[CopperFox]

I'm assuming when you say you use Tungsols and Shuguangs in combination you mean having 2 Tungsols and 2 Shuguangs to make up the 4 power tubes? This is going beyond where I've gone, I haven't looked into this exhaustively, only tried a couple of combinations which weren't very promising.

Yes, such as this:



Here's the 6SL7-Ts as drivers and Tung-sol 6080 and Shuguang 6N5PJ white top as power tubes. My most used combination although I usually have the rear tubes the other way round with the Tung-Sols on the left.

Did you notice any difference in sound characteristics such as harmonics with the film cathode caps? The lytics may output more harmonics I don't know.

The sound is cleaner and clearer with the film caps, without that additional edginess and without the softer bass that the electrolytics had. I don't know how those would be described in terms of harmonic distortion. Haven't seen any measurements that would have compared electrolytics vs film caps cathode bypass and don't have an oscilloscope. Would be interesting!

 

[CopperFox]

The 200uf cathode bypass caps that I now have are working great and since they took a long time to burn in I probably won't try other values there. Probably not for the PSU ouput caps either. One thing that I had thought up about those is that it might be possible to install the 200uf value caps there with shorter wires, which might or might not have lower total resistance as the wiring is probably something like 1 mOhm per centimeter. The Wima DCL-MKP datasheet specifies the 400V 400uf version as having 1 mOhm ESR and 400V 200uf having 1,2 mOhm ESR so the wiring can have higher resistance than the caps themselves.

I realised yesterday that with the preamp outputs removed there is now more room in the case to rearrange the 400uf PSU output caps with shorter wires so I did that:
(so will not likely put 200uf caps there)



It was also possible to rearrange the cathode bypass caps and power tube decoupling caps for shorter total wiring length. Looks almost tidy! Also used some rather thick silver wire for those. A keen eye may also notice the PSU electrolytics have been replaced... have had those there for a couple of weeks. will comment more on those later.

I also put the impedance mod back on after coming to the realisation that the reason it didn't work before were likely the 100 Ohm resistors at the output. Now with those removed it does seem to work.

 

[baronbeehive]

Thinking of trying these, cheap for Shuguang!

https://www.ebay.co.uk/itm/265064525748?hash=item3db713afb4:g:0DwAAOSwvVxgNPXC

I assume these are the clear tops you mention? Sound like a good replacement for Tungsols if there aren't any NOS available.

Edit: I agree with Maxx I never saw any resistors on the output wires and I replaced those.

 

[CopperFox]

Thinking of trying these, cheap for Shuguang!

https://www.ebay.co.uk/itm/265064525748?hash=item3db713afb4:g:0DwAAOSwvVxgNPXC

I assume these are the clear tops you mention? Sound like a good replacement for Tungsols if there aren't any NOS available.

Edit: I agree with Maxx I never saw any resistors on the output wires and I replaced those.

The site where they are most widely available probably is Aliexpress where they can be often found for ~$20 a pair. The reason they're sometimes listed as "Dawn" or "Dawning" tubes there is that it is the literal translation of the brand's name. And the model name is 6N5P to be specific, and the J in the end is a military-grade designation. They don't seem to be a very popular tube as they do sound too soft in most amplifiers. Seems to be more supply than demand for them.

The TS 6080s do also still seem to be available at places like ebay.

One power tube that I haven't tried are the Tung-Sol 5998s, which seem to be the most exepnsive ones here. I also haven't tried any of the other US/EUR big bottle tubes (RCA, Philips GE etc.) yet either.

 

[CopperFox]

I have spent at least $2k back then on every 12pin and 6pin, and even some 8pin with adapters, rolling everything into the driver stage.

The problem was that the stock bias was set so low, that all the tubes were away from there ideal performance. The result was a more pronounced tube differences when rolling.

That being said, only one tube rose from the ashes... The Tung-sol 6c8g tube (with adapter for top). My choice anyways.
It is older version tube than 6SL7, and had the higher "MU" output level.
I had also optimize my bias at that point.

How does the Tung-Sol 6C8G sound - what makes it stand out, does it have some sort of particular emphasis, and how's the bass with it?
I could try them at some point... wonder if it would work ok with the 470R cathode resistor value or need 1K or something.

 

[baronbeehive]

The site where they are most widely available probably is Aliexpress where they can be often found for ~$20 a pair. The reason they're sometimes listed as "Dawn" or "Dawning" tubes there is that it is the literal translation of the brand's name. And the model name is 6N5P to be specific, and the J in the end is a military-grade designation. They don't seem to be a very popular tube as they do sound too soft in most amplifiers. Seems to be more supply than demand for them.

