Friday, April 06, 2012

OAFS redux

A couple of days ago I decided to take another look at the Spectrum Communications Off-Air Frequency Standard (OAFS). It had been suggested that its failure to work might be the result of a solder bridge or similar error. I had a good look at the board using a high power magnifier and found a pair of pads that were suspiciously linked to ground. A moment's work with the desoldering braid and sure enough there was a solder bridge exactly the width of a PCB track.

Having found a fault I was optimistic that the board would work. The setup adjustments were completed OK. But instead of hearing BBC Radio 4 in the speaker as the instruction sheet suggested I received a loud heterodyne with some speech faintly in the background, like listening to an AM signal in SSB mode with the BFO a couple of kHz off-tune.

I was looking at the Spectrum Communications advert in Practical Wireless to check how the ferrite rod was mounted and noticed that the description said "Background heterodyne whistle at 2kHz confirms lock condition." That is exactly what I was getting. Odd that the instructions didn't mention it though. Nevertheless I gave a cheer and went ahead with installing the board in its box.

My happiness was short-lived when I put my frequency counter on the output. It was 10MHz sure enough, but it was not phase locked to anything. I was only receiving the output of the uncontrolled 10MHz crystal oscillator which could be tuned a few tens of Hz either side of 10.000MHz. No adjustment I could make would cause lock to occur.

Comments made to my original post about this suggested that I might have problems with the OAFS as I am not in a good location to receive a strong signal from BBC Radio 4. I'm unhappy with the amount of time I've wasted on this. I think it would be best to write it off and forget about it. I'd rather not be bugged by it or have it taking up scarce space in the G4ILO shack. If anyone would like to have it and see if they can make it work then it's yours for the cost of the postage.

6 comments:

G6BHB said...

Hi Julian, If you let me know the cost of postage I will send you cheque.

John G6BHB.

G0FTD said...

Julian. The bit about the uncontrolled 10Mhz output has me wondering if you are suffering what I had.

The 10MHz output from the OAFS contains harmonics, only a few db down on the fundamental.

The result is that the frequency counter gets a bit confused.

I made very simple 3 pole LPF which I fixed on the inside of the OAFS and as a result all is now ok.

Andy

Julian Moss said...

Andy. All the outputs show the same frequency numbers shifted by factors of 10 so I think I am measuring the true frequency.

I suspect that a weak Radio 4 long wave signal may be the root of the problem. I tried coupling in an active SWL antenna and the heterodyne went to the background and Radio 4 came up to comfortable listening strength.

I could also hear a sort of interference effect which might have been the "mush" Martin G4FUI mentioned in his comment to my first OAFS post. So for all I know it may be capable of working properly and I just live in the wrong place!

Jeremy said...

Hi Julian
I'll pay postage and test it out for you and get it working if you like?

I already have an OAS that I designed and built in the early 1990s. I don't live far from Droitwich but even so an OAS does ideally need an RSSI indicator and a locking 'progress' indicator to show if it is going to have a chance of locking. Also, they do take quite a while to achieve lock. Mine takes about 30 seconds?

Julian Moss said...

Thanks, chaps. I'm pretty sure now that the OAFS is working as the designer intended. The problem being that the designer lives on the south coast and presumably gets a better Radio 4 long wave signal than I do. It's some consolation that the poor performance is not due to a fault with my construction.

This is not something I would want to use as an external frequency reference for my K3 which was the original purpose of it (though I lost interest since finding I have a brain tumour.) The reference frequency jitters: my frequency counter reads 10.000.001 but the Hz digit goes 1,2,3,1,2... whereas measuring the rubidium standard it is completely stable.

Jeremy said...

I wouldn't really recommend using a fully working OAFS as a reference clock for the K3 anyway as I'd expect the jitter and phase noise to be degraded due to the limitations of the 198kHz OAFS system. Also, you will run the very real risk of annoying lock dropouts if there are any static crashes or glitches that upset it. I only use mine for calibration

purposes. i.e. I calibrate a high quality 10MHz OCXO with it plus I calibrate the reference oscillators in my signal generators and analysers with it. Mine is stable to <<0.1Hz short
term (for counter purposes) and I'd expect the Spectrum OAFS to be similar. This makes it ideal for calibrating stuff or as an external reference for a 9 digit frequency counter.