advertisement


'itran' as a test bench RF sig gen

Jim Audiomisc

Jim Audiomisc

pfm Member
Someone elsewhere has told me they used an 'itran' and an SDR to do some stereo FM tests years ago. I don't know anything about 'itrans' so anyone here done any tests on them or using them who can say more?
 
FWIW from simple searches these devices look like being cheap-and-cheerful audio-> VHF modulators for use with car radios. i.e. not likely to have a level of performance above 'crappy'. They also seem to be illegal in the UK (?) But someone is claiming to have used one for making low distortion measurements on FM RXs. But now seems reluctant to give details, so I'm not sure if they may be speaking spheroids...
 
I've tried a Belkin Tunecast to generate local signals for testing. It's adequate but the inclusion of an ALC means that trying to do anything useful with crosstalk separation in the MPX decoder is difficult to say the least.

I've been trying the SDR approach on and off for ages now using a LimeSDR Mini. Gnuradio is very capable but there are a lot of gotchas e.g. their wideband FM modulator contains a low pass filter that isn't immediately obvious and prevents anything above 15kHz getting through. This, of course, prevents you feeding it with a an MPX signal and expecting stereo out.

As luck would have it, I was exploring this again yesterday and came across this paper: https://www.google.com/url?sa=t&rct...iplexing.pdf&usg=AOvVaw0-TNv1OvoF7kQS5Kdc-ds1

This is very easy to implement in GRC but the lack of reconstruction filters results in a lot of IM distortion when you look at everything in the digital domain. I'm going to have a play at getting this into a tuner later.
 
Thanks. :) I've had a quick search but can't find any real technical data on the Tunecast. I'm trying to do two things here.

One is simply to see what level of 'quality' they can produce so as to find one that will act as the best possible modulatable FM generator. The other is then to get one to experiment with as a possible FM signal generator.

The challenge is that I'd want one whose performance allowed it to operate as a low distortion and noise stereo FM source for tests. By 'ALC' do you mean an automatic level control? if so, it will mess up many tests.

FWIW I don't use the GNU radio software but wrote my own for the FUNCube Pro Plus so I know what the software is doing. So I'd write my own for the generator control to run at the same time.

I've known the duality between time-division-mpx and frequency-division-mpx for some time. Used to teach it as part of some of my old courses. In fact IIRC Mullard pointed out during the period when the UK was decising what stereo radio system to adopt. :)

TBH I have my doubts that any consumer 'itrans' would be able to test something like a CT-7000 as it would add too much other problems like not having a correct response, etc. But someone has claimed they got results way below the 0.1 - 0.3% level for 30% mod at a few hundred which made me interested. Particularly because limited bandwidth impies this can't happen unless you 'cheat' somehow. (The person who made the claims has been shy since about giving me actual details of what items they used, which may tell us something...)

If all else fails, I'll revent to my 'Plan A' which is - should I live long enough - make a simple varactor controlled circa 100MHz oscillator and carefully measure its modulation curve. The drive it with a predistorted MPX pattern generated by the controlling program. That seems to me a better tack. But of course I might change my mind when I tried it. 8-]

Don't know anything about the LimeSDR.
 
Arkless Electronics said:
https://www.americanradiohistory.com/UK/Elektor/90s/Elektor-1993-05.pdf

A S/H Radiometer SMG1 or 40 would be the ideal...
Click to expand...

To be honest I've lost count of the number of hours wasted getting SDRs to work, flow graphs working, forgetting about it for six months, and re-learning it all that I'm seriously just going to buy a secondhand FM stereo RF function generator. I was spoilt at Naim, we had the full Soundtech 1020 setup. Now that was a lovely piece of kit with the matching 1701 distortion measurement system. I retired the ST1701 and replaced it with an Audio Precision Portable One and drove the GPIB from a PC. I got all the tuner test menu driven and automated and cut test time to a 3rd of what they were. The only other guy who knew as much about the tuners, and who taught me, is now my colleague at Witch Hat.
 
Alas, I'm nowdays limited by two factors.

1) I can't really build very complex circuits anymore. Partly due to hands/eyes deterioration, partly down to lack of any workbench or kit beyond a weller and some cutters/pliers . i.e. no scope, etc.

2) Space and a sane chief executive who'd frown at my having things like Radiometers about in addition to all the other clutter I have. She'd tell me to throw out something else I want.

So in terms of making circuits I'd limit myself to a simple VCO in a tiny box using just a varactor control. Then drive that using a DAC. You may recall I wrote a program some time ago to generate a MPX waveform from a stereo wave file. So I could then drive the simple VCO with that. And add predistortion to correct if the VCO wasn't linear.

i.e. these days I find doing all the complex bits by writing a program to control simpler kit. I'd me more likely to make something like this
https://www.simplecircuitdiagram.com/2740-frequency-modulated-oscillator-for-fm-transmitter/
and then correct its behaviour by initial calibration tests.


Is there a VCO IC that makes it even simpler?

Measurement of the results isn't a problem as I have an SDR and wrote my own 'scope/analyser' etc to use it and get the data. Can also do more dedicated system for FM tests given a suitable VCO.
 
MJS said:
To be honest I've lost count of the number of hours wasted getting SDRs to work, flow graphs working, forgetting about it for six months, and re-learning it all that I'm seriously just going to buy a secondhand FM stereo RF function generator. I was spoilt at Naim, we had the full Soundtech 1020 setup. Now that was a lovely piece of kit with the matching 1701 distortion measurement system. I retired the ST1701 and replaced it with an Audio Precision Portable One and drove the GPIB from a PC. I got all the tuner test menu driven and automated and cut test time to a 3rd of what they were. The only other guy who knew as much about the tuners, and who taught me, is now my colleague at Witch Hat.
Click to expand...

