Portsdown boxing/build thread

Discussion about this major DATV Project. See https://wiki.batc.org.uk/The_Portsdown_Transmitter
radiogareth
Posts: 1215
Joined: Wed Jan 06, 2016 9:46 am

Re: Portsdown boxing/build thread

Post by radiogareth » Sun Nov 25, 2018 5:44 pm

Without the filter board inline I've been checking the outputs from my 4 way switch, using this little eBay power meter, good for +5dBm max.
2018-11-24 18.57.58.jpg
2018-11-24 18.57.58.jpg (95.89 KiB) Viewed 4261 times
Changing the indicated frequency changes the displayed value so I can but 'hope' that its a reflection of the manufacturer's published sensitivity scale??
I get the following readings:
Direct from the modulator 4m 843mV, 2M 800mV, 70cms 670mV and 23cms 490mV.
After passing through the attenuator and 4 way switch AND ATTENUATOR (SET TO ZERO) I get understandably a bit less: 4M 814mV, 2M 720mV, 70cms 620mV and 23cms 475mV. These seem appropriate.
I understand the need for filtering the squarewave from the DDS prior to feeding the modulator. I'm not sure why there is a 19dBd amplifier in the output from the filter chain, especially as '23 passes through directly' and it seems to work amp-free (ie at -19dB). Its quite a lot of gain but presumably all accounted for further down the line on the modulator board (off to look up modulator circuit diagram now to try and understand why). Its (the filter board) out of circuit ATM anyway as I seem to have expired the 4 way switch and possibly the amp, which, BTW, Farnell are now listing (ADL5602) as awaiting stock. Digikey have plenty ATM.
Gareth

radiogareth
Posts: 1215
Joined: Wed Jan 06, 2016 9:46 am

Re: Portsdown boxing/build thread

Post by radiogareth » Thu Dec 06, 2018 12:35 pm

I needed a compact 100 Watt dummy load for testing so I bought one of these : https://www.ebay.co.uk/itm/DC-3GHz-50oh ... 2749.l2649 hoping it wasn't a piece of junk......
Works well and measures 50 ohms too. Bolted to an old PC CPU heatsink that had a solid copper core all the way through. Some only have a short slug. Drilled 2.5mm on 18mm centres and tapped the two holes M3. BNC socket from stock, job done. Running 100 watts HF intot it produces warmth, but with fan running on 12 volts it keeps it all at sensible temperatures.
Gareth
2018-12-06 11.31.10.jpg
2018-12-06 11.31.10.jpg (58.25 KiB) Viewed 4160 times
2018-12-06 11.30.57.jpg
2018-12-06 11.30.57.jpg (103 KiB) Viewed 4160 times

G3UEQ
Posts: 93
Joined: Sun Jul 26, 2015 10:28 am

Re: Portsdown boxing/build thread

Post by G3UEQ » Thu Dec 06, 2018 9:13 pm

Hi Gareth
I have same power module but not tried it yet, have you any idea how the 4 way switch expired? Do I need to watch out for anything. Monitoring o/p on my spectrum analyser I get similar results.
73
Andy

G3UEQ
Posts: 93
Joined: Sun Jul 26, 2015 10:28 am

Re: Portsdown boxing/build thread

Post by G3UEQ » Thu Dec 06, 2018 9:23 pm

Correct me if wrong, but the LO filter amplifier is in circuit even on 23cms, its just not filtered as are the other lower bands, I got a lot of harmonics particularly on 4m from the o/p of the modulator board.

73
Andy

G8GKQ
Site Admin
Posts: 2798
Joined: Sun Mar 22, 2009 1:21 pm

Re: Portsdown boxing/build thread

Post by G8GKQ » Thu Dec 06, 2018 9:34 pm

Hi Andy

The LO filter only filters the signal going into the Quadrature modulator so that it is able to correctly extract the I and Q signals with exact 90 degree phase shift.

