Homebrew SSB Transceiver
 

Hi Peeps...... Now the Exciter and receiver sections are complete... I am turning my attention to the PA. I bought a 150W Linear kit from Ebay... and its still to be constructed...however one or two questions have now cropped up.
1. Meter. At present my meter is monitoring the RF level to the input of the second mixer... i.e post 9 mhz filter..thus I get an idea of signal strength on receive.. and on TX the modulation or drive "if you like" . Should I monitor the PA Current ? or what other points do you suggest.

2. Power switching. I have bought a SMPS with an output of 12V @ 20A. On transmit would you switch the DC Output or the AC input ? Obviously the DC end would require BIG contacts... however the relay is connected it will have a delay of 1 second to allow the aerial and other relays to have operated.

The exciter has a dedicated psu...

Your thoughts would be appreciated

Re: Homebrew SSB Transceiver
 

With an SSB rig, if you're not talking the PA is only drawing bias current so 99.9% of the time you wont be switching large currents - DC side switching is fine.

10amp relay with two sets of contacts paralleled.

Metering - nobody meters anything except SWR and sometimes power out the spout these days.

Re: Homebrew SSB Transceiver
 

Or a medium sized MOSFET.

Re: Homebrew SSB Transceiver
 

If you're going to use a MOSFET watch out for a little gotcha. If there's a big cap on the output of the power supply and one on the input of the amp, the current dump from the former to the latter via the MOSFET can exceed Ids of a lot of "ballpark" MOSFETs and let the magic smoke out.

I would just leave the PA with standing current running and just switch the RF in on TX. I'm doing this with my slightly lower wick 25W CW amplifier. Idle current is about 100mA though the PA IRF510.

Re: Homebrew SSB Transceiver
 

Modern rigs just switch the bias and lower level Tx paths and leave full +V on the PA these days.

Actually, I think a lot of older rigs did as well.

Re: Homebrew SSB Transceiver
 

Hmmmm... Great food for thought peeps... Many thank Mr Bungle.. Terry, and Merlin..The SMP is a commercial LED PSU..Thats it then..... No DC switch/relay....Leave it idling.. and just switch the drive... I will still need an antenna relay.. but thats no problem..re metering..Just "sniff" the output.

Re: Homebrew SSB Transceiver
 

The 150w spec on the linear may also be a bit pie in the sky, depending on what the output transistor/FET is.

Especially as the frequency goes up.

Re: Homebrew SSB Transceiver
 

I agree Terry... I am not intending using full power.. whatever that is anyway.. as my antenna is a half 5rV dipole and on 80m is a bit lacking.. I have a KW antenna match.. but I realise its not optimum.. so I will limit the power anyway The transceiver is 80 / 40 / 20 .. Onward and upward.

Re: Homebrew SSB Transceiver
 

Another thing to watch out for is the potential for HF PSU switching noise on the output of the PSU powering the PA.

As you're using a high current LED supply the output might not be adequately filtered for radio purposes which could lead to spurious generation of an unacceptably high level centred around your main signal.

Re: Homebrew SSB Transceiver
 

That's a good point. When I looked at mine under ~4A load on a scope there was 100mV p-p noise which was pretty bad. I couldn't tune it on my receiver however which was probably good. The final build has a Schaffner mains filter and then a fairly hefty pi filter on the output of the power supply. The pi filter is a bit of FR4 with a couple of low ESR electrolytic caps paralleled with 3 100nF ceramics and a hand wound FT82-43 with about 50 turns on it ~ 1mH. All glued down and then conformal coated. Can't see anything at 5mV/div on the scope

Re: Homebrew SSB Transceiver
 

Hi Wendy,

May I enquire: what method of SSB generation are you using?

I plan on doing some bench experiments (no radiations!) with SSB and the Weaver method struck me as about the easiest to implement.

Re: Homebrew SSB Transceiver
 

Hi Karen... I used Double balanced modulators.. the MC1496.... with a 9 mhz (ish) carrier and a 9.00128 filter. The ( mhz ish) oscillator is switched with caps to drag the frequency +/-
I would be interested in your results....

Re: Homebrew SSB Transceiver
 

Hi Mr Bungle.....and Tim.... Grrrrr... Many thanks for the info guys... I will load up the outputs and scope the voltage... THANKS both for your inputs..As it happens I have enough space on the top of the chassis to fit RF Bandpass filters, the SMP and now.. it seems a probable DC filter..
Nothing is EVER simple LOL.

