GS-31B 50mhz Amplifier Project..

[VK3ZL]

I have been fairly occupied these past couple of months building another GS-31B 2 meter amplifier and also developing a 70cm amplifier...The 70cm unit is on going at the moment..

I am half way through a 6 meter GS-31B amplifier at present using the Russian GS-31B...This will be a general purpose amplifier designed to run the VK legal power limit with a good reserve so that the tube is just loafing along..I prefer to use a tube with twice the Anode dissipation needed rather than something that is running close to the edge of it's capabilities...This way if the amplifier is used as designed the tube should last many years..

I will include a couple of pics of the progress to date, adding a few more as I near completion..I make the amplifier box by hand with simple tools..These days I have the aluminum sheet cut to my specifications at a local sheet metal fabrication shop.I bring it home and put it together carefully using Aluminum angle, dozens of rivets and lots of patience...It takes me about two days to build up a case...I have developed a standard design for myself so I have learned most of the pitfalls putting cases together..One must hasten slowly to do a good job..

This amplifier will use a custom built Toroidal HT transformer and filament transformer...High Tension will be 3500volts and the transformer can handle a KW of power comfortably..It uses a simple L/R soft start....Cooling is a Papst 28 volt all metal fan rated 250cfm in free air which is adequate to cool the GS-31B tube..The RF deck will be sealed and pressurised, air runs through the Tube cooler and a Teflon chimney on the top cover..The rest of the amplifier will be straight forward, a tunable T input network and a Pi-L output...I add the L section as additional harmonic suppression...

Three pics show the completed amplifier case, front panel cut out ready for painting and populating and some of the wiring just started...

I love doing this work...Keeps a man in his 70's occupied and out of other mischief....Enjoy.. Bob..VK3ZL..

[VK4KHZ]

Keep up the good work Bob! I admire your passion for homebrewing such well designed and expertly completed amplifiers.
I look forward to seeing the final results of your 70cm version.

Regards

Shane VK4KHZ

[VK2ZRH]

Brilliant stuff, Bob.

How about you teach a few of us the techniques of how you put those cases together, beginning with "measure twice, cut once".

73, Roger Harrison VK2ZRH
PS: I have ignored any techno-rectitude comment re "mhz" vs MHz.

[VK7DX]

Yet another work of art from VK3ZL.

Onya Bob

cheers
Frank

[VK3ZL]

Hi all...I have been keeping busy with this amplifier these past few days and I am nearly finished the hard work, wiring, RF deck etc, so starting to see the light..I am into my second week doing a few hours each day..The major job with this kind of project is the wiring looms....This amplifier has four power supply circuits, Filament transformer and associated circuits, 24volt relay circuit, HT circuit and the cooling fan supply...

Top picture is the loom ready for sorting out and wired to the front control panel..I always double check everything..All wiring is color coded..

Picture 2 is the completed job..I like to try and keep everything neat and tidy..

Picture 3 is the right hand side of the front sub panel showing the W4ZT bias board and bias resistors etc..

Picture 4 is the left side of the sub panel showing the simple step start circuit for the HT...I just use a 50 ohm 50 watt resistor across a 240 volt relay in these amplifiers...It gives a distinct delay as the HT supply charges up before shorting out the resistor...

There are two meters in the amplifier, 1 amp for the Anode current and a 500ma meter for Grid current...These meters are already calibrated with their own shunts and I just put back to back 1N5408 diodes and a 0.01 mfd disc ceramic across each meter..

In the center of the sub panel is a small relay which shorts a bias resistor when PTT is activated putting the amplifier into operate mode..

Once all the circuits are wired into the amplifier one has to do the traditional smoke test....I use a small variac when first testing everything....I put a 1 amp fuse in the mains input and bring the mains up slowly and test each circuit switch on to make certain that it is operating correctly...I test the high tension separately isolated from the wiring loom first, also using the variac to bring it up to full potential slowly....Once I am satisfied that the rectifier stack is working properly I wire the HT transformer into the step start and with an 8 amp fuse in the mains input switch it on again....I try to use as light a fuse in the amplifier mains input that will withstand the first inrush current through the step start...When the amplifier is running 400 watts it is only drawing a few amps from the mains..I often see commercial ham amplifiers with 20 amp fuses in the mains input and wonder why so high...If anything major goes wrong something is going to be cooked big time...Or, someone is running more power than is allowed..Ha...

I will post the next phase of this amplifier as soon as I finish it off...Hopefully the parts will arrive soon....

Enjoy..Bob..VK3Z

[VK3ZL]

Thanks for the encouraging comments and emails...If anyone would like one of these amplifiers, 6 or 2 meters I can build one for you...

Bob..VK3ZL..

