9Xtreme Upgrade Board for 9x Radios

[MikeB]

I think we can say IT WORKS!

I've just mounted the 9Xtreme in a 9X, with a brand new M64 main board. Switched on and it found the board, detected it wasn't flashed correctly, flashed it, then started working.
It is reading the sticks and trims. I've still to plug in the main 12-way connector from the back, I'm testing with my external PSU.
Backlight (single colour) working.

I'll proceed to testing everything out now. I've just got to undo one or two mods I have on this radio, like the RS232 buffers on the telemetry.

Mike.

[jhsa]

Great.. I think you will have a new favorite radio

Thank you.

João

[magnetus]

Mike thanks for your hard work. I know this is in the early stages still but is it there a possibility to upgrade the LCD screen, maybe in a DIY transmitter, to a 212x64 pixels or anything different than the stock LCD Display. Also is there going to be LUA support for this product?

thanks again.
Miguel

[MikeB]

Changing the display is not straightforward. We are using the stock AtMega64 to drive the display. There are, however, possibilities, providing you can get the display connected.
I may be able to add LUA support, but that will be further down the line when all else is working.

I have the radio now closed up, using the battery instead of the bench power supply, except for the USB and serial debug cables. I'm now finding various small bugs, like some switches are backwards, and some menus are not quite right.
The bootloader (both horizontal trims inwards) is working and I've updated the firmware using it.
Just sorting the options for the 3-colour backlight, Steven needs this working to test the actual backlight.

Mike.

[jhsa]

I thought we didn't need the trim inwards rain dance anymore i ersky9x On the pro and skyboard we just plug the cable and turn the tx on. Can't it be like this also with this board?

For the backlight maybe create 3 sliders for the intensity of the 3 colors? That would probably make it easier to select the color we want for the backlight, perhaps along with some presets.
Another option for the backlight might be needed on the Voice /Alarms menu? We could also control it with alarms, voice, or even just by itself..

Thanks

João

[MikeB]

Bootloader access is a function of the processor, not the firmware. With the STM processor, when powered off, as soon as you plug the USB cable in, it start in STM DFU mode.

Fancy control of the backlight will wait until all the normal functionality has been tested.

Mike.

[jhsa]

Bootloader access is a function of the processor, not the firmware. With the STM processor, when powered off, as soon as you plug the USB cable in, it start in STM DFU mode.

So, nothing you can do then to make it behave like the skyboard?? There is no reset pins, or whatever the skyboard has? the pro has the reset button for eventual critical situations.
It's very cool not having to remember ONE MORE key(s) press..

Fancy control of the backlight will wait until all the normal functionality has been tested.

Mike.

Of course But it can be discussed

João

[MikeB]

No, the Atmel processor has a flash bit you set to force it to start in normal mode. I have some sneaky code that can drop out to the ROM bootloader. The STM only operates using hardware, so there isn't anything I can do.

Latest: A DJT module in the module bay is controlling a servo.
Next up: wire up the telemetry and test that, and also a XJT module.
I've just checked all the menus and fixed all the oddities I can see, so we are looking good. Still things to test, like trainer function when operating as a student (powered by plugging in the trainer cable).
I also haven't implemented the extra method (non-spring pins) for sending the pot and switch status from the M64 to the 9Xtreme.

Mike.

Edit: DJT telemetry now working. XJT in PXX mode and SPort telemetry OK.
Edit: Trainer output now working. I'm considering turning both modules power off when in student mode.

[MikeB]

So DJT and XJT all working, with telemetry, also trainer output.

Since the 9X I've put the board into had the ELE switch changed to a 3-position switch, I've started to handle extra digital inputs. All the trim inputs to the M64 are available as extra inputs, so I've grabbed one of them, already being sent to the 9Xtreme board, and used that as the extra input needed. All works fine. This is an enhancement, so I'll not be adding much else like this until the other basics have been checked.

Mike.

[jhsa]

It would make it easier to find bugs if other people that spend a lot of time flying were also testing the board and the firmware. with only 2 people doing it it is difficult to catch all of the possible problems, as some might happen only under certain conditions.. I speak based on the experience while testing the 9XR-PRO.
I will probably be glad to help with the enhancements when I have my board. I can do a lot of flying now, but soon it will be winter here I guess the boards will still take a few weeks until they are ready to ship

João

[MikeB]

One problem with the prototypes is lack of all the extras like mountong spacers and screws, tri-colour backlight, USB connector, and all the other specific cables, with their correct connectors, like the telemetry cable.
This makes it tricky to install.

Mike.

[jhsa]

Yeah, I understand. And prototypes are expensive to be making many.

When testing the PRO, I remember logging at least 30 hours of effective flight time only with 2 of my models, since you implemented logging around March 2014, until the radio was released a few months later
It was a lot of flying. And it was fun

João

[magnetus]

(((So DJT and XJT all working)))
Will the 9Xtreme board be able to handle DSM2/DSMX, UHF and FASST protocols ?

[jhsa]

FASST? I doubt it

[s_mack]

Hey Mike, I just noticed that on my prototypes (probably the same with yours), they put the "wrong" chip on! But in a good way. Look carefully, it has 1M flash, not 512K. HOWEVER, that may be something they did for the proto due to availability, so we can't count on it. Still... temporarily, at least, ours is bigger than theirs is

[s_mack]

João... regarding testing of the 9xr. I don't know, but I'd guess that perhaps that was a round later than we're at now. They probably already had gone though a round or two of internal testing. That's where we're at now. I'm hoping it is the only round, mostly based on the fact we're not new at this. The physical aspects are all extensions of the proven SP and tEZ board and the bulk of the circuitry and firmware are expansions/translations from existing mature projects (I'm probably over- simplifying it, of course). While I'm elated things are going as well as they are, I'm not really surprised given Mike's experience and Mad Playa Skillz!

