Is there a way to calibrate RF module?

Hi!

When I try to see which freqs are active, I see that freqs show not exactly. For example, 433000 show as 432999 and etc. The different is 1 KHz. I tried it with two different transceivers and it show 1KHz difference too. Is there a way to correct it?

First I’d be curious of your method of testing.

The receiver you’re using, if not the Flipper, it’s been recently calibrated itself?

If the Flipper is the receiver, has your transmitter recently been calibrated itself?

1khz is still in the realm of acceptability for any sort of issue really.

I tried three tranceivers - Yaesu VX6R, Yaesu FT-857 (TCXO) and DM-1801 (yes, I know about Bao :wink: ). All attempt show the same difference. I noticed the same issue in other messages on this forum. Yes, I understand that 1KHz difference is not a problem when we look at 400MHZ, but I think that an ability to correct this difference may be useful.
Screenshot-20221125-130524

Test with another firmware. Try with official and un. see if it helps.

Sounds like you may be a Ham, or know one. A common Amateur radio concept is the golden screwdriver. That’s when someone has a good piece of gear but tinkers too much and messes the whole thing up. The display only shows 3 digits of precision. I suspect that value is a float with 6 digits of precision. That is probably 299.999999MHz. I think the feature you really want is rounding.

You are right, I am a Ham. And I’m IT engineer too. I don’t know realization of frequency computing in this case, but if it is rounding issue, it’s a real bug.

1 Like

My theory is based on programming experience. The smallest float value in C is six points of precision and I see lot’s of numbers ending in .999999 after a calculation. I’m waiting on my Flipper so I can’t test that theory yet. I suspect that if you connect to your Flipper over SSH you would see the full precision float value.

Result of 434 MHz testing:
2022-11-27_09-45
I think it is a averaging and rounding bug.

and I don’t think there is any error. the chip is rebuilt not in the full frequency range, but in a multiple of how many hertz (you need to look in the datasheet)