your calculation still ignores the 27mA which the OP thinks his amp draws from the 210v rail ... so 35 minus 27 gives just 8mA through the regulators ... not on the high side at all, just barely meets the 5mA min spec ...
1. If the amplifier circuit draws more current, then reduce the two 8000 Ohm resistors.
Make it so the amp current draw and the excess current from the resistors leaves enough current left over to put the regulators into the middle of their current range spec.
2. If the OC3s fire at 115V, than 115 + 115 + 20V margin = 150V needed for a sure startup.
3. Yes, I said OD3, and gave that as an example, but then talked about the same principle applies to the lower voltage regulators.
Make it so the amp current draw and the excess current from the resistors leaves enough current left over to put the regulators into the middle of their current range spec.
2. If the OC3s fire at 115V, than 115 + 115 + 20V margin = 150V needed for a sure startup.
3. Yes, I said OD3, and gave that as an example, but then talked about the same principle applies to the lower voltage regulators.
We just talk about how to address the occational flicker in the regulator.
Has nothing to do with the amp as such.
If your numbers of the load on the 219v rail (27mA) and the voltage before the regulator (348v) are correct then there is about 8mA flowing through the VR105s. Spec for OC3 is 5 ... 40mA, so the "flaw" if we want to call it so - is that its a bit close to the min.5.
Now, let mains voltage be low and/or amp current be just 3 mA higher, and the current through the regulator will fall below the 5mA min.
Furthermore my experience with such regulators is that they tend to flicker when operated close to min specs, which also causes small current fluctuations which in turn produce noise on the power rail.
My experience is that 2x the min current at low mains voltage and max load current works best for me.
Yes my suggestion would be to lower the series resistor from 4000ohm to approximately 2700ohm. That doubles the current through the regulators to about 15mA nominal and 10mA worst case.
Simplest way to do that is to solder one 10 kohm resistor (3 Watt) in parallel to the existing two 8 kohms.
Alternatively replace the two 8 k with 5.6 k (5 W each). Or use a single 2.7k (10 W).
Or do nothing, sometimes new regulator tubes after long periods of storage just require some time to run in ....
Has nothing to do with the amp as such.
If your numbers of the load on the 219v rail (27mA) and the voltage before the regulator (348v) are correct then there is about 8mA flowing through the VR105s. Spec for OC3 is 5 ... 40mA, so the "flaw" if we want to call it so - is that its a bit close to the min.5.
Now, let mains voltage be low and/or amp current be just 3 mA higher, and the current through the regulator will fall below the 5mA min.
Furthermore my experience with such regulators is that they tend to flicker when operated close to min specs, which also causes small current fluctuations which in turn produce noise on the power rail.
My experience is that 2x the min current at low mains voltage and max load current works best for me.
Yes my suggestion would be to lower the series resistor from 4000ohm to approximately 2700ohm. That doubles the current through the regulators to about 15mA nominal and 10mA worst case.
Simplest way to do that is to solder one 10 kohm resistor (3 Watt) in parallel to the existing two 8 kohms.
Alternatively replace the two 8 k with 5.6 k (5 W each). Or use a single 2.7k (10 W).
Or do nothing, sometimes new regulator tubes after long periods of storage just require some time to run in ....
I would just look for any ripple or disturbances in the output that suggest the flicker affects the output. I have seen some instances where the flicker will be around the mica insulators and really affect anything.
I'm on the side of being suspicious about 'that 27 mA'. Basically, we don't have a whole-schematic circuit to make our bets and conjectures from. So a here-say of 27 mA is as probable (to me) as 10 or 15 mA. Need the diagram!
Your point however is well taken.
I personally don't like using VR tubes as in-circuit power regulators.
Not at least as this case.
I rather like having them work at a modest current flow (15 to 20 mA, depending on the model), to establish a fixed voltage reference. Tap it via mid-value (100 kΩ) resistor, feed a capacitor (0.47 µF), and now that to the gate of a MOSFET source follower.
Voilá.
The VR sees absolutely constant current flow, right dead-center at its sweet spot. The downwind circuit sees rock-solid constant voltage. Any 'hum reduction' from regulation is good to go.
One could get all fancy and employ an error-amplifier to further stiffen the constancy of the delivered power, but again in my experience heroic measures are way more intellectually interesting than necessary or practical.
I hope the OP is happy with the discussion and rdux opinion(s).
And it still would be nice to see a whole-circuit diagram.
⋅-⋅-⋅ Just saying, ⋅-⋅-⋅
⋅-=≡ GoatGuy ✓ ≡=-⋅
Last edited:
I'm gaining a LOT of good info from all of you!
I've messaged the guy to ask his permission to post his schematic.
I've messaged the guy to ask his permission to post his schematic.
I haven't seen any datasheet that specifies it, but if that 100 nF that is apparently the maximum recommended load capacitance is for a single VR-105, then reducing C3 to 47 nF as there are two VR-105s in series would be a logical first step.
So, I experimented with my breadboard version. Without input or load on the output there was just the tiniest flicker in the downstream tube. When I added the 10K resistor in parallel, the flicker disappeared and both tubes brightened considerably. Things are tight in the compartment in the built unit so I'll probably replace rather than add another resistor.
Thanks
Thanks