Tell me about this optional capacitor

[JMan]

Hi All,

I've been working on a BA6A using the now-out-of-production Drip board (wait wait, please don't run away, this isn't a "help with my build" post!). I've traced the board and it is quite faithful to the original schematic, although it "simplifies" the power section by opting to run all the 6.3v heaters on AC only.

I'm trying to understand the one other thing on this board that deviates from the schematic. There is an optional 0.1uF film capacitor that he has added between pin 5 of the OD3 regulator tube and ground (I've drawn it into the schematic, see below). Now, he did get back to me when I wrote him to ask about this. His explanation in two separate emails was this:

"skip the .1 normally on a vr tube data sheet the maximum capacitor value after the vr tube is .1 otherwise it forms a relaxation oscillator and the tube trys to kill its self . i added it cause it can help stabilize/filter hf noise etc ."

"Add the .1 it’s fine - not essential though"

Now, I get it from a practical perspective -- I can add it or leave it out, and if I add it maybe it helps with noise. I also googled voltage regulator + capacitors to try to get a fuller picture, since PSU design is but one of my multitude of blind spots. But I'm wondering if some of the folks here have any tidbits of knowledge (even just trivia, I just want to learn) about this sort of thing -- why would it have been omitted originally, does it have only positive effects if implemented within the value limits that Greg alluded to, am I a complete moron for not knowing what this was on sight? Perhaps it's as simple as it seems, but every time I make that assumption, that's the time when it's actually super complex.

 

[Bo Deadly]

I would leave it out to start. Apparently it was added to filter out noise. R43 and the 0.1uF make a simple RC filter. But there are two problems with that. One is apparently it can cause oscillation because the regulator tube is teetering on high impedance. The other is that on startup, you will get high current through R43. So you cannot go higher. But for now I would just leave it out, get it all working as well as you can given that its a D**p board and then get a spectrum of the noise floor and then, if it doesn't look good, you could try tacking it in and measure again and compare.

 

[JMan]

Thanks for the info and advice, Bo! I will leave it out for the time being.

I'm kicking myself for not recognizing what I now see is very obviously an RC filter. I've been in this hobby (and electronics in general) only a couple of years and sometimes my brain still just doesn't put those things together.

 

[abbey road d enfer]

Tube regulators work by ionization of a gas. In order to start ionization, a voltage must be applied, that generates an electrostatic field. When ionization has started, the gas starts conducting,. If there was no resistor to protect it, the current would be too high and result in overheating and destruction of the tube. That's why a series resistor is a requisite.
Putting a capacitor across the tube indeed creates a relaxation oscillator. It is quite possible that 0.1uF results in a highly damped circuit, so the oscillation would be quite small, but that's not my idea of a regulated voltage.

 

[swpaskett]

Admittedly out of my league here. I stopped adding to tube knowledge after 9th grade shop class. This thing being a voltage regulator, though, won't it eventually see capacitors somewhere down the line whether or not there is one here? Is it then incumbent on the designer to keep the total C load on the VR tube below 0.1, or just near the tube? This line of thinking is antithetical to us transistor hacks who hang caps near every active device and blowing what is left of my mind. Please help me out, guys.

 

[Bo Deadly]

This thing being a voltage regulator, though, won't it eventually see capacitors somewhere down the line whether or not there is one here?

No. VR tubes can't handle capacitance in parallel. As already stated, it makes an oscillator. When the cap charges up to the tube forward voltage, the tube impedance drops and the cap discharges through it. But there is some hysteresis to it. Not sure about the exact mechanism but probably has to do with ionized gas being lower impedance so when current is flowing it keeps the tube conducting until the voltage drops below the forward voltage. But at some point the voltage drops low enough that the tube does finally stop conducting and switches back to high impedance. Now the cap charges up again and you have an oscillator.

 

[swpaskett]

That was quick. Thanks Bo. I think I get it.

 

[GJB17]

Hi All,

I've been working on a BA6A using the now-out-of-production Drip board (wait wait, please don't run away, this isn't a "help with my build" post!). I've traced the board and it is quite faithful to the original schematic, although it "simplifies" the power section by opting to run all the 6.3v heaters on AC only.

I'm trying to understand the one other thing on this board that deviates from the schematic. There is an optional 0.1uF film capacitor that he has added between pin 5 of the OD3 regulator tube and ground (I've drawn it into the schematic, see below). Now, he did get back to me when I wrote him to ask about this. His explanation in two separate emails was this:

"skip the .1 normally on a vr tube data sheet the maximum capacitor value after the vr tube is .1 otherwise it forms a relaxation oscillator and the tube trys to kill its self . i added it cause it can help stabilize/filter hf noise etc ."

"Add the .1 it’s fine - not essential though"

Now, I get it from a practical perspective -- I can add it or leave it out, and if I add it maybe it helps with noise. I also googled voltage regulator + capacitors to try to get a fuller picture, since PSU design is but one of my multitude of blind spots. But I'm wondering if some of the folks here have any tidbits of knowledge (even just trivia, I just want to learn) about this sort of thing -- why would it have been omitted originally, does it have only positive effects if implemented within the value limits that Greg alluded to, am I a complete moron for not knowing what this was on sight? Perhaps it's as simple as it seems, but every time I make that assumption, that's the time when it's actually super complex.

View attachment 86623

Hi... from down here in AUS. I have a DRIP ba6a unit belonging to a mate that I am trying to sort... mainly the metering circuit... or lack thereof! It is vastly different to the RCA original. Do you happen to have a schematic of the DRIP unit that I could use to track down what I need to do. My mate purchased an original RCA meter... but it does not seem as sensitive as the DRIP unit and is only reading around 50%. I need to find a means of sorting this. Thanks.

 

[firerunner]

Hi... from down here in AUS. I have a DRIP ba6a unit belonging to a mate that I am trying to sort... mainly the metering circuit... or lack thereof! It is vastly different to the RCA original. Do you happen to have a schematic of the DRIP unit that I could use to track down what I need to do. My mate purchased an original RCA meter... but it does not seem as sensitive as the DRIP unit and is only reading around 50%. I need to find a means of sorting this. Thanks.

I have the Drip BOM with a schematic but it's just the original schematic with the selected/implemented components highlighted.
If you don't need the tube check features you could possibly wire the meter offboard, the original connection (Points A and B) parallel the 16ohm resistor (R4) on the cathode of v1 with a 100ohm trimmer in series with the + of the meter and an 82ohm (R60) in series with the - of the meter. DYOR but I think it should work. Any pros please correct me if I'm off.