Is anyone familiar with the function of this part of the JVM410H(JS) capacitor?

[SunTzuBean]

Hi all,

I am trying to figure out what the purpose of this capacitor is in the input gain stage of the JVM410HJS. The 410H also has this capacitor, but it is 220pF vs. the HJS' 1 nF. It's circled in red here:
I've never seen a capacitor connected to both the cathode and the plate at the exact same time!

And while we're at it, I want to be sure I understand the function of the FET that is also connected to the cathode. It looks like it's configured to be a switch, in order to switch R181 in and out of the circuit.

I've made some mods to my HJS in order to make the clean more like the 410H, the one channel which I believe to be superior. However, I think that most of the clean tone comes from the negative feedback arrangement of the 410H, and I would lose basically all of what makes the HJS sound so much better to my ears in many ways. I am thinking of reverting everything back to stock except for changing the clean channel EQ slope resistor back to 100k because it increases highs and reduces bass, which is one of the problems of the stock clean to my ears. I may also keep the 220k plate resistor on the gain stage prior to the cathode follower because it significantly increased the volume from the clean channel, allowing me to turn the gain lower and keep the clean channel even more clean.

 

[PelliX]

I've never seen a capacitor connected to both the cathode and the plate at the exact same time!

By the looks of it, I'd say it's rolling off a certain amount of high end (without doing the math, maybe even beyond the audible range) and preventing oscillation.

 

[Spanngitter]

By the looks of it, I'd say it's rolling off a certain amount of high end (without doing the math, maybe even beyond the audible range) and preventing oscillation.

Exactly, this type of cap is called snubber cap. It can be there to prevent oscillation but also round off the ice pick on hights.
For the standard JVM it's more an osciallation prevention, for the JHS (due to it's size) also shaves off highs.
I am just too lazy to calculate but the formula is f = 1/(2*pi*R*C)

 

[Pete Farrington]

I want to be sure I understand the function of the FET that is also connected to the cathode. It looks like it's configured to be a switch, in order to switch R181 in and out of the circuit.

The FET is switching C153, the partial cathode bypass for V101a.

Ignoring the circuit parts not shown, the low pass filter corner frequency with a 1nF snubber will be about 4kHz, with 220pF about 18kHz.

 

[FleshOnGear]

Out of curiosity, what is the difference between doing this with the cap vs. just attaching the cap to ground, or attaching it to the grid.

 

[PelliX]

Technically both would (and do) work. I'm wondering what this would do in the way of feedback.

 

[Pete Farrington]

Out of curiosity, what is the difference between doing this with the cap vs. just attaching the cap to ground

Pretty much the same. Anode to cathode at the socket terminals eliminates wire and HT decoupling inductance.

or attaching it to the grid.

That’s somewhat different. It becomes Miller capacitance, the cap value becomes multiplied by the stage gain (eg roughly x 60) and loads the grid.

 

[Browneyesound]

It’s the “tone suck” cap.

 

[SunTzuBean]

Exactly, this type of cap is called snubber cap. It can be there to prevent oscillation but also round off the ice pick on hights.
For the standard JVM it's more an osciallation prevention, for the JHS (due to it's size) also shaves off highs.
I am just too lazy to calculate but the formula is f = 1/(2*pi*R*C)

It looks like it's in a negative feedback arrangement, cancelling out frequencies according to that formula

 

[mickeydg5]

Hi all,

I am trying to figure out what the purpose of this capacitor is in the input gain stage of the JVM410HJS. The 410H also has this capacitor, but it is 220pF vs. the HJS' 1 nF. It's circled in red here: View attachment 127918
I've never seen a capacitor connected to both the cathode and the plate at the exact same time!

And while we're at it, I want to be sure I understand the function of the FET that is also connected to the cathode. It looks like it's configured to be a switch, in order to switch R181 in and out of the circuit.

