Fender bias cap question

[trem]

In the link below, there is a 50/70 cap where you would suspect in the bias circuit.
But there is also the same value on the other side of the 3k3 resistor to ground.
What is that cap doing.?
Not sure i have seen this before

Thank You


https://schematicheaven.net/fenderam...a270_schem.pdf

[Chuck H]

I think it's a "decoupling" capacitor. The ideal impedance for the bias voltage would be zero. This is actually a common practice on most bias circuits. Though admittendly not all demostrate it. But it IS best practice to have a low impedance circuit directly on the DC source for the power tube grids.

[Justin Thomas]

Lots of Fenders have both caps. I don't understand a lick of all that impedance stuff, but I do get smooth filtering. I just look at the bias supply as a miniature power supply. Something in me says that riple in a bias supply is about as nice as ripple on a first gain stage: blech!
Actually I think my 61 Concert might have 2 caps. Dang. Now I hafta look.

Jusrin

[trem]

I think it's a "decoupling" capacitor. The ideal impedance for the bias voltage would be zero. This is actually a common practice on most bias circuits. Though admittendly not all demostrate it. But it IS best practice to have a low impedance circuit directly on the DC source for the power tube grids.

Hey Chuck -

Sorry, this is a little above my pay grade.
Are you saying that cap presents a lower loading onto the power tube grids via the neg bias supply.

To my knowledge i have never seen this before. If you take that cap away does it make a notable difference when played.?
Thank You

[Chuck H]

It's not really lower loading. The bias supply node is also the 0VAC node for the power tube grids and the 100k bias feed resistors are the grid load. The second cap provides more filtering (as Justin noted) and a lower impedance source (0VAC) for the DC bias voltage (-48VDC) to the grids. If you removed the cap there would be less filtering and some interaction between the grid signals elevated on the 3.3k resistor. Most bias supplies, though not all, do use two caps and one is nearly always right at the VDC source for the bias feed resistors.

I know that earlier Fenders used a single cap. It looks like the circuit was changed when the balance circuit was added. Maybe the added cap was needed to incorporate both the balance circuit and the trem switching. I didn't evaluate for that, but maybe.

EDIT: Well I just ran some sims (as I'm prone to do) and it looks like there is not really any significant interaction between the grid signals when the bias feed resistors are elevated on the 3.3k resistance. In fact you have to go much higher than 3.3k before it seems to matter. So I have to assume the added cap is just a matter of more filtering. So that would be two 50uf caps in a pi filter type arrangement. This would be A LOT more filtering than the old BF design. Probably overkill and now I wonder about it too. But not that much because there were a few differences in some SF designs that seem to be trying to improve what was working fine in the BF designs.

[trem]

Awesome..!!!
Thank You So Much

[Chuck H]

I just look at the bias supply as a miniature power supply.

Just to be pedantic I want to note in the amplifiers we deal with the bias supply is a voltage supply. Though that's probably what you meant. A power supply, to me, implies it must carry some amount of current. Which bias supplies don't.

[Helmholtz]

To be pedantic as well, the bias supply is loaded by the resistance which connects to ground (15k).
So it supplies a little power (~0.15W) .
If there was no current to ground, the filter cap(s) could made be very small.

[Chuck H]

To be pedantic as well, the bias supply is loaded by the resistance which connects to ground (15k).
So it supplies a little power (~0.15W) .
If there was no current to ground, the filter cap(s) could made be very small.

Love it And I was thinking about the filtering relative to current even as I was writing that the bias supply isn't a current source. But it seems it actually IS because it's voltage divided. Great consideration and info.

Though it still seems that a 50uf/3.3k/50uf pi filter is a lot of filtering for this circuit. Relatively high value caps in bias supplies have been questioned on the forum before with implications of long charge times. I think I ran some sims once upon a time and found it to be true. IIRC some known circuits took more than ten seconds to stabilize. Since this is roughly equal to the warm up time for most filaments maybe it's not a problem.

[Helmholtz]

What matters for filter effect as well ramp-up time is the R*C product (aka time constant), where R is the equivalent source resistance.
When R is larger, C can be smaller.

[Justin Thomas]

Ah, go ahead & be pedantic, my pedants. I get accused of that all the time over on De Flarpenbloober when I tell people ALL amps can have their power tubes biased & it's just a matter of how easy the manufacturer makes it. Besides, I'm here to learn AND share what I've learned.

It's good to see that such "pedantry" actually sparks honest quality discussions around here that can benefit all of us. Unlike that Other Social Media place & its ilk...

Now, I'm gonna go figure out a way to run some kind of load off a bias supply so that I can be just as pedantic & call it a power supply because it supplies current AND provides my negative voltage. And I still might mess around with a small-signal duo-diode to rectify my bias supply to make my nook-proof toob-amp. Maybe I'll even blow something up while I do; it's been awhile.

Jusrin

[Justin Thomas]

Relatively high value caps in bias supplies have been questioned on the forum before with implications of long charge times. I think I ran some sims once upon a time and found it to be true. IIRC some known circuits took more than ten seconds to stabilize. Since this is roughly equal to the warm up time for most filaments maybe it's not a problem.

Me: Last time my power tubes lost bias (when my Bassman blew (like, it literally spilled its guts) its bias cap) it took five minutes to hum loud enough for me to hear. Twenty seconds to "stabilize"? Bah! This is why I don't spend stoopid money on no fancy power toobz!

Jusrin

[Chuck H]

Now, I'm gonna go figure out a way to run some kind of load off a bias supply so that I can be just as pedantic & call it a power supply because it supplies current AND provides my negative voltage.

Sorry, Randall Smith beat you to it. And knowing him it's probably patented

Mesa uses a PI circuit in some of their amps that is a constant current source at the tail while providing bias for the PI circuit. It works a treat too. My old Subway amp used that circuit and it's one of the only things I never changed.

[Helmholtz]

Being more pedantic:
1) As explained above, your typical bias supply delivers current and voltage. No voltage divider can work without a current.
2) The so-called constant current source neither delivers current nor power. It's just a large impedance in series with the tube which dictates the current.

[Tom Phillips]

When you upgrade the bias supply circuit of a classic black face Fender amp from a single stage filter to a two stage filter you will easily hear a reduction in the hum level at the output. (Assuming that there is not something else wrong with the amp that is causing an increase in hum level) The change is significant but not dramatic or necessary for stage use. However, players who use the amp for recording and recording engineers often appreciate the improvement.