Now from what I have read, you should pull up to a 45* line and roll right. If the plane walks to the right, I have too much down aileron? Just sounds backwards to me, because if I roll to the right, with the left aileron going down, wouldn't that make the plane walk to the left with too much down aileron, more drag and so forth? What am I not understanding?

I dont do aerobatics, but my son has some fast 1/2 A fighters that he incorporated differential
aileron into, and complained the rolls were sloppy.
I told him to get up real high and try deadstick rolls, they were perfect.
I think torque and/or spiral airflow is the issue

G'day Mate,
Follow this trim chart, it may help.
http://www.palosrc.com/instructors/trim-chrt.htm

No, you got it right. Differential (more up aileron than down) is supposed to counter the drag of the down going aileron by moving the up going ditto more, enough to equal the drag of the down one. You will only need it if your model yaws with aileron movement. Fly the model away from you so you can see straight up the tail. Apply aileron (say right) If the model yaws left you will see the tail swing right, and you will need to add more up aileron than down. If the tail swings left, ie the model yaws right, then either you have too much differential, or you need to add more down movement than up to the ailerons. Most aerobatic types don't need it, most biplanes and high wing types need a lot.
Evan, WB#12.

G'day Mate,
Differential actually reduces the amount of down aileron, & leaves the up aileron as is.
The method described in the trim chart is easier to understand. You fly the plane towards yourself, pull up until vertical, release sticks, then half roll, then you are looking at the top of the plane.
That way, Left is left & right is right.

Trouble with that idea is, with most models that will actually need differential, they will have stopped flying, stalled, and be heading straight back down again, possibly autorotating. It is much easier to see if you are standing behind the model when it is flying, then any tail swing indicates that there is some yaw, and you can then try changing the relative movements of the ailerons. The SIAS thing.
Evan. WB#12.

Thanks, so you are saying that by leaving the up aileron alone, your roll rate should stay the same, which is another reason to adjust the down aileron and not adjust the up one to compensate for the lowered aileron. Doe's that make sence?

There's a good chance your roll rate will increase.

And actually, the reason to reduce the down deflection instead of increasing the up deflection is practicality. You're already using a fair amount of the UP that you can get. And you don't need the amount of DOWN you've already got.

Truth is, if you change the airplane to get your differential, you're changing both at the same time. But if you're doing the changes in a computer radio, I'd make the change by reducing the down first. That would be helping the servos out with their leverages. It'd be easier for them to do the down motion. If you increase throw with a computer radio, you wind up giving the sero a tougher job. Ain't a big deal, but every little thing helps.

Makes sence now, thanks.

If your aircraft has a full symmetrical airfoil aileron deferential will be a problem when you are flying inverted.

The above method is specialized for pure aerobatic planes, and can be used to find the right amount of differential to use for Comp-Arfs and such that use top-hinged ailerons. The aerodynamic differential that exists as a result of the non-symmetrical hinge location (or wing incidence, etc) will cause heading deviations that can most easily be seen and corrected using the above test. It is NOT a suitable test for the amount of differential to use for a Cub or a PT-19. For the latter airplane types I fly straight away from me in level flight at a moderate cruise speed, and use rudder at lower speeds to make up the difference.

Differential is any case with different amounts of up and down. It doesn't have to only refer to reducing the down. What if I wasn't using all the throw and I chose to increase the up? The effect would be the same, and it would still be differential aileron throw.

Even on a positive 45d up-line, the wing is positively loaded, though not as heavily as in level flight, and adverse yaw works in the conventional way, i.e. too much down aileron causes yaw in that direction. But here's a twist: In a 90d up-line, the nose will yaw slightly toward the aileron with the greatest deflection, or presented area. So if both ailerons move up more than down, in this case the nose will yaw right when applying right aileron, and left when applying left aileron, ceteris paribus. The above method uses a 45d line as a compromise between level flight and a 90d upline. I think a 60d upline might also be a good compromise, since it will result in even less differential, which suits me. But if you have an airplane that is symmetrical with neutral incidences, and symmetrical surface hinging, just set the aileron throws as equal as possible, then fly the thing. These effects cannot be eliminated for all airplanes and flight conditions; they can only be made small enough that they are not noticeable to the judges under most conditions.

Incidentally, I don't use the above method because of the relatively strong P-Factor effect that exists in a 45d upline. You end up ignoring P-Factor, while deciding how much differential to program. What about the 45d inverted uplines? Now you've fouled things up, though maybe not to a large degree. Full-scale acro pilots (I'm one of them) don't spend time trying to optimize aileron differential and rudder mixing. They spend time practicing using the rudder to correct for these effects. That way, with practice, they can use the right amount of rudder for any flight condition, instead of trying to find an automatic compromise that only works perfectly in one flight condition. I know this flies in the face of the conventional wisdom that you have to set up and trim your airplane the way the big dogs do to hope to compete in pattern/IMAC. To each his or her own!

Differential is different amounts of up and down. It doesn't have to only refer to reducing the down.
Differential is different amounts of up and down. It doesn't have to only refer to reducing the up.

It also might refer to rigging the model.
It also might refer to changing the throws with a computer radio.
Lots of little details

to add--
the lighter the wing loading - the less problem you will see ( wing flies at lower AOA)
also
the higher the power loading ( or power applied) the less problem you will see. (wing flies at more constant AOA)
why?
if the required trim drag is constant - there is less need for trim or relative aileron difference .
example - pull a vertical UP line -- if speed stays constant thru the up line roll- the plane more easily retains original path.
many models slow on vertical up lines. for really good aerobatics - you want enough power to hold or accelerate on any line.
Cubbys simply can't do this so the desirable aileron deflections actually CHANGE thru a roll.
Unless you are in a vertical dive -which ain't a good idea .
BTW, I am getting a new H9 scale 1/4 size CUB - what a knockout!
It even has a posable pilot -with his hand in his lap
I will take his hand out of his lap----