AliExpress $30 35MHz-4.4GHz ADF4351 Touchscreen Signal Generator

[bb1]

Here is short report on the output signal waveforms and spectra of AliExpress touchscreen ADF4351 board.

[bb1]

The lowest ADF5451 frequency permitted by AliExpress touchscreen is 35 MHz.
Steady state spectrum near it is shown on attached picture.
Another attached figure shows how output frequency is changed in first 80 uS after the output is enabled.
 

[bb1]

The first picture attached here.

[maximvs]

I wonder if it is possible to take out the 25Mhz crystal and always provide an external RF source.
I have a 10Mhz. clock (GPS Sync), which has a very pure output, and I'd like to use that.
My question is:
    I read the specs on the ADF4351 and it is not clear if the reference signal
    should be square wave or could be a sine wave. Any thoughts ?

I might buy another unit and do a test but I'd like to hear what you guys think.

[radiolistener]

There is more interesting ADF4351 development board on aliexpress:

https://www.aliexpress.com/item/33032121416.html

Two ADF4351 + IAM81008 + AD8307 + STM32F103C8T6 on the board.

First ADF4351 is used as LO and second ADF4351 is used as tracking generator. Two ADF4351 allows to avoid spurs for bode plot.
Can be used as signal generator, spectrum analyzer and tracking generator.

Price is 24 USD. I just ordered one, it looks very interesting as a signal source and bode plot for 35M-4400M.

[BugKiller]

Entirely coincidentally, just before I saw this post I ordered the "MAX2870 23.5-6000MHz RF Signal Source Generator Module" for US$52 (I have background with Maxium based products). Should be interesting to compare.

I am like a kid in a candy store on AliExpress, always running a multi-page Wish List.

[maximvs]

Unbelievable! I just ordered one. I'll do some testing when it gets here and post my results.

[mawyatt]

Just received the ADF4351 Touchscreen board we ordered awhile back. A quick test and it seems to work, had a problem with Setting menu but its resolved.

Here's some quick results at various frequencies. Not too bad, but the frequencies are low (~-9ppm @ 200MHz) so evidently the reference XO is low. Now that the "setting" menu is active you can change the Ref Clock but only 3 decimal places of precision, not enough resolution to correct the -9ppm frequency error accurately   

Best,

[bob91343]

What about amplitude calibration?  Modulation?

[bb1]

Karl-Heinz Schweikert found commands to control the board via USB serial port.
I copy his post here:

Here the secret of USB serial control:
Baud:115200, Data:8, Stop:1, Parity:none, Type:Hex, Send on Enter:none
Point: send 11 bytes B0..B10
with B0:ad; B1:01=ADF4350,02=ADF4351; B2:01=point; B3:01=0dB,02=-3dB,03=-6dB,04=-9dB
B4-6:fRef (03 d0 90 = 25.0000 MHz); B7-9:fPoint (03 0d 40 = 200.000 MHz); B10:sum(B0..9) & 255
Stop: send 3 bytes B0..B2
with B0:ad; B1:ff; B2=ac
Sweep: send 18 bytes B0..B17, with B2:02=sweep; (others same as Point) 
B7-9:Start; B10-12:Stop; B13-15:Step; B16:01=kHz,02=MHz; B17:sum(B0..16) & 255
Return on success: ad 66 13
Example:
ad 02 01 01 03 d0 90 00 88 b8 54 = ADF4350,Point,0dB,fRef=25.0000,f=35.000
ad ff ac = Stop
ad 02 02 01 03 d0 90 0a ae 60 0c 35 00 00 13 88 02 0b = ADF4351,Sweep,0dB,fRef=25.0000,f=700.000..800.000MHz, Step 5.000MHz
Nice to use a small Python script to set a defined frequency, and e.g. read  a filter response with logarithmic RF detector+Arduino, before switching to next frequency. Mixed operation with remote und manual works fine.

[mawyatt]

What about amplitude calibration?  Modulation?

Amplitude at 300MHz is -9.07, -6.15, -3.20 and -0.04dBm @ 4 amplitude settings which should be 3dB apart. A sweep to 3.2GHz produces reasonable flatness, can't go past 3.2GHz as I have no means to evaluate performance. There is no modulation capability, but has a SMA input for an external reference. The internal 25MHz XO reference signal is supplied to this SMA, and checking the PCB it seems you can reroute the external SMA by changing a couple SMD components.

This is basically a PCB with a ADF4351 chip without any additional chips in the RF chain, so expect typical ADF4351 data sheet performance. Not a bad little device for ~$30 IMO.

Best,

 

[bob91343]

Without some means of modulation this thing has limited application.  One could build an analog amplifier with AM capability but FM is another story.

