Deviation for Devo F4?

Arnold wrote:

On 22 April 2016, at 15:56, SirDomsen wrote: ... Wiring as in the pictures. Tried the latest nightly - the switches work now as expected...

I used SirDomsen's pix from this post (#47103) as my guide for connecting switches SWA & SWB, but I get the trim beeping. I connected two 3-way switches. The first attempt was without diodes and the result was that the radio would startup in USB mode unless I centered the switches first, then it would start normally, but I couldn't get past the trim beeps with every button press. So, the second attempt was with diodes and this time the radio starts OK, but still getting the trim beeps and won't proceed.
So, how can I get rid of the trim beeps? I find no clues from reading the thread.

> Edit: I changed the switches. Now have 1-3 way and 2-2 way, just like Sir Domsen. But Still have those beeps.

I attach a better pic , recheck wiring, pay attention on unwanted GND touches at the switches (had that on my Devo10). Note that the black wire goes to the middle pin (PB8)

I've used trim and turned the transmitter off with trims changed and it hasn't crashed... I don't usually do it because I tend to replace trim with subtrim once on the ground, but on the odd day I've forgotten to... in fact, I have yet to experience any of these FS problems you guys talk of...

What power supply are you guys using? LiPo or AA??

2 Cell Lipo, 2x18650

SirDomsen wrote: I attach a better pic , recheck wiring, pay attention on unwanted GND touches at the switches (had that on my Devo10). Note that the black wire goes to the middle pin (PB8)

Yes, saw that, thanks. Your new photo confirms that I have the correct connections. So, I must have bridged something somewhere. Using a magnifying headset, I've spotted some questionable solder joints which I'll redo.
Just to confirm the Hardware.ini comment, I entered, "extra-switches=3x1, 2x2", right?
Once I get this sorted out, I'll look into soldering to the MCU for other switches. A 3-position FMode switch appears to be challenging if even possible.

Switch problem solved. The white wire was poorly soldered to the board. Fixed now. Moving on.

A 3 position FMODE switch is possible, just pretty hard, as is the MIX 3 way switch. A good way to solder to the MCU is to cover the adjacent pins you are not going to be soldering to with come kapton tape, as it's resistant to the heat of the soldering iron. I've done the 3 position FMODE switch, so I know it can be done and the connections are correct, but I changed it back to a 2 position as that's the switch I wanted there.

In fact, it would be easier to remove the 2x2 you have just done and and wire it into the 3x1 you want to use for FMODE, and feed those 2x2 from, say, the GEAR and the AIL DR points on the MCU, as those solder points are not right next to each other, where as the FMODE in 3 way switch mode does have 2 solder points right next to each other, which is the hardest thing to solder... and the end result will be the same; two 3-way switches and two 2-way switches... and less of a soldering headache.

I have two 3-way switches and three 2-way switches on my F4, and I didn't have to solder any adjacent pads on the MCU, I just soldered individual ones.

Epitaph wrote: A 3 position FMODE switch is possible, just pretty hard, as is the MIX 3 way switch.

OK, on the F4 schematic I see only pin #51 for FMode and no others. How do you get 3 switch positions from one pin? ... or did I miss something? For Mix I see pins #55 (switch position 0) and 56 (switch position 2). And while we're talking about MCU pins, what's up with those two Hold pins, #52 & 53? I believe you mentioned previously that neither of those worked for you and you did the work-around to the throttle pot instead.

About soldering to the MCU pins...
I no longer have the benefit of keen vision or steady hand of my youth, so the risk of bridging adjacent pints could mean "game over" for my F4 unless there is a way to repair that with solder wick or something. I've seen the video where the tech girl describes the process using Kapton tape and I believe I'm up for that. But still, I hesitate...
Maybe I'll warm up to the idea by first soldering pin #2 for the 3rd module of the 3-in-1 module, as you suggested. Wish me luck!

SirDomsen wrote:

sensaurabh wrote: Yeah, that is what I usually do. I use the buildfs to directly create the devo.fs on the TX instead of saving it and transferring it over. I do keep a copy of all the changes that I make on the TX on the buildable rootfs I flash from. It's just that sometimes, even if I do not make any changes, the filesystem crashes. I somehow suspect that the TX saves the trim settings when it shuts down which causes writes to the filesystem and so now I trying to cultivate the habit of returning all the trims to neutral before switching the TX off.

Can confirm, when you trim your quad, it's most likely that the FS crashes - can you confirm that if not trimmed, it does not crash? I think i never recognized that relationship between trims and crash...

That is correct. Changing settings in model config and saving trim values at the time of shutdown 'sometimes' corrupts the filesystem.

The schematic isn't quite right... it turns turns out that pin 51 and 52 are for FMODE0 and 2 (#51 is for FMODE2 and #52 is for FMODE0). On the schematic it's marked up as being for HOLD, and it's incorrect... I found this out when I was trying to find HOLD, using pin 52 and found the toggle between FMODE0 and 1. So, if you put a 3 way switch with the piddle pin to ground, and the 2 extremes to pins 51 and 52 you toggle between FMODE0 (pin 52), FMODE1 (no connection) and FMODE2 (pin 51).

