XK K110 (S-FHSS) build for Devo 7E and Devo 10

OK, I'll check again - SeBy DocKey should have the model and can confirm it's S-FHSS compatible or not.

Yes ... I have a more or less working profil for the 7E for the X100

nice one guys, much appreciated

HappyHarry wrote: nice one guys, much appreciated

right ruddler trim tor 3G/6G .... and HOLD1 for 3D

thanks again seb but i'll be switching those out to use fmod for 3G/6G and one of the 3ways for triggering 3dmode, hold is always throttle hold for me, it's a habit from the heli's and whenever i look to be going in muscle memory triggers the hold switch so i can't use it for anything else lol

@victzh I'll get some pictures when I am back at home, RX, TX or both? There is a PA on the TX side but not on the RX side.

PA on TX only is normal. Both is preferable - to estimate the place of soldering. Not that I'm going to buy this particular equipment but I can advise you what's easier.

And may be make its own thread in Protocols section of the forum.

The Radiolink RX is made up with a CC2500 daughter board plugged onto a main board. The pins are marked on both sides.

On the Radiolink TX side:

I just love when everything is so well marked and nice and big - so easy to connect, you don't even need to solder to it.

Hi Victzh,
I got 2 binding problems report from Taiwan & China RC forum
#1 : DEVO 12S user , XK K120 & K110 , can't binding.
#2 : DEVO 10 user , XK K120 , he have 2 K120 , one of them can't bind and the other OK.
One of China user asked XK company , XK told him it's "frequency shifting" (I'm not sure I use correct wording or not?)
it seems like frequency difference problem, do you have any idea about it?
Is it possible that XK X6 TX (or RF module) has any special design or wilder frequency tolerance ?
All of them can bind with XK X6 TX, no problem.
Thanks!

CC2500 used for S-FHSS has quite strong requirements to the quartz quality than other chips. Manufacturers of cheap models don't bother to use crystals with high precision - as long as they are from the same batch they will match and so will be able to communicate. I thinks that what's they meant under "frequency shifting". XK module does not have wider tolerance, it just matches the model's crystal.

We already faced with this in FrSky protocol and implemented software correction to the frequency, fortunately CC2500 allows it.

I don't know how well this works, but it's a matter of enabling a couple of parameters. Also, I wonder where their CC2500 modules are from and how do they look?

victzh wrote: CC2500 used for S-FHSS has quite strong requirements to the quartz quality than other chips. Manufacturers of cheap models don't bother to use crystals with high precision - as long as they are from the same batch they will match and so will be able to communicate. I thinks that what's they meant under "frequency shifting". XK module does not have wider tolerance, it just matches the model's crystal.

We already faced with this in FrSky protocol and implemented software correction to the frequency, fortunately CC2500 allows it.

I don't know how well this works, but it's a matter of enabling a couple of parameters. Also, I wonder where their CC2500 modules are from and how do they look?

Thanks for your explanation.
The DEVO 12S user use 3in1 RF module and I will ask DEVO10 user for his RF chip soon.
BTW, is it possible to use X6 TX stock RF chip to solve this problem ?

dc59 wrote: BTW, is it possible to use X6 TX stock RF chip to solve this problem ?

@dc59, since you are also active in 5imx, there is a post in the Chinese language reporting that using stock RF module works: bbs.5imx.com/forum.php?mod=viewthread&tid=1200794

I think so, but I'm not sure how they handle PA/LNA (if they have one).

That is bad news. Basically it means if the CC2500 modules from ebay do not work, we have to buy a stock XK TX first. The stock TX has been the show stopper for me in buying any XK birds. Don't plan on getting them until they are devo manageable.

BTW, my analyzer is tracked at Toronto this morning, 400km from Ottawa. I probably would get it this week. However, I have no CC2500 modules yet unless I borrow one from my radioLink or Skyartec stock TX. I haven't received my new devo 10 or H11D. Cannot do much with the new analyzer unless I try it on the L6F which has an unknown RF module.

How much can a crystal be off???

Can the "Tuning" options used in the FrSky protocol be used for this frequency shifting?

@wasp09 - no it's not THAT bad news.

If you look into CC2500 datasheet you'll see the recommended crystal accuracy is +-40ppm. It allows to keep the central frequency (around 2.4GHz) at +-96KHz. If you use cheap crystal, say 100ppm it becomes 240KHz. Channel bandwidth at datarates less than 100kbps is around 240KHz - it becomes harder for the chip to compensate for such large deviation. So the less data rate the more receiver sensitivity, but on the other hand the stricter requirements for the crystal. It can explain why some protocols are less sensitive to this effect.

You can add tuning and try to offset the central frequency for your TX one way or another until it binds. As @mwm mentioned, it was used with FrSky, I just don't know how effective. The code is there, it's not enabled yet.

rbecq wrote: @dc59, since you are also active in 5imx, there is a post in the Chinese language reporting that using stock RF module works: bbs.5imx.com/forum.php?mod=viewthread&tid=1200794

Thanks rbecq , I think I missed that post.

victzh wrote: @wasp09 - no it's not THAT bad news.

If you look into CC2500 datasheet you'll see the recommended crystal accuracy is +-40ppm. It allows to keep the central frequency (around 2.4GHz) at +-96KHz. If you use cheap crystal, say 100ppm it becomes 240KHz. Channel bandwidth at datarates less than 100kbps is around 240KHz - it becomes harder for the chip to compensate for such large deviation. So the less data rate the more receiver sensitivity, but on the other hand the stricter requirements for the crystal. It can explain why some protocols are less sensitive to this effect.

You can add tuning and try to offset the central frequency for your TX one way or another until it binds. As @mwm mentioned, it was used with FrSky, I just don't know how effective. The code is there, it's not enabled yet.

Hi Victzh,

Is it possible to enable tuning option on S-FHSS protocol and make a DEVO 12S test build, my friend can try to confirm it ?