The modulator itself re-introduces a lot of the harmonics, as will any power amplifier that you put after it. Hence the recommended configuration is to have a good low pass filter (or band-pass) after the final PA. No real point in filtering any harmonics out before then.

Dave

radiogareth
Posts: 1215
Joined: Wed Jan 06, 2016 9:46 am

Re: Portsdown boxing/build thread

Post by radiogareth » Fri Dec 07, 2018 9:37 am

Andy, not sure what killed the 4 way switch, but not reading the data sheet killed the amp after it. I think 'ham fisted' dc checks on the switch path may have contributed, or something.
I really need to see what's coming out (and so know its working) but waiting on another synth board. Plenty of other projects to divert me though....
Gareth

G3UEQ
Posts: 93
Joined: Sun Jul 26, 2015 10:28 am

Re: Portsdown boxing/build thread

Post by G3UEQ » Sun Dec 09, 2018 7:10 pm

G8GKQ wrote:
Thu Dec 06, 2018 9:34 pm
Hi Andy

The LO filter only filters the signal going into the Quadrature modulator so that it is able to correctly extract the I and Q signals with exact 90 degree phase shift.

The modulator itself re-introduces a lot of the harmonics, as will any power amplifier that you put after it. Hence the recommended configuration is to have a good low pass filter (or band-pass) after the final PA. No real point in filtering any harmonics out before then.

Dave
Hi Dave, yes, realise the situation and currently building LPFs!
Andy

G3UEQ
Posts: 93
Joined: Sun Jul 26, 2015 10:28 am

Re: Portsdown boxing/build thread

Post by G3UEQ » Sun Dec 09, 2018 7:14 pm

radiogareth wrote:
Fri Dec 07, 2018 9:37 am
Andy, not sure what killed the 4 way switch, but not reading the data sheet killed the amp after it. I think 'ham fisted' dc checks on the switch path may have contributed, or something.
I really need to see what's coming out (and so know its working) but waiting on another synth board. Plenty of other projects to divert me though....
Gareth
OK Gareth, just wanted to be sure I didn't do something silly. Just found my power meter as per yours is duff!
Re projects, I have just been given a poorly R1967 (RAF AR88), so possible winter diversion!!

Andy

radiogareth
Posts: 1215
Joined: Wed Jan 06, 2016 9:46 am

Re: Portsdown boxing/build thread

Post by radiogareth » Mon Dec 10, 2018 6:12 pm

OK, new DDS board arrived today so after convincing myself that it was the problem, I swapped them over. Bingo, normal service is resumed, almost......
I'm happy that both my attenuator is working and that the output port socket is working if I feed the signal generator direct to them. I'm also very sure that my now SECOND filter board isn't. As I built the second board I checked the DC path through the resistors/inductors/tracks and I am happy with my soldering. But as an example, my 24cm signal reduces by -35dB between the input track and the RF switch input. Something I'm sure should NOT be happening. I have a third board, so I might just build that one too. I'm not 'new' to SM and have worked with electronics as a hobbyist all my life. Can I DC check the 4 way switch or will my multimeter subject the poor thing to excessive V? The datasheet doesn't seem to indicate DC properties, unless its embodied in RF descriptions.
Guidance anyone?
Thanks
Gareth

radiogareth
Posts: 1215
Joined: Wed Jan 06, 2016 9:46 am

Re: Portsdown boxing/build thread

Post by radiogareth » Tue Dec 11, 2018 11:17 am

So a new day....googling dbm to volts gives 7dBm (listed in the specs) as 1.4V p-p. Quite likely that my DMM gives more than that as it will light a blue LED when on diode test so I'll measure V out on the various ohms setting before I try that. Before I cripple another chip I'm going in search of greater magnification to check my soldering, again....Very handy that my local Lidl have handheld magnifying glasses (x2 and inset x4) with two level cold white LED lighting available at the moment. Soldering looks fine, but will re-flow just in case.

Post Reply

Return to “The Portsdown Digital ATV System”