Pin 53 doesn't seem to want to do anything, and I've tried various configurations, even bridging it with pin 52 before I found that 52 was really for FMODE. I just can't find HOLD at all on this transmitter. That pot work around I changed for the AIL DR switch and just use that as if it was HOLD, because the workaround works great for normal planes, quads, etc. but not so good when you need to have throttle hold at 0 instead of -100.

As for the soldering issue, there is another way that might make it easier, and that is physically separating the pin from the board, as it's not doing anything on the board, they are not connected to anything.

Pin 16 for the 3-in-1 module is pretty easy to solder as it's the end pin, so just solder it from the side instead of head on, and that will give you more surface to solder to.

Epitaph wrote: The schematic isn't quite right... it turns turns out that pin 51 and 52 are for FMODE0 and 2 (#51 is for FMODE2 and #52 is for FMODE0). On the schematic it's marked up as being for HOLD, and it's incorrect... I found this out when I was trying to find HOLD, using pin 52 and found the toggle between FMODE0 and 1. So, if you put a 3 way switch with the piddle pin to ground, and the 2 extremes to pins 51 and 52 you toggle between FMODE0 (pin 52), FMODE1 (no connection) and FMODE2 (pin 51).

Nice! Good find.

As for the soldering issue, there is another way that might make it easier, and that is physically separating the pin from the board, as it's not doing anything on the board, they are not connected to anything.

Sounds dicey...

Pin 16 for the 3-in-1 module is pretty easy to solder as it's the end pin, so just solder it from the side instead of head on, and that will give you more surface to solder to.

Yeah, pin 16. I should have said, ",,, pin A2".
Anyway, mission accomplished, as Geo Bush said...
I think it was more difficult placing the tape precisely and stabilizing the wire securely than actually soldering. A small amount of tinning of the wire and a couple taps with the iron on the target and it was done. Now, I have all 3 additional modules in my F4. Whew! OK, next.

Stumbling along here.

Back on 16 April, 2014, PhracturedBlue, when he spoke about adding switches to the 7E wrote: ... In the switch mod what you are doing is making adding a switch that looks like a button. To save wiring, the buttons are actually wired together in a grid (much like your keyboard) to allow a few wires to detect many more buttons. The downside of this, is that there are many cases where holding multiple buttons simultaneously can't be detected. Since we're treating a switch like a button, it is as if we are holding it down forever. This would prevent other buttons from being detected if the diode wasn't there.
The diode only allows current to flow in one direction. This allows us to read the buttons in one direction and the switches in the opposite direction. Basically it means activating a switch doesn't affect our ability to read the buttons.

... The answer I was looking for.

Wondering why the extra switches I added are not showing up anywhere. I entered the linein the Hardware.ini file. My added modules are enabled, but not the switches. What's wrong?

I think as the 2x2 ones are technically a 3x1 switch that has been "split in half", it still counts as a 3x1 as far as the software is concerned, in which case the line would have to read

extra-switches=3x2

When you write 2x2, the software understands it as two 2-way switches, but one as SWA and the other as SWB... then you're telling it that there's also a 3-way switch, so it's thinking "hang on a minute, if one of the 2-ways is A, and the other is B, then what can I honestly call the 3 way one?"

Correct. Unfortunately there is (yet) no way to add a SWC, which would be needed if using 3 switches. So you have to use - as Epitaph already wrote - 3x2 instead.

Thank you both. I'll make the change.
In the mean time, I need someone to proofread the attached PDF file of the schematic for the F4. I made corrections to the file based on Epitath's comments about the error for pin 51 - it was mis-named "Hold" and I changed it to "FMode2". While making that change, I also changed the name of pin 52 to "FMode0", again based on her comments. I also changed the names in the GPIOC section for C10 & C11.
If it checks out OK, I'll see if I can work out how to replace the incorrect version in the Wiki. The PDF is too big for the forum to handle as an attachment, so get it from my DropBox here:
dl.dropboxusercontent.com/u/13050475/SchemaDevoF4v2.pdf
PS. I use an uninfected Mac... trust me ...

If you want to add something, pins 2 and 3 (PC13 and PC14) are written down as simply "D/R"... Pin 2 is actually AIL DR1, and pin3 is GEAR1, both of which are simple 2 position "on/off" switches.

Also, if you want to change the definition of pin 16 (PA2) for where a 3rd module would go (activated in the hardware.ini file as "A2").

Epitaph wrote: If you want to add something, pins 2 and 3 (PC13 and PC14) are written down as simply "D/R"... Pin 2 is actually AIL DR1, and pin3 is GEAR1, both of which are simple 2 position "on/off" switches.

Thanks. I made those changes and replaced the PDF in my DropBox folder, here:
dl.dropboxusercontent.com/u/13050475/SchemaDevoF4v2.pdf
The PDF has the same name as the previous version.

Also, if you want to change the definition of pin 16 (PA2) for where a 3rd module would go (activated in the hardware.ini file as "A2").

I didn't do this mainly because I don't have a font that matches the style of the text in the file. Actually, there are no alphabetical characters in the PDF at all. All text was converted to vector graphics by someone, probably to avoid the issue of users not having the appropriate font. So, to make the edits all I did was copy & paste, not actually typing anything.

FDR wrote: Would you correct the wiki page too?
www.deviationtx.com/wiki/hardware/devof4

Done!