So what we have is pretty much just an oscillator that could be used as the starting point for a signal generator.

[ledtester]

I wonder if it is possible to take out the 25Mhz crystal and always provide an external RF source.
I have a 10Mhz. clock (GPS Sync), which has a very pure output, and I'd like to use that.
My question is:
    I read the specs on the ADF4351 and it is not clear if the reference signal
    should be square wave or could be a sine wave. Any thoughts ?

I might buy another unit and do a test but I'd like to hear what you guys think.

https://youtu.be/UFp4GFxFYco

and the previous video:

https://youtu.be/SNuPzkJFMwY

[radiolistener]

There is more interesting ADF4351 development board on aliexpress:

https://www.aliexpress.com/item/33032121416.html

Two ADF4351 + IAM81008 + AD8307 + STM32F103C8T6 on the board.
Price is 24 USD. I just ordered one, it looks very interesting as a signal source and bode plot for 35M-4400M.

I got it. Here is my first impression and some measurements for this unit:
https://www.eevblog.com/forum/rf-microwave/sdr-rtl-nwt4000-hackrf-what/msg3376978/#msg3376978

At a glance 45-50 dB dynamic range up to 1.5 GHz and about 30-40 dB up to 4.4 GHz. But response is not flat above 1.5 GHz.

I got it for 20.49 USD by using discount and coupons. I expected a little better performance, but still not so bad for the price.

The output impedance according to my measurements is about 40 Ohm. It works from 33 MHz to 4400 MHz and has pretty flat amplitude level. tested it from 35 to 1.6 GHz with RTLSDR. At 1.6 GHz frequency it has some little drift for about 500 Hz/minute, but I'm not sure if the source of drift is LTDZ board or RTLSDRv3 itself.


Input impedance is about 29 Ohm and SWR is about 1.5 - 1.7. but very flat from 40 to 1500 MHz. Don't have equipment to check higher frequency. It looks that there is needs to remove 50 Ohm resistor on the input before mixer.  But it looks that output level will be too high for the input and it will require 10 dB attenuator for bode plot.

[radiolistener]

The lowest ADF5451 frequency permitted by AliExpress touchscreen is 35 MHz.

I tested LTDZ and it looks that the lowest frequency that ADF4351 can produce is 33 MHz.

Here is 33, 34 and 35 MHz from ADF4351 on the oscilloscope:

[maximvs]

I made a video on my mods to the ADF4351> This is my first video so I am sorry for the "not great" quality. I will make an update though. Here it is:

[mawyatt]

Nice work!!

Do you have the ability to do some narrow band measurements like I've posted a few threads above, especially since you are using an external reference which I assume has excellent phase noise?

I was impressed with what the ADF4351 can accomplish. Kind of a "poor persons RF signal source", doesn't have any modulation capability, nor large output range scaling, but not bad for ~$30 IMO.

Best,

[EV]

I bought also this touch screen generator. I attached some picture of its square waves. They are taken from my Tektronix 7904  scope with 7T11 and 7S11 (with S-4) plugins installed.

[EV]

I attached some more pictures of wave forms. Amplitude is 250 mV on 100 MHz square wave but it is 450 mV on 4 GHz sine wave. Amplitude setting is the same in the generator all the time.

It looks that the peak to peak amplitude is nearly same on all different frequencies.

[EV]

As the rise time of the generator is about 70 ps (look at attached picture) the BW of the generator is 350/ 0.070 = 5 000 MHz. It means that this generator can show square wave correctly  to 5000/5 = 1000 MHz and more accurately only to 5000/10 = 500 MHz. You can notice this on the pictures of previous post.

[ledtester]

Tony Albus has just done a review and comparison of a MAX2870 touchscreen module:

https://youtu.be/DFtC7icVr-E

Apparently uses exactly the same PCB as the ADF4351 offering.

[EV]

The rise time of this generator is so fast (70 ps) that it can be used to measure scopes rise time.

I attached a picture where can be seen that rise time of my Tektronix R7103 scope with 7A29 and 7B10 plug ins is about 290 ps.
So the BW of this scope is about 350/290 = 1.2 GHz.

[EV]

I attached a picture where can be seen that the rise time of my Tektronix 2465B scope is about 0.84 ns.
So the BW of this scope is about 350/0.84 = 417 MHz.

[andymok]

I attached a picture where can be seen that the rise time of my Tektronix 2465B scope is about 0.84 ns.
So the BW of this scope is about 350/0.84 = 417 MHz.

I am also looking one for calibrating my 2445B
Does anyone know whether ADF4351 or MAX2870 module has better Fq accuracy?

[bob91343]

I see AliExpress no longer has this item avalable.