Progress with the GU74B amplifier

After the summers antenna work it was time in October to pick up where I left with the amplifier last spring. I'm in to the second year of the construction now. But that is OK. The original time plan was two years. I have a small kid, and a lot of "ordinary" work to do. So the construction has to be done on my spare time.

The picture above shows the RF tank nearly completed. The coil mounted vertically on the bottom of the chassis is the 30-80 meter coil and the one above is the 10-20 meter coil. The latter is wound with 5 mm soft copper tubing from Biltema. Not visable on the photo is the L-coil mounted on the back of the band switch. The L-coil consists of two T200 ferrite cores with teflon insulated wire.

On 80 meter the tank was measured to Z=50 with X=3 on 3.750 MHz with capacitors set to calculated vaules. So the tank seem to resonate just fine. The only problem now is to overcome the limitations of the plate capacitor. The minimum capacitance jumped up to 40 pF when stray capacitances were added. This will present me with problems on the upper HF bands where a capacitance of 20-30 pF is needed. I played with different solutions like adding another L-section before the 10 meter coil or switch to a vacuum varaible. But I finally decided to find another air variable capacitor with two stator sections.

I'm also not happy with the loadline calculated for the tube when I was new to this. The original loadline translates into a plate load resitance of 1700 ohms. A plate load resitance of 2200-2400 ohms would be more correct for my project. So, I will most likely end up redesigning the tank circuit anyway. This higher plate load calls for a capacitance on 10 meter under 20 pF. A new capacitor is needed for sure!


Input network. Componentes are soldered on the underside of the tube socket. The tube will be grid driven with grounded cathode. The cicuit is very simple. I expect instabillty on the higher HF bands. But I will deal with that when time comes.

Connections to the grid box are made with feedthrough capacitors. RF input signal is to be fed through the SMA contact.

View of the RF deck from above. The empty space in the upper left corner will be occupied by the 160 meter coil.

Close up view of the 30-80 meter coil. I use pieces of delrin rod as insulating spacers.

Close up view of the RF-output. The RG-142 leads to the RF relays. The RF choke ensures that the antenna output is kept on DC ground. The choke also short circuits DC to ground if some component failure results in high voltage at the output.

Environmentally friendly radial system at SM0NOR

Environmentally friendly radial system at SM0NOR

In preparation for the contest season it is time to go over last years radial system in the lawn. According to plan all the wires layed out last September"disappeared" in the grass and became invisable by midsummer the following year. So, no problem there. However, recent remaking of the garden cut several wires and thus making the radial system inadequate. Above is a picture of the new biodegradable radial pins from DX Engineering. They claim them to dissappear in one year... in time for the grass roots capture the wires! Amazing, I'm proud to be the owner of the first environmentally friendly radial system in the neighborhood.

Thursday May 27th.

Now everything is ready for the concrete truck. The truck will arrive Monday morning at 7 am. Wow, I'm excited!

Here is a picture of the hole with rebar cage and support for the sleeve. The sleeve is a piece of bended sheet metal in which the ground post will slid in. It absolutely essential that the sleeve is positioned exactly vertically, or the mast will lean.

Plate and load capacitor installed.

Toroiods for 160 meter (the big one) and L-coil

And now a little side track. This is my new 12 meter antenna mast to replace the roof mountings. Eventually I will have a Steppir in the top.

The 7 meter bottom section lying on the ground. The 5 meter top extracts by means of a winch.

Rotor cage

Ground post. When retracted, the mast fold over the ground post for easy access to the antenna mount.

The dig! It will be 1x1x1 meter when ready!

Work in progress. The making of an air wound pi coil!

It has been a while since last post. The build has not stalled but the process has been very slow lately. The reason is that I have had too much work to do the last months. I have not been able to spend any time in the work shop. However, this Easter weekend gave me the opportunity to to do some work. The goal for this weekend was to make the 30 to 80 meter pi coil. The raw material is 3 mm copper wire which should be more than adequate for 1 KW in SSB service. Further on, this size should be able to withstand continuos service as well

The completed coil mesures 70 mm OD, 84 mm in length and has a total number of 18 turns. The measured inductance is 11,18 micro Henry and the calculated value is 12, 54 micro Henry. The needed inductance for the Pi-L tank is 11,04. So the result is perfect!

