Tuning the vacuum tube oscillator / Tesla Coils / Forums

Forums

4hv.org :: Forums :: Tesla Coils

« Previous topic | Next topic »

Tuning the vacuum tube oscillator

1 2 next

Move Thread

LAN_403

Dr. Dark Current

Sat Oct 31 2009, 06:23PM

Registered Member #152 Joined: Sun Feb 12 2006, 03:36PM
Location: Czech Rep.
Posts: 3384

Hi, I recently got two PL509 horizontal output pentode tubes (40W max. on the anode, 1.4A anode peak current, anode peak voltage max. 8kV, data

), and was wondering how to get the most out of them - in other words, which circuit to use and how to tune it for the best output power/anode loss ratio.

I'd like to use them on a TV flyback to power a "retro" Jacob's ladder.

I plan to power the anode from a voltage quadrupler (ca. 1200V DC).

As far as I understand, there are two circuit options (G3 always connected to cathode):
1) Tie G1 and G2 together and connect them through a parallel RC network to the feedback coil. The resistor needs to be a high-power type.
2)Connect G2 through a high-power resistor to anode supply and input the feedback through a parallel RC network to G1. The feedback resistor doesn't need to be physically large.

Which circuit would be more efficient?
In both cases, would it be benefical to put any "bias" supply between the feedback coil cold point and ground/cathode?
Any other elements to put in the circuit to reach better efficiency?
How much peak-to-peak voltage should the feedback coil give?
How to calculate the RC feedback network?
What can I use G3 for when not connected to cathode?
I'd like to produce the "buzzing arc" effect - can I modulate the tube e.g. with a little mains transformer and if so, where should I put it?

This is the first time I'm doing something with tubes, so sorry for so much questions

Myke

Sat Oct 31 2009, 08:40PM

Registered Member #540 Joined: Mon Feb 19 2007, 07:49PM
Location: MIT
Posts: 969

I think it would be best to use something as an oscillator tube and something else as a small buffer to drive the grid. In total, two to 3 tubes. The tubes that work well for oscillators are the 6AU6 and the 6BE6. There are many other tubes that work well for oscillators but these are the common ones I know of. I don't know the configuration but you could use a R/C network to get lower freqs and an LC network for higher freqs. I used the clapp oscillator topology before and it worked well.
You can either modulate the oscillator tube though the screen grid (3rd grid in the 6BE6) or you can modulate the screen grid of the power tube. You can use a transformer for the modulation or just a DC blocking cap.

Proud Mary

Sat Oct 31 2009, 08:55PM

Registered Member #543 Joined: Tue Feb 20 2007, 04:26PM
Location: UK
Posts: 4992

Excellent stuff and very sensible too, Mr KV.

If you're a serious valve/tube player, I'll help you every inch of the way, so long as you remember your debt to this site, when it comes to the Grand Christmas Appeal.

Dave Marshall is the state-side Treasurer for New World contributions.

So your first concrete ally in this evacuated world is PL509 data sheet.

But before I go on, I should point out that I have boxes and boxes of sweep tubes
two or three times more powerful than PL509, which could be yours
for a few dollars (plus postage) contributed to the 4HV.org.

Dr. Dark Current

Sat Oct 31 2009, 11:00PM

Registered Member #152 Joined: Sun Feb 12 2006, 03:36PM
Location: Czech Rep.
Posts: 3384

Hi Harry

Well I'm not sure I understand everything you're trying to say (English is not my native tongue), but - My main field is power semiconductors, I just wanted to learn a bit about operation of vacuum tubes in power oscillators and how to tune them for optimum performance (if it is VTTC, flyback driver or anything else), and how does changing the values of components affect operation. I believe the goal here is to get the oscillator work in Class C ?

What inspired me into this is that I found a pack of vacuum tubes on a local auction server for almost nothing. They were mostly some small (signal?) tubes for which I don't have any use, I took it because of the PL's and just want to experiment a little.

I'm not planning to get a lot of tubes, I just liked the idea of a tube powered Jacob's ladder "because it would look cool"

As for your offer, I'm afraid the shipping of such fragile items to central Europe would be Huge, but maybe I'm mistaken?

Thanks

-J.M.

Mads Barnkob

Sun Nov 01 2009, 07:54AM

Registered Member #1403 Joined: Tue Mar 18 2008, 06:05PM
Location: Denmark, Odense C
Posts: 1968

Steve Ward got a excelent VTTC faq on his site that you should take a look at, it was a great help when I built my first VTTC.

I have also made a single PL36 (under half the power of a PL509) VTTC a while back, never got to document it, I used a mot for the high voltage, but at full power it would turn red in seconds, it could run stable with a cool plate when giving out 1+ cm brush discharge.

It was very crudely tuned with changing tank capactitor value, so it was very far from optimized.