[VK3ZL]

Hi all, Monday and still held up waiting for parts...I have most of the amplifier completed so it isfrustrating when held up..

Anyhow, a couple of pics, Plate choke and the HT transformer and a pic of the HT power supply compartment...

Plate chokes are a mystery to many hams and selecting the right materials can be a problem...I wind all my own chokes on machined, hardened Polyethylene stock....For 50Mhz I use a 3/4 inch diameter former and 50 close wound turns of 20 gauge formvar coated copper wire...The trick is how to attach the wire to the choke former for a neat job....Well, I have found that two circlips which are a firm fit around the former the best method...I clean away the coating from the circlip mounting holes and treat it with soldering fluid then tin it to make a good solder joint for the choke wire and connections...With circlips you don't have a shortened turn on the choke...Pics show before and finished choke..

The large HT Toroidal transformer is rated at 1300vac @ 1amp CCS...I have these custom built to my requirements..I use a voltage doubled rectifier stack and amplifier HT is 3500vdc peak...More than adequate power for this amplifier..

Last pic is a look into the HT power supply side of the amplifier....The small power supply at the back is to run the amplifier cooling fan...The supply is simple and slightly over engineered but with 24 vdc there is more than enough air to cool the tube up to 1kw pep output..So running at the VK legal limit everything is just coasting along...

Well hopefully the final components will arrive soon and I will complete this job and post pics of the RF stages and the front panel..

Enjoy...Bob..VK3ZL..

[VK3ZL]

Saturday 5th....Some parts finally arrived yesterday so I was able to complete the HT supply and the top RF deck..
The amplifier uses a simple Pi-L network output...The GS-31B tube shows about 5pf or so to it's surrounds so I have a bit of leeway for Plate tuning...The pics show a simple Flapper capacitor setup..The flapper is actuated with a threaded shaft from the front panel and a piece of Polyethylene at the flapper end..I am using 0.32mm brass plate for the Flapper..This is nice and flexible and tunes smoothly...The Loading capacitor is an ordinary 160pf unit...The Plate coil is made from 4mm solid copper wire with a smaller L coil at the output end...This setup tunes perfectly through 49 to 54mhz....

The HT rectifier is a straight forward Voltage doubling circuit..There are 18 1N5408 rectifier diodes and 8 450volt 470Mfd electrolytic capacitors...There are 16 150K 2 watt metal film equalising resistors in pairs across the electrolytic's ..This HT power supply produces around 3500vdc peak just sweet for the GS-31B tube..

I have now completed the bulk of the amplifier...All electronics have been fully tested and function perfectly....Tomorrow I will complete the T input network and this amplifier is ready for RF..

Next lot of pictures will be the completed amplifier fully tested and running....Enjoy..Bob..VK3ZL..

[VK4WTN]

Another fantastic amplifier Bob an absolute credit to you
regards Wayne VK4WTN

[VK2OMD]

...
I have now completed the bulk of the amplifier...All electronics have been fully tested and function perfectly....Tomorrow I will complete the T input network and this amplifier is ready for RF..
...

Nice work!

Is there a DC ground on the Pi-L circuit to provide some safety in case of coupling cap failure? It isn't very obvious.

What Insertion VSWR do you achieve in bypass mode?

Cheers
Owen

[VK3ZL]

Hi Owen, Good question and the answer is No...The Load capacitor spacing will only support around 1Kv and if the Plate coupling cap shorted the 3500vdc would short across the Load cap to ground blowing the Glitch resistor and probably the main fuse...A choke across the output socket is good to prevent static buildup on the transmission line but little else...Info, Tom W8JI..The Plate DC blocking capacitor is 390pf 15kv..I have never seen one fail..

In bypass mode vswr will be less than 1.2-1...That's why I use bladed relays on input and output for 6 meters....Relays with internal leads can create reactance on the transmission line, ( vswr bump) ...This can also be easily cancelled using a small 12pf variable trim cap from the line to ground mounted on the input relay...I have used this method from time to time and it works fine..

Bob..VK3ZL..

[VK6OX]

Bypass VSWR in the amp that Bob has supplied me Owen is 1.12:1

The amp I have utilises 2 x GI-6B beasties which are pretty rugged!
And yes, I know that has no relevance to bypass VSWR.....just thought I'd mention it nonetheles!!

Keep up the good work Bob, the 31B would look nice in my shack HI!!

[VK3ZL]

Gidday Andy..I am always learning new things with each amp I build..This job is a twin to the last 2 meter amplifier so already has a future home..I still havn't built one for myself though...That's a "round tuit" project...All fine here BTW....Bob..

