Throttle Doesn't Go Full Range Unless "Tricked"

[JNOV]

After a "clean" calibration, my throttle doesn't report full range. For instance, after a normal calibration, if I go into IL-2 and max my throttle, the "HUD" display will read 104 or 105, rather than 110. This is not a surprise, because it shows up the same way when I test the calibration in CM. I've always suspected that this is a bug in the calibaration routines. In fact, the only way that I can get the throttle to generate 0-255 is if I don't push the throttle all the way forward during calibration.

Has anyone else experienced a similar problem? Does anyone have a suggestion to fix it?

Thanks for your help.

- JNOV

[Revvin]

Can't say I've experienced this problem, have you tried re-installing the drivers, or perhaps using CH delete and installing again?

[JNOV]

I just noticed a similar problem in the harware forum. It seems that a number of folks are having a similar issue with the throttle. For all of my other axes (I don't know about the toe brake or Fighterstick throttle axes, because I've never paid much attention), the calibration routines work perfectly, in the way that Bob describes in the thread in the hardware forum. I wonder if there isn't a snag somewhere in the throttle calibration code? Or maybe a calibration error that only affects non-centering axes?

- JNOV

[JNOV]

Revvin:

Thanks for the reply.

I've been fighting (not really "fighting," just compensating for, really) this problem since the CM 2 beta, and it has followed me through any number of re-installations, deletes, new USB cards, and even a new PC.

- JNOV

[Hercules]

I have the same throttle problem and just read somewhere else on-line that the problem could be caused by having another game port joystick plugged into your computer. I have another one plugged in so I am going to unhook it and reboot. I will be sure to post here if it solves the problem.

[Bob Church]

It sounds like the throttle is moving the pot past the end of the active area a little bit. Go back to the Test/Calibrate screen. Watch the raw data display and see what the values are doing. It sounds like they're going to some minimum value, ideally 0 (0 is max throttle) and then as you push it further forward the value goes up a little and backs the throttle off. See what value the raw data display gives you with the throttle hard forward, then run through the calibration again but when it tells you to go full forward, only go far enough that the raw value is still slightly greater than the "bounce back" value you see at full forward, e.g. the throttle hits a minimum value of 2, for example, then bounces back to 4. When you go to calibrate the forward position, set it for a value slightly above th 4, maybe 6 or 8, and click the button there. See if that will bring it into line for you.

Having a second throttle can cause problems in Longbow II (I think it was LB2), where it would only go to 56% if there was a second throttle. Also in the MS sims (FlightSim and CFS) where a second throttle axis still controls the throttle and any noise on it, a little spike or whatever, can kick the throttle back to the value that the second throttle sent. Those are easy, just go into the Control Assignment screen and make sure you've only got one throttle actually assigned.

- Bob

The StickWorks
http://www.stickworks.com

[JNOV]

Bob:

Thanks for the reply. I put my controllers in direct mode and selected the test/calibration dialog for the Pro Throttle. As I pushed the throttle forward, the raw data decreased smoothly to--but did not go below--12 or 13. The "bounce" effect you describe is absent. Also, I have only one throttle.

Thanks again - JNOV

[MichaelCHProd]

Those numbers are not cool. Shoot me an e-mail so I can get this fixed.

[Bob Church]

JNOV,

Okay. Well, I tried it here and I don't seem to have trouble with the 13 for min. Mine normally runs 0 to 217, but if I set it to 13 during calibration I seem to get full range at 13 after the calibration is complete, e.g. a raw 13 will produce a scaled 0. If you can't see a bounce, the only other thing I can think of would be that the sim does something at full throttle that causes a little voltage drop to the sticks. Otherwise the bounce ought to show up in the Test/Calibration screen.

Calibrating short like you (and I) mentioned shouldn't be necessary, it should get to zero scaled when it gets to whatever the minimum value was that you clicked on during calibration, e.g. if you only went to 13 min and click the button, then 13 should scale to 0. The scaled value shouldn't come back up unless the raw value comes back up. The scaling itself is just a simple multiplication and it's the same for all the non-centered axes, though the throttle is probably the only one that doesn't go 0 to 255 in most cases, anyway. For the others, the scaling routine just ends up being a multiplication by 1. Still, truncation error sucks the value down, you'd think it would pull it off minimum throttle and toward maximum throttle if the scaling were behind it.

Anyway I see Michael is on the case. He should be able to sort it out for you. Let me know how it turns out. If nothing else works, I can always increase the deadzone at the end of the throttle travel a little, the short calibration would happen automatically that way.

- Bob

The StickWorks
http://www.stickworks.com

[JNOV]

Thanks, Michael -- I'll send you an e-mail.

Bob:

It's nothing peculiar about any particular sim. I've seen the same thing in at least a dozen. The thing is, after I calibrate, I see the non-full-range response when I test the throttle axis of my program (e.g., the "Z" axis of CM Device 1) in the CM test/calibrate dialog. (Oddly, the other end of the throttle axis, which produces a raw input of about 236 gets calibrated to a perfect 255.) Given that, the sims respond as I would expect.

It seems really peculiar that the bottom end gets fixed (by calibration) exactly, but the top is completely unaffected (i.e., the top end shows 12 raw and 12 after calibration). Is there some sort of range check in the calibration code that the short full-range count (236-12 = 224) of my throttle is violating?

I just did a test (sorry for the stream of consciousness here) to try out my short full-range theory, and I think there's something to it. If I push REALLY hard on my throttle (I never tried this until just now), I can get it to register about 6 and 241 raw. If I use those stressed values to calibrate, the calibration works as expected. It's still no good, though, because I can only get 0 and 255 by pushing REALLY hard on the throttle, which is impossible to reproduce reliably (not to mention uncomortable) and likely to damage the throttle.

- JNOV

[Bob Church]

Well, it wouldn't be the first time I've made a coding error. :) It is pretty straightforward, though. Before I look at it, I'm just a little confused about where the values you mention are coming from. If you look at the ProThrottle with the CM in Direct Mode, the Raw Data display shows what's coming back from the hardware. Those are the values I was talking about. Calibration won't affect those, really. If you make a map with just the ProThrottle in it, default assignments (straight through) and download it, then you'll see the Control Manager Device when you go to test. The Raw Data display there shows what the scaling produces.

If you get a chance, I'd like to know what range the ProThrottle gives in Direct Mode and what range the Control Manager Device gives in mapped mode for the throttle axis. In both cases, it's the number from the Raw Data display after you've done a calibration that's of interest. Maybe I can see where the problem is coming from. If you want, record the two Raw Values when you click the button on the throttle, too, and I can tell what the calibration values were. If there's an error, it has to show up if I manually run through the calibration routine.

WRT the minimum difference (236 - 12), there's not one. The routine checks that the values are in order, that max is bigger than min, but insofar as a minimum separation, it doesn't care.

- Bob

The StickWorks
http://www.stickworks.com

[JNOV]

Bob:

Thanks for the reply.

I'm sorry if I was unclear about the source of the data. What I meant was that the raw data was 236-12 before and after calibration in direct mode and 255-12 after calibration in mapped mode. But, I just recalibrated and for the first time ever, I am seeing 255-0 after calibration in mapped mode. (I performed the procedure twice to be sure.)

I have no explanation for this, as I have done this procedure at least 50 times before. The only thing that I can posit is that it has something to do with me cranking on my handle to get the raw value down to about 6 earlier (althought I didn't do that this time; I just recalibrated at 236-12 raw as usual). This doesn't make any sense to me, so I'm stumped.

Nonetheless, it seems ok now. Oh, well, better lucky than good.

Thanks again - JNOV

[Bob Church]

>> ...better lucky than good... <<

Words to live by. :) It is kind of strange that it would start up like that. The 236-12 range shouldn&#39;t be any problem for the scaling calculations, but it should produce the 255-0 when you go back and look at the CM device. How it got the 12 scaled is a mystery. For it to match the raw data (12 in, 12 out) you&#39;d have had to have essentially a zero raw value when you calibrated the full forward position, otherwise there would have been a shift. I dunno, maybe a loose pot connection that started making good contact again when you pushed hard on the handle?

In any case. what you&#39;re seeing now is what it should do - the raw data (Pro Throttle data) isn&#39;t real critical but the scaled data (CM Device data) should cover the range. Keep an eye on it, if it goes out again, maybe we can catch it in the act.

Anyway, I&#39;m glad to hear it&#39;s working again!

- Bob

[Kakuli]

Hello, i have (i think) this problem also. I have CH Pro Throttle USB for about 8 monthes, and in the last week i have a problem with it, it keeps on moving up in the game. I mean when i pull down all the way the handel, in the game (FS or Lo-Mac when in the VC) it stays a bit up and moves up slowly a bit. I can see this movement allso in the CH control managar v3.00, and i try to clibrate it again and again, but still the throttle will not stay down. I didn&#39;t understand how to trick this problem, please help. ;)

[JNOV]

Kakuli:

The problem I was having was slightly different than yours, I think. It sounds like you are experiencing "drift" on the throttle axis. That is, the axis value is changing on its own without any input from you. My problem wasn&#39;t drift. The problem I was having was getting the throttle axis to calibrate so that it would produce a full-range output after calibration. The "trick" was to leave the throttle handle short of full range during the calibration routine so that what the calibration routine through was full-range deflection was actually less.

That "trick" might help with your "it stays up a bit" problem (did I just type that?), but it won&#39;t help with drift. Michael or Bob Church might be able to give you some advice, but the drift indicates to me that you may have a bad (or just dirty) potentiometer in your throttle.

Sorry I can&#39;t offer any real help,

JNOV

[MichaelCHProd]

The pots should not be worn or dirty after 8 months, unless this is an "outdoor" throttle :)

What chipset is your USB host controller? Are you using a USB hub? Any chance I can get you to plug it into another PC to see if it does the same thing?

[Kakuli]

I have ASUS P4C800DELUX with Intel 875 chipset, i don&#39;t know what is a USB hub :blink: , but the CH products are conected directly to the mother boards USB, with out any device that provides more USB. I think that JNOV is right :"the drift indicates to me that you may have a bad (or just dirty) potentiometer in your throttle" because when i move the handle all the way up and than down the problem "disappears" for a while, like that by moving the handle i "clean" the throttle. Thank you. B)
Edit:
Intel to PCI to USB Enhanced Host Controller - ICH5
Intel® 82801EB USB Universal Host Controller - 24D2
Intel® 82801EB USB Universal Host Controller - 24D4
Intel® 82801EB USB Universal Host Controller - 24D7
Intel® 82801EB USB Universal Host Controller - 24DE
(Logitech USB Camera Pro 4000)
NaturalPoint SmartNAV
USB Root Hub
USB Root Hub
USB Root Hub
USB Root Hub
:blink:

[JRH147]

I&#39;ve been having the same problem as of late - my solution has always been to "scrub" the throttle forward and back quickly, then my throttle will go up to 110%. Is this a bad "pot", and if so, how do I go about fixing it?