Hi Steve I've enjoyed following your pursuit of the GS-35B Amp. To directly address one point you made. As you increase the Bias Voltage you push the Amp closer to Class C. As a result it will become LESS linear and more dirty. It will take more power to drive it and it will increase efficiency. If you want to improve signal quality you should reduce the Bias Voltage as long as you can accept the increase in idling current. I'm not sure what the tuned circuit on the Grid or the cathode will do. My first siggestion would be to try a traditional Pi-Network circuit on the input with a Q between 2 to 5. If that does not produce satisfactory results I would then look in to an RF Negative Feedback Circuit between the Grid and Cathode. Keep in mind this will also increase the drive power requirements. BTW you are doing just fine with the your Plate Voltage supply. The manufacturer of the tube recommends not to exceed 4000 Volts although I have read of EMEers pushing the GS35B as high as 4500 Volts. The tube is rated for 1.4 Amps of Plate Current and the Maximum Voltage on the Plate at that current should be no higher than 3800 Volts. At those numbers running 60% efficiency you will be pushing more than 3 KW out. Overall you are in very good shape Steve. One other point I would suggest is to lose the variac on the Plate Supply Transformer. It is another core that has to be energized and will adversely affect voltage regulation. It would be an interesting test to see what the unloaded and loaded plate voltages are with and without the variac. Your new transformer seems to be doing a good job on its own. 73s and Best to you in the New Year Bob K6TR
Hi Bob, thanks for the comments. I have put the amplifier back to grounded grid mode and made a couple of other changes that now give me a nice linear sinewave output. There is a later video showing this. I have had 5kV on the anode with no idle current and it withstood that with no problems. I think I may have to use a slightly higher anode voltage to get my 1500W output. The mains voltage here is only 234V so I have to wind the variac up to 110% to get 1500W output. The Goldilocks voltage sits between the voltages I can get from the transformers I have. With the variac I can see precisely what Goldilocks voltage is required... it's about 3800V under load for 1500W output.
Bob you point out some good staff,but he want to avoid input tuned circuit,aka input low pass filter...but due to reactances and capacitance from driver interact with cathode,and linearity suffer...and amplifier is harder to drive,so its standard that input must be tuned ,or pre tuned with low pass filters for each band,like in all commercial amplifier...to cancel inductive ractance and capacitiv reactance,plus driving koax,low pass filters must be close to cathode...negative feedback resistor in series with input can improve linearity,but is not practical when all 100w drive is needed..it is useful when drive hi mu triode like 3CX1500A7..only 60w of drive power is needed for same power like the GS35B medium mu triode...so you safely drive 3CX1500A7 with 100w with increased linearity....you dont want to destroy 1500 US dollars EIMAC tube,,,...negative feedback is useful when using tetrodes or pentodes passive grid driven to linearize them ,and lower gain ,so no need for neutralization,5 to 20 ohm resistor from cathode to ground....yep,increased resting current improve linearity,easier to drive..also lower sag on anode voltage....sure ,there are also other staff to consider,output tank q factor......regards
@@vk6cs456 if you not have any idling current,you run pure class B...so tube not work linear,distortion is present,you must have around 150 to 200 ma idling current to be in CLASS ab1...that will eat some voltage drop...anode voltage stability is essential for GG linears to be more linear....
Thanks Steve for sticking with the old girl, really interesting what your doing . Just a suggestion , but why not ask Mr BBI in the good old US of A if he's had any experience of this phenomenon ? After all he pumps some wind out of his amps........ Regards Carl.
Nope I've never seen a proper spectrum display of the output of those mega watt solid state amps, but he dose WHISTLE a pretty tune down the microphone 😚 . Regards Carl.
No spectrum or scope displays. He tells us they're good though... None of these guys show anode current or grid current or any waveforms, just output power. All of the interesting stuff is missing.
Great to see that you are more concerned about the linearity of the output, rather than the total power! Well done!
Maybe I'm in the minority... but yes, I do care about the technical quality of my station's output.
Interesting to see the progress and learn about how Tube amps really work.
Thanks Ade. I'll upload another video to show the changes I made.
Std By.