[ShowMaster]

Being in Los Angeles, the weather is usually good for year round flying. I'm ready when the board is delivered and ready to install.
Great on all the progress you're making Mike and Steven.

[Kilrah]

(((So DJT and XJT all working)))
Will the 9Xtreme board be able to handle DSM2/DSMX, UHF and FASST protocols ?

Those are function of whatever module you use, not the radio itself.

Envoyé de mon SM-G920F en utilisant Tapatalk

[MikeB]

The code is all present to support DSM 'hack' modules (I don't have one to test with), and also the HK DSM telemetry module, and the, as yet unreleased, DSM module from another manufacturer.
Any module that fits in the module bay, and supports PPM should also work.
You may fit a DHT 'hack' module internally. I don't have the code quite sorted to use a DSM 'hack' module internally yet, I'm using the code from the Taranis and that has a XJT permanently internally, so there was never a requirement to support anything except PPM and PXX. It shouldn't take too long to add though.

I've managed to allow all 8 trim inputs on the original board as extra switch inputs, selectable for THR, ELE, AIL, GEA, RUD, PB1 and PB2 switches.

The target is to get working firmware with all basic functionality operating so the production boards are ready to go. Clearly, there may be later updates. With the bootloader all working (I'm now using that as the method of updating the firmware over USB), subsequent firmware updates should be easy.

Mike.

[ShowMaster]

Is it too early to ask if the first boards shipped to early buyers will be only require component or simple trace mods to be upgraded?
Is there an early option to also pay for a newer board at a reduced price if too many changes would be required? I'm thinking early contributors are helping to get this product to reality and it'll go through design changes once it's out there.
Hopefully any changes will be user upgradable if one has the soldering skills.
Then again that's why these first beta boards are being sorted out now, to avoid issues.
I'm just suggesting I'm willing to spend some extra $.
I don't need another TX but I can't resist trying another new DIY type upgrade for a 9X, the Swiss Army knife of Rc!

[MikeB]

Steven will confirm, but the plan is that boards shipped will require NO mods. There are only a couple of circuit mods needed on these prototypes, and some re-positioning of connectors to make it easier to fit.

I'm just working on the 'backup' method of sending data from the M64 to the 9Xtreme. I'm about 90% of the way there, the M64 is sending the data, the 9Xtreme is not quite detecting it yet. I added a manual option for updating the M64 as I needed an easy method while in development.

Mike.

[ShowMaster]

Thanks for the update Mike.

[s_mack]

I'm not really sure I understand the question. Just to be clear, Mike and I have prototypes but nobody else does. If those are the "first boards" you're referring to, then it isn't an issue. If you mean the first production run and any problems it may have, there's no way to really answer that. If we see the problem, it will be fixed now. If something slips by then there's no way for us to prepare for its fix now (because we don't know about it). You may be asking about expansions? Some will certainly require soldering of some sort because all we can really do is give access, not prepare for every thing someone may want to do with it. Feel free to re-word if this doesn't address your concerns.

[s_mack]

Steven will confirm, but the plan is that boards shipped will require NO mods.

Yeah, that's the idea of course. Any changes we determine are necessary will be reflected in the production run, but obviously mistakes can happen.

The only thing is that two of the accessories... the backlight and the speaker... are being made by a different factory than the main one. Let's say the boards are done in a month and the speaker guy tells me he needs 3 months. In that case I'm not going to hold up the project for 2 months for something that isn't critical to the function of the board. I know people would probably prefer I send the speaker out later for free, but at the discounted price I can't do that, which is why I've tried to be careful to say "I plan to include" the backlight and speaker. In that eventuality, I will probably have to provide those accessories for sale at a later date (likely free + s/h). Understand that this should be a low-probability outcome. I do expect it to be a complete kit, I just can't guarantee it.

All the cables and things that are necessary will be included, and they're made by the same people as the board so it shouldn't be an issue.

[magnetus]

is it there the possibility to have a "kit" without soldered connectors? or the connectors will be soldered at the manufacture process, I am planning to use mine with my 9XR....

[s_mack]

I'm not sure that's possible. Physical fit might be way different, connections are all different. If you're up for that though, then you're qualified to remove connectors No, they won't be available in unfinished form.

If you do have the skills and equipment to hack this thing into a product it wasn't designed for, let me know by PM.

[magnetus]

pm sent

[ShowMaster]

Yes, the questions were about production runs not prototypes. From what you've posted I think I understand. No shipments until all the features are functional. Thanks

[MikeB]

The 'backup' method is now working. It isn't as fast as the direct (via pogo pin) method. That sends the data 100 times per second, the 'backup' method only manages 33 per second, so it is really only a backup method for use if the direct method fails. I may look to see if I can speed this up, but there are technical reasons why it is slower, so any speed up will be by encoding the data in some way. For now, since it is working, I'll move on to other testing.
I just now need to detect the direct method has failed and swap to the 'backup' method.

Mike.

[Daedalus66]

I really appreciate the chance to get in early on this project -- and at a reduced price! I will fully understand if problems do crop up and I'm confident Mike and Steven will provide a fix if needed. I hope others who participate in this first round will appreciate the opportunity and value the experience. It's great fun to be watching this process now and it will be even better to get the board up and running. But no great rush for me.


Sent from my iPhone using Tapatalk