[coinmaster]

The goal here was to get as much B+ as possible, give the tubes as low current as possible without any SQ drawback. We work with what we have here, though 250+ V B+ would of been much better, of course.

Just make a voltage doubler. Only requires a few parts.

APPJ's is speed and sharp transient leading edges which the LD does not quite have,

Film caps in the mk6 power supply solves this.

Oh you mean the CCS mod, I don't know if that's part of a LTP on the LD, waiting to hear more!

it is.

 

[baronbeehive]

...
Film caps in the mk6 power supply solves this.
...

....yes but I have no heavy articulated truck to deliver it.....!

 

[MrCurwen]

I don't exactly recall what was in the PSU, but I'm sure it could just be tossed and put a simple RCRC with small R (like 100 or 200 R) and a simple 2 FET reg. Then you'd have more or less about the same B+ voltage, which is fine.

Use cheapest resistors and 105 degree 47µF electrolytes. Absolutely no difference in sound quality either theoretically or observably is going to be happening. It's a constant current draw stage at that point and behind a regulator. Put one 1 to 4 µF film cap after the reg for stability. Russian MBGO 4µF recommended. No observable sound quality difference either way, this cap is for stability and nothing else.

Then just put the gyrators in there. The input stage only needs to provide a small voltage output because of the nature of the amp as a whole. No need for 70V voltage headroom, 10 or 20 will suffice, 50 will be overkill.

With gNFB and adding stuff inside the loop stability is always a concern. But I guess there's only one way to find out; stability cannot be simmed.

 

[MrCurwen]

After the mods described in my previous post having the input stage inside the loop is actually only hurting the sound quality of the input stage. It is however still required for making the output stage have a useable output impedance.

 

[SonicTrance]

Edit: Oh you mean the CCS mod

No, I mean this: https://www.head-fi.org/threads/modern-balanced-tube-amp-build.852879/

But, yes. I also recommend the CCS mod for the MK6!

I guess this is one of those which simmed better in spice than it performed in real life.
If I inject a signal in one phase and say the output is 2 Vpp. Then the output of the phase without input signal is about half, so 1 Vpp. Still much better than before but not "forced balanced". But when listening to music we have input signal in both phases, of course.

All in all I think it sounds much better than before! As it should.

 

[baronbeehive]

After the mods described in my previous post having the input stage inside the loop is actually only hurting the sound quality of the input stage. It is however still required for making the output stage have a useable output impedance.

Thanks, we will look into this. As I said it is not a problem for me because I like a non fatiguing sound over headphones, nevertheless to get the most definition it may need to be done.

 

[SonicTrance]

I don't exactly recall what was in the PSU, but I'm sure it could just be tossed and put a simple RCRC with small R (like 100 or 200 R) and a simple 2 FET reg. Then you'd have more or less about the same B+ voltage, which is fine.

Have you looked at any of the pics I've posted? There's absolutely no room for any reg's or gyrators and most definitely not any russsian caps, lol! Sure, I could rebuild the psu filter, but without the reg there's no point.

 

[coinmaster]

Have you looked at any of the pics I've posted? There's absolutely no room for any reg's or gyrators and most definitely not any russsian caps, lol! Sure, I could rebuild the psu filter, but without the reg there's no point.

You wasted all the space using an elaborate ccs scheme instead of a 2 terminal scheme. I also said there are low profile gyrator pcbs you can buy. You seem to be ignoring me so I'll just leave it at that.

 

[SonicTrance]

You wasted all the space using an elaborate ccs scheme instead of a 2 terminal scheme.

Well, I had all parts on hand so...And I trust MrCurwen's advice as it has worked out great for me so far.

I also said there are low profile gyrator pcbs you can buy.

Do you have any pics with these boards fitted inside the MK6?

After the mods described in my previous post having the input stage inside the loop is actually only hurting the sound quality of the input stage. It is however still required for making the output stage have a useable output impedance.