First step of the coil making. About 40 turns coiled up on a 60 mm steel tube. I used only 18 turns.

Four polycarbonate ribs are cut to size. They will support the coil and keep the pitch of the turns through out the length of the coil. Polycarbonate (Lexan) is a very good plastic materiel, far superior to the more common acrylic. It is a lot easier to work with machine tools since it does not splinter as easy.

The supporting ribs are drilled to accommodate the coil.

This is the time consuming part! The polycarbonate ribs are threaded on the coil, turn by turn!

The coil is ready! Nice and shining. It almost looks factory made.

After making this 11 micro Henry coil I realize that the amplifier housing does not have room to accommodate the 160 meter pi coil nor the L-coil for all bands if I make them as traditional air wound coils. For this I need to investigate how to use toroids in high power coil design!

Gu74B amp project progress!

It was I while ago I had some updates here. This is because of several reasons; one of them I will tell about further down. The other reasons is that I had a hard time finding meters for a good price. I searched Ebay and other sites hoping for some nice looking meters to show up for a bargain. Finally I gave up and bought two brand new SIFAM meters. Althoug they look nice, they were also very expensive. The other reason for the delay has been my fear of making square holes! Back in the days I never succeeded, but this time I actually got it right! The secret? The right tools and patience!

Here is the front panel with almost all holes done. What is missing are three shaft holes for the tank capacitors and band switch. As you can see, I succeeded very well with the square holes. The meters are beautifully aligned and so are the rocker switches. The over all finnish of the front panel may look ugly in this phase. But you should understand that the grey surface with blue lines is a protective plastic film on the aluminum. Once taken off, I expect a shiny and bright aluminum surface. I plan to anodize and print labels on it. Does anyone know where I can get the sheet anodized in the Stockholm area? Another alternative is to order a custom made front panel from the guys at Front Panel Express. It is very easy, you draw your panel with their software and you get the price instantly. You then upload the file to there workshop in Germany and about a week later you get a beautiful anodized front panel with holes and prints for about €120. But, I don't think that will be necessary!

Of course the panel meters are back-lit! (I will customize the scales later).

Detail picture of the wiring. I only use teflon or silicon rubber insulated high temperature wire. It is wonderful to work with, no accidentally melted plastics while soldering. But more costly!

This is the 900 volt AC post. I could not find high voltage rated posts, so this is a 600 volt post that I elevated from the chassis by two pillars. Will I have a flash over between the connectors here? I don't think so. Here it is "only" 900 volts AC and about 8-9 mm between the cables. But I'd be happy for any comment on this. Also the teflon wire is only rated at 600 volts. To increase the insulation factor I have led the cable through a piece of teflon tubing. Maybe this is over-kill, but it feels good!

I've got some comments about the RF deck being empty so I thought I might upload some pictures of things that are going in there in the next couple of months. The main reason for the RF deck being empty is that I'm bent on having completed the PS 100% before starting with the fun parts. Back in the days when I built 813 amplifiers I always started withe the big tubes and capacitors and when it finally was time for the PS I was so eager to get it working that the PS ended up being a very hasty work. But I have learnt that a good PS is really one of the most important things both from a performance and a security stand point. Hence, I will not go into building the RF deck until the PS is fully operational. But, in the mean time, enjoy some pictures of the "big things":

Tank coil for 10-20 meters. I wound with 4,5 mm copper tubing

Very nice NOS ceramic band switch with silver plated contacts. I got from Ralf at QRO-shop in Germany. (http://www.qro-shop.com/)

Tune capacitor. 100o pF, 1,5 mm spacing. I bought it second hand from a amateur in Skåne.

Custom made Plate capacitor from Ham Center in Italy. 23-320 pF, 3 mm spacing.

A weeks work on the GU74B Kilowatt amp project

The work has slowed down somewhat since the last post due to other things. But last week I felt that it was necessary to start again. I'm starting to realize that this is actually a hugh project! Some things just takes a lot of time and you don't really see any progress. But this undertaking is like any big project; it is about commitment and persistence! It is not always a perfect day to do some work on the amplifier. But when you feel like that you have to go there and at least do something. Otherwise there is a big risk that the project ends up on the shelf. I'm, however, determined that this is not going to happen.