Hi Steve I've enjoyed following your pursuit of the GS-35B Amp. To directly address one point you made. As you increase the Bias Voltage you push the Amp closer to Class C. As a result it will become LESS linear and more dirty. It will take more power to drive it and it will increase efficiency. If you want to improve signal quality you should reduce the Bias Voltage as long as you can accept the increase in idling current. I'm not sure what the tuned circuit on the Grid or the cathode will do. My first siggestion would be to try a traditional Pi-Network circuit on the input with a Q between 2 to 5. If that does not produce satisfactory results I would then look in to an RF Negative Feedback Circuit between the Grid and Cathode. Keep in mind this will also increase the drive power requirements. BTW you are doing just fine with the your Plate Voltage supply. The manufacturer of the tube recommends not to exceed 4000 Volts although I have read of EMEers pushing the GS35B as high as 4500 Volts. The tube is rated for 1.4 Amps of Plate Current and the Maximum Voltage on the Plate at that current should be no higher than 3800 Volts. At those numbers running 60% efficiency you will be pushing more than 3 KW out. Overall you are in very good shape Steve. One other point I would suggest is to lose the variac on the Plate Supply Transformer. It is another core that has to be energized and will adversely affect voltage regulation. It would be an interesting test to see what the unloaded and loaded plate voltages are with and without the variac. Your new transformer seems to be doing a good job on its own. 73s and Best to you in the New Year Bob K6TR
Hi Bob, thanks for the comments. I have put the amplifier back to grounded grid mode and made a couple of other changes that now give me a nice linear sinewave output. There is a later video showing this. I have had 5kV on the anode with no idle current and it withstood that with no problems. I think I may have to use a slightly higher anode voltage to get my 1500W output. The mains voltage here is only 234V so I have to wind the variac up to 110% to get 1500W output. The Goldilocks voltage sits between the voltages I can get from the transformers I have. With the variac I can see precisely what Goldilocks voltage is required... it's about 3800V under load for 1500W output.
Bob you point out some good staff,but he want to avoid input tuned circuit,aka input low pass filter...but due to reactances and capacitance from driver interact with cathode,and linearity suffer...and amplifier is harder to drive,so its standard that input must be tuned ,or pre tuned with low pass filters for each band,like in all commercial amplifier...to cancel inductive ractance and capacitiv reactance,plus driving koax,low pass filters must be close to cathode...negative feedback resistor in series with input can improve linearity,but is not practical when all 100w drive is needed..it is useful when drive hi mu triode like 3CX1500A7..only 60w of drive power is needed for same power like the GS35B medium mu triode...so you safely drive 3CX1500A7 with 100w with increased linearity....you dont want to destroy 1500 US dollars EIMAC tube,,,...negative feedback is useful when using tetrodes or pentodes passive grid driven to linearize them ,and lower gain ,so no need for neutralization,5 to 20 ohm resistor from cathode to ground....yep,increased resting current improve linearity,easier to drive..also lower sag on anode voltage....sure ,there are also other staff to consider,output tank q factor......regards
@@vk6cs456 if you not have any idling current,you run pure class B...so tube not work linear,distortion is present,you must have around 150 to 200 ma idling current to be in CLASS ab1...that will eat some voltage drop...anode voltage stability is essential for GG linears to be more linear....
@@vk6cs456 5KV with no idle current? Don´t expect any linearity!
looks like the output tank do not have enough Q to smooth the some wave from the tube since it operates at half wave amplification... 73 IU3IRR
Hmmm... I had put a line through the problem being output tank related. But I'll calculate the Q, thanks for the suggestion. 73.
Seems as if this would be a better tube for digital modes than for AM.
Thanks Steve for sticking with the old girl, really interesting what your doing .
Just a suggestion , but why not ask Mr BBI in the good old US of A if he's had any experience of this phenomenon ?
After all he pumps some wind out of his amps........
Regards Carl.
He does, but he never shows output waveforms does he...
Nope I've never seen a proper spectrum display of the output of those mega watt solid state amps, but he dose WHISTLE a pretty tune down the microphone 😚 .
Regards Carl.
No spectrum or scope displays. He tells us they're good though... None of these guys show anode current or grid current or any waveforms, just output power. All of the interesting stuff is missing.
Yes... that would be interesting for sure.
@@vk6cs456 they are CBers,they are not interested in linearity,only more juice LOL
Your output pi-net might be totally wrong, at least it looks very strange to me. Linearity will be better cathode driven.