[VK2OMD]

Hi Owen, Good question and the answer is No...The Load capacitor spacing will only support around 1Kv and if the Plate coupling cap shorted the 3500vdc would short across the Load cap to ground blowing the Glitch resistor and probably the main fuse...A choke across the output socket is good to prevent static buildup on the transmission line but little else...Info, Tom W8JI..The Plate DC blocking capacitor is 390pf 15kv..I have never seen one fail..

In bypass mode vswr will be less than 1.2-1...That's why I use bladed relays on input and output for 6 meters....Relays with internal leads can create reactance on the transmission line, ( vswr bump) ...This can also be easily cancelled using a small 12pf variable trim cap from the line to ground mounted on the input relay...I have used this method from time to time and it works fine..

Bob..VK3ZL..

I have not seen Tom offer that advice, and I note that in two PAs that I have had that I understand from him that he designed there are RF chokes from the output network to ground. It is a very common practice.

Yes, I was going to mention that you can compensate those relays. My article at http://vk1od.net/transmissionline/conce ... /index.htm explains a technique. A short piece of RG58 coax as an o/c stub is very effective, and with higher voltage rating than most inexpensive trimmers or caps.



Above shows an example of compensation of a power relay, in this case Insertion VSWR is less and 1.1 at 60MHz. It is intended for HF in this case and Insertion VSWR at 30MHz is 1.03, for 60MHz you would make the leads shorter as in your pics and the Insertion VSWR would be improved).

Owen

[VK2OMD]

...A choke across the output socket is good to prevent static buildup on the transmission line but little else...Info, Tom W8JI...

Hi Bob,

I was a little surprised to see that quote, and as I mentioned, Tom has produced designs that do include the 'safety' RF choke in the output network.

I asked Tom if the he had said that, or if he had been misunderstood / misquoted. His response was "I would never say anything like that. Where did it appear?? ".

Back to the RFC, not only does it (should it) provide a low resistance path to ensure that serious failure of the blocking cap results in a fuse or resistor failing to make the system safe, but it also provides a current path in the event that the blocking cap develops even a very low leakage. In the latter case, were it not for the RFC, a DC voltage slowly develops on the high side of the load cap that eventually / periodically breaks down (often with an audible tick), and recharges and so on. This latter effect may occur even in standby mode, but with RF on, the DC bias compromises the power handling capacity of the network..

There is good reason to use the 'safety' RF choke, I recommend you consider fitting the 'safety' RF choke to your design.

Cheers
Owen

[VK3ZL]

Hi Owen, You are correct..I will add a choke today..Spoke to Tom and he said that I was probably confused over a quote regarding the choke as a static drain...Thanks for bringing this to my attention...

Bob..VK3ZL..

[VK2OMD]

....Tomorrow I will complete the T input network and this amplifier is ready for RF...

I am intrigued by your choice of a T input network. Is that a low pass T?

A Class B (or near Class B) cathode current has high second harmonic component, and if the cathode circuit does not offer a low impedance at the second harmonic (and higher ones), then cathode degeneration results in anode current more like Class A and the attendent drop in efficiency. The amplifier looks like Class B with no signal, but looks like Class A with full signal.

One way of achieving minimal cathode degeneration at the second and higher harmonics is a low pass pi network placed real close to the cathode. Even intervening transmission line can spoil behaviour.

Owen

[VK3ZL]

Hi again Owen, I have always used the T input network on my 6 and 2 meter amplifiers with no problems and good efficiency..I note the English Discovery amplifiers also do the same and most of my ARRL handbooks also..

Bob..

[VK2OMD]

Hi again Owen, I have always used the T input network on my 6 and 2 meter amplifiers with no problems and good efficiency..I note the English Discovery amplifiers also do the same and most of my ARRL handbooks also..

Bob..

Oh well, if you are interested in some more techical discussion, see http://vk1od.net/tx/concept/SpoilingClassB.htm .

I don't put a lot of credibility in the ARRL. They make plenty of mistakes, and they don't fix them.

Eimac/CPI hint at a little issue in their C&FPT "In grounded-grid circuits, the cathode lead inductance must be minimized to prevent loss of drive power before it reaches the grid-cathode region of the tube" but their reasons show a shallow understanding of the issue.

You might argue that a T network adds a bunch of inductance in series with the cathode, and either that it is really good because it is part of the network, or really bad because it breaks EImac's rule. The real issue is if you do not facilitate Class B waveshape (ie its harmonic content), you increase conduction angle and obtain the 'benefits' of Class A.

I mentioned the harmonic content of anode current due to Class B, but note that the cathode current in a GG triode also has substantial odd harmonics due to the shape of the grid current waveform.

Owen

[VK3ZL]

Thanks for the link Owen..I will peruse it carefully...Bob..