Page 2 of 4

Re: Virtual Trackgen

Posted: Sat Apr 06, 2019 12:06 pm
by PeterW
I have installed Version 1.0 Alpha and running on a RSP 2 works great.

Now i have downloaded Virtual Trackgen and Com0com
I have Virtual COM3 and COM4 on Win10/64 computer and running Virtual Trackgen in VS 2015.
The GUI comes up and I can set COM3 in Trackgen and COM4 in Spectrum analyser .

In the Virtual Trackgen program I got "CONNECT" message and I can send to Analyser Trackgen V1.0, it comes up in the textbox in Spectrum analyser

But the Track gen is not starting up. shows RED and off
Do I have to do any more settings ? I thought I could communicate with Trackgen and se commands and so in Virtual Trackgen
Or have i misunderstood the concept ?

Regards
Peter

Re: Virtual Trackgen

Posted: Sun Apr 07, 2019 5:04 am
by SteveAndrew
PeterW wrote:I have installed Version 1.0 Alpha and running on a RSP 2 works great.

Now i have downloaded Virtual Trackgen and Com0com
I have Virtual COM3 and COM4 on Win10/64 computer and running Virtual Trackgen in VS 2015.
The GUI comes up and I can set COM3 in Trackgen and COM4 in Spectrum analyser .

In the Virtual Trackgen program I got "CONNECT" message and I can send to Analyser Trackgen V1.0, it comes up in the textbox in Spectrum analyser

But the Track gen is not starting up. shows RED and off
Do I have to do any more settings ? I thought I could communicate with Trackgen and se commands and so in Virtual Trackgen
Or have i misunderstood the concept ?

Regards
Peter
Hi Peter

It looks as if you have established a comms link between the analyser and virtual Trackgen. The analyser is expecting to see "CONNECTED" back from either the Arduino, or virtual Trackgen. Type CONNECTED into the box on virtual Trackgen and click "Send to analyser". This will (should) complete the handshake between the two, the red patch should now change to green.

Virtual Trackgen is extremely basic and was written just to demonstrate how to connect your application to the analyser. Make sure you check the Trackgen section in the manual. There have been some changes and additions to the command set.

Let me know how it goes, or if you have any other queries.

Steve

Shows response from the analyser after virtual Trackgen issues the CONNECTED response.
Virtual Trackgen-1.JPG
Virtual Trackgen-1.JPG (23.85 KiB) Viewed 47921 times
The tacking generator is started on the analyser - this shows the data sent from the analyser to virtual Trackgen.
VirtualTrackGen-2.JPG
VirtualTrackGen-2.JPG (27.68 KiB) Viewed 47921 times

Re: Virtual Trackgen

Posted: Sun Apr 07, 2019 8:40 am
by PeterW
Hi Steve

Yep it work´s now , typed CONNECTED in capitals, i read that other commands to DDS could be either low or upper case.

Also I hooked up a AD9850 and programmed a ProMini with the nano code and that worked also but i had to run Spectrum analyser as Administrator
otherwise it would not connect to the com port.

Waiting for a ADF4351 to arrive I will try to adapt to a PIC.

//Peter

Re: Virtual Trackgen

Posted: Sun Apr 07, 2019 11:26 am
by SteveAndrew
Hi Peter

Good news indeed. I'm glad you got it sorted out and running. I'll have to check, but I think CONNECTED can be in lower case, the same as all commands - Try it.

I'll be very interested to see how the ADF4351 works out.

Good luck

Steve

Re: Virtual Trackgen

Posted: Mon Apr 08, 2019 1:21 am
by SteveAndrew
Hi Peter

I've just checked, the analyser limits the maximum frequency that may be sent to Trackgen to 40MHz for the AD9850, and 90MHz for the AD9851. I'll add a 'Custom' type to the DDS list in the options dialogue, this will unlock the frequency limit.

Steve

Re: Virtual Trackgen

Posted: Tue Apr 09, 2019 6:46 pm
by PeterW
Hi Steve

Ok about the frequency limit, i am waiting for the board it was out of stock, probably here about 2-3 seeks so no hurry.

BR
Peter

Re: Virtual Trackgen

Posted: Wed Apr 10, 2019 6:27 pm
by cmr77
I also looked at the ADF4351 and it did look like a great solution until you read the small print... 2.2-4GHz sin wave output which is good but 35M-2.2G is a square wave which is less useful. Perhaps someone can come up with a suitable filter to turn the 35M-2.2G square wave into a suitably clean sine source which has a reasonable amplitude balance over the entire range.

The 1.0a Spec An release looks really good good although for me the waveform stops redrawing all of the time. Windows 7 Ultimate on a i7 4770K (if that's of any use)

Re: Virtual Trackgen

Posted: Thu Apr 11, 2019 7:58 am
by SteveAndrew
cmr77 wrote: The 1.0a Spec An release looks really good good although for me the waveform stops redrawing all of the time. Windows 7 Ultimate on a i7 4770K (if that's of any use)
There is a known bug that causes the display to freeze. It will occur on wide spans where the selected span will require a change of frequency during the sweep. If the frequency crosses a band boundary during a sweep, then the display will freeze. This only affect s spans greater than 1MHz and has been fixed. Below is a list of the band boundaries for the various RSP modules.

RSP1A and RSPduo

0-2 MHz
2-12
12-30
30-60
60-120
120-250
250-300
300-380
380-420
420-1000
1000-2000 MHZ

RSP1 and RSP2

0-12 MHz
12-30
30-60
60-120
120-250
250-300
300-380
380-420
420-1000
1000-2000 MHz

Steve

Re: Virtual Trackgen

Posted: Thu Apr 11, 2019 9:40 pm
by cmr77
Thanks Steve,

You are quite right - within the bands all is well and the display updates without any problems - a useful bit of information to know and we look forward to the next release - it's a great piece of work and I now wonder if I really need an old (and of course large and heavy) Agilent box.

I'm tempted to get one of the ADF4351 board and have a play with some filtering as it's a cheap device.

Carl

Re: Virtual Trackgen

Posted: Fri Apr 12, 2019 5:38 am
by AussieSusan
cmr77 wrote: I'm tempted to get one of the ADF4351 board and have a play with some filtering as it's a cheap device.
Agreed that it is a cheap device but how do you intend to go about the filtering.
The 'simplistic' approach is a LPF that would cut out the higher frequency harmonics of the square wave. But the issue I see is that the LPF would need to be tuneable.
At the low end of the frequency range - lets say 50MHz for an example, - the harmonics are at 150Mhz, 250MHz,350MHz etc. All of these are perfectly valid frequencies that the device can create as the fundamental frequency. Therefore any LPF that would work at the low end of the frequency range would remove all higher fundamentals as well.
Trying to make a tuneable LPF or BPF that covers such a wide range of frequencies would be an engineering challenge I suspect.
The data sheet for the device really only talks about how to use the device for 2.2GHz and up frequency output and you need to use the internal 'divide by x' register for the lower frequencies which is why the signal is 'squared up'.
I think that this chip would be very useful for the range it is intended to be used, but that other DDS chips would be better at the lower frequencies.
Susan