tube theremin power supply

[charbot]

I apologize for the OT post, but I need some advice and this is by far the most helpful diy/music stuff forum Ive found.  I'm building a Doug forbes tube theremin:     http://www.thereminworld.com/files/Projects/Forbes/theremin.gif
...or rather  rebuilding   one.  I guy I work with gave me one that he had tried to build about 10 years ago.  He didn't have much luck with it and he gave it to someone.  Some how it made its way back to him and was being stored on an old tugboat.  He need to clear out his stuff and gave it to me under the conditions that I would not give it back.  good deal, right?

anyway, it was sort of  a mess; so I made a new chassis, took apart and desoldered everything, and rebuilt it in the new box. Heres the problem:  The heaters are getting power and the tubes glow but the oscillators don't get power because  the 1 K resistor in the power supply circuit gets fried every time.  The rectifier tube appears to work and puts out 283v.(pin 1)
I have a Thordarson power transformer , sec#1 250v @70 and with no load measures about 283v. Thats about 10% above the  rating, not great, but not unusual(from what I understand.)  I checked the wiring repeated for hours last night, so I'm pretty sure everything is hooked up properly. The resistor that burns out is  1K 1watt.  The only difference between  the schematic and mine is I am using 22uf 350volt filter caps instead of the listed 20uf @ 300v.  Could this be it?
Doug forbes' site is no longer up and theres not a whole lot of info out there on this one, except: http://mrr3000gt.mystarband.net/MRT/index.htm
this guy also had trouble with the power supply also and there is only one topic on the doug Forbes theremin in the theremin world forums.  Someone there mentioned only being able to get the heater circuit to work but he also had the wrong oscillators or something.   Others have gotten this circuit to work, so I'm sure that problem is either with me, my parts, or both.

If anyone has built one of the things, has any suggestions/ solutions, or can direct me to where I learn more, I would be very appreciative. Thanks in advance!
.   

[charbot]

a few more things:
heres a link to the text that goes with the schematic. http://www.thereminworld.com/files/Projects/Forbes/theremin.txt

The guy who gave it to me said that he thought that he had trouble with the power supply, and he believes that he did it originally with out the 6x4 tube and used a bridge rectifier... when that didnt work he tried the tube.   I suspect that the transformer may be the problem...I was looking at the data for the 6x4 and it says the max current should be 70ma.   That is what my transformer's secondary is rated.  Is this cutting it too close?   How about a new transformer, like 250v 60ma?   This is the only thing Ive come up with so far

[charbot]

ok,  this might be it.     I found this page about building tube power supplies. http://members.nuvox.net/~zt.robgrow/circuits/powersupply/regpowersupply.html
 
Should I be using a 5 w resistor?  is that it?   shoud I get 5w zeners to replace the 1w ones?   I hope thats it.

[R.G.]

What voltage do you get across the zener string before the 1K burns out? Is it the intended 246V or is it much lower? And what is the voltage across the 1K resistor before it burns out?

If the "246v" is very low, something is shorted or in error in the circuit, not the 1K resistor or the rectifiers.

Open the connection at "A", and measure voltage again, both across zeners and across 1K resistor. Did the voltages change? How much?

This info will let us separate the power supply as a problem from the circuit as a problem.

[charbot]

Usually by the time i pick up the meter, the resistor is already smoking, so I havent yet checked the voltage across the zeners.  Maybe Ill get the meter dmm ready and get my girlfriend to turn it on.   Ill post the results  soon. thanks.

[R.G.]

If it happens that fast, you have something badly wrong.

Before you do anything else,
(1) Leave the line cord unplugged from the wall.
(2) Hook the negative lead of a 9V battery to ground
(3) Hook the positive lead of the battery to pin 7 of the rectifier tube.
(4) Measure the voltage to ground on both sides of the 1K resistor and report that back.

I suspect that the zeners are installed backwards. This needs to be eliminated before you do anything else. If the zeners are installed correctly, the voltage will be 9V on both sides of the 1K resistor. If the are backwards, one side will read close to the battery voltage and the other will read about 2 to 2.5V. You also need to be sure that all of the zeners are installed in the same direction, like the schematic shows. One of the backwards or shorted will cause problems as well.

[charbot]

ok, I disconnected the A, clipped on the tester, and tried it again.  the resistor blows as soon after about 15 secs, as the rectifier tube warms up.  I can detect no voltage through the zeners or any where after the resistor, before it gets fried.  before the resistor I measure 382v,( I think that I posted earlier that it was 283v)...
to sum it up... it has the resistor burn up problem with and without A connected.  and Im out of 1k resistors!

[charbot]

  I have the zeners in a chain , all silver stripes are oriented toward the power.

[jrc4558]

Isn't that what the girlfriends are for, anyway? 
You probably have a short between ground and point A. (or a very low resistance path).  :icon_exclaim:WITH POWER OFF, measure the resistance between point A and ground.

[R.G.]

I can detect no voltage through the zeners or any where after the resistor

And we have a winner! If one side of the 1K has voltage and the other doesn't, and the A connection is open, that leaves only ... the zeners.

They're either shorted, possibly shorted around by poor wiring, or installed backwards.

and Im out of 1k resistors!

No problem. What kinda resistors you got? Two 2K's in parallel are 1K, as are four 4.7K's (about...) and frankly, there is no magic associated with "1K" if the circuit is not supplying anything but zener diodes.

