marcushh777
Well-known member
Very nice feature; similar to range checking feature... IMHO Spektrum should also have a low power setting for use with the Sim WS2000.
marcus
TX16s with USB connection as controller
- Thread starter hsamuels
- Start date
marcus
marcushh777
Well-known member
On the surface, this is a mostly correct statement; however, it is incomplete. The 'controller' you refer to (within the Sim) is what RealFlight calls a 'Radio Profile'. There are actually two (2) ways to create a new Radio Profile:
(1) ... as you describe above; reset and modify an existing Radio Profile and then "Save As..." giving it another name.
(2) Create a Radio Profile in a simple text editor (like Windows Notepad) from a template.
No. (2) above is an interesting option. The Rf9.x Radio Profiles are held in:
C:\Users\<userid>\Documents\RealFlight 9\Radio Profiles\
The Radio Profiles are simple text files, virtually self explanatory, easily edited, and saved as any name you wish. Once the Radio Profile interface is understood its much easier IMHO to create and edit new profiles with Notepad directly, rather than using the "Select Controller Editor". The only downside to this approach is that the specifications are sometimes misleading and inconsistent (this bug(s) has already been reported and discussed)
What I did on my system was to create a template Radio Profile; I copy the template profile to the Radio Profiles directory (folder) with the "new name," and then I tweak it. ? I only need open the "Select Controller" dialogue to 'select' the new profile and I need not mess around with the editor ( I find the editor annoying; this will change if, and when, KnifeEdge or Horizon Hobby decide to get busy and clean up the labels and inconsistencies ).
marcus
Mysterious-er and mysterious-er. Works with Taranis and Jumper just fine. Wonder if N8LBV will get similar problems with his Radiomaster?
I had myself convinced that reloading Win drivers would be the magic. Can't really see where if working fine in Win, it wouldn't in RF. Which also doesn't seem to point toward the radio as the issue. The Win calibration would compensate for any deviation in centering/pulse widths that might be present.
Don't see anything in the custom radio profile that would induce the centering bias. All I can guess is that somehow the Win calibration is not being saved and/or not being implemented to what is fed to RF.
I had myself convinced that reloading Win drivers would be the magic. Can't really see where if working fine in Win, it wouldn't in RF. Which also doesn't seem to point toward the radio as the issue. The Win calibration would compensate for any deviation in centering/pulse widths that might be present.
Don't see anything in the custom radio profile that would induce the centering bias. All I can guess is that somehow the Win calibration is not being saved and/or not being implemented to what is fed to RF.
Hi Marcus and N8LBV,
I have a DX9 G2 Black Edition, and that followed my original DX8. I do know those radios quite well. I got theTX16s with the simulated carbon fiber faceplate, added the aluminum folding handle, which is highly recommended for fitting into this case with room to spare for a tool kit. For the price, I was expecting a lower-quality feel. My gadget freak son just did a comparison, without knowing the provenance of either. He says the TX16s plastic 'feels' higher quality. The switches have a more tactile feedback, but that may be due in part to my having used the DX9 for a while. The TX16s feels more solid - it's about 900g vs 800g for the DX9, plus another 50g for my Crossfire module. And he was insightful enough to notice that the rounded grips on the back of the DX9 may be more ergonomic, which I agree with.
I'll admit that there is a learning curve with OpenTX, but I love the ability to upload my own model images to be viewed on the color screen. When I go back to the DX9 it seems positively antiquated. Marcus, if you can find somebody with a TX16s, ask if you can hold it and check out the quality. OpenTX is a bit frustrating at first, because simple things like making elevons in Spektrum involve making your own mix in OpenTX. But if you think the wing is, say, very pitchy compared to its roll rate, you can change that mix to your heart's desire. With Spektrum, it's take it or leave it.
The multimodule does include the ability for DSM2 and DSMX, but I don't believe it is possible to adjust the power level. My external Crossfire module does have that ability. In fact, it can also adjust its own power level depending on the RSSI, which is a neat trick. Unfortunately when you turn on the Crossfire module, and you haven't yet powered the receiver, the Crossfire module is at max power, searching for a model that isn't providing any telemetry. And you can tell it is doing that because it makes a little noise - really!
I don't think latency is much of an issue. HID devices have around 10ms of latency anyway. You can bind to the WS2000 in DSMX 11ms. The WS2000 is also an HID device, so it adds another 10ms. But I haven't ever noticed an issue with it, even when running FPV quad sims flying through gates.
I have had issues trying to bind to Spektrum receivers when the radio is too close - and they tell you that. But I've never had a problem with the WS2000, which is about a meter away with my current setup, and maybe a little less with my previous setup.
