6146 HFVVTC
Marko, Sat Dec 03 2011, 12:06AMHi guys
Long time ago I was given some 6146's (actually their 12V version, 6883) and tried building a VTTC from them; due to their low voltage rating they didn't seem very useful though.
After digging them out last days I decided I might at least give a proper funeral to them... so I rigged up a "plasma tweeter" circuit (though with no audiomodulation yet) and a variac-controlled MOT for power supply. I ended up producing some nice HF sparks, 8.5Mhz resonator in the video.
The variac fuse ultimately limited the power throughput due to particularly nasty power factor and saturation of the MOT caused by using a half wave rectifier and a quite big filter capacitor bank.. I replaced the fuse with 10A one and briefly gave the coil some more juice.
I tried a bit longer run (on much of variac's unhappines) and could see the plate of one tube getting dimly red... so the current sharing wasn't so great after all. Input voltage was about 1kV and power quite unknown but assumed to be 300-500W considering the size of the spark. The spark squegs a bit as it's unhappy by antenna being a bit too far away in this video.
Now, some questions:
How much can I overvolt these tubes without blowing them up too fast? I thought to slam 3kV at them at some point just to satisfy my destructive curiosity.
Yet I don't really even know what is the overvoltage failure mode of electron tubes - I know arcs can from between anode and grids but how lethal this is anyway? I'll be using protection resistors on the anode supply in future.
Regarding the current sharing issue, I assume the easiest way to fix this would be providing variable screen biases to each tube (which are currently connected to same supply over 47K resistors)
I'm not sure how would this fare along with audiomodulation (which would involve modulating the screen bias) - I hope I wouldn't end up requiring a separate modulator for each tube?
Cheers,
Marko
Re: 6146 HFVVTC
Mads Barnkob, Sat Dec 03 2011, 08:38AM
That is a lovely little coil, about overvoltage the greatest damage is if an arc burns hole in the grids.
Balancing two tubes can be mysterious and i found the easiest way was to tap the feedback coil rather than giving each tube a grid leak circuit. Take a look here
Mads Barnkob, Sat Dec 03 2011, 08:38AM
That is a lovely little coil, about overvoltage the greatest damage is if an arc burns hole in the grids.
Balancing two tubes can be mysterious and i found the easiest way was to tap the feedback coil rather than giving each tube a grid leak circuit. Take a look here
Re: 6146 HFVVTC
Steve Conner, Sat Dec 03 2011, 11:45AM
Nice work Marko! See, I told you those tubes would withstand a MOT
To stop the saturation, could you reconfigure the power supply as a doubler and feed it 120V?
Steve Conner, Sat Dec 03 2011, 11:45AM
Nice work Marko! See, I told you those tubes would withstand a MOT
To stop the saturation, could you reconfigure the power supply as a doubler and feed it 120V?
Re: 6146 HFVVTC
Marko, Sat Dec 03 2011, 02:18PM
Hi guys
So, a bank of resistors on the plate supply should do well enough to stop the large arc explosions in the tubes? What kind of resistors would you recommend though, I thought of something like a series of several 4.7 ohm 2W carbon film resistors (they have the film in a little helix around them)
Mads: not sure why I didn't think of that before - it should be fairly easy to use a separate antenna for each tube! If I at least had 3 similar telescopic antennas around.
Interesting thing was - I firstly used a small resonator wound with enamelled wire for this coil, and it tended to heat up like hell. The white resonator was apparently PVC insulated wire wound on a PVC form that I had around for ages - I thought it would be useless for a HFVTTC but in reality it tends to stay very cool compared to the first one and produces much better output! (THe first resonator ended up being used as a DC link choke BTW).
Is it true, by the way, that an ideal DC link choke in this application would be an exact copy of resonator itself? What's the theory behind that?
Regarding the power supply issue, the obvious first thing to do is to stop using half wave rectifier (btw, how safe it is to disconnect the mot ground leg from the core and use a bridge rectifier with grounded - ? )
Still if I run the mot offline (with or without some buck transformer) I think the voltage is going to be too high for these tubes anyway.
My small variac is too weak to drive a MOT and I wouldn't be bothered to drag the 50kg one anywhere.
As an alternative choice for power supply I thought about royer powered from few tens of VDC driving a PC PSU transformer in reverse, the output of which feeds a CW multiplier which steps it up to the required few kV.
