FY326 Q7 "red board"

Hello hexfet.

Thanks for looking at the data.

I have only had my logic analyser for less than a week, so I'm still getting to grips with it. I went to 'Options' in the top RH corner, then 'Export Data', which brings up a dialogue box that has CSV as an export format option, so thought I had it covered!

I have captured data from two different quads and transmitters, and the RF channels are dependent on which transmitter is being used (the other transmitter uses channels 0a, 1b, 2c, 36, 43 ).

It would be great to get the red board FY326 Deviated so we can find out what it's really capable of!

Steve

The +-20 dead zones on the pitch, roll, and yaw functions are slightly noticeable as a lack of 'responsiveness', but are not nearly as troublesome as the dead zone at the lower end of the throttle travel. I did notice, however, that the quad's motors do not start moving until a (decimal) value of 30 is reached on the throttle byte. Is this an indication that there is something else going on at the receiver end?

I had assumed that the centre-stick dead zones were to mask poor quality gimbals in the transmitter, but may be wrong about that. [Edit: I have since noticed that the dead zone in the pitch control is not even close to being symmetrical - using transmitter 'A' there is hardly any dead zone in the forward pitch direction, and a considerable amount in the other direction. This cannot be equalised using trim. Transmitter 'B' is a little better, but still not perfect. It has been noticed that the FY326 is less prone to 'calibration issues' than similar quads. I wonder if the wide dead zones at the centre of RH stick travel are the reason why.]

Byte 13 depends on the quad used.
Byte 14 depends on the transmitter used.
I have seen (decimal) values of 85 and 170 on byte 6.

There is no third control that I'm aware of.

About the motors not starting until 30 - would you please check what effect the throttle trim has on this value? Might tell something about how the receiver handles that trim. What is the usual effect of the deadzone in the throttle - do the motors smoothly slow to a stop but there's still travel left in the stick, or the motors cut power suddenly?

Here's info from the captures in the two links posted. Could you also post these values for the other quad (with the 0a, 1b, 2c, 36, 43 RF channels)?



Leaping straight to conclusions to be disproven later...
- packet bytes 0, 7, 8, and 14 are fixed and differ by tx - the "txid"
- byte 6 in the tx bind packet varies by tx (but is not transmitted as part of the data packet so won't consider it part of the txid)
- byte 13 in the rx bind and tx data packets varies by rx (possibly calculated from tx bind packet byte 6 and/or txid?)
- bytes 1, 6, and 13 are used as a "handshake"
-- tx bind packet always sends 0x55 in byte 1 and 0xaa in byte 13
-- rx bind response has txid in same bytes as tx bind packet, 0x55 in byte 1 and 0xaa in byte 6. New value in byte 13 (rxid?)
-- tx data packet keeps 0xaa in byte 6 and sends rxid in byte 13
- the RF channel cycles through [0..4] in the upper nibble
- the RF channel lower nibble is taken correspondingly from [tx bind byte 6 lower nibble, tx bind byte 6 upper nibble, tx bind byte 7 lower nibble, tx bind byte 7 upper nibble, tx bind byte 8 lower nibble]

Hello,

I've haven't been able to get the throttle trim to have any effect at all (on anything!). Others on the RC forum have claimed to have witnessed effects on the throttle response, but I'm skeptical, having tried it myself and failed. Balancing the motors just on the point of turning and then adjusting the trim seems to make no difference. The motors start suddenly as the throttle stick is advanced, BTW.

To avoid any confusion, I have designated the transmitter I used for the captures already posted as 'transmitter A', and my other as 'transmitter B'. Likewise, the quad used up to the current time is 'quad A', the other, obviously, being 'quad B'. There's also robca's quad and TX (may I suggest 'C'? - I apologise for 'pushing in', but have already marked my quads with a Sharpie without thinking about courtesy!)

I'll work on the data you asked for and report back...

No worries. Thanks for making the captures.

It would be interesting to see if the trims affect the cyclic controls around center stick. Would be more difficult to tell but maybe if the throttle was just at take off point so that a roll/pitch command would lift one side of the quad...

With the quad's motors set at the point where it's just about to take off, the pitch and roll trims work as expected, tilting up the appropriate side of the quad.

The throttle trim, on the other hand, doesn't seem to do anything at all.

I hope that tells you what you need to know.

The transmitter's plastic case is a familiar off-the-shelf item used for transmitters supplied with several different quads, so there was no alternative for the designers but to have a button in the 'throttle trim' position. The buttons modify bytes in the TX payload, so could be used for something useful in a different model or future product, I suppose - camera, headless or something?

Taking-off with full yaw trim (in either direction) does not produce any yaw.

Here is a capture of the bind sequence for the 'B' transmitter and 'B' quad (hence the 'BB' in the file names). I hope that is what you wanted. If not, please let me know (I'm new to this, so may have misunderstood what you meant). I trust the csv file format is correct this time.

I should point out that transmitter B has had a hardware modification that means that the throttle byte never has a value less than 14 (hex), at least after binding (ie, is never '0'). This was an early attempt my me to modify the throttle response. The fact that I can still stop the motors is notable, as mentioned in an above post.

There is another small quirk of the red board FY326. As the motors are armed by setting full throttle and then minimum throttle, the transmitter beeps twice, but the 'low throttle' beep does not occur at the same throttle position that actually arms the motors. It is possible to hear two beeps, but the motors are still not armed (until the throttle is reduced even a little further). I don't know if this is significant, but thought it was worth mentioning while I was thinking of it.

Pressing throttle trim (+) sets byte 1 to 0x80

Pressing throttle trim (-) sets byte 1 to 0x40

Pressing LH stick sets byte 1 to 0x04 (this can be added on to either of the above)

I have seen other values on byte 1 but have yet to figure out how they got there.

Switching the transmitter off then back on re-binds, but also seems to re-set bit 1 to 0.

Byte 6 is initially set to 0xAA, but if the transmitter is switched off then back on again, quad is still bound, but byte 6 is re-set to 0x55.

S.Giles wrote: Pressing throttle trim (+) sets byte 1 to 0x80

Pressing throttle trim (-) sets byte 1 to 0x40

Pressing LH stick sets byte 1 to 0x04 (this can be added on to either of the above)

I have seen other values on byte 1 but have yet to figure out how they got there.

Switching the transmitter off then back on re-binds, but also seems to re-set bit 1 to 0.

Byte 6 is initially set to 0xAA, but if the transmitter is switched off then back on again, quad is bound, but byte 6 is re-set to 0x55.

Could be related with the fact that there are reports about a version with red board and Headles mode?
Maybe the transmitter is sending the Headless mode command, but only work in some quadcopters.

My FY326 isn't one that came with a yellow 'headless mode' sticker on the box, but I tried putting it in headless mode with the 'throttle trim' button anyway - no luck.

Someone recently reported that headless mode worked well on his FY326, which suggests a board revision. I should find his post and ask him what he did to enter headless mode.

That's my thought. Maybe the TX's internal firmware has the Headless mode command, but only some Quadcopters (with yellow sticker) have the feature active at internal firmware.

A long LH stick press (high beep) adds 0x10 to byte 1 of the TX payload.

Another long LH stick press returns it to its previous value.

(I only just became aware this function existed).

I've added code here that builds and has a chance of binding. If you set the fixed ID to 762704 it will use the same txid as transmitter A. If you don't have your build environment set up yet let me know what tx you have and I'll upload a test build.

Not sure what to make of the throttle trim changing byte 1. You may be right about the same tx being used on multiple models.

I will add another channel for the LH stick press, but it would be best to know what it's used for. We're trying to head to standard channel assignments for these things. Currently I have flip on channel 6 and expert mode on channel 10.

I'm conflicted about adding the expert mode control. What effect does it have on the quad? I favor the approach of always setting the quad to it's highest performance mode and using deviation mixers to scale back. But it's possible the bit has some other effects at the receiver, and it seems wrong somehow not to be able to control every bit the receiver pays attention to.

Expert mode just seems to make the RH stick more responsive. There aren't any special features (flips, for example) that are not available in 'standard' mode, as far as I know.

I (for one) would be more than happy to do away with standard mode, and just scale back the full response to taste. Isn't having two pre-set modes just a cost-driven simplification of what Deviation can do so much better? I always use expert mode anyway (and it's a pain always having to remember to push-in the LH stick on the stock transmitter).

I have a Devo 7e being shipped as we speak (did I mention I was new to this? ), so would someone who already has a suitable setup (and knows what they're doing, unlike me!) be willing to step in and help at this point?

As S.Giles mentioned, he has not received his 7E yet and no build environment. I can test a build but have also no "build environment". Perhaps Robca does but i'm not sure. I guess Durete has such environment for the 7E and is willing to build it or someone else? Sorry

I compiled latest code for FY326 from Hexfet and uploaded to my personal Dropbox.
Devo 7e build:
Removed. Obsolete build.
As you know, I don't have this quadcopter so can't test. I'm only trying to help/learn as much as possible, because the 7e build require an "special" environment config to compile properly

Thanks, I will try with the fixed ID to 762704 etc.
Flip channel 6 and high rates channel 10.

As a first short test, it binds with that id 762704
Pitch ok Aileron is reversed and rudder too but I cured that in my ini.

First powered on the quad, switching on the 7E and it binds and then needs the throttle stick up and down before motors can start turning.
It auto rebinds after switching 7E off and back on, meaning just give throttle and it takes off again.

The rudder seems faster to the left and to the right seems slower (on my beated up FY-326)
The throttle dead zone is not much but once reached the motors just cut out abruptly like we are used too but needs
more tests outside because after changing the ini it seemed better.
Flips, I still need to check but wanted first to be sure the controls worked normal.
For the rates I have to check again my ini.
After changing ini, channel 6 flips are working and rates on channel 10 too. The yaw speed is proportional to the rate.
If using my trim buttons it seems to do nothing on mine (I have for channel 3 "usetrim=0" but for yaw, pitch and roll they are defined
but the problem is perhaps in the ini.

Already happy

Someone on the RC Forum ("splodgie") has posted that (on the model that supports it) headless mode is activated by pressing the (+) throttle trim and RTH using the (-) throttle trim. He also mentions owning a version where the throttle trim buttons actually work as a throttle trim, but doesn't go into detail. Any 'red board' transmitter will bind with any 'red board' quad, if I understand his post correctly.

So it looks as though the 'throttle trim' buttons increment byte 12 in case they are used to implement a trim function, and also modify byte 1 for when special functions are implemented instead.

Test ini for the Devo 7E to use with the FY-326 Red Board test build from Durete's post.
(For what I noticed during test see my previous post)

The ini,
Channel 6 flips is on the HOLD switch.
Low and higher rate is om the FMODE switch.

After bind (on mine) it needs the throttle stick Up and Down just like with original tx.
It is a test ini so somebody can check it out and make improvements.
-

Removed; obselete