So do the following:
(1) CHECK THE WIRING OF THE ZENER DIODES. Find out if they have a dead short around one or more of them by using the ohmmeter section of your DMM, and also check their direction with the diodes range of your DMM's ohmmeter setting. They will conduct one way, but not the other if they are correct. The band on the package is equal to the bar on the schematic. Don't turn it back on until you're sure it's wired 100% to the schematic and have personally ohmed out the zeners and their wiring.
(2) Leaving the AC cord unplugged, try the 9V battery trick on the circuit when you think it's perfect. If you read battery voltage on each side of the 1K (well, it's gonna be 2K, or 3.3K, or whatever you have) then and only then are you ready to try it on the AC line again.
(3) Once you get battery voltage each side, remove the battery, leave connection A unconnected and connect your DMM to the circuit at the zeners. Blip the switch on and off immediately. Did the DMM meter reading change? Blip it on and off a little longer. Walk your way into seeing if the voltage across the zeners ever comes up. If it has passed the battery test, only some issue with one of the zeners will foul you up then.

[charbot]

 thanks,  Looks like I have an evening of testing awaiting me. I meant that I ran out of 1w resistors, but Ive found a few more.  do I even need 1 w anyway?   How is the required resistor wattage determined?  if watts is voltage x current, wouldnt I need a resistor in the 15-20w range?  Sorry if this is such a  simple question.
Because I learning by myself, I guess there are a few gaps in my understanding. Thanks again, I really appreciate the help.  Ill post back after I try your suggestions

[charbot]

ok, I think that my zeners are fried.   They measure any where from .5- 1.5  ohms ....both directions!...I guess I could have cooked them while soldering, but could something else have done this?
Can anyone direct me to somewhere that explains, on a basic level, how the zener shunt works... I think that Im getting more confused.

[R.G.]

All diodes have a reverse voltage breakdown limit. Up to that voltage, they keep current from flowing. Above that limit, they let current through, but keep the same voltage across the diode. It's like their reverse voltage is a bucket that keeps the electricity in until the level gets over the edge, then all the electricity flows out over the bucket edge. The level (voltage) of water doesn't get any higher, but lots of water can flow over the edge.

As long as the current going through the diodes in the reverse direction is not so big that the current times the reverse voltage that they're holding off (that is, the power they have to dissipate, voltage times current) doesn't burn them up, they can run that way forever. So the way to use zeners (i.e., reverse breakdown diodes) is to use a voltage supply that's bigger than their voltage, plus some current limiting part between them and the voltage. The voltage supply gets them high enough to break over and the current limiter keeps them from dying.

If I'm reading your schemo right, the zeners are rated at 82V, 1W. That means they have a max current of 1W/82V = 12ma. If you get more than 12ma through them, they WILL burn out.

If the power supply is operating the way it's supposed to, the circuit uses some current through the 1K resistor, and the zeners use the rest. We can guess what the currents are: the power filter caps are rated at 300V, so it's less than that, and the zeners come up to 246V, so the voltage across the 1K resistor must be intended to be less than 54V. If that's the case, the current through it is less than 54ma. And there's a problem. Whatever current that the circuit does not eat, the zeners have to eat. So if you have about 50ma of current available, the circuit does not use at least 38ma, then the zeners eat more than 12ma, and they fail shorted. When that happens, the 1K resistor now has the full power supply across it and it starts burning out. Next to die would be your rectifier tube.

Things like not having the tubes in the sockets when you turn it on would cause the little horror story I just recounted. Bottom line? It's a bad power supply design. It's not forgiving of any error.

Here are your options.
(a) get rid of the zeners and put in a bigger resistor to drop the voltage at A to 246V under load.  Works, but clumsy.
(b) Add a power MOSFET to the circuit to change the zener shunt regulator setup into a series regulator. This does not have the sudden-death failure modes and should materially improve the performance of the circuit because it will have a lower output impedance than the zeners by themselves.
(c) If you are squeamish about a silicon MOSFET power regulator because you're overcome with the "no silicon inside" stuff from the theremin web pages, you could do much the same with a tube, but the performance would be worse.

I would add a "D" option as well. I personally would split the power supply after the choke into two separately decoupled paths, one for the volume side and one for the pitch side. Theremins are notorious for having problems with pulling into synchronism, and splitting the power path will materially improve the low frequencies that you can get out without pulling.

Did I mention that I've built a few theremins? 8-)

[charbot]

too bad I didnt read your latest post before and trying it again with new diodes.  yep, they burned out again and then the resistor went... just like you said.     
I am goning to go with option "c". I dont care what's insidem, as long as it works! After it works (if it works) ill try  splitting the pitch and volume power paths like you suggested.   
could I use something like this?
http://members.nuvox.net/~zt.robgrow/circuits/powersupply/regpowersupply.html
I dont feel confident enough to whip something up myself, so I need to find a proven design before heading back to the electronics store. I think that Ill build the power supply on perf board so it will be easier to experiment with.
once again, Thank you, R.G.
because you have taken the time to help me out and explain things, this might actually become a working theremin, and not just a box of junk that gets passed along to some other fool. 

[charbot]

how's this:


Im using the forbes power supply up to the choke, and used the regulator from the link above.
R1 will be about 1k and R2 will be about 10K.  the extra filter cap after R2 will be  22uf 300v , like the others. The zener diodes voltage will be about 246v.
the text on the link reccomends using 5 watt resistors and zener diodes for extra precaution. Is that necessary?