100w presence cap

[benoit]

The ODS-101-HRM schem says it is supposed to be a 1uF. Tough to find that, esp in 6ps. How important is that value?

Also, what is the Voltage rating supposed to be on that one?

Finally, how much leeway do I have in changing the cap value?

[heisthl]

1uf can be electrolitic, 10v rating is enough. 25v is more common. Connect the negative(-) side towards ground. Bear in mind this cap bleeds off the frequencies so they won't get fed back negatively. So if you make it too large in capacitance value the presence control doesn't do anything. Too small and the presence on 10 is like an ice pick. Try for a happy medium between gain(resistor values i.e. 4k7/390 will have less gain than 8k2/390), sensitivity (ratio of the resistors i.e. 4k7/390 will be more dramatic in the control sweep than 8k2/390) and desired effect (cap values i.e. want more mids, use a larger cap). Although cap values can also affect gain perception and sensitivity.

[Fischerman]

Bear in mind this cap bleeds off the frequencies so they won't get fed back negatively. So if you make it too large in capacitance value the presence control doesn't do anything.

I thought making it larger just made the Presence control act more like a full-range boost instead of just a high-frequency boost. But I don't think I've ever gone larger than 1uF.

[heisthl]

The presence knob doesn't boost anything it only takes frequencies away. If a frequency can go to ground via the pot/cap it doesn't get taken away.

[Bob-I]

I use the poly cap in that location, but any voltage is fine.

I've also found that I like a smaller value cap. I find the 1uF a bit harsh in the upper mids when cranked. A .68uF sounds better to my ears, adding more in the highs and less harshness.

[glasman]

The presence knob doesn't boost anything it only takes frequencies away. If a frequency can go to ground via the pot/cap it doesn't get taken away.

Take a closer look at the schematic. The presense cap is in parallel with the 390 when the presense control is at zero ohms. This effectively makes the AC reactance (390 || (Xc 1uf) lower increasing the power amp gain at frequencies above the breakpoint. Xc drops as Frequency is increased.

Gary

[Fischerman]

The presence knob doesn't boost anything it only takes frequencies away. If a frequency can go to ground via the pot/cap it doesn't get taken away.

Yes, I understand how a Presence control works. It's shunting higher frequencies in the NFB to ground. Bigger cap allows more (lower) frequencies to get shunted to ground...and thus not fed back...and thus the power amp has more gain at those frequencies. That's why I was saying that a bigger Presence cap makes the Presence control 'respond or act' more like full-range boost (where as you said a bigger Presence cap makes the Presence control 'not do anything'). A bigger cap doesn't make the Presence control do less or 'not do anything'...just the opposite...it makes it 'do more'.

[dogears]

I agree with Matt here. Bigger cap makes the presence control allow more power amp gain at lower frequencies. If you put a 22uf there for instance, you'd pretty much have open loop when the control was cranked.

The presence knob doesn't boost anything it only takes frequencies away. If a frequency can go to ground via the pot/cap it doesn't get taken away.

Yes, I understand how a Presence control works. It's shunting higher frequencies in the NFB to ground. Bigger cap allows more (lower) frequencies to get shunted to ground...and thus not fed back...and thus the power amp has more gain at those frequencies. That's why I was saying that a bigger Presence cap makes the Presence control 'respond or act' more like full-range boost (where as you said a bigger Presence cap makes the Presence control 'not do anything'). A bigger cap doesn't make the Presence control do less or 'not do anything'...just the opposite...it makes it 'do more'.

[heisthl]

The presence knob doesn't boost anything it only takes frequencies away. If a frequency can go to ground via the pot/cap it doesn't get taken away.

Take a closer look at the schematic. The presense cap is in parallel with the 390 when the presense control is at zero ohms. This effectively makes the AC reactance (390 || (Xc 1uf) lower increasing the power amp gain at frequencies above the breakpoint. Xc drops as Frequency is increased.

Gary

My scope always shows more output at all frequencies when I have no NF. Do I have a scope problem?

[glasman]

The presence knob doesn't boost anything it only takes frequencies away. If a frequency can go to ground via the pot/cap it doesn't get taken away.