My goal is not to mimic the modern LTP. I already built my dream amplifier so there's no point in trying to make the mk6 sound just like it (I realize it never will with the WCF output stage). If other people are willing to extend the bottom of the chassis or make a separate box for the input stage, that's totally understandable. After all that was my initial goal before I built the modern LTP.

 

[coinmaster]

Well, I had all parts on hand so...And I trust MrCurwen's advice as it has worked out great for me so far.

It's an upgrade to be sure, I still don't trust the b- connection though. Any harmonics that get through the current source will be heavily amplified. Like I said you might want to scope your b- supply to see how dirty it is. The dirtier it is the less impedance the current source will provide and the more unwanted harmonics will leak into the input stage.

Do you have any pics with these boards fitted inside the MK6?

I only applied my improved wcf mod to my mk6 a while back. My headphones bit the dust after years of abusing them in experiments so I didn't see the point in doing any more mods. I'll soon be repairing them while I build full scale prototypes of one of the amps I designed.
That being said the PCBs are mostly flat, the mosfet is the exception but it can be bent back. http://www.bartola.co.uk/valves/for-sale/gyrator-pcb/ I would be surprised if you couldn't squeeze it in there.

My goal is not to mimic the modern LTP. I already built my dream amplifier so there's no point in trying to make the mk6 sound just like it (I realize it never will with the WCF output stage). If other people are willing to extend the bottom of the chassis or make a separate box for the input stage, that's totally understandable. After all that was my initial goal before I built the modern LTP.

The WCF may not be technically optimal but it does a lot of things well in OTL and has the hidden advantage of how it sounds when mixing tubes. There are ways to make the wcf and the "modern LTP" perform much better without the need for feedback, I'll soon find out how this better performance translates into sound in the next few weeks.

 

[SonicTrance]

Any harmonics that get through the current source will be heavily amplified. Like I said you might want to scope your b- supply to see how dirty it is.

I have looked at the B- using FFT on the scope. Looked very clean.

I would be surprised if you couldn't squeeze it in there.

Remember, you need four of those for the input stage. Probably heatsinks on the FET's as well. I could most likely build p2p gyrators smaller than that pcb, especially if not using a heatsink. Anyway, I'm not all that interested in installing gyrators and grid drivers in the mk6. I might hook something up externally because it's fun to experiment.


When listening to music I have my other amp.

 

[coinmaster]

Remember, you need four of those for the input stage. Probably heatsinks on the FET's as well.

No need for heatsink. I have 4 of those pcs I used for other projects. They would fit nicely even with 4.

I have looked at the B- using FFT on the scope. Looked very clean.

What are you using to read fft? Oscilloscope ffts are not designed for audio work.

Anyway, I'm not all that interested in installing gyrators and grid drivers in the mk6. I might hook something up externally because it's fun to experiment

Voltage doubler would probably work for you then.

 

[SonicTrance]

No need for heatsink.

Yes, I realized there’s no need for heat sinks as the dissipation would be very low in the mk6

Oscilloscope ffts are not designed for audio work.

Yes, FFT on the scope is what I use. Range is from 1Hz to many MHz.

Voltage doubler would probably work for you then.

It would definitely fit inside if I take out the existing filter. Question is how the PT(s) would react to a doubler.

 

[coinmaster]

Yes, FFT on the scope is what I use. Range is from 1Hz to many MHz.

Only the most expensive scopes have FTTs that aren't garbage, they are added to the scope as an afterthought. FTTs are process intensive, you need a dedicated audio spectrum analyzer to get a good reading, or at the very least a sound card. What scope are you using?
It may be easier to just look at the B- in AC scope mode. You don't need an FFT reading to know if the supply is clean or not.

It would definitely fit inside if I take out the existing filter. Question is how the PT(s) would react to a doubler.

They wouldn't. At the current you need the caps don't have to be huge.

 

[SonicTrance]

It may be easier to just look at the B- in AC scope mode.

There's only about 100 mVpp of ripple on the B-. Same on the B+ for the output stage as well as the B+ for the input stage. About 100 mVpp of ripple on all of them.