Hmh what is the packetlength define you added?

Hmh reminds I need to make that preamble longer. ..

That is my special test version, derived from your code before you added the bind goodie and other stuff. Nothing "flight worthy". I do use a 6 byte packet to test code. I dont mind sharing it, its just to experimental/bad to be intresting.

protocol
// 2 bits rf header seem enough to me
// protocol idea:
// 3 bytes per 2 channel. 9 bit data. 3 bit channel indicator. 12 bits * 2.
// 6 bytes -> 4 channels per message.
// no header. no tail.

// tx decides which channels need update
// determine old-new difference for each channel.
// send channels with biggest difference first.
// send all channels at slow rate, maybe reserve 1 channel data
// for scanning trough all 8 chan, and only send 3 diff channels

I have got the byte stuffing code, but no good / usable channel selection code yet.

RX --------------
{
int rxed[4];
unsigned char rxedno[4];

rxed[0] = rx_buf[0] + ((rx_buf[2] & 0x08) << 5);
rxed[1] = rx_buf[1] + ((rx_buf[2] & 0x80) << 1);
rxed[2] = rx_buf[3] + ((rx_buf[5] & 0x08) << 5);
rxed[3] = rx_buf[4] + ((rx_buf[5] & 0x80) << 1);

rxedno[0] = (rx_buf[2] & 0x07);
rxedno[1] = ((rx_buf[2] >> 4) & 0x07);
rxedno[2] = (rx_buf[5] & 0x07);
rxedno[3] = ((rx_buf[5] >> 4) & 0x07);

cli();
PPM[rxedno[0]]= (rxed[0] << 1);
PPM[rxedno[1]]= (rxed[1] << 1);
PPM[rxedno[2]]= (rxed[2] << 1);
PPM[rxedno[3]]= (rxed[3] << 1);
sei();
}

TX ---------------------------
{
int tosend[4];
unsigned char tosendno[4];
unsigned char i;

for (i=0;i<PPM_CHANNELS;i++) diff[i] = PPM[i] - OldPPM[i];

// sort diff.
actchan = 0;

for (i=0;i<PPM_CHANNELS;i++)
{
if ((diff[i] >= 3) | (diff[i] <= -3))
{
tosendno[actchan] = i;
actchan++;
OldPPM[i] = PPM[i];
}
if (actchan >= 4) exit;
}

cli(); // disable interrupts when copying servo positions, to avoid race on 2 byte variable
tx_buf[0] = (tosend[0] & 0xFF);
tx_buf[1] = (tosend[1] & 0xFF);
tx_buf[2] = ((tosend[0] >> 5) & 0x08) | ((tosend[1] >> 1) & 0x80);
tx_buf[2] |= (tosendno[0] & 0x07) | ((tosendno[1] & 0x07) << 4);
tx_buf[3] = (tosend[2] & 0xFF);
tx_buf[4] = (tosend[3] & 0xFF);
tx_buf[5] = ((tosend[2] >> 5) & 0x08) | ((tosend[3] >> 1) & 0x80);
tx_buf[5] |= (tosendno[2] & 0x07) | ((tosendno[3] & 0x07) << 4);
sei();

}

Got it.

I though of doing something like that decided not to. Under extreme conditions when maybe every 10th packet gets thru it is best to use single packet.

Currently it makes a guess jump twice but I'm going to extend that to full hopping cycle as the timing is accurate enough for that. Also I'm going to add satellite rx support really soon.

Originally Posted by kha

I'm going to add satellite rx support

I think ppm-sum satellite would be great. Use a frsky and openlrs receiver in parallel. 2.4g and 433m.
Redundant. Top of world. No-one else does this

Originally Posted by Cesco

I think ppm-sum satellite would be great. Use a frsky and openlrs receiver in parallel. 2.4g and 433m.
Redundant. Top of world. No-one else does this

Well quite easy to do actually, just need a reliable way ro get fail info from secondary.

But I was more thinking of native openlrs slave connected via i2c this is quite easy to do and very little additional code.

Edit: actually rssi/validity from secondary is not mandatory. When olrs rx fails it goes to failsafw which is ppm passthru if we have valid ppm at input. I will add that.

Edit2: I need to do some rework to have one timer doing ppm in and out, of course the ppm in failsafe is active only when in ppm out mode. Nice thing is that this is completely separate from satellite mode thus both can be used at the same time.

There is actually few devices that can do 433/2.4 redundancy i.e RVOSD with its dual PPM inputs.

No videos of this system fpv

Sounds like it is very far from plug and play
And will be discontinued soon

Originally Posted by David22

No videos of this system fpv

Sounds like it is very far from plug and play
And will be discontinued soon

I'm not sure what you're saying, or trying to say here, or even why you would believe this?

The modules work straight out of the box. I've been testing mine over the past couple of weeks with great sucess in stock form. Although I fly in a pretty unpopulated area it's nice to know that:

1. I have full control over the unit, pick my own frequencies according to the lcoal radio landscape, and that I can fly safely with other users with a quick drop into the code. That's not a huge issue, unless you're scared of a bit of tweaking.

2. That despite having an established user base (as it's openLRS remember) and declining enthusiasm on some forums, now HK has released a low-entry-cost version that there's ongoing development and new interest in improving an excellent codebase.

You just have to take it for what it is. It's not a high powered LRS, with the price tag to match. It's extraordinary good value for a hugely flexible 100mW solution.

Try googling "openlrs" for some examples of videos

@henkster, what range did you got and when you say stock you mean not uploading any new version or same version of the software to the RX or TX?

Yesterday I finally managed to test mine and like what happened to my friend it lost signal and went into failsafe after no more than 30-40mts. I had made a dipole for the RX and using it made little or no difference.
Assuming that there is no problem with the hardware, they seem to be running on the minimum power setting (1dbm) as this was the range I was getting before while testing some other boards I have with the same hardware.
Making them unsuitable to fly with stock software, well if you keep flying in <40mts radius it should be fine.

mts?

Yep 40 meters, and luckily I did the range test before flying if not I'd have crashed for sure...
I've set the alarm to be on when in failsafe so I could notice when it lost signal and that was it. Tried it even with FPV gear turned off just in case...


Oh and a word of caution about the failsafe. When it reconnects it puts channels at 50% for 1 or 2s something not good if you testing it with a multi copper and you're not aware.

Originally Posted by IceWind

Yep 40 meters

Which firmware did you use?
40m is indicating something is wrong. I get more than this when i have no antenna attached at all in the receiver. Check your connections, antennas, etc.

Stock fw. All was checked. I suspect they send them (or at least some), set to 1dbm.
I'm waiting for the 3.3v ftdi to load something different.
If HK wouldn't have been dumb and used 5v I would have done that already.

agreed that this indicates a problem, I get much more range with video gear running and installed very close to the aircraft r/x (quadcopter). I don't think that they would ship with 1/100th the output power for any reason, you must be doing something wrong or there is an intermittent hardware problem like I had (prevented me from flying until fixed though).

Inside my transmitter, my rfm22b was barely soldered in a couple of places causing it to fail most of the time. Perhaps this is your problem and your movement causes it to loose contact.

I hope hobbyking will fix these issues like they did with the orange rx dsm2 module.