Re: Homebrew SSB Transceiver
 

Well back to it... after a short break and hollybobs. I have a slight problem stopping me completing the project.
The PA output filter. The common thought is to use Low Pass filters to eliminate out of band spurious signals, and thus I made LP Filters for 80 and 40 metres with reasonable success. However the 20M filter is causing me concern.
I have made quite a few different filters based on designs from the interweb... mostly the Butterworth configuration but I cannot get the insertion loss at 14 Mhz to be better than 13db, and the fall off is quite shallow, rather than a reasonable "cliff edge" on 80 and 40, where the insertion loss is 2 db. All the charts give values of L1/L3 = 0.6 uH and L2 0.9 uH in round figures with C1/C4 100pf C2/C3 390pf or similar.
Should I consider a Bandpass filter in this location ?? or will I have trouble matching the 50R in / Out impedance.
I add the pcb and schematic for your info.
L1/2/3 are 50/6 torroids. The caps at present are disc ceramics.. no point in buying until I know what values are suitable.
Your comment will be most welcome

Re: Homebrew SSB Transceiver
 

Sorry I forgot the schematic

Re: Homebrew SSB Transceiver
 

Your filter values are way different to mine on 20m. I wonder if there's a mismatch between stages. I've got a 3dB point of 16.4 MHz, 30dB down at 22.5MHz. Insertion loss at peak is around 1-2dB. My calculated values were normalised for off shelf capacitors. I can't remember where I got them from, possibly GQRP club. I've run the inductors through toroids.info based on the values presented:

c1 - 180pF
c2 - 390pF
c3 - 390pF
c4 - 180pF

l1 - .773uH - 14t T50-6
l2 - .904uH - 15t T50-6
l3 - .773uH - 14t T50-6

Terminating 50 ohms at one end shows a flat VSWR and 50 ohms in the pass band on my MFJ259 analyser. Swept with my AWG and power meter and it's about right. I think it was about 50KHz off. I didn't check return loss as VSWR was good enough to worry about dB here and there.

I used T37-2 toroids, which was a bad idea as they get rather hot!

Re: Homebrew SSB Transceiver
 

The current G-QRP club technical sheet for LPFs is here http://www.gqrp.com/technical2.htm

Tried and tested designs and values

Re: Homebrew SSB Transceiver
 

Hi Mr Bungle. I will return to it this am with the values you suggest.... I used T50-6 due to the power levels of approx 45 W
Hi Martin. I used this data for one of my trials...and it gave me similar results to others..however..... its not a "sprint"... I need to get it right.... so time will be taken.

THanks both for your comments

Re: Homebrew SSB Transceiver
 

I think the default Fc for a Butterworth will be at the -3dB point so if you design for Fc= 15MHz the filter will already have 3dB loss by 15MHz. I can't explain why you are seeing >13dB loss unless you have a component or circuit issue. The PCB layout is a bit odd with inputs crossing outputs and lots of long link wires but I don't think this will cause the 10dB extra loss.

A Butterworth filter won't give a steep rolloff rate and I would have gone for a classic cheby LPF and you could expect over 50dB rejection by 30MHz. For a 100W+ PA I would have gone for T80-6 toroids but for low duty cycle SSB you will probably be OK with T50-6 toroids.

I'd expect a 7th order LPF to have about 0.25dB insertion loss at 14.3MHz and 50-55dB rejection by 30MHz if built with good components.

Re: Homebrew SSB Transceiver
 

If they get beyond about 100degC then you could enter a scary place where the coil Q will start to fall markedly. This is due to increased copper and core losses (at higher temperature) and it is possible to hit a runaway condition where the Q falls (so the losses increase) so it gets hotter and so the Q falls some more etc etc until the wire and the toroid can end up looking like it has been smoked in barbecue coals. Usually this is a fairly slow process requiring a full cw test signal into an antenna load with a poor match. It might take a few minutes to get it to the runaway condition but the coil can end up being so lossy (when hot) it resembles a low resistance dummy load. When this happens, a 50-100W transmitter will be able to properly fry a small toroid into a charred mess :)

This effect is more common in steep elliptic filters when the operator transmits very close to the filter cutoff and I've seen it the most in highpass filters where the inductor is fitted as a shunt component.