ERSKY9x Rev. B1 User Area

[cre8tiveleo]

This is the easiet 9x install ever!!!

Ya think?!

This is the way the 9x variaions of tx's should have come. The power in your hands is unlimited!

[Rob Thomson]

I know there are certain features missing.. but over-all it is a fully functional system!

The only thing not there at the moment is the ability to program from the desktop

[bertrand35]

Thanks, I will be looking at it more, I have had a quick look. I like the idea of operating asynchronously, particularly since we can control when power goes off so we can finish writing safely. As to what is needed, I don't know what people will want to use the SD card for. We will need it to store voice files for spoken alerts etc. They will be too big for anything else. I'm just looking to get the er9x audio working, close to testing something. Then I can cut a slot to put a SD card through and write the low level code to drive it.

Mike.

Also I am ok to help for the EEPROM editor on the computer. The conversion is almost ready in companion9x, but I need a way to transfer it to the Tx, do you have an idea of a way to do it just as easily as we can do in eePe and companion9x ...

Bertrand.

[Kilrah]

Should all be done directly through USB somehow...
Maybe turn the 9X into a standard HID?

[MikeB]

Bertrand: See viewtopic.php?f=7&t=4&start=390#p7732
We should be able to use SAM-BA dll, and the applet I wrote to use the USB to read and write the EEPROM.

Mike.

[bertrand35]

Excellent! I missed this post. Let-me have a try!
Bertrand.

[hpihl]

Ok, got the board yesterday and had a quick test tonight with Optrex LCD. Some cutting needed to get the LCD correctly aligned but effort was well paid as it's far better than the stock display. I'll do some flight testing with EasyGlider during easter weekend and will report later.

[Reacher10]

I hooked up a vibrating motor to the haptic header. Is this correct? The header has constant voltage coming out, the motor just runs. I didn't check what the voltage is.

[Clivew]

Yes it does, I don't think haptic is implimented yet!

[MikeB]

I'm working on the audio/haptic now. The constant output is a test to check the hardware is working correctly. Audio may be available shortly, haptic will be a day or so.

Mike.

[Romolo]

Do you think it will be possible to use this bluetooth module ?
http://www.miniinthebox.com/serial-blue ... 03544.html
It seems pretty cheap considering free shipping

[bertrand35]

Mike I just succeeded to backup the EEPROM from the ersky9x board and load it into my simu. Everything fine, I can see my models. Now I need to find the way to load the DLL or the exe from inside companion9x.
Bertrand.

[MikeB]

Bertrand: Sounds good.

Romolo: Could be, the chips look the same as those on the RN42. The RN42 defaults to 115200 baud, this one to 9600.

Mike.

[MikeB]

Revision 44 is available. I've added the audio, one or two minor bugs still to sort out. Not had time to do the haptic bit yet.
There is also a BATTERY status screen, long-up goes to this, then up to STAT and STAT2. The mAh monitoring now operates in tenths of mAh, any previosuly stored value will now be wrong, its ten times too low!. MENU clears the value back to zero on this screen.
CPU temperature is also displayed, internal sensor.

ROB: you created the sounds, do they sound correct? I've never done the speaker mod on er9x so I don't know what they should sound like.

Mike.

[Rob Thomson]

Will test tomorrow and let you know


Sent from my iPhone using Tapatalk

[PNaz]

Tried the new audio in R44. I have an 8 ohm speaker in the ERSKY9x and a piezo speaker mod in the ER9X. Sounds on the ERSKY9X seem to be significantly higher pitch.

Paul

[th9xer]

I have searched but couldn't see it mentioned. What do I connect to the backup battery port ?

[pmackenzie]

^ +1

What and why.

[Reacher10]

I don't think the battery back up has been enabled yet.

Mike if I didn't know any better I'd say your having fun with all the new found room you have to program.

The sound version is nice. I'll hear this sucker from 100 yards away when the inactivity alarm goes off.

Very nice indeed!

Thanks

[pmackenzie]

What is the battery intended to back up?
If I remove the main pack, will it lose all the model setups?

[Reacher10]

There will be a real time clock as well as keeping track of total time radio is turned on and what ever you can think of. There was discussion of this early on.

You won't lose your models if it's not hooked up.

[pmackenzie]

So more of a RTC battery than a memory backup one.

[Rob Thomson]

Mike.. just tested the audio

My initial thoughts are that it is running a little fast?

The essence of the audio is in place - but it does not sound quite right.

Am I also correct - the audio you added is the frsky internal audio mods? For the alarms? Currently I believe the audio still needs updating on the trim sticks and other events?

As a point of obeservation - before we go too far on the audio stuff.

I have personally think the overall audio implimentation on Open9X is better that on er9x. The er9x version was a bit of a comprimise to allow it to fit into the small epprom space we had free. Is there any value in reworking this for ersky9x at this early stage?

Rob

[MikeB]

Pat: We have this second processor, ATTINY167, that provides extra I/O and has a 32.768 kHz for a RTC. It is this device that uses the battery backup. It will use the I2C bus to communicate with the main processor. At present it has no code written, not enough time to do it.
Model data and general settings are all stored on the external EEPROM (flash device) and do not need battery backup.

Mike.

[MikeB]

Rob: Sure it can be reworked. I've mapped the existing er9x code onto the ersky9x hardware. The bottom level driver sends an actual frequency and a time, computed as a whole number of complete cycles to the hardware. Changing the frequency part way through, using the freqIncr value, probably upsets this timing as at a higher frequency each cycle is shorter. I've also had to map the requested frequency to an actual frequency (multiplied by 30.5).
It may well be it needs a bit of a rethink as the new hardware provides different (and better) functionality, and can do more without processor intervention. e.g. A whole cycle of a sine wave is sent with only a single interrupt at the end.

Mike.

[Rob Thomson]

Sounds good!

I will try find some time today/tomorrow to start looking at the actual code - and tweaking the existing events to sort the sound out. I assume that you have the same function calls for the sound events?

If this is the case - I will just need to tweek them to suit the new platform.

Rob

[MikeB]

We do have the same function calls to the audio routines, and the same queueing, the interface to the lower level is different. I'm wondering if it might be worth changing the frequency value. I have a table of 'piano frequencies'. We could specify the frequency as an index to this table.
The queue audio.cpp could be longer, we are not short of memory now!
No problem with changing the model structure or the general structure. It is easiest to just make them longer and add new items at the end. These will autpmatically be filled with zero when a model is loaded, then saved to the EEPROM when saved. Users should not notice any problems. We have up to 4K for each model and 4K for the general.
I'll see if I can make the bottom level handle changing frequencies while keeping the time duration constant.

Mike.

[Rob Thomson]

I'll see if I can make the bottom level handle changing frequencies while keeping the time duration constant.

That would be nice. Will give a pitch control

[Mechcondrid]

ok just found a bug in the frsky telemetry internal alarm functions
if the alarm is set and goes off say for the rssi alarm it will go off if it goes below the specified signal strength but once the signal goes back up the alarm wont turn off until a reboot

[MikeB]

Could you be more specific please?
The RSSI alarm, as on er9x, is not handled internally (yet). Er9x didn't really have enough storage to keep the values in memory. Ersky9x has, but the code need changing.
I don't think the internal alarms actually sounded either, they were commented out.

Mike.