Ok, let's g over to pictures and tell what has been going since the last post!

Overview of the amplifier. PSU compartment and RF deck.

From left to right:

The first boards are the two stacked G3SEK tetrode boards. Mounted on the sides are two power resistors that are part of screen grid circuit.

The board in the middle is the new redesigned voltage doubler. When the voltage doubler was completed the first time i realized it was too big and heavy (see earlier post for picture). With 10 400 volt NOS capacitors it was so heavy that the circuit board was bending under the weight. The new board is designed around 8 450 volt low-profile capacitors I found on Mouser. They were expensive at almost 6 Euros a piece. But it was well worth it, given the fact that the result made it possible to build everything smaller which in turn will make it possible to accommodate a second HV transformer later on.

To the right you can see a smaller toroid that carries 18-0-18 volts for the control boards and relays. Also here you can see two hefty relays with associated power resistor for HV control and soft start.

Side view of the PSU

On this side view you can see that everything is actually built on two floors. The two stacked G3SEK boards are clearly visible with a large heat sink beneath. Then the voltage doubler and then, to the far right transformer for control voltages and power entry module.

On the second floor I have bolted the HV transformer to the bottom of the chassis together with terminal blocks and 230 volt mains wiring. As you can see here, there is room for another toroid exactly like the first one, which would make it possible to upgrade the system to 1600 VA input instead of the current 800 VA. 800 VA input is basically fine for 1 Kw+ PEP in CW or SSB. But should I like to operate in 100% duty cycle I need more iron! I don't know if I ever go there, but it is nice to have the option.

Next step now is to hook up the G3SEK boards to 18 volts and start testing it. Now there is also an awful lot of boring wiring to do. I think next post will come sometime during the christmas holidays.

Control Boards for the 1 KW GU74B amplifier

The last two weeks I have been busy assembling the Tetrode Boards. They have been designed by Ian, G3SEK and they provide a lot of functions to the amplifier that makes the whole construction a lot easier! On the two boards there are supplies and regulators for screen grid and bias, HV fault and tube protection circuits, ALC, and TX/RX sequencing.

These are the assembled Tetrode Boards.

They are 5x4 inch and you can buy them either as empty boards and provide components by your self. Or you can buy them as a kit and solder everything together. You can red more about them on Ian's web page: http://www.ifwtech.co.uk/g3sek/. Ian also sells a Triode board.

Next step is to fit the boards together with transformers and the HV board into the PS compartment. More pictures will come in about a week!

Major steps on the Gu74B 1 Kilowatt amplifier

Some major steps has been taken since the last post. The structural part of the chassis assembly is finished. (When you build an amplifier; actually half of the work is in the workshop). What is missing now are the holes in the front panel for switches, shafts and metering.

I will also start posting parts of the circuit diagram below.

Overview picture of the amplifier:

I have started with parts of the Mains wiring, mounted the blower and the socket in the pressurized sub chassis. The air intake is in the upper right corner, just to left of the Power Entry Module. The blower the then drags air through the PS compartment, and then through a hole in the wall between PS and the RF deck. The air exhausts through the tube cooling fins and up through a chimney and finally through a hole in the top cover.

Detail pictures:

The blower is obtained from the very helpful people of Airflow UK; http://www.airflow.co.uk/. It is rated 31 CFM at 0.5 inch back pressure, which is 50% above the tube specification. I hope this will be enough considering that the airflow will be somewhat hindered by all the components in the PS compartment and in the RF deck.

All wiring follows European standard; blue for Neutral and brown for Live!

Parts of the Mains wiring. All connections are shrink tubed to prevent accidental contact!

Safety interlock switch. If the top cover is removed. HV is instantly disconnected. But of course; you should never take of the top cover on a live High Voltage Power Supply! Still you also have to think of the capacitor bank which can take up to a minute to fully discharge. But more on that when we come into the filtering section!

Heavy Duty relays in the Mains circuit to cope with high current surges!

