How to check what hopping frequencies a radio is using?

[robismyname]

FHSS channel detection

I have a 2.4GHz radio that uses fhss. There are about 10 networks for this radio. Each network has a unique hopping pattern. The spec doesn't include the frequencies they use per network.

My question is, is there some sort of device or technique available that can tell me what hopping frequencies the radio is using?

I tried max/hold with a SA and the hops are so fast it impossible to record a value.

 

[biff44]

Re: FHSS channel detection

The cheapest way to figure out the frequencies is to use a delay line discriminator and an oscilloscope. As the channels hop, you will see a staircase pattern on the scope.

 

[robismyname]

Re: FHSS channel detection

you will see a staircase pattern on the scope.

Do you happen to have a screensot of the staircase that I canuse as a reference?

 

[biff44]

Re: FHSS channel detection

See attached. This is what ONE hop would look like. If you kept the time base longer, you would see multiple hops, each to a different channel frequency center. It helps if you have a scope with a long record length, and it helps a lot if you can hook up to the unit directly to get a good trigger.

 

[robismyname]

Re: FHSS channel detection

it helps a lot if you can hook up to the unit directly to get a good trigger.

So that would me hook the o-scope up to the radio pin that does TX and not the antenna port?

 

[biff44]

Re: FHSS channel detection

I mean, if you have access to the inside of the radio to figure out when a new frequency hop is about to occur, you could trigger off of that digital signal.

 

[robismyname]

Re: FHSS channel detection

I mean, if you have access to the inside of the radio to figure out when a new frequency hop is about to occur, you could trigger off of that digital signal.

I do have access to the radio. The pinout is attached.

So do I just place one channel of the oscope on pin2 and one channel of the oscope on pin 7?

 

[biff44]

Re: FHSS channel detection

You missed the part about the delay line frequency discriminator. Check this out, figure 2.

https://www.rfcafe.com/references/electrical/ew-radar-handbook/mixers-frequency-discriminators.htm

you can use a mixer in place of the phase detector.

 

[robismyname]

Re: FHSS channel detection

you can use a mixer in place of the phase detector.

I got a mixer: minicurcuit ZFM-15+

I got a power divider: minicurcuit ZAPD-4

But how do you make the inductor delay line?

What value inductor should I use? 580nH?

Added after 3 hours 19 minutes:

you can use a mixer in place of the phase detector.

discriminator has 2 inputs and 2 outputs. Mixer has two inputs and one output. If I use a mixer, the mixer output goes to channel 1 of the o-scope. What goes into ch 2 of the oscope?

Added after 29 minutes:

you will see a staircase pattern on the scope.

Should it look like this?

 

[biff44]

Re: FHSS channel detection

Yep, that looks about right. Unfortunately, you seem to have a lot of frequency deviaiton at each hop channel center, so it will be a little hard to figure out the exact frequency centers, but you have the idea now. IF there was some way to turn off the modulation, you would get a cleaner scope picture.

I probably would not use an inductor, but a length of 50 ohm transmission line. I would choose the length to be X-45 degrees at the low end of the frequency band, and X+45 degrees at the high end of the frequency band.

If you have access to a CW synthesizer, you can calibrate the frequency discriminator output voltage vs. input frequency. Just make sure your cal syntheszer has the same RF power level as the radio you are trying to measure (the cal is RF power sensitive too).

Good luck.

 

[robismyname]

Re: FHSS channel detection

so it will be a little hard to figure out the exact frequency centers, but you have the idea now.

Please explain,how would you calculate the frequency from one of the shown step-pulses provided in the image? Use f=1/T? but it looks like all steps are the same distance apart. Does each step-pulse provide a individual frequency?Please provide more info so that I can understand.

IF there was some way to turn off the modulation, you would get a cleaner scope picture.

This radio uses a LMX2330A dual synthesizer. I have access to the radio boarad to turn it off. Will this help?

If you have access to a CW synthesizer, you can calibrate the frequency discriminator output voltage vs. input frequency.

I have access to a HP 8648C signal generator. But how do i integrate the signal generator with the frequency discriminator assembly? Where would the signal generator hook up in order to calibrate the frequency discriminator?

 

[Mazz]

FHSS channel detection

Since you have direct access to RF synth, there could be a simpler way to check hopping sequence.

Monitor the SPI of synth and decode its commands to check programmed freq vs. time.

I hope it can help.

Mazz

 

[biff44]

Re: FHSS channel detection

See attached. The output of the mixer a voltage that corresponds to a frequency. You need to figure out the calibration for the discrimator. First slowly scan your frequency source from the low to high end of the band of interest. You will get a changing output voltage at the mixer IF. If it looks roughly like what is drawn below, then you just set the frequency to a known cw, measure the voltage, move the frequency, measure voltage again, until you have a calibration table written down.

If you do not see the waveform shown, you have to play with the discrimator values. If you see a multi humped, semi sine wave, then you have too long of a delay line--shorten it up. If you see almost no change in the mixer output voltage, you do not have enough length in the delay line. If it looks like the delay line length is right, but you do not get a monotonic "line" but a single hump, add or subtract small lengths of coax line to recenter the discrimator output.

When you are done calibrating the discriminator, hook up the WLAN card and what you see on the oscilloscope (dc coupled) can be converted to absolute frequency using your lookup table. Just make sure the power level used during the calibration is the same power level used by the WLAN card when it is actually transmitting (not the average power level when it is on/off).

And yes, if you can decipher the digital commands to the PLL chip, and know the clock frequency, you could figure it all out mathematically without RF testing.

 

[robismyname]

Re: FHSS channel detection

If it looks like the delay line length is right, but you do not get a monotonic "line" but a single hump, add or subtract small lengths of coax line to recenter the discrimator output.

The results of my 1st attempt of the discriminator calibration table suggest that I peaked at 2446MHz (see atttached).

I then added about a 1/4" of coax (6' + .25") and I miraculously peaked at 2400MHz (good thing) but only staying "monotonic" up until 2456MHz. At 2456MHz the curve began to go more positive (starting to creat another peak). (see attachment).

Im changing frequency vs. voltage too too rapidly. At this rate I could never finish going through all the frequencies (2400-2480MHz) with having the curve change directions.

I started out with a 50 ohm transmission line of length 6' because I wasn't very clear on the X-45 degrees and X+45 degrees cable length concept.

Any/All suggestions are greatly appreciated.

 

[robismyname]

Re: FHSS channel detection

Monitor the SPI of synth and decode its commands to check programmed freq vs. time.
Mazz

I appreciate the feedback.

What pin(s) of the LMX2330A should I monitor on the o-scope? Data?,LE?Clk?

Can you provide an example of how to decode a command to check programmed freq. vs. time?

I.e. what do I expect to obtain after I monitor the SPI and how do I convert the SPI information to a freq.

 

[Mazz]

Rob

it will be too long to go through the detail of the meas, but all the informations you need are in the datasheet (pages 10 to 13).

In the SPI, you should monitor the data pin, using LE as trigger (active low).

The SPI should have most of the registers with always the same values (ref dividers, charge pump) and some varying hop by hop.

I hope it can help.

Mazz