# voltage regulator tubes

#### rkoonce

Can someone explain the procedure to calculate the series resistor for a VR150 tube feeding a preamp using about 15 mA. I'd like to run the VR tube at about 22 mA. My raw power supply will be around 190-200 volts.

Thx!

#### v4lve lover

The VR150 or 0D3 tube needs about 180 volts to start up, so a 190-200V supply is cutting it close. In any case. Use ohms law.

Operating the tube at 22mA and a suppy voltage of 200V

(200-150V)/(0.022+0.015) =1.35K

power dissipated will be 50*0.037=1.8W take a 7W resistor.

Beware that these tubes where built to provide reasonably well regulated supplies over a wide range of currents (5-40mA) and that the exact voltage varies somewhat between samples. These tubes can regulate anywhere between 145-155V depending on current .
Also beware that you cannot decouple the tube with more than 100nF of capacitance or it might turn into a relaxation oscillator.

Last edited:

#### Eli Duttman

The VR150 or 0D3 tube needs about 180 volts to start up, so a 190-200V supply is cutting it close. In any case. Use ohms law.

Operating the tube at 22mA and a suppy voltage of 200V

(200-150V)/(0.022+0.015) =1.35K

power dissipated will be 50*0.037=1.8W take a 7W resistor.

Beware that these tubes where built to provide reasonably well regulated supplies over a wide range of currents (5-40mA) and that the exact voltage varies somewhat between samples. These tubes can regulate anywhere between 145-155V depending on current .
Also beware that you cannot decouple the tube with more than 100nF of capacitance or it might turn into a relaxation oscillator.

I'm cautious. 0.68 μF., for noise suppression, is about as big as I'll go. Getting a relaxation oscillator, instead of a shunt regulator, is most vexing.

Perhaps the OP can source a VR150 that contains a bit of radioactive material. Those specimens were developed to ensure the tubes would reliably "strike", in the dark.

Last edited:

#### GoatGuy

Me personally? I'd put a 5 kΩ resistor on that regulator and not use its regulated voltage directly. Instead, run it to the base of a HV transistor in series with the HV leg, to deliver around 0.9 volts less than the reg-tube nearly-constant-current stable volts… Nice mid-to-large value capacitor on the emitter and you have a cheaper'n'dirt regulated HV supply that will last a lifetime (lower reg current.) Just saying… remember that power supplies are just power supplies once they're bulked up enough to have fat reservoir capacity and darn low surge impedance.

[goat hides behind a tree, expecting fusilage of eggs, tomatoes and last year's sushi…]

GoatGuy

#### tomchr

Paid Member
I'm cautious. 0.68 μF., for noise suppression, is about as big as I'll go. Getting a relaxation oscillator, instead of a shunt regulator, is most vexing.

LOL.

[goat hides behind a tree, expecting fusilage of eggs, tomatoes and last year's sushi…]

So let me take my share of the sh*tstorm. A zener stack with an emitter/source follower or a Maida regulator would provide a much more stable voltage than any VR tube. Just saying...

If the OP insists on using a VR tube, I suggest feeding it with a CCS. The IXYS 10M45S could be used. That would make it possible to dissipate the power into a heat sink, so bulky resistors can be avoided.

Tom

#### pmillett

I'd just use a 1.5k 5W wirewound (or the closest thing I had available) and be done with it. The exact current though the VR tube is not that important. Who cares if it's 146V or 152V? Why make it overly complicated?

#### vega65

as for cap size over the vr tube, and zout:
kolbrek.hoyttalerdesign.no/index.php/tubes/testing-power-supplies

#### rkoonce

I'd just use a 1.5k 5W wirewound (or the closest thing I had available) and be done with it. The exact current though the VR tube is not that important. Who cares if it's 146V or 152V? Why make it overly complicated?

Thanks Pete. That's where I'm headed. I just wanted some validation of my calculations.

#### Enzo

Also, look up some regulator gas tubes in the RCA manual. Whichever one is first - 0A2 maybe? - has a little tutorial on designing for the things.

Zeners do not glow pretty purple as a rule, making them less desirable than gas regulators.

#### rickmcinnis

VR tubes used as a series regulator?

My curiosity has the best of me and I hope one of you who know how these things work will answer my question.

We always see these used used as shunts.

Is there any way to use them as a series element? The cathode connected to the plate of a regular tube?

If using the VR75, as an example, would it simply remove approx. 75 volts of the voltage needed on the plate and pass the difference along? Or would it fail to operate at all? Or only pass 75 volts?

The DECWARE fellow is using these in a new version of his venerable SV83 amplifier and he says something about using the tube's gaseous separation of cathode and anode as a "galvanic isolator" from the worst noises of the power supply - which makes a certain kind of sense yet I have never seen them used this way nor read of them being used this way. I do not think he said galvanic isolator - I used the term to describe what I think he is alluding to. I am speculating since I do not see how used as I have always seen them used they would do any isolation at all.

Is such a thing even possible? Or are these strictly limited to be used as shunts to ground?

Any thoughts will be greatly appreciated and a hearty guffaw, or three, would be perfectly acceptable.

#### v4lve lover

Me personally? I'd put a 5 kΩ resistor on that regulator and not use its regulated voltage directly. Instead, run it to the base of a HV transistor in series with the HV leg, to deliver around 0.9 volts less than the reg-tube nearly-constant-current stable volts… Nice mid-to-large value capacitor on the emitter and you have a cheaper'n'dirt regulated HV supply that will last a lifetime (lower reg current.) Just saying… remember that power supplies are just power supplies once they're bulked up enough to have fat reservoir capacity and darn low surge impedance.

