The 4 m Forum

[G4IPZ]

I wonder if anyone out there happens to have a copy of the instructions/circuit diagram or any technical blurb on the G3WPO 46PA 4 metre PA.
I have just bought one (as seen) from EBAY. It works sort of but even without drive it's running way too hot because it's pulling 3A on standby.

There is no RF switching but as well as the phono socket for the PTT line there is also a 3.5mm jack socket with a feed to a preset resistor which seems to feed a pair of diodes from the TX out. It may be some sort of power monitor connection.

I know it's probably just the PA device going west but there is one rather charred looking 1/8w resistor of unreadable value whose job (and value) I'd like to know before trying another output transistor.

Any help will be much appreciated.

Oh the device is an MRF1946A

Slim G4IPZ

[GW4RWR]

I've emailed your ilizarov email with a zipped 46pa package.

Otherwise, in the absence of a circuit diagram I'd replace resistors, intending to get the standing current down. Start by terminating i/p and o/p with 50ohm loads. Get the transistor cut off (base < 0.6V), then play with the values to increase the standing current.

Good luck!

[GI0GDP]

The best thing to do here is remove all the bias components, diode, 100R resistor i think and caps and inductor. The fact that MRF PA transistor is drawing 3A is a good sign its OK.
The bias network in this PA is not PA device adjustable or temperature tracked and is generally a bad design.
My humble advise is to use and emitter follower to bias the PA device.
This will allow easy selzection of bias current ( say 50 to 100mA)and will be temperatute tracked.
Go to the 75 watt PA eslewhere on the site and lift the TIP31A and it's associated parts and apply to your PA.
This will also ensure that the IMD products are tolerable.
regards
Geoff Pike
GI0GDP

[G4IPZ]

Firstly thanks to GW4RWR for the circuit info it help immensely. There was a burnt out resistor (R3 on the diagram) which is a 10R resistor, DC isolated from ground by a 0.1uf cap. It seemed odd but as this is obviously part of an rf bypass circuit warning bells about parasitics began flashing.
The bias network is (feebly) temperature tracked by having a 1N4001 forward biassed providing the DC voltage to the PA device base, but with the diode physically on top of the transistor housing so as the device heats up, the diode conducts more and the bias falls. In theory!
I changed the diode anyway but added a brass strap over it to keep it pressed against the transistor for better thermal contact.
Before I put DC on the thing again I detuned all the trimmers to minimum value. On powering up with no rf drive there was no massive current draw, just about the amount I'd expect for the power LED and a forward biassed diode fed by a 220R resistor (different from the circuit original of 100R)
I then started playing with the trimmers and sure enough I found a couple of places where the current draw would jump to about 1 to 2 amps.
So parasitics it was.
I added a few small extra decoupling caps but also added a screen isolating the input and output tuned circuits. This time when powered up there were no points where the current took off.
Next step was to try it with drive from my ARAC transverter and again, with RF drive as well as DC, no massive sudden current draw.
After retrimming the caps I reckon the thing is giving me about 20W out wirh 1W FM drive and it doesn't run hot, merely slightly warm after 5 minutes keydown.
I am suspicious that the device used in mine (MRF1946A) may have more gain than that used in the original (TP2330) hence the instability and self oscillation.
So thanks again for the help and advice and at some point Geoff I agree that I need to put a proper bias circuit into the thing and will have a look at the one you recommend.

Cheers for now

Slim
G4IPZ

[GI0GDP]

I have built a few of these and they all suffered from parasitisc oscillation, the snubber network often melted the resistor. I wolud recommend say 0.1uf and 220/330R in series from collector to base to tame the LF gain in these devices. As i have said before it is way easier to build 4M amps using HF devices rather than VHF devices,
regards
Geoff Pike

[G4IRX]

Some useful notes, thanks guys.

I got the 6m version of this PA along with the transverter at a local rally recently. I've just started converting it to 4m, have ordered S18 coils and capacitors from Crickewood Electronics, noting that the S18 white is S18 grey in their catalogue.

My amp has a SD1274 fitted (30W@160MHz) and the 10R resistor is burned out. Will try Geoff's suggestion of the 0.1µF and 220R between C and B of the device to see if it tames it.
_________________
Nick G4IRX
Nottingham IO92JV

[G3PTU]

Probably the pair diodes etc is an ALC line to the prime mover what ever it was.
As said the problem with some VHF design // transistors is that they go off at some indedterimate frequency far removed from the desired.
The diode arrangement of bias temperature stabilisation works ok as long as the device does not get VERY hot and even then I would put some heat transfer gunge between the diode and the transistor case. Somewhere I have seen graphs of diodes used as thermistors in Solor Heating panels and they have a useful slope up to a point, then flatten out.
I always worked on the thing should be kept cool, the bias should be 'Stiff' and plenty of decoupling on supplies, multiple caps and ferrite etc. Snubbers work wonders.
Before collecting my Bus pass early VHf radio PA's could be terrible to tame. We had some Marconi things made some where or other on licence and these appeared to be all different in behaviour on a day to day basis.
David G3PTU