Voltage rating for caps around the NFB and PI region?

[demonstratedspatulas]

I usually see a cap going from the PI to the presence control. What sort of voltage rating is needed for this cap?

In addition, I'm wondering about the voltage ratings for these other caps in the same region:

* Cap going to the "non input" of the PI. I wish I could remember the name of this one! Usually with a value larger than the coupling cap for the input grid of the PI.
* Cap in series with the NFB. Surely this can be less than 100VDC right? (As the output from a tube amp should never come close to that)
* Cap connected in parallel with (ie wiper and one side of) variable NFB pot.

I know the voltage ratings for everything else, and I always leave a safe margin of error if I'm unsure (ie defaulting to 630V) but it would be nice to know more specifically about these components.

[pompeiisneaks]

any kind of coupling caps are almost always 600 if not 1000vdc, as they're in the anode path of the high voltage. NFB is post the actual high voltage and only linked to what comes out of the tap of the output transformer, and the other side of the coupling for the PI, so it's AC more than DC. I think the AC rating can be upwards of 70 to 80 VAC, but not sure on DC, someone can probably correct me if wrong on either. It depends on the amp and designed output side of the OT.

~Phil

[Stevem]

Caps have a AC voltage limit also which most times is 1/2 there D.C rating.

The danger is that a low D.C. Voltage rating cap can have a big ac signal voltage swing across it.

[sluckey]

Fender used 200v caps.

[demonstratedspatulas]

Rated 200 VDC makes sense if the AC voltage rating is 1/2 the DC voltage rating, and at the same time pompeiisneaks mentioned that the AC voltage rating is around 70-80 VDC.

This rating would be for all caps mentioned?

I can certainly overspec but I'm curious to learn here, and not to mention larger voltage rating usually means a larger, more expensive cap.

Thank you!

[pompeiisneaks]

no, coupling caps usually need to be around that same 600-1000 types used in most other stages of the amp. I was only mentioning the lower voltages for the presence cap.

600 is usually fine unless in an amp running much hotter like 500 to 600V say for 6550's or other tubes that like it higher voltage. Then the 1000 becomes a smarter choice, as spikes on line voltage can cause it to jump pretty high.

~Phil

[demonstratedspatulas]

no, coupling caps usually need to be around that same 600-1000 types used in most other stages of the amp. I was only mentioning the lower voltages for the presence cap.

600 is usually fine unless in an amp running much hotter like 500 to 600V say for 6550's or other tubes that like it higher voltage. Then the 1000 becomes a smarter choice, as spikes on line voltage can cause it to jump pretty high.

~Phil

Ah sorry, I meant for the four locations I mentioned in my original post:

* Cap going from the PI to the presence control.
* Cap going to the "non input" grid of the PI.
* Cap in series with the NFB.
* Cap connected in parallel with (ie wiper and one side of) variable NFB pot.

Thank you!

[pompeiisneaks]

no, coupling caps usually need to be around that same 600-1000 types used in most other stages of the amp. I was only mentioning the lower voltages for the presence cap.

600 is usually fine unless in an amp running much hotter like 500 to 600V say for 6550's or other tubes that like it higher voltage. Then the 1000 becomes a smarter choice, as spikes on line voltage can cause it to jump pretty high.

~Phil

Ah sorry, I meant for the four locations I mentioned in my original post:

* Cap going from the PI to the presence control.
* Cap going to the "non input" grid of the PI.
* Cap in series with the NFB.
* Cap connected in parallel with (ie wiper and one side of) variable NFB pot.

Thank you!

I guess maybe it depends on the specific amp, but I normally only see one cap related to the presence circuit. I've not seen the three different ones you're referencing above. (NFB, PI to Presence, and NFB Pot are all the 'presence circuit' and at the same basic potential)

I think any caps related to the PI are generally kept at that 600V because oftentimes it's cheaper to by in bulk of the same kind than having a single cap at a lower voltage potential just for an area that 'shouldn't' see higher voltages. All of those you mention seem to me to probably be fine with a 200VDC rating (or around 100VAC) but I've rarely seen that, and almost always see the 600V caps in all coupling situations. Maybe it's an inexpensive way to protect other things in the case of one of the caps shorting out and sending 600VDC across to other components in that same area as well?

~Phil

[gingertube]

Just a correction to info above.
Caps have a DC voltage rating - in simplest terms, that is also their Peak AC Voltage Rating

So
Their AC rating in Volts RMS = their DC rating divided by root 2 ( Same as saying x 0.707 of DC)
Their AC rating in Volts peak = their DC rating
Their AC rating in Volts peak to peak = 2 x their DC rating

Cheers,
Ian

[sluckey]

... All of those you mention seem to me to probably be fine with a 200VDC rating (or around 100VAC) but I've rarely seen that, and almost always see the 600V caps in all coupling situations. Maybe it's an inexpensive way to protect other things in the case of one of the caps shorting out and sending 600VDC across to other components in that same area as well?

Take a look at this popular schematic. What do you see around the PI?

https://el34world.com/charts/Schematics ... _schem.gif

Most all Fender amps that use a LTP PI follow that same cap scheme. Look at this '64 Deluxe Reverb...

[roberto]

I'll try to split the AC and DC component for each:

* Cap going from the PI to the presence control.
It depends on the kind of configuration.
With 12ax7 we'll have around 10V DC across the 4k7 resistor with presence maxed.
AC component will be few volts (consider that the worst case can be considered a 100W tap at 16 Ohm will have 40 Vrms swing, then there's the NFB voltage divider).


* Cap going to the "non input" grid of the PI.
Old Fender UL amps with 12at7 will have around 100V DC across it.
AC component will be higher than the previous one, so 200V is reasonable.

* Cap in series with the NFB.
Talking about DC, one side is grounded while the other one is +20 to +30V.
AC component will be few volts (consider that the worst case can be considered a 100W tap at 16 Ohm will have 40 Vrms swing, then there's the NFB voltage divider).

* Cap connected in parallel with (ie wiper and one side of) variable NFB pot.
This is commonly called depth.
Talking about DC, if we have a cap in series with the NFB both ends are at same potential.
AC component will depend on how you set the control. If maxxed at 1M with 100k feedback, consider that the worst case can be considered a 100W tap at 16 Ohm will have 40 Vrms swing.