Take a closer look at the schematic. The presense cap is in parallel with the 390 when the presense control is at zero ohms. This effectively makes the AC reactance (390 || (Xc 1uf) lower increasing the power amp gain at frequencies above the breakpoint. Xc drops as Frequency is increased.

Gary

My scope always shows more output at all frequencies when I have no NF. Do I have a scope problem?

No scope problem and yes if you open loop the power amp you will have max gain. I think of it like this, the 4700 is Rf , the 390 is Ri.

The gain (not full formula, 1st approximation) would be.

Gain = Rf / Ri.

If you make Ri smaller the gain will increase. As Ri in an amp is the combination of the 390 in parallel with the 2K+Xc(1uf). As frequency is increased the net Ri is decreasing. I have tested this many times on a spectrum analyzer and you can see the increase at the breakpoint.

Gary

[jelle]

Heistl,

you wrote:
'Bear in mind this cap bleeds off the frequencies so they won't get fed back negatively'

When the high freqs bleed through the cap to ground then these freqs will be excluded from the NFB loop and will sound louder compared to the lower freqs that are still in the NFB loop.

[ayan]

The presence knob doesn't boost anything it only takes frequencies away. If a frequency can go to ground via the pot/cap it doesn't get taken away.

Take a closer look at the schematic. The presense cap is in parallel with the 390 when the presense control is at zero ohms. This effectively makes the AC reactance (390 || (Xc 1uf) lower increasing the power amp gain at frequencies above the breakpoint. Xc drops as Frequency is increased.

Gary

You both are saying the same thing: the presence control doesn't BOOST anything, it just reduces the amount of UNBOOST of the power amp through the negative feedback loop.

Someone mentioned getting a bottom end "boost," for which you do the analogous of the presence control arrangement. Instead of a cap to ground to decouple of the highs from the NFB loop, you put a cap in series with the feedabck wire to decouple the lows form the NFB loop. You can start with a .005uF cap, and if you put a pot (wired as a variable resistor) in parallel with it, you can take it out of the loop and have a "resonance" control.

Gil

[ayan]

The gain (not full formula, 1st approximation) would be.

Gain = Rf / Ri.

Gary

[ I am bored today, so here's my take on this. ]

I would rather call unity "gain" the configuration when you have no feedback loop. Then, a 4.7K Rf and a 390 ohm Ri gives you a transfer function of:

G = 1 - Xi/(RF+Xi+4)

and obvisouly the gain is always less than or equal to 1, since we're talking about a NFB loop.

Xi is the capacitive (= no reactance = -j/[2Pi()fC] = 1/[j2Pi()fC]-- for the math-oriented minded) impedance formed by the 390 ohm resistor in parallel with the series of the presence pot (depends on the setting) and the presence cap. 4-ohms is the tap you took the feedback from.

In the best case scenario, where Xi is infinitely large (huge cap), the second term goes to zero and the gain goes back to one. Of course, talking about "gain" when referring to a power amp is not really all that useful since the mission of that section is really to "transconduct," i.e. generate enough current based on a voltage model, vs amplifying a voltage. But, as we put a feedback loop on the primary, talking about "Gain" is once again fair game.

Gil

[Fischerman]

FWIW, the Bogner 'Excursion' switch works on the NFB and the 'Loose' setting is a sort of a fixed 'Resonance' setting. Basically the NFB resistor is a 100k (with a typical later Marshall Presence control, i.e. BM Presence/PI) and that is the 'Tight' setting. The next setting (Medium?) adds a 220k resistor in series with the 100k for a lot less NFB. Then the 'Loose' setting is a 1M resistor and 470pF cap in parallel, then all that in series with that original 100k NFB resistor.

The bass boost you get from any sort of 'Resonance' control is pretty loose and can easily get boomy if not careful. And it reduces the damping but this is where it gets over my head...but you do get to use cool terms like 'slew-rate'...so that's nice.

[jelle]

"But, as we put a feedback loop on the primary, talking about "Gain" is once again fair game. "

Is this because the phase inverter has gain and is in the loop?

Jelle ( kind of bored too... )