This would then be problematic due to requirement of a low voltage, high current supply (could use a car battery!).
I could perhaps also run off a single phase while using halfwave multiplier at 50Hz (nasty!)
I've also been thinking about the possibilities of improving the oscillator circuit itself... commonly a the antenna, a 20k pulldown resistor beneath it and the RFC choke are the classic main parts (+ anti-squeg resistor/inductor networks).
The 22ohm anti-squeg resistors I'm using are likely quite inductive which is another problem. It's hard to find carbon composition or metal film resistors of such resistance and wattage here!
Regarding screen biasing, is it better to hold the screens at constant regulated voltage or use reistors in series instead?
I'm thinking to replace the bias resistors with a modulated regulator based on a HOT or a HV N-channel mosfet, which I can then feed audio to.
(would probably be cool to build the modulator entirely with tubes, I'm not sure how to though!)
Marko
Marko, Sat Dec 03 2011, 02:18PM
Hi guys
So, a bank of resistors on the plate supply should do well enough to stop the large arc explosions in the tubes? What kind of resistors would you recommend though, I thought of something like a series of several 4.7 ohm 2W carbon film resistors (they have the film in a little helix around them)
Mads: not sure why I didn't think of that before - it should be fairly easy to use a separate antenna for each tube! If I at least had 3 similar telescopic antennas around.
Interesting thing was - I firstly used a small resonator wound with enamelled wire for this coil, and it tended to heat up like hell. The white resonator was apparently PVC insulated wire wound on a PVC form that I had around for ages - I thought it would be useless for a HFVTTC but in reality it tends to stay very cool compared to the first one and produces much better output! (THe first resonator ended up being used as a DC link choke BTW).
Is it true, by the way, that an ideal DC link choke in this application would be an exact copy of resonator itself? What's the theory behind that?
Regarding the power supply issue, the obvious first thing to do is to stop using half wave rectifier (btw, how safe it is to disconnect the mot ground leg from the core and use a bridge rectifier with grounded - ? )
Still if I run the mot offline (with or without some buck transformer) I think the voltage is going to be too high for these tubes anyway.
My small variac is too weak to drive a MOT and I wouldn't be bothered to drag the 50kg one anywhere.
As an alternative choice for power supply I thought about royer powered from few tens of VDC driving a PC PSU transformer in reverse, the output of which feeds a CW multiplier which steps it up to the required few kV.
This would then be problematic due to requirement of a low voltage, high current supply (could use a car battery!).
I could perhaps also run off a single phase while using halfwave multiplier at 50Hz (nasty!)
I've also been thinking about the possibilities of improving the oscillator circuit itself... commonly a the antenna, a 20k pulldown resistor beneath it and the RFC choke are the classic main parts (+ anti-squeg resistor/inductor networks).
The 22ohm anti-squeg resistors I'm using are likely quite inductive which is another problem. It's hard to find carbon composition or metal film resistors of such resistance and wattage here!
Regarding screen biasing, is it better to hold the screens at constant regulated voltage or use reistors in series instead?
I'm thinking to replace the bias resistors with a modulated regulator based on a HOT or a HV N-channel mosfet, which I can then feed audio to.
(would probably be cool to build the modulator entirely with tubes, I'm not sure how to though!)
Marko
Re: 6146 HFVVTC
Steve Conner, Sun Dec 04 2011, 11:58AM
Yes, a second identical resonator would make a good DC link choke. The argument goes that the system operates roughly at the quarter wave resonant frequency of the coil. So the DC link choke will also be roughly quarter wave resonant. Therefore if you put a short circuit (RF bypass cap) at the power supply end, it will be transformed into an open circuit at the plate end. In other words, the choke will just deliver DC without interfering with the circuit's RF behaviour.
Steve Conner, Sun Dec 04 2011, 11:58AM
Yes, a second identical resonator would make a good DC link choke. The argument goes that the system operates roughly at the quarter wave resonant frequency of the coil. So the DC link choke will also be roughly quarter wave resonant. Therefore if you put a short circuit (RF bypass cap) at the power supply end, it will be transformed into an open circuit at the plate end. In other words, the choke will just deliver DC without interfering with the circuit's RF behaviour.
Print this page