creativity or should I say fiddling capability which, and it should be said, has it's charms too.

I've only had two engines in 10 years that have had any type of imbalance issue. One was cured with the simple addition of a reversed air scoop installed over the opening in the diaphragm plate. The other, an early BME 115 would not have mattered what the user did externally because there were design issues with the induction issues.

My feelings are that diaphragm carbs on two strokes, although generally quite adequate and generally effective, are just a little too slow in their reactin times to effectively compensate for changes in pressure. Those fooling with tuned exhausts have it a little more difficult. Most of our engines sound like they are a little unstable in various rpm bands that are otherwise steady in the trotlle input. From what I've experienced, injection will be the only way to completely stabilize power output in our little engines but it won't come cheap. Injected engines are rock solid in all rpm bands with absolutey no "wavering" like that found in carbuerated engines.

The mixture control solution you described would only be effective if the mixture required adjustment. That is typically not the case since once correctly tuned the Walbro and Tillotsen carbs have an extremely wide altitude operational range. What many may perceive to be an engine going lean on long uplines may in fact be a need to retard the timing from the advanced position it is in. As an engine loads up and heats up there are changes within the cylinder taking place (minor alterations in port timing due to changes in heating mixture densities and pressures) that are minimized, or eliminated, by retarding the timing until the temps and loads reduce a little. So we have a two fold problem to deal with, and most don't know about, or recognize, the second half of the problem. One thing is for certain, we will not be getting ignitions with a split timing curve from any of the current manufacturers. At least not at our price levels.

Most Walbro carbs do not have accelerator pumps either, so you have to compensate the mixture if you want to have a brisk accelerating.

The main problem with membrane carbs on single cylinder engines (or opposed twins) is that the carb membrane admits a little bit of fuel EACH stroke. The main needle opens and closes,bounces on the seat, and the fuel admission rithm may get out of tune with the engine rpm. This can only be solved with a large plenum chamber, which however opens up a new can of worms.
Another problem to address is the idle transition mixture control. Walbro type carbs (walbro, tillotson, bing, dell'orto etc) have a strange way of brake air mixing into the fuel flow. At one time the second or third idle jet acts as brake air jet, the other time it acts as idle jet. This makes tuning a membrane carb very difficult.
In float carbs there were well defined and well separated items like slide chamfer, slide needle, main jet, air jet and float height to play with. Still quite a few variables, but a mechanic savvy to his engine would have the tools in his box to tune the carb for the day, morning or evening. If I play with the seemingly simpler walbro, it often gives me fits, because one change influences almost the whole adjustment range. No wonder there are so many Walbro numbers available, without any system in the numbering.
Some folks at Walbro seem to know what they are doing, but the best they can do is make a new setup for a new engine. Or so it seems.

Edit:
Grammar.

Seeming designed to provide average performance under average conditions. In a way not a bad design criteria because it addresses most conditions. Those seeking the highest possible performnce will never be satisfied with less than optimal equipment. The only issue is whether or not they are willing to pay the price for better equipment. Judging by the various threads regarding engine pricing, they will not be, but will still demand the best of the best while castrating any that fail to ante up for free.

Even most engine manufacturers shy away from the cost to get a carb just as it should be for their engines. Instead they pick one from the many many available, and hope it will work out ok. In the process, they even look at the competition, and use the carb already in use hoping it will be the right one for their own use.

av8tor, What kind of data link did you use? Just curious! Capt,n

My ears....

AV8TOR

I just use my ears. Maybe tom is right. I just work with what I have and it seems to solve the issue.