I've made some mods to my HJS in order to make the clean more like the 410H, the one channel which I believe to be superior. However, I think that most of the clean tone comes from the negative feedback arrangement of the 410H, and I would lose basically all of what makes the HJS sound so much better to my ears in many ways. I am thinking of reverting everything back to stock except for changing the clean channel EQ slope resistor back to 100k because it increases highs and reduces bass, which is one of the problems of the stock clean to my ears. I may also keep the 220k plate resistor on the gain stage prior to the cathode follower because it significantly increased the volume from the clean channel, allowing me to turn the gain lower and keep the clean channel even more clean.

It indicates NF, not found, not used.

If it were used with the value you specified then if would indicate 1n.

 

[Jon Snell]

Sorry to burst the fragile bubble but NF means Not Fitted.

 

[PelliX]

Sorry to burst the fragile bubble but NF means Not Fitted.

Mornin' Jon, I see we've joined the condescending club...

Correct, though.

 

[Pete Farrington]

It looks like it's in a negative feedback arrangement,

No, it’s a simple passive low pass filter.

 

[Jon Snell]

No, it’s a simple passive low pass filter.

Correct, (if it was fitted of course). 🤣

Morning PeliX, (someone had to point NF out). 👍

It was not my intention to be distainful.
Appologies if that is how it sounded.

 

[PelliX]

All good, Jon - maybe I had a lump or two too many of cynicism in my coffee!

 

[FleshOnGear]

Correct, (if it was fitted of course). 🤣

Morning PeliX, (someone had to point NF out). 👍

It was not my intention to be distainful.
Appologies if that is how it sounded.

Last time I was distaintful, I couldn’t sit for a week!

 

[SunTzuBean]

No, it’s a simple passive low pass filter.

With the output impedance of the 12AX7 (38k)? I used that to verify the calcs you did earlier in the thread and I got the same results you did. But why wouldn’t the phase inverted plate signal interact with the in-phase cathode signal?

I’m not an expert at amp design by any means but to me it just looks like the amplified signal at the plate would interact with the signal at the cathode, especially since that cap is connected to the cathode side of the tube… I’m trying to learn!!

 

[SunTzuBean]

Correct, (if it was fitted of course). 🤣

Morning PeliX, (someone had to point NF out). 👍

It was not my intention to be distainful.
Appologies if that is how it sounded.

The cap is definitely fitted, and it measured about 1 nF with my multimeter. But the more you know! I never like on schematics when something measures 1 *something* and they don’t bother to put that one, they just say “NF” like this

Beyond that, I have reverted the previous mods I made and all that’s not stock is the 100k slope resistor for the clean channel (changed to 56k) and the last gain stage’s plate resistor is still at 220k. This increases the compression of all the channels a little bit (most importantly, the clean channel) while still maintaining the lovely, open OD sound of the higher gain channels. The clean sounds less anemic and is plenty punchy and warm.

 

[Pete Farrington]

why wouldn’t the phase inverted plate signal interact with the in-phase cathode signal?

Good point, sorry, I was assuming that the cathode bypass cap was engaged (and hence for AC signal analysis, it’s assumed to be held at 0V AC).
When it isn’t, the numbers change a bit and yes, there will be a bit of negative feedback. My thinking is that the high anode impedance, low cathode impedance, and low cap value will result in the effect of the NFB being fairly low, and so the low pass filter effect will probably still dominate what’s happening.

The cap is definitely fitted, and it measured about 1 nF with my multimeter. But the more you know! I never like on schematics when something measures 1 *something* and they don’t bother to put that one, they just say “NF” like this

It’s very ambiguous, but on balance, the lack of a cap numerical value preceding the ‘NF’ probably indicates that it means ‘not fitted’, rather than ‘nano farads’.
If your amp did have the cap fitted, then it may have been made to a different revision of the schematic.

 

[mickeydg5]

Mornin' Jon, I see we've joined the condescending club...

Correct, though.

I think I said it first.