Experiments with 18 GHz TWTs getting 4 Watts on 24 GHz by PAoEHG.
With the 24 GHz designs which are mostly used, the output power is limited to a maximum of around 200 mW. Most active amateurs have around 80 till 100 mW output using two MGF 1303 in parallel.
Very few have more power using dedicated MMIC devices, these devices deliver output power in the region of 700 mW until 1.5 Watt. Unfortunately these devices are very seldom and quite expensive. The future will bring a change and more of these MMIC amplifiers will become available for amateur radio experiments.
Getting more output power is the best possible way to achieve greater distances on this band. The presently used output power until 100 mW is enough to make contacts over 80 km and with ducting up to 300 km. With 10 to 16 dB more power, possibilities arise for rainscatter contacts and troposcatter contacts over greater distances.
When I became active on 24 GHz, I wondered if 18 GHz TWTs could be used for 24 GHz. I got hold of several wideband TWTs with a working range of 10 until 18 GHz. The power range of these tubes where from 200 mW until 20 Watt at the original working frequency range.
The first tube I started my experiments with was a tube manufactured by Teledyne MEC type
MTI 1017. The tube which was complete with power supply, gave around 2 Watt at 18 GHz with more than 50 dB gain. After getting this tube to work on its normal frequency I tried the tube on 24 GHz. I made special arrangements on the power supply to get higher output voltage on the tube.
Normal supply voltages are heater 6.3 V; grid -8 V; anode 1.5 kV; helix 1.9 kV. Changes on the power supply gave a tuning range of the helix voltage until 2 kV and the helix until 2.5 kV. Within these voltages the helix current was still in an acceptable range of maximum 1.5 mA. The grid voltage could be changed as well from almost 0V until -40 V.
Experiments with this set-up with the power meter directly connected to the output SMA connector and about 10 mW of input power took quite some time. After some time I reached a point where 1.5 Watt output came out of the TWT which looked to be on 24 GHz. The output was quite linear to the drive power which gave me the impression that it was working well.
After adding a 24 GHz waveguide on the output with a tuned filter to 24192 MHz the output was gone. With this new set-up I again tried to tune the power supply in a way that I could measure output power on 24 GHz. Unfortunately I was not able to find a point in which the tube was working on 24 GHz.
Most possible explanation for the output power I found was that a spurious product from my 24 GHz transmitter in the frequency range until 18 GHz was fed to the TWT. Due to the high gain of this amplifier ( when peaked with a specific supply voltage more than 70 dB) the spurious product was amplified to the 1.5 Watt which I measured.
After this failure I started with the 20 Watt tube experiments.
This was quite more complex because I did not have a power supply for this tube. After a lucky visit to a surplus market I found an complete TWT amplifier from Hughes. This unit had a TWT with a working frequency until 12 GHz.
Unfortunately the power supply in this amplifier was not working. Quite some time was spend in finding the cause of the power supply failure. Several transistors replacements looked to solve the problem, but then the real cause of the failure was found. The high voltage transformer was defective. A short circuit between two of the high voltage windings indicated that the transformer was destroyed. The whole high voltage unit was sealed in silicon and had to be removed very carefully.
When the transformer was removed it was obvious that all the windings where damaged. Possibly the temperature became much to high which caused the short circuit.
After a long period trying to get the transformer fixed, I found PAoJGF to be able to rewind the transformer for me. When this was ready the power supply was working correctly and it was tested using a dummy load made of many power resistors in serial.
I tried to get the original 12 GHz tube to work with the power supply. This worked very well indicating that the power supply failure was the reason that the amplifier was on the surplus market.
Being certain that the power supply was working correctly I put in the 18 GHz tube. With a dummy power supply load, made with 23 resistors from 1k8 with 5 Watt dissipation, the output voltage was tuned to the desired voltages of the tube. After connecting the tube the anode current was not good and changes where made on the heater supply. The specified heater current was not reached with the specified voltage of 6.3 V. Raising the heater voltage to a level for nominal heater current the anode current was also in the specified region. Driving the tube with a 15 GHz signal it was easy to get output power from it. Then I tried connecting the output of the tube to a waveguide transition without giving drive power to the tube. The noise power from the amplifier was easily monitored on my 24 GHz receiver, with more than a meter distance between the output transition and my receiver.
This indicated that the tube was doing something on 24 GHz. Then I connected the output to my power meter and driving the tube with 24 GHz drive power. With about 10 mW of input drive and after some fine tuning of the power supply voltages a solid 2 .5 Watt output came out of it.
Then I tried to put in the waveguide filter to be certain that the output was on 24 GHz. The difference which I measured was exactly the same as the insertion loss of the filter.
The next problem was to measure more than 2 Watt output power on 24 GHz, because my attenuators where not able to handle more power. For this reason I had to make a coupler and a dummy load.
With the maximum available drive power I had, which was 50 mW, the output power I measured was 4.1 Watt. After a few minutes this power destroyed my home made dummy load which was not able to handle it for a longer period.
My finger in front of the radiating open waveguide was heated that quickly that an impression of real output power could not be missed.
Now the most difficult part has to be done. How to get the TWT operational in my mast mounted station. For this I am working at the moment and I expect it to be operational by the end of this year.
I still want to do some extra experiments with the tube in and output matching which will possibly give some extra gain or output power.
The TWT which I use is from Hughes and originally working as an instrumentation amplifier from 12 to 18 GHz. The type of the tube itself is 848 HD with power supply voltages of : heater 5.9 V (in my set-up the heater voltage must be higher around 7.5 V but with nominal heater current of 700 mA); anode voltage : 50 V (is grid voltage) Cathode voltage : - 3.6 kV; Cathode current 48.8 mA; Helix voltage : 1.8 kV; Max. helix current 6 mA in my tube 1.2 mA.
Hughes TWT 848 HD on testSome time later I did more testing with a similar tube and with more available drive power. With help off a directional coupler I could make a real power measurement. With about 0.2 W drivepower I could manage 12 Watt output, this I used for my EME experiments on 24 GHz
Tips for other experiments with 18 GHz tubes not known to work on 24 GHz.
First of all get the tube operational at its normal working frequency and check output power is correct. Then monitor with a 24 GHz receiver the output noise of the tube without input drive. If you dont hear it then the tube voltages can be changed within the range that the maximum helix current is not exceeded. Some tubes are very tolerant and allow very large changes in helix and cathode voltage. Be certain not to exceed the maximum helix current for long periods, because this is the most possible way to destroy the tube.
If noise output from the tube is clearly present then connect it to the power meter and give small input drive. Be careful not to destroy the attenuator or the power meter. If output is available put in a filter on 24 GHz to check that it is actually on 24 GHz. If radiating from open waveguide, be very careful because the radiation power might be dangerous to your body.
Possible other 18 GHz TWTs can be used on 24 GHz and for this a want to suggest anyone having 18 GHz tubes to try and get it running on 24 GHz. Ruud PE1BTV has a small tube operational from Teledyne-MEC type M5977 which gives 210 mW output with 4 mW input. The original specifications of this tube are 8-18 GHz with G-max of 60 dB and P-Max of 2 Watt. After changes of the voltages on the tube it worked on 24 GHz.
As far as I know now there are several other amateurs which have a TWT operational on 24 GHz. The very challenging period for 24 GHz with more power is started. Lets hope that it will have similar effect as it had on 10 GHz.