I got a schematic somewhere, ill find it if you want to take a look :)

pl36 vttc at 2300VAC in, 2 second run or so.

Proud Mary

Sun Nov 01 2009, 08:08AM

Registered Member #543 Joined: Tue Feb 20 2007, 04:26PM
Location: UK
Posts: 4992

Hello JM/Herr Kilovolt.

I now have three Thermionic ships in tow (think: deep sea tug!) so there is no problem adding you to the list! Welcome aboard!

CZ is now part of the EU, so postage costs are not important. The main thing is for you to give whatever you can to the 4HV Christmas Appeal - if a service is worth having, it is worth paying for. Even one single American dollar is one step away from having the 4HV site fouled by Viagra advertisements.

Step One - The Variable Frequency Local Oscillator(VFO) (RF pentode)

Step Two - the Buffer-Amplifier (low-mu HF tetrode)

Step Three: The Power Amplifier. (beam output tetrode/s - e.g. 2*PL509 or more.)

Please remember that it takes a long time to think through and type out my opinions and answers.

The best way of proceeding further is for you to send me your circuit diagram for a 500kHz VFO using an RF pentode.
It doesn't matter very much if you make mistakes, so long as you give it your best shot, lad. (So none of this rubbish about connecting g1, g2, and g3 in parallel!!!)

Lastly, I use modern technical English. For example only, when I write 3K3 I mean three thousand three hundred ohms. 3M3 means three million three hundred thousand ohms, and so on)

So that's enough from me, lad. If you want to start with my Variable Frequency Oscillator (VFO) group please tell me soon. There are no fees nor charges beyond (in your case) EU postal charges.

your pal,

Cap'n Christopher D. Eng. etc

Steve Conner

Sun Nov 01 2009, 09:49PM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

I'll try to answer Dr. kV's questions in some random order.

The problem with TV sweep tubes is that they need a screen voltage much lower than the plate. Driving the screen from the plate supply needs a great big resistor and wastes a lot of power. You're also tempted to decrease the resistor, because this dramatically increases the power output, but at the risk of melting the screen grid.

Tying the screen grid and control grid together isn't such a good idea either, because the control grid will overheat before the screen grid gets enough drive. The control grid in TV sweep tubes isn't designed to handle any significant power: to get high transconductance it's made of fine wires and spaced very close to the large red-hot cathode, so it already runs dangerously hot.

But I've seen VTTCs built successfully with both methods. The best way is a separate screen supply, though.

You can get the buzzing arc by powering the whole thing off unsmoothed DC. If you need to smooth the plate supply because you're using a voltage multiplier, then take an unsmoothed screen supply from a separate transformer, and the output will be modulated just as if you'd modulated the plate. Or maybe you can figure out how to tap halfwave unsmoothed supply from the first stage of your multiplier.

With a separate screen supply you must avoid turning it on when there's no plate voltage present. If the plate isn't able to collect electrons, the screen will get them all and burn out. If you think this might happen, then you need to limit the current somehow.

A negative bias voltage on the control grid is nice, but not essential. You can operate the tube in self-bias mode where the grid current generates the required bias. This mode is more risky, though, because the tube will conduct fully and melt down if the oscillations stop for any reason. The control grid doesn't need any great amount of drive, maybe about 100V peak-peak and 50V of DC bias. So it would swing between 0V and -100V. Remember these tubes were engineered to be easy to drive for cost savings in TVs.

The separate VFO trick works, and I've seen it done, but at the end of the day, feedback oscillation makes for a simpler circuit to build, if a harder one to analyze. For the low frequencies used by flybacks, you could use an astable multivibrator made from an ECC83, and that could probably drive the sweep tubes directly.

Did you see my vacuum tube plasma globe?

Dr. Dark Current

Sun Nov 01 2009, 10:27PM

Registered Member #152 Joined: Sun Feb 12 2006, 03:36PM
Location: Czech Rep.
Posts: 3384

Thanks Steve, that cleared some things up a bit for me. But I'm still not sure about the grid feedback network - the capacitor just blocks DC, or does something else too?

Proud Mary

Mon Nov 02 2009, 03:03AM

Registered Member #543 Joined: Tue Feb 20 2007, 04:26PM
Location: UK
Posts: 4992

Wouldn't it make more sense to listen to women to whom thermionics is second nature?

So let's see a rough circuit diagram and component values which I'd sketch out
in three minutes for you on a restaurant table cloth?

Dr. Dark Current

Mon Nov 02 2009, 09:22AM

Registered Member #152 Joined: Sun Feb 12 2006, 03:36PM
Location: Czech Rep.
Posts: 3384

Okay so I have been thinking about the vacuum tube flyback driver circuit yesterday and combining various circuits together, and this is the result.

[edit] changed g2 resistor value
[IMPORTANT EDIT] The capacitors must be rated at 400V, not 200!

1 2 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob