5F2-A with VVR

[Leevi]

I share here the schematic of the amp I built recently.
The circuit is based on 5F2-A with an added VVR.

I made the following observations:

1. I tried first to regulate the power amp only but that didn't work at all.
The amp started to overdrive already in low volume levels.

2. Then I changed the regulation concerning also the preamp and it started to work better but
on lower B+ setting when turning the volume from clock 7 to 12 the volume pot crackled.
I investigated that little bit and noticed that there was a low DC voltage before the second preamp stage.
Then I added an extra 22n (+ground resistor 1M) cap before the second preamp stage.
I just wonder the DC there.  It was not a bad cap since I changed them but didn't get rid of the crackling.

3. I replaced the expensive MOSFET NTE2377 (~$14-15) with a much cheaper IRFP450PBF (~$1-2) which
seems to work fine.

4. My understanding is that a tube rectifier doesn't bring any added value for this?

/Leevi

[FYL]

1. I tried first to regulate the power amp only but that didn't work at all.
The amp started to overdrive already in low volume levels.

The output tube is slammed too hard. You should add a master volume between V1b and V2 or scale the whole amp.

2. Then I changed the regulation concerning also the preamp and it started to work better but
on lower B+ setting when turning the volume from clock 7 to 12 the volume pot crackled.

Pretty normal. Preamp stages can show some DC on their grids at low VVR settings.

3. I replaced the expensive MOSFET NTE2377 (~$14-15) with a much cheaper IRFP450PBF (~$1-2) which
seems to work fine.

The IRF is perfect in this amp. The NTE (or a bigger IRF) is a must only when B+ > 450V.

4. My understanding is that a tube rectifier doesn't bring any added value for this?

Yup. Or in any class A amp, where current draw is pretty constant.

[Leevi]

The output tube is slammed too hard. You should add a master volume between V1b and V2 or scale the whole amp.

Yes. I just wonder why many prefer the solution where only the power amp is regulated.
I have seen many comments on that regarding 15-18W P/P amps?

/Leevi

[Willabe]

Yes. I just wonder why many prefer the solution where only the power amp is regulated.

So the tone/voicing of the preamp stays the same.

But, if you VVR the preamp stages too then you change the whole tone/voicing of the amp as you dial down the VVR. You may or may not like this/find this usable tone wise.

The _ idea_ of VVR, Power Scaling, is to be able to get the same cranked amp sound but at lower volume levels.

You should be able to dial in your tone with the VVR full up and _then_ dial down the over all volume with the VVR and _then_ adjust the drive compensation to the output tubes (master volume) back to the same tone/distortion level you started with.

You can also VVR the PI or in this case the output tube driver (V1b) and this will be an automatic drive compensation pretty much on it's own. KOC says this works pretty well up to a certain dB level. But _now_ you won't get PI/driver distortion. This may or may not be a usable sound for the player.

One more thing to consider is as you turn down the VVR the amp is drawing less power from the PT, so it unloads and B+ to non VVR'ed tube stages goes up changing the tone/voicing of the preamp. It can/will get cleaner, brighter and louder depending on how much the B+ rises, which will also create a dead spot at the beginning of the VVRs pots rotation.

KOC uses a voltage clamp on the non Power Scaled, VVR tube stages to stop this effect from happening.


                                       Brad       

[Leevi]

The _ idea_ of VVR, Power Scaling, is to be able to get the same cranked amp sound but at lower volume levels.

Exactly
In order to reach that I think the VVR is not enough. When changing B+ the biasing of the tubes should be
changed as well. Maybe there are more advanced power scaling mechanism published. I'm not familiar with Kevin O'Connor's Power Scaling system which should be more advanced than a simple voltage regulator. If someone knows
more about that please share it.
/Leevi

[HotBluePlates]

The _ idea_ of VVR, Power Scaling, is to be able to get the same cranked amp sound but at lower volume levels.

Exactly
In order to reach that I think the VVR is not enough. When changing B+ the biasing of the tubes should be
changed as well.

Since the amp in question is cathode-biased, then the bias is changed along with the B+.

You turn the B+ down, then the tube draws somewhat less current (primarily due to the drop in screen voltage). That reduced current equates to less cathode current, and therefore less voltage drop across the cathode resistor.

Therefore, the bias is reduced along with the B+.

[Leevi]

Code: [Select]

Since the amp in question is cathode-biased, then the bias is changed along with the B+.
Is it that simple?
I measured the voltages from the amp in three different VVR positions (off-middle-full):

Plate   Cathode  Cathode current
30       1.2         2mA       
177     8.8         18mA
318    17.2         37mA

Since the cathode resistor is in all positions the same 470 Ohm
the tube doesn't operate with a same current which means it has an influence on the sound.
I think the tube operates in totally different area  depending on the position of the VVR pot?

The same happens in the preamp as well.

/Leevi

[Willabe]

In order to reach that I think the VVR is not enough.

FYL said you need to add a master volume to keep the power tube from being _ slamed _ with to much signal.

The output tube needs less signal to drive it as you turn down the Power Scale knob. Loose Change has posted about this a good number of times here along with others, FYL, HBP....

Try it and see how much it helps?

Is it that simple?

I think the tube operates in totally different area  depending on the position of the VVR pot?

No.

I'd have to go through my KOC - TUT books to answer you better, but IIRC, what KOC writes about this is that the _ main point _ is that your keeping the power tubes _ transfer curve _ the  _ same _ as you dial down the PS/VVR. That's why it works as far as keeping the output tube distortion (transfer curve) the same.  HBP already said since this amp is cathode bias it will self adjust on it's own.

FYL, HBP, RicharD, Dummy Load, PRR, come to mind, or any of the other better understanding plate curve guys can explain/show you what I'm trying to say.

Please try the post driver master volume and see how it works for you?


                        Brad          

[ampgeek]

Just thinking out loud here.

Would a separate VVR feeding the pre-amp stages and PI, along with a PA VVR, work equally well (or better?) to a PPIMV?

Cheers,
Dave O.

[Leevi]

HBP already said since this amp is cathode bias it will self adjust on it's own.

I'm not questioning the theory written by KOC and other guys but would like to get answers
to real practical cases:

If I bias a power tube hot using the maximum B+ let's say 318V and 470 Ohm cathode resistor
the tube works then on its maximum limit which means a certain character to the sound.

If I then decrease the B+ to 177V the tube will work much colder since I have not changed the
cathode resistor. The power will of course drop and tube starts to operate in different area.

I used the Weber bias calculator to illustrate that:
http://www.webervst.com/tubes1/calcbias.htm

If I set the maximum B+ 318V and the measured cathode current is 37mA I get the plate dissipation ~12W.
If I set B+ to 177V and the measured cathode current is 18mA I get the plate dissipation ~3W.
If I in the latter case had changed the cathode resistor to a lower value so that the cathode current is still the
37mA I would have reached ~7W plate dissipation e.g. little bit more than half of the maximum.
But what I find important is that the tube is then working equal hot.

Maybe I'm totally wrong with my message. please correct me.

/Leevi

[FYL]

Would a separate VVR feeding the pre-amp stages and PI, along with a PA VVR, work equally well (or better?) to a PPIMV?

I've built a dual-scaled fixed-bias 6L6GC amp a while ago. Scaling the pre brought it in Mesa territory, with some ugly preamp disto. Not worth the bother IMO. An OD stage is a much better sounding proposition for the same cost.



 

[HotBluePlates]

I'm not questioning the theory written by KOC and other guys but would like to get answers
to real practical cases:

If I bias a power tube hot using the maximum B+ let's say 318V and 470 Ohm cathode resistor
the tube works then on its maximum limit which means a certain character to the sound.

If I then decrease the B+ to 177V the tube will work much colder since I have not changed the
cathode resistor. The power will of course drop and tube starts to operate in different area.

Don't think "colder" at least not yet.

If the B+ was still 318v, and you lowered the tube current from 37mA to the 18mA you get with 177v, then yes you'd be running the tube colder.

Notice that when you drop the B+ from 318v to 177v (roughly half), the tube current also changes by a simlar amount from 37mA to 18mA (again, roughly half). Cathode voltage, which is the same as bias in this case, also dropped by about half.

That seems like a roughly same operating point, but as though the output tube were smaller. If you were inspecting tube curves and drawing a loadline for your amp, it would look like a loadline of the same slope (same primary impedance) sliding left as the intersection on the plate voltage axis happens at a lower and lower plate voltage.

To stay at the same relative operating point on your shifted loadline, you need to reduce the bias somewhat as the supply voltage drops.

As the bias gets smaller, the absolute size of the input voltage needed to drive the output tube to distortion gets smaller. A peak input signal that momentarily equals the bias voltage (+37v peak on a tube with -37v of bias) drives the tube to full output power and commences grid current.

So as your supply voltage drops, and the resulting bias across the cathode resistor drops, the input signal needs to get smaller as well to maintain the same point on the volume knob that distortion occurs. When KOC adds a master volume to a VVR/PowerScale circuit, it's to turn down the distortion as the output stage is scaled smaller.

Scaling the whole amp attempts to automatically compensate for this, by making the preamp appear "smaller" too, and reduce its signal level driving the output stage.

[Leevi]

Thanks HBP for your clarification!
/Leevi