Speaking of setups, my tower gaming system (bought from a guy who needed it out of his life so he could graduate from college) drives a 43" 4k monitor, with the graphics quality set to max. And it sits on the desk, and I'm in my desk chair. Maybe not the comfort of a La-Z-Boy, but it's pretty immersive.
In any case, I'll come back here if I find the solution to the USB problem. Thank you all for your help!
I have a DX9 G2 Black Edition, and that followed my original DX8. I do know those radios quite well. I got theTX16s with the simulated carbon fiber faceplate, added the aluminum folding handle, which is highly recommended for fitting into this case with room to spare for a tool kit. For the price, I was expecting a lower-quality feel. My gadget freak son just did a comparison, without knowing the provenance of either. He says the TX16s plastic 'feels' higher quality. The switches have a more tactile feedback, but that may be due in part to my having used the DX9 for a while. The TX16s feels more solid - it's about 900g vs 800g for the DX9, plus another 50g for my Crossfire module. And he was insightful enough to notice that the rounded grips on the back of the DX9 may be more ergonomic, which I agree with.
I'll admit that there is a learning curve with OpenTX, but I love the ability to upload my own model images to be viewed on the color screen. When I go back to the DX9 it seems positively antiquated. Marcus, if you can find somebody with a TX16s, ask if you can hold it and check out the quality. OpenTX is a bit frustrating at first, because simple things like making elevons in Spektrum involve making your own mix in OpenTX. But if you think the wing is, say, very pitchy compared to its roll rate, you can change that mix to your heart's desire. With Spektrum, it's take it or leave it.
The multimodule does include the ability for DSM2 and DSMX, but I don't believe it is possible to adjust the power level. My external Crossfire module does have that ability. In fact, it can also adjust its own power level depending on the RSSI, which is a neat trick. Unfortunately when you turn on the Crossfire module, and you haven't yet powered the receiver, the Crossfire module is at max power, searching for a model that isn't providing any telemetry. And you can tell it is doing that because it makes a little noise - really!
I don't think latency is much of an issue. HID devices have around 10ms of latency anyway. You can bind to the WS2000 in DSMX 11ms. The WS2000 is also an HID device, so it adds another 10ms. But I haven't ever noticed an issue with it, even when running FPV quad sims flying through gates.
I have had issues trying to bind to Spektrum receivers when the radio is too close - and they tell you that. But I've never had a problem with the WS2000, which is about a meter away with my current setup, and maybe a little less with my previous setup.
Speaking of setups, my tower gaming system (bought from a guy who needed it out of his life so he could graduate from college) drives a 43" 4k monitor, with the graphics quality set to max. And it sits on the desk, and I'm in my desk chair. Maybe not the comfort of a La-Z-Boy, but it's pretty immersive.
In any case, I'll come back here if I find the solution to the USB problem. Thank you all for your help!
Attachments
Yes, in fact I calibrated it again yesterday after updating to 2.3.11. Since it works fine with the WS2000, I think the radio itself is fine. I'll get some popcorn while N8LBV tries his experiment.
marcushh777
Well-known member
I had another idea; my hypothesis is that this is a bug in the OpenTx driver. I am wondering if (for some reason) the trim tabs are offset incorrectly to the far right... um, 64%.
Could it be that the problem is corrected by moving the trim tabs back to the left to 'center' all sticks?
Is that much motion 'left' for the trim tabs to make this possible?
I think at this point, if you really want to get the usb to work, contact the OpenTx folks and open an incident (because you can reproduce the issue). I'm very much interested in N8LBV experiment. Any time you have a solid failure, which can be reliably reproduced, its almost always software (one way or the other).
Best of luck !
marcus
Could it be that the problem is corrected by moving the trim tabs back to the left to 'center' all sticks?
Is that much motion 'left' for the trim tabs to make this possible?
I think at this point, if you really want to get the usb to work, contact the OpenTx folks and open an incident (because you can reproduce the issue). I'm very much interested in N8LBV experiment. Any time you have a solid failure, which can be reliably reproduced, its almost always software (one way or the other).
Best of luck !
marcus
Marcus's suggestion sent me down the trim road. I was able to get two of the channels to show as 64% in RF, with full trims, even after calibrating. But only a couple - calibration compensated for the others, though. I even mis-calibrated my radio, to see if that would do something - nope, RF calibration corrected for that.
So, let's make sure we're starting as simple as we possibly can. Attached is the most basic 16 channel configuration possible for the TS16. You'll need to load it into Companion (the PC interface for OpenTX), then copy it over to the actual model list uploaded from your radio. Careful - if you just download straight to the radio, it will overwrite all your other models, leaving just the one I supplied.
It has absolutely nothing for rates, no reduced throws, trims have been completely disabled, etc. Servo centering is at Spektrum default of 1500 ppmus. Servo throw is default OpenTX, which is a bit more than Spektrum uses, but has been shown to initially be fine for Real Flight in the past.
Just to be complete, recalibrate your TX again before trying it out in RF.
Try it out, setting up a new controller again in RF, and let's see how it behaves. File attached. Change the extension from ".txt" to ".otx" to get it into Companion.
So, let's make sure we're starting as simple as we possibly can. Attached is the most basic 16 channel configuration possible for the TS16. You'll need to load it into Companion (the PC interface for OpenTX), then copy it over to the actual model list uploaded from your radio. Careful - if you just download straight to the radio, it will overwrite all your other models, leaving just the one I supplied.
It has absolutely nothing for rates, no reduced throws, trims have been completely disabled, etc. Servo centering is at Spektrum default of 1500 ppmus. Servo throw is default OpenTX, which is a bit more than Spektrum uses, but has been shown to initially be fine for Real Flight in the past.
Just to be complete, recalibrate your TX again before trying it out in RF.
Try it out, setting up a new controller again in RF, and let's see how it behaves. File attached. Change the extension from ".txt" to ".otx" to get it into Companion.
Right, thank you! I use that for the range test, and it didn't occur to me that it can also be used for home simulator use. On Spektrum radios, you put it in that mode, then hold a button to reduce the power. It's supposed to ensure that you don't actually take off in low power mode.
OpenTx gives you a warning if you are in low power mode when you power on or select the model with low power.
Last edited:
Here's a screen shot illustrating my centered trims. You can tell they are centered because they have two bars ||. If the trims were off-center, there would only be one bar. I could trim them back, and I think it's possible to adjust the trim range in OpenTx. That may solve the 64% issue, but not the crosstalk or mixing that also happens, but I haven't described very much. I look forward to N8LVS's test as well.
Attachments
That's thinking outside of the box!
In the Windows USB Game Controller calibration, I turned on 'Display Raw Data' and everything went from 0 to 2047, with a midpoint of 1024. I think that's another data point that says that whatever the radio is sending can be calibrated in Windows before it gets to Realflight.
In the Controller section of the game Settings check the Deadband Percentage is set to 0.
marcushh777
Well-known member
Yeah, that "cross talk" has got to be a bug in OpenTx.
I forgot to ask you, is the mixing in any way logical, for instance, like aileron|rudder, or flap|elevator?
If its random mixing there has to be a bug in there somewhere... guess we wait for N8LVS
I'm astonished! After calibration, all 8 proportional axes are at 50%. Possible differences between our setups include:
- I use TAER, you used AETR. I don't know if that makes a difference, although the Windows setup starts with the D-Pad. The D-Pad apparently is two of the 8 proportional axes. Could this be a problem?
- You assigned 8 real proportional controls (sticks, 2 knobs, and two sliders) to the first 8 channels. I had been assigning 3-pos switches to some of the proportional channels (triple rates, flaps, flight mode, etc). I wonder if that makes problems for RF's calibration routine.
- What I was calling mixing between channels, or crosstalk, appeared during the calibration of the 8 proportional channels on your setup. But it disappeared once all 8 axes had been calibrated.
This is fantastic progress. I'll try copying it and reassigning a 3-pos switch instead of a slider or pot.
Regards,
Howard
Attachments
Channel order doesn't matter, since you assign it to be what you want in RF anyway. I just used default "Futaba" order, 'cause I'm old and don't like change. If you want to play, change the sources from proportional to switches and vice versa, and I'll be it will be just fine, also. (But say if Elevator is now using a 3 pos switch it will only be all up, neutral, or all down). But neutral will be neutral. All sources (sticks, sliders, knobs, switches) in OpenTX use the same range of values to start, just that some are variable, and some are fixed.
I only got the 64% in an earlier version, by cranking the trims all the way over on all 4 channels. But after calibration in RF, it only persisted on two of them. The file I sent had the trims completely disabled. You could baby-step back in, and set the lines back to "own trim", and see if that makes a difference.
Next steps would be to create Inputs, so you can do rates, etc. and use them in the Mixer instead of the "real" sources.
If you want, attach a copy of one of the "bad" setups here, and I'll be glad to see if anything looks off. Just change the type to txt, so it can be attached.
But change the dead band setting noted previously to 0 anyway. That also gives a lot of heartache....
Ted