Had a full Audio Precision System One with PC control at my last employer... Very nice.. but I still like an all analogue traditional setup. I simply have no wish for print outs of results etc. If it's important I'll write it down.
I use home made very low distortion oscillator (i don't know how low as it's too low to measure with anything I have here) and a home made distortion analyser designed by Ted Rule (also have Marconi Instruments and Dymar distortion analysers but they are nowhere near as good as the DIY one). The noise floor of this is 0.0014% and the distortion of my oscillator is way below this as the needle doesn't shift from the noise floor when it is directly connected to the analyser! I could get the noise down considerably from this but it's pretty pointless. An amplifier with 0.001% THD will register quite clearly and let you know that ie it is somewhere around the 0.0008 - 0.001% ish level and I have no need or wish to measure any lower than this. There are analogue techniques I could use to add another zero when testing at specific spot frequencies but it all gets pretty academic at that low a level:)
 
Jim Audiomisc said:
Alas, I'm nowdays limited by two factors.

1) I can't really build very complex circuits anymore. Partly due to hands/eyes deterioration, partly down to lack of any workbench or kit beyond a weller and some cutters/pliers . i.e. no scope, etc.

2) Space and a sane chief executive who'd frown at my having things like Radiometers about in addition to all the other clutter I have. She'd tell me to throw out something else I want.

So in terms of making circuits I'd limit myself to a simple VCO in a tiny box using just a varactor control. Then drive that using a DAC. You may recall I wrote a program some time ago to generate a MPX waveform from a stereo wave file. So I could then drive the simple VCO with that. And add predistortion to correct if the VCO wasn't linear.

i.e. these days I find doing all the complex bits by writing a program to control simpler kit. I'd me more likely to make something like this
https://www.simplecircuitdiagram.com/2740-frequency-modulated-oscillator-for-fm-transmitter/
and then correct its behaviour by initial calibration tests.


Is there a VCO IC that makes it even simpler?

Measurement of the results isn't a problem as I have an SDR and wrote my own 'scope/analyser' etc to use it and get the data. Can also do more dedicated system for FM tests given a suitable VCO.
Click to expand...

Bound to be various VCO IC's around but prob need multiplying up to wanted F. A PLL would also do it as would a reactance modulator to phase modulate and then use suitable pre-emphasis before a multiplier chain.

There are some very cheap DDS units available from China which may be FM able directly and controllable from PC...
 
Arkless Electronics said:
Had a full Audio Precision System One with PC control at my last employer... Very nice.. but I still like an all analogue traditional setup. I simply have no wish for print outs of results etc. If it's important I'll write it down.
I use home made very low distortion oscillator ...
Click to expand...

I can understand what you wrote as I still miss some of the kit I had when at Armstrong!

However for the last few deades I've tended to find writing software and generating/analysing LPCM via DACs and ADCs works very well indeed. I'm also well happy with the SDR I use. (OK I calibrated this using an HP precision sig gen I borrowed from my old research group. 8-]) Results here

http://jcgl.orpheusweb.co.uk/funcube/FUNcalibrate.html

Although in practice of late I've mostly used it to check whats on the UK DVB bands at UHF. Not audio. However it means I can measure VHF signals OK. But lack a decent FM stereo source. I can sort out the stereo bit if the FM sig gen has good wideband mod, though.

Alas, chasing the bloke who made the claim about the 'itrans' just gets him dodging the questions. So I look like getting nothing from him to back is assertions - which fly in the face of both FM theory and what I got when measureing tuners many years ago.
 
Just to add in case someone else is interested - or has already used the following item and can comment:

I've talked to some people using the same SDR RX as myself and they suggested I consider this for generating stereo FM
https://www.mouser.co.uk/ProductDet...MIi-uqlaCN6AIVyLTtCh3W8wDWEAYYASABEgJhI_D_BwE

Apparently changing a config file will widen the range to 70MHz - 6GHZ (!) And it seems to be able to operate as a source/sink for a 20Msample/sec I/Q stream, so wideband accurate modulation should be possible.
 
It's a good device is the ADALM-PLUTO - a friend of mine demonstrates radio with it to ham radio Foundation license trainees. It has a lot more community support than the LimeSDR-Mini I'm using. I didn't realise it could go down to VHF frequencies. I'll take a closer look
 
It would be useful to know someone can confirm it can be tweaked to cover the VHF Band II for stereo FM as that's the main reason I'm intending to buy one! I asked about this topic on the 'FUNCube Pro Plus' email list and people there suggested the PLUTO. As it stands, it looks ideal for my purposes if it covers this and higher frequencies as I can then use it as a standard RF tool for tests and measurements, not just for my initial task wrt FM radio.
 
Update: From what I've been told, tweaking successfully to get wider operation is fairly reliable. And it looks like the device can work nicely once I can get my head around libiio. is there a particular email list or forum for the device? I'd like if it exists to find a good 'ab intio' book or guide on this approach as the info on the web tends to be 'bitty' and scattered, with differing approaches - adds a layer of confusion to my attempts to understand how to get it going and do what I'm wanting.
 
Thanks for that reference. Interesting. :) Although at present I've put getting the AP on hold because of being distracted by challenges that are more basic - i.e. getting food and trying to avoid covid!
 
You must log in or register to reply here.


advertisement


Back
Top