[goat hides behind a tree, expecting fusilage of eggs, tomatoes and last year's sushi…]

GoatGuy

Actually a pretty decent idea, TIP50 is the cheapest NPN i can think of that has decent SOA for the job.

You could take a high transconductance pentode or triode to buffer the voltage, if you take a TV tube with good H-K isolation you dont need another supply. The benefits of that is that you can RC filter the G1 voltage.

#### Wavebourn

You could take a high transconductance pentode or triode to buffer the voltage, if you take a TV tube with good H-K isolation you dont need another supply. The benefits of that is that you can RC filter the G1 voltage.

You can take even higher transconductance MOSFET and use it as a source follower, feeding well filtered by RC voltage from the VR tube. It is what I do in my stellar quiet preamps.

#### Magz

I'm cautious. 0.68 μF., for noise suppression, is about as big as I'll go. Getting a relaxation oscillator, instead of a shunt regulator, is most vexing.

Perhaps the OP can source a VR150 that contains a bit of radioactive material. Those specimens were developed to ensure the tubes would reliably "strike", in the dark.

.I assume you mean .068uF, correct?

#### kd1yt

My curiosity has the best of me and I hope one of you who know how these things work will answer my question.

We always see these used used as shunts.

Is there any way to use them as a series element? The cathode connected to the plate of a regular tube?

If using the VR75, as an example, would it simply remove approx. 75 volts of the voltage needed on the plate and pass the difference along? Or would it fail to operate at all? Or only pass 75 volts?

The DECWARE fellow is using these in a new version of his venerable SV83 amplifier and he says something about using the tube's gaseous separation of cathode and anode as a "galvanic isolator" from the worst noises of the power supply - which makes a certain kind of sense yet I have never seen them used this way nor read of them being used this way. I do not think he said galvanic isolator - I used the term to describe what I think he is alluding to. I am speculating since I do not see how used as I have always seen them used they would do any isolation at all.

Is such a thing even possible? Or are these strictly limited to be used as shunts to ground?

Any thoughts will be greatly appreciated and a hearty guffaw, or three, would be perfectly acceptable.

They need a certain minimum current flowing through them to regulate well, and also have a maximum current flow that they can tolerate. A changing series-only load would seem hard to match to either one of those requirements. "Galvanic" sounds like BS in this context, and I am understating my reaction in that regard.

#### dgta

Hahaha, yes, galvanic isolation by means of a device that conducts is beyond BS.

#### jan.didden

Paid Member
A tube used as a series regulator needs to have at least three electrodes - input, output and control electrode. Triodes and up can and often are used as series regulators.

I agree with you that the 'isolation' argument doesn't hold water.

Jan

#### Lampie519

These tubes need a "ignite" voltage higher then their working voltage. If the load will not draw enough current at warming up (and it will not as no voltage will be offered) it will not ignite. You could bypass the tube with a resistor to make sure there will be some current flow to overcome this issue.
In case the tube ignites it will then drop the voltage (75V in case of a VR75 etc.) but there will be no galvanic insolation as the gass will conduct.

Servicing the Leslie 147 Amplifier - Benton Electronics

#### Attachments

• 147regulator.gif
17.7 KB · Views: 269
Last edited:

#### NickKUK

I looked at implementing an SMPS in tube form.

It is possible to create a tube SMPS. Isolation-wise the tube has some but you need a tube that will cope with mains transient spikes without arcing over. It can then regulate quite happily (6as7 was originally a pass regulator IIRC) otherwise.

There are very good tube regulators using a tube for both in-series and shunted that get into the mV.. however into the uV is something that would need something special or solid state.

#### rickmcinnis

This is not something I intend to do - I am just trying to understand what Mr. Deckert is doing.

Herb Reichert says the thing sounds good - and Reichert always makes it plain he would prefer "good sound" over "accurate" sound. I like something in between. Whatever that may be ...

I do understand how typical regulators work in series or shunt - but was simply wondering if one of these things would work at all with the anode attached to a voltage source and the cathode attached to a plate. In this case an EL84. No grounding other than the cathode of the EL84 or any tube for that matter.

Of course, enough voltage would be applied to the anode. I was wondering if the thing would conduct the voltage at all in a series connection AND if you supplied it with 250 volts would 175 volts be the result or would a VR75 still only allow 75 volts to pass (until it ceases to work).

The reason for my curiosity is first and foremost I DO NOT KNOW - but something said in the DECWARE site made it sound like he was doing something like this with his new version of his amplifier antecedents of which have been on the market for a very long time.

He is using two VR75 tubes - I first assumed these were in series and used as they are demonstrated in the application guides I have seen but then 150 volts does not seem enough for an EL84 - much lower than I have seen elsewhere which made me wonder if he was using them in an unconventional way.

I am not interested in doing this so I am even less interested in better ways of making a series regulator since I know there are plenty of ways to make series regulators. Just trying to find out if the device would work AT ALL (conduct electricity) if used in this manner.

One would have to think if you could get the voltage to jump across that gaseous gap, even though this in itself would produce noise, that it would leave most of the noise from the line in the gas.

Not sure how much regulation would ensue.

I cannot conceive that a typical shunt connection would do anything to isolate the line which Mr. Deckert says is happening with references to the sound being like listening at 2 to 3 AM in the morning which would imply some kind of isolation from the line.

I am not trying to belabor this and certainly not arguing with anyone here but my curiosity remains curious.

Last edited: