VR150 use

[Andrewbee]

Hi all,

Quick ?,

I am rebuilding my RH84 (EL84 Pentode SE) and am considering regulating G2.
I have never used Gas tubes before, only zeners but have a VR150. So, I am thinking about the VR150 and was wondering if there would be a problem using it with SS rectification for the B+.
I have been doing some reading on them but never saw mention of this.

IOW, is there an issue with the gas tubes being "hit" suddenly with voltage?
I considered this since the gas tubes are from an era before SS. I did read somewhere that they do not like being turned on and off and on again quickly.

Thanks.

[Tom Bavis]

No, no problem. Make sure the dropping resistor can take the initial surge - it could be several times the normal dissipation, depending on supply voltage and loading. A VR tube works much like a Zener, but you shouldn't bypass it with more than .047 uF or so - it may oscillate.

[kevinkr]

I use VR150 to regulate the solid state rectified bias supply in my 300B SE amplifier. (I want the supply bias to come up almost instantly. No problems in 4 yrs.)

You need to make sure that there is sufficient voltage for the tube to strike properly and in this case something like 220V should be sufficient. Minimum current through the VR tube should not be less than 5mA, and no more than 40mA before your screens start to draw current. You want to make sure that at amplifier clip there is still enough current (5mA or greater) available to keep the VR tube from extinguishing. Some tweaking will be required.

One very important thing to mention is that the screen voltage should not be there before the plate voltage as this can have catastrophic consequences for the screens in your output tubes. The reverse is not an issue.

I have seen VR tubes break into relaxation oscillations with as little as 0.047uF - 0.1uF across the tube, and generally use no more than 0.022uF in parallel with the VR tube.

[Wavebourn]

With no bypassing cap it is no surge current. Opposite, to ignite it you need a bit more of a voltage that is needed to keep it glowing.

Originally posted by kevinkr

I have seen VR tubes break into relaxation oscillations with as little as 0.047uF - 0.1uF across the tube, and generally use no more than 0.022uF in parallel with the VR tube.

It depends on the current: if it is too low the toob discharges the cap then stops conducting until a voltage on it is back enough to ignite it.

[kevinkr]

Originally posted by Wavebourn
With no bypassing cap it is no surge current. Opposite, to ignite it you need a bit more of a voltage that is needed to keep it glowing.



It depends on the current: if it is too low the toob discharges the cap then stops conducting until a voltage on it is back enough to ignite it.

My experience too, however the current levels are higher than you would expect.. I've had problems at 15 - 20mA.. Something to do with the negative resistance characteristic of the VR tube.

[Wavebourn]

Originally posted by kevinkr


My experience too, however the current levels are higher than you would expect.. I've had problems at 15 - 20mA.. Something to do with the negative resistance characteristic of the VR tube.

May be your output stage helped it to oscillate causing a B+ sag?

[kevinkr]

Originally posted by Wavebourn


May be your output stage helped it to oscillate causing a B+ sag?

Nope, totally independently powered, and as I only build SE amplifiers these days I use them solely for bias supplies. Above some threshold capacitance level it no longer matters what the current is - all that happens is the relaxation frequency shifts around.. (more current = faster repetition rate.) Too much current and the tube enters the region where the behavior is more analogous to an arc than a glow discharge and the tube fails quickly at that point.

Have done this experiment on the bench with tightly regulated power supply as the source, just cap, resistor, and VR tube. Outcome is the same - there is always a point where regardless of the current through the tube above a specific capacitance the regulator becomes a relaxation oscillator. (Generally with values of 0.1uF and above.)

[kevinkr]

Some useful information on VR tubes in general is here:

http://en.wikipedia.org/wiki/Voltage_regulator_tube

[Wavebourn]

Originally posted by kevinkr



Nope, totally independently powered, and as I only build SE amplifiers these days I use them solely for bias supplies. Above some threshold capacitance level it no longer matters what the current is - all that happens is the relaxation frequency shifts around.. (more current = faster repetition rate.) Too much current and the tube enters the region where the behavior is more analogous to an arc than a glow discharge and the tube fails quickly at that point.

Have done this experiment on the bench with tightly regulated power supply as the source, just cap, resistor, and VR tube. Outcome is the same - there is always a point where regardless of the current through the tube above a specific capacitance the regulator becomes a relaxation oscillator. (Generally with values of 0.1uF and above.)

You are right; a cap discharge current causes some dynamic negative resistance of SG-4S. Confirmed right now on the bench.

[rdf]

Consider driving the VR tube with a CCS. The IXYS works a pip.