Circuit diagram of the Mains wiring:

Here is the first part of the schematics. I will add more as we get further into the project. Here is how it works:

When the Mains switch in the Power Entry Module (The Power Entry Module contains switch, fuses and EMI filter) is turned on power will be present in the control transformer, blower and mains indicator on the front panel. In this stage the amplifier rests until the HV ON switch on the front panel is activated. When HV ON is switched, 12 volts DC will be present on the first relay. When the relay is actuated 230 volts will be present on the HV transformer primary. But the current will actually start flowing in two steps. For the first half second the current will be limited by the resistor in series with the Live lead to the transformer. When the capacitor is charged, after about 0.5 seconds, the second relay will actuate and bypass the resistor. The purpose of this step start is to take stress off from the large capacitor bank in the DC circuit of the Power Supply.

At this stage; grid voltages will be applied before the cathode is heated. I will have to adress that later with some kind of delay circuit.

The circuit diagram is done by the help of TinyCAD. An Sopen Source software available at: http://tinycad.sourceforge.net/

Update on the Kilowatt amp project - bench testing of voltage doubler

Here is a picture of the voltage doubler board under bench testing. For the test, the transformer primary is fed with a 11 volts AC source.

NEVER do a bench test like this with full High Voltage!!!!!

The transformer is rated 900 volts at 0.8 Ampere, with 230 volts on the primary. With 11 volts AC on the primary the calculated DC output after voltage doubling and filtering should be 133 volts. On the picture you can see the DVM reading 126.8 volts. Very close to the predicted value!

The total capacitance of the capacitor bank is 47uF and the voltage rating is 4000 KV. Under no load conditions, that will give a safety margin close to 50%. I'm contemplating reducing the number of capacitors to 8 in order to save som space. In that case, I would have a safety margin of 18% under no load conditions. I'd be happy to take comments on that!

I have also continued with the metal work of the cabinet. It is almost done, and I will update with pictures in the next few days.

A weeks work on the chassis for the Kilowatt amp project

The first step in building an amplifier is to make the cabinet. The cabinet is made from scrap aluminium sheets and angle stock. The total cost of the aluminium was about 40 Euros. The size 17x15.5x7.5 inches. It took about a week of evening work to complete it.

In the upp right corner you can actually see the very first component fitted! Yes!!! It is a double pole mains switch with EMI filter

Update on the Kilowatt amp project

The most important update is that I in the final moment decided to scrap the double triode project based on GI7B and go for a tetrode design instead. The main reason is that I felt that no one knows about the supply of GI7Bs in the future. It is no doubt a powerful and cheap triode. But at the same time, very few comercial builds use the tube. So I decided to go for the more common tetrode GU74B/4CX800 which is used in many comercial builds. I ordered one from Dr. Alex in Ukraine. The service was excellent and the price fair. The tube arrived only a week after order. Here it is:

The tube arrived in a sealed plastic bag with original documentation in Russian. The document was stamped 1986/01/16. In the same package I also bought a socket and a ceramic chimney:

Contact with FT5GA - Glorioso DX Expedition 4th mot wanted DXCC

I've heard FT5GA on many occasions on 20-17-15, but I have not had the time to chase them. But today I managed to log FT5GA on 17 meter. The Pile up was huge, but I got through after about 15 minutes. I think I was lucky because many strong signal were calling everywhere 5-20 up. I just selected a spot about 7 KHz up where it was empty and suddenly I came through. Now, lets see if I can log them on 20 and 15 too! And, perhaps 12 and 10 opens up?

This is how my "score-card" looks:

Construction of a 2xGI7B Power Amplifier

After spending the first 6 month of this year colleting parts it has become time to start the actual building. This is a mockup to decide the size of the cabinet. The tube in the upper right corner is not the 2xGI7B. I'm still waiting for the delivery. I put in the GS35B since the mockup looked empty without a tube. The transformer delivers 900 volts at 0.8 ampere ICAS, which I calculate will be enough to drive the amplifier to 1 kw in SSB service. With a voltage doubler the DC ouput will be around 2500 volts. The transformer was ordered from a factory in Czech Republic. The plate capacitor measures 22-320 pF and is also specially made for this project. It was orderd from Ham Center in Italy. The tune capacitor is a 1000 pF surplus capacitor. Maybe it is a little bit too big physically. I might find a smaller one for the final design. The next step is to etch the voltage doubler board and assemble the G3SEK Triode Board