VR tubes, minimum current

[tomchr]

Folks,

I'm toying with the idea of using a voltage regulator tube as the reference voltage in a voltage regulator. Specifically, I'm thinking to use a 0A3. I like the looks of it....

The maximum current of the 0A3 is about 30~40 mA if I recall correctly. But is there a minimum current that's needed to ensure stable operation?

~Tom

 

[aardvarkash10]

yes, and a minimum voltage difference too for both operating and striking (start-up). On the spec sheet usually.

 

[mastodon]

Minimum current for regulation is 5mA for the 0A3, as per the Tung Sol datasheet.

 

[DF96]

Did you choose the 0A3 for its appearance or performance? Is this art or engineering?

 

[Lingwendil]

Did you choose the 0A3 for its appearance or performance? Is this art or engineering?

I personally use them for both, they're a nice big zener that looks great, is more rugged, equivalent noise specs (often better IME )and cost is usually about the same when purchased as surplus.


Back on the OP's question, run 'em at around 10 ~ 15 mA and you're set. For even better performance use a CCS.

 

[bob91343]

A zener diode will not do as well; it will have a temperature coefficient that may bother you. VR tubes are very good in this respect.

 

[Wavebourn]

Temperature effect of Zeners when used to power tube stages, even directly copuplet, does not bother me at all. But tubes look better.

 

[tomchr]

Did you choose the 0A3 for its appearance or performance? Is this art or engineering?

Yes!

I love the orange glow. It's purrrdy...

10~15 mA. Oof! I'm coming down from 600+ V. That's a lot of power in the series resistor.

~Tom

 

[Wavebourn]

5 mA is enough.

 

[zigzagflux]

Especially if you supply it with an ultra-simple CCS, maybe with the 900V rated IXYS device (no need for cascoding), the high voltage isn't really a problem. Firing is all but guaranteed even with 5 mA when you use a CCS.

 

[Knarf]

If 600V starting voltage doesn't do it, then the VR tube probably won't start at all.

However, if you do happen to find yourself short on starting voltage some day, then it is possible to 'cheat' by temporarily illuminating the innards with a blue LED for a few seconds after power-on. VR tubes are light sensitive.

 

[tomchr]

5 mA is enough.

Especially if you supply it with an ultra-simple CCS, maybe with the 900V rated IXYS device (no need for cascoding), the high voltage isn't really a problem. Firing is all but guaranteed even with 5 mA when you use a CCS.

660 V worst case --> (660-75)0.005 = 3 W. Not horrible, but still an amount of power that will need to be dissipated safely.

Actually, my two remaining brain cells got to work and figured out that there's a 70 V tap on the mains transformer I'm using (Antek AS-4T430). That could easily be used for the reference voltage. 70 V AC should give about 90 V rectified DC. That may not be enough to light the lamp, though. The starting voltage is 100 V. But I could double the 70 V... Options, options. It would eliminate the need for high-voltage CCS'es and result in about 0.5 W of dissipated power in the CCS. That's more like it.

~Tom

 

[Lingwendil]

660 V worst case --> (660-75)0.005 = 3 W. Not horrible, but still an amount of power that will need to be dissipated safely.

Actually, my two remaining brain cells got to work and figured out that there's a 70 V tap on the mains transformer I'm using (Antek AS-4T430). That could easily be used for the reference voltage. 70 V AC should give about 90 V rectified DC. That may not be enough to light the lamp, though. The starting voltage is 100 V. But I could double the 70 V... Options, options. It would eliminate the need for high-voltage CCS'es and result in about 0.5 W of dissipated power in the CCS. That's more like it.

~Tom

Voltage doublers make the the reflected current draw twice what it would normally be, make sure the trafo is able to pull that current through that winding, usually it's fine. IMO, voltage doubler plus CCS should be a great implementation, try it and tell us how you like it.

 

[Wavebourn]

Peak current in doubler's diodes would be twice higher, but power dissipation still the same. I would consider this option.

 

[tomchr]

Peak current in doubler's diodes would be twice higher, but power dissipation still the same. I would consider this option.

Especially since the 70 V winding will only carry the 5 mA for the 0A3. Well... Possibly in addition to 20 mA for the input stage of the amp, but that's a story for another day... Still. That's small potatoes for a 400 VA transformer.

Yeah. I'll let you know how it goes.

~Tom

 

[Wavebourn]

Yeah. I'll let you know how it goes.

Don't forget to take pictures in darkness!

 

[tomchr]

Don't forget to take pictures in darkness!

Oh, yeah... Glow pictures are a must for this one.

By the way. The general recommendation seems to be to put a small cap in parallel with the VR tube. If I drive the VR tube by a CCS, is the normal 56~100 nF cap still recommended or should a different value be used? After all, I'm interested in a voltage reference not an oscillator...

~Tom

 

[zigzagflux]

Currently using 8 glow tubes in my setup between preamp and amp. Routinely use 56 - 68 nF with no issues. They are all CCS loaded.

 

[Lingwendil]

Oh, yeah... Glow pictures are a must for this one.

By the way. The general recommendation seems to be to put a small cap in parallel with the VR tube. If I drive the VR tube by a CCS, is the normal 56~100 nF cap still recommended or should a different value be used? After all, I'm interested in a voltage reference not an oscillator...

~Tom

With a CCS, I'm usually inclined to omit the caps, but many leave them in for good measure.