Protocol Stacks

More
09 Aug 2012 12:53 #1003 by FDR
Replied by FDR on topic Protocol Stacks
Probably would be more consistent/informative, if you would reset all the trims...

Please Log in or Create an account to join the conversation.

More
09 Aug 2012 13:41 - 09 Aug 2012 13:42 #1004 by PhracturedBlue
Replied by PhracturedBlue on topic Protocol Stacks
Your logs show the maximum range to be 0x000 to 0x3ef, so I've changed the protocol code to match.

using the 0x3ef max with my 2801Pro Rx:
2801 protocol: center = 2.0msec with +/- 0.5msec throw
2601 protocol: center = 1.2mec min=1.5 and max=1.8, so this is definitely wrong

using the value of 0x190 that I got from renatopub:
2801 protocol: center = 1.7msec with +/- 0.2msec throw

I found a spec that at 100% I should expect:
center=1.5msec with +/-0.4msec throw

So, I must have something wrong in the protocol since it isn't possible for me to achieve those values with the signals I'm sending.
Last edit: 09 Aug 2012 13:42 by PhracturedBlue.

Please Log in or Create an account to join the conversation.

More
09 Aug 2012 13:51 #1005 by wuselfuzz
Replied by wuselfuzz on topic Protocol Stacks

FDR wrote: Probably would be more consistent/informative, if you would reset all the trims...


Didn't pay attention to those, just did some quick dumps while rushing to work. I'll do some more investigation later.

1.5msec +/-0.4msec sounds about right. Most servos could do quite a bit more than that, I'd expect to see 0.9msec - 2.1msec full range with everything maxed out. (at 150%, 0.4 * 1.5 = 0.6, 1.5-0.6 = 0.9, 1.5+0.6 = 2.1.)

Please Log in or Create an account to join the conversation.

More
09 Aug 2012 14:01 #1006 by FDR
Replied by FDR on topic Protocol Stacks

wuselfuzz wrote:

FDR wrote: Probably would be more consistent/informative, if you would reset all the trims...

Didn't pay attention to those, just did some quick dumps while rushing to work. I'll do some more investigation later.

I'm interested in that, because all the four trim values are different in your logs, and I don't know if the trim value is reflected in the channel value too, or it is sent only separated.

I would like to see a channel value at min/max and mid stick with zero/min and maxed trims. (That's 9 combinations... ;) )

Please Log in or Create an account to join the conversation.

More
09 Aug 2012 14:17 #1007 by PhracturedBlue
Replied by PhracturedBlue on topic Protocol Stacks
Ok, I figured it out. I misunderstood renatopub's logs. In the wk2801 protocol, byte 13 is the sign bit.
when the channel is at 0, the value is 0. a value of 0x190 is either +/-100% where '+/-' is set by the relevant bit in byte 13

Please Log in or Create an account to join the conversation.

More
09 Aug 2012 14:31 #1008 by FDR
Replied by FDR on topic Protocol Stacks

PhracturedBlue wrote: Ok, I figured it out. I misunderstood renatopub's logs. In the wk2801 protocol, byte 13 is the sign bit.
when the channel is at 0, the value is 0. a value of 0x190 is either +/-100% where '+/-' is set by the relevant bit in byte 13


Was that 0x190 with travel adjust maxed out or is that just the 100% ?
It will be interesting to compare the max values with the WK2402, which has no such travel adjustment...

Please Log in or Create an account to join the conversation.

More
09 Aug 2012 14:58 #1009 by PhracturedBlue
Replied by PhracturedBlue on topic Protocol Stacks

FDR wrote: Was that 0x190 with travel adjust maxed out or is that just the 100% ?
It will be interesting to compare the max values with the WK2402, which has no such travel adjustment...

That is 100%. I was able to get the pulsewidth to 2.5msec with the RX 2801 which would be 250%
The firmware is currently only capable of driving +/-100% though.

We'll likely need to see the servo PW and packet data together on the 2402 for a channel sweep with and without trims as FDR mentioned to figure out what is actually going on.

Please Log in or Create an account to join the conversation.

More
09 Aug 2012 15:03 #1010 by FDR
Replied by FDR on topic Protocol Stacks

PhracturedBlue wrote: The firmware is currently only capable of driving +/-100% though.

...and I guess that is not even the real 100%, because on the CB100 I barely could see the servos move, and the max throttle couldn't lift off the heli.

Please Log in or Create an account to join the conversation.

More
09 Aug 2012 15:07 #1011 by PhracturedBlue
Replied by PhracturedBlue on topic Protocol Stacks

FDR wrote:

PhracturedBlue wrote: The firmware is currently only capable of driving +/-100% though.

...and I guess that is not even the real 100%, because on the CB100 I barely could see the servos move, and the max throttle couldn't lift off the heli.

Yes that is what we're talking about.
I have a fix for the 2801 that gets it to a real 100%. The code I checked in for the 2401 MAY get you there too, but we'll need to wait for wuselfuzz's data to see where we stand.

Please Log in or Create an account to join the conversation.

More
09 Aug 2012 15:14 - 09 Aug 2012 20:53 #1012 by wuselfuzz
Replied by wuselfuzz on topic Protocol Stacks
Throttle min/max is up there. I also have dumps for aileron min/max and rudder min/max available. I think both channels were set to reverse in the TX2402 for my UFLY (which died in a horrible... HORRIBLE chicken dance a couple of weeks ago. I think I plugged the ESC into the wrong rx output, and those Walkera ESCs don't have a "Throttle to zero to arm" safety). I just wanted to set the UFLY up for a buddy to try out. Anyways, enough talk, raw data:
removed to avoid confusion, MSB first/LSB first mixup, see below

As stated earlier, both channels might be reversed in the Tx, checking later.

/edit: oops, and reduced.
Last edit: 09 Aug 2012 20:53 by wuselfuzz.

Please Log in or Create an account to join the conversation.

More
09 Aug 2012 15:15 #1013 by FDR
Replied by FDR on topic Protocol Stacks
Sorry, it's my english...

I would like to say that "it is not even close to the real 100%"...

Please Log in or Create an account to join the conversation.

More
09 Aug 2012 15:16 #1014 by PhracturedBlue
Replied by PhracturedBlue on topic Protocol Stacks

FDR wrote: Sorry, it's my english...

I would like to say that "it is not even close to the real 100%"...

With the fix I checked in an hour ago that is still true?

Please Log in or Create an account to join the conversation.

More
09 Aug 2012 15:23 #1015 by PhracturedBlue
Replied by PhracturedBlue on topic Protocol Stacks
wusselfuzz,
there is no need to send all channels at the moment. i already have the mapping. what I need to see for a single channel is the zero, min, max with the trims zeroed, and zero/min/max with the trim maxed out. We should probably collect the same with the trim at full-min as well so we can see how sign is handled.

I would like to see both the data packet and the servo Pulse-width at these points.

I don't care about reverse (as long as we have servo pulsewidth we can figure that out)

Please Log in or Create an account to join the conversation.

More
09 Aug 2012 15:25 #1016 by FDR
Replied by FDR on topic Protocol Stacks

PhracturedBlue wrote:

FDR wrote: Sorry, it's my english...

I would like to say that "it is not even close to the real 100%"...

With the fix I checked in an hour ago that is still true?

I don't know, I'm still at work, and I'm not sure I will be able to check that this evening. I spent too much time with it, and my wife got across me...

Please Log in or Create an account to join the conversation.

More
09 Aug 2012 15:35 #1017 by PhracturedBlue
Replied by PhracturedBlue on topic Protocol Stacks

FDR wrote:

PhracturedBlue wrote:

FDR wrote: Sorry, it's my english...

I would like to say that "it is not even close to the real 100%"...

With the fix I checked in an hour ago that is still true?

I don't know, I'm still at work, and I'm not sure I will be able to check that this evening. I spent too much time with it, and my wife got across me...

I know the feeling :)
Anyhow, the new code does increase the range substantially for the 2401 protocol. Whether it is right or not, I do not know.

Please Log in or Create an account to join the conversation.

More
09 Aug 2012 15:36 #1018 by wuselfuzz
Replied by wuselfuzz on topic Protocol Stacks

PhracturedBlue wrote: there is no need to send all channels at the moment.


That was just data I had available from the dumps in the morning. I'm home in about an hour and can dump more data then.

The RX2435 normally has an ele-ail-ext pot and integrated 3-axis-gyros. I removed the gyro board from the rx, though, and grounded the ADC input of the ext pot. This should result in max servo throws (confirmed by a test servo). I hope that the missing gyros don't affect outputs too much.

Please Log in or Create an account to join the conversation.

More
09 Aug 2012 17:07 - 09 Aug 2012 20:54 #1019 by wuselfuzz
Replied by wuselfuzz on topic Protocol Stacks
this is so awkward... see below.
Quick checks for Throttle 11%, 12%, 15% also yielded 910us duty cycle. Didn't extract SPI data for those.

However, at 16% Throttle with 0% throttle trim, the pulse width rises to 920us.

=> 15% Throttle = 50% Throttle trim in Walkera land?

/edit: Measured with 10MHz sampling rate, dT=0.1us
Last edit: 09 Aug 2012 20:54 by wuselfuzz.

Please Log in or Create an account to join the conversation.

More
09 Aug 2012 17:32 - 09 Aug 2012 17:33 #1020 by PhracturedBlue
Replied by PhracturedBlue on topic Protocol Stacks

wuselfuzz wrote:

Throttle 50%, Throttle trim 100%:
df 00 2f ff a6 5f 40 c8 ff 93 86 62 a0 62 34 38 

Throttle 50%, Throttle trim 0%:
df 00 8f ff a6 00 40 c8 ff 91 86 62 40 02 49 6b

Can you check these again?
The 1st looks like 100% throttle, 0% trim
The 2nd looks like 25% throttle, 0% trim

Is it possible there is some CCPM mixing going on here, and we're not getting linear throttle control?

Edit: Rudder is probably the best channel to look at for a heli Rx I guess.
Last edit: 09 Aug 2012 17:33 by PhracturedBlue.

Please Log in or Create an account to join the conversation.

More
09 Aug 2012 18:15 #1021 by FDR
Replied by FDR on topic Protocol Stacks

PhracturedBlue wrote: Is it possible there is some CCPM mixing going on here, and we're not getting linear throttle control?

The WK2402 does't have mixing.

PhracturedBlue wrote: Edit: Rudder is probably the best channel to look at for a heli Rx I guess.

Not really...
Even the older models have headhold gyro, which effects the rx's rudder output...

Please Log in or Create an account to join the conversation.

More
09 Aug 2012 18:31 #1022 by PhracturedBlue
Replied by PhracturedBlue on topic Protocol Stacks

FDR wrote:

PhracturedBlue wrote: Is it possible there is some CCPM mixing going on here, and we're not getting linear throttle control?

The WK2402 does't have mixing.

PhracturedBlue wrote: Edit: Rudder is probably the best channel to look at for a heli Rx I guess.

Not really...
Even the older models have headhold gyro, which effects the rx's rudder output...

Right...4-channel doesn't have pitch control... :oops: . But the 4channel does have heading hold? how do you control it? That would need to be a 5th channel (even if it is not proportional) right?

Please Log in or Create an account to join the conversation.

Time to create page: 0.090 seconds
Powered by Kunena Forum