The TS 6080s do also still seem to be available at places like ebay.

One power tube that I haven't tried are the Tung-Sol 5998s, which seem to be the most exepnsive ones here. I also haven't tried any of the other US/EUR big bottle tubes (RCA, Philips GE etc.) yet either.

Right, they could be worth a try to replace give my RCA's a rest! I sat up and noticed when you mentioned the soundstage! The softness agrees with some of the negative comments I have heard about Shuguangs and why I steered clear but as you say with the mods, especially the film caps then that could make the difference.

I've tried Tungsol 7236's, found them very analytic so they went straight out, also tried WE 421A's which I liked a lot. They're similar to TS5998's in that they are quite relaxed and flowing and airy but are a bit sharper in attack I think with harder hitting bass. The other ones I had were Bendix 6080WB's with graphite plates which I also liked. I had 5 or 6 sets of these but eventually sold them as I always liked the common or garden RCA's. And after the mods the sound is so good that it didn't really matter what tubes were there because they were all elevated in sound IMO.

Also tried GE 12SL7's, same as 6SL7's but with adapters, typical GE sound very clean and clear but a tad one dimensional.

GEC's are one of the best, very detailed apparently, one of the members here had them but they were incredibly pricey and rare for me to bother with.

Edit: that TS6C8G must be good if Maxx raved about it! I haven' tried it. The cathodes won't make much difference in sound just very fine tuning. I found the 470R to be the best but anything up to about 1K would be fine.

 

[Maxx134]

One power tube that I haven't tried are the Tung-Sol 5998s, which seem to be the most exepnsive ones here. I also haven't tried any of the other US/EUR big bottle tubes (RCA, Philips GE etc.) yet either

For the WCF to perform optimally, the tubes should be if the same model number. I never mixed tubes with different model numbers, as I didn't think possible unless similar.
Also there was another mod to switch bias for other power tubes.

Check your anode resistors in case of excess burn from unbalanced tubes, but if your replacement resistors are wirewound resistors, you should be ok.

How does the Tung-Sol 6C8G sound - what makes it stand out, does it have some sort of particular emphasis, and how's the bass with it?
I could try them at some point... wonder if it would work ok with the 470R cathode resistor value or need 1K or something.

They had the best transparency, with lively and balanced sound, while also being good at top and bottom FR extremes. I just used adapters for them.

As for those

 

[Maxx134]

So as they now worked in my system I'd rank them as follows: First, Audio Note Niobium non-magnetic - these are worth their price here and they are in a different league to the rest. Second, the Kiwames - they have good punchy and deep bass and are relatively cheap too even if they don't have the very best clarity. Third, the AN tantalums which had many things in common with the Kiwames but were worse. Fourth the TKD metal films which have a crisp character but are low on bass. Fifth and last the Ohmite Audio Golds which didn't really stand out with their characteristics

This is excellent info! Good work!

Do the WCF caps need to be "audio caps" or just have the best electrical properties for power delivery - (how much) does the audio signal pass through them?

Yes as @baronbeehive stated very well, they are smack in line between the output tube triodes, so they need to be quality.
A good point was made about the capacity. We have not tried larger WCF caps. What size are yours. I only tried .33uf, but according to Baron, we might see current gains with larger size, although I do believe the current is mainly a tube limitation.

Also, what about the eight small Wimas that are located between the power tubes and PSU section? At least some people have replaced those with something else, such as the Mundorfs here. Is their function power delivery or something else?

Those are strictly PSU caps design to keep the circuit from having any noise like hum or oscillation. They will not affect sound, only isolation.

Would the opamps' biasing function somehow rely on feedback from the headphone output? And if yes, could it be that having the impedance matcher attached somehow cancels some of this?

Nope. Opamps only affect DC-offset, which would result in a pop sound (when headphones insertion) if they were off by any small amount.

I've been looking into this too. This effect happens with the B10 impedance matcher connected regardless of the tubes. At startup the vu meters both show 60 ma, but the left then begins slowly approaching 40 ma or lower and the right side begins approaching 80 ma or higher. This doesn't fix itself (but instead gets worse) if I leave the amp running for a long time. It also happens only with the impedance matcher attached, and is reset (so that the readings are 60 and 60) if the amplifier is turned off and on again

My guess is that the tubes are having a bad reaction and causing the meters to slowly change.
Check for tube thermal runaway by swapping tubes to other meter(other side).

PSU caps

That may be a bad idea to put film caps where the 4 big PSU electrolytics are..
The reason is that we do not want such a fast load on the bridge Rectifier. This is why there are those huge resistors under PSU section, to absorb the current surge when you turn on the unit...

I thought it was established early in the thread that the opamps are used for biasing here?

And that the blue trimpots are there to set some sort of starting point for the biasing, which will be affected by feedback from the output? (Is any of this correct? )

And if so, would the values from the trimpots somehow be multiplied or magnified in absence of normal/expected feedback from output?

In that case, some sort of imbalance in the trimpot adjustments could cause growing imbalance in the meters and also be redeemable by adjusting the trimpots.

But I don't know how to adjust the trimpots... how should that be done?

Check the schematic and you will see it does take a bit from the outputs as there is a very high value resistor at the output (on the board). This DC-offset it critical, as if it was off by a large amount (enough to move meters), then it would create a loud pop and trigger the output relays.
So this is not happening and most likely no one ever has to adjust the trim pots.
It can be "fine-tuned" if you hear a tiny click/pop when plugging in headphones, but try it after it is warmed up, not when just turned on as that's when circuit settles.

Edit: This may also explain why the Impedance matcher didn't work on the LD

I don't know why, but I have permanently installed different anode resistors to my output tubes to change the impedance out from a calculated 600 ohm to a 50ohm headphones impedance, so the calculations to change output impedance work.
It changed the sound of a stock HD800 from a more mids, to a bit recessed mids, which is how it sounds with solid state.
At the time I changed my amp, It was tuned for the "Ether" planar, which was around 50ohms. So output was calculated for a 50 ohm load.

What impedance load have you tried set it to? It cannot be zero as it must optimally be the setting of the headphones impedance.
I thought the mk6 was already set for around 50ohm load?

 

[baronbeehive]

For the WCF to perform optimally, the tubes should be if the same model number. I never mixed tubes with different model numbers, as I didn't think possible unless similar.
Also there was another mod to switch bias for other power tubes.

The only person who did that was coin I think!

Re: the power tube mod I tried this, and the only effect I found was an increase in volume which gave a perceived improvement in sound, but when the volume was adjusted back I didn't really find any difference so I would think that the primary benefit of this would be to get the tubes operating at their optimum current, but some may find it preferable to leave the 5998's to operate at half current and therefore much cooler.

 

[baronbeehive]

Yes as @baronbeehive stated very well, they are smack in line between the output tube triodes, so they need to be quality.
A good point was made about the capacity. We have not tried larger WCF caps. What size are yours. I only tried .33uf, but according to Baron, we might see current gains with larger size, although I do believe the current is mainly a tube limitation.

A further point we need fast acting caps here so something like Mundorf Supreme SIO because they are film caps, being metallized PP, as well as being audio caps.

I believe the MKVI is less susceptible to current fluctuations than the MK8? I never saw the meters fluctuate on power surges/dips so it would be difficult to check visually with different WCF cap sizes. Maybe though I just don't play it loud enough to see these performance effects.

I don't know why, but I have permanently installed different anode resistors to my output tubes to change the impedance out from a calculated 600 ohm to a 50ohm headphones impedance, so the calculations to change output impedance work.
It changed the sound of a stock HD800 from a more mids, to a bit recessed mids, which is how it sounds with solid state.
At the time I changed my amp, It was tuned for the "Ether" planar, which was around 50ohms. So output was calculated for a 50 ohm load.

What impedance load have you tried set it to? It cannot be zero as it must optimally be the setting of the headphones impedance.
I thought the mk6 was already set for around 50ohm load?

I changed mine to 38 ohm load for the HE-500's. CopperFox changed his even lower for his very low impedance Monolith headphones.

 

[CopperFox]

Right, they could be worth a try to replace give my RCA's a rest! I sat up and noticed when you mentioned the soundstage! The softness agrees with some of the negative comments I have heard about Shuguangs and why I steered clear but as you say with the mods, especially the film caps then that could make the difference.

Seeing that the decoupling caps and film caps for psu output improved the 6N5Ps' sound, I (only) now decided to look at their data sheet to see if there's something that might explain that. It looks like their plate resistance would be 450 Ohms and transconductance 4450 micromhos, while most 6080 / 6AS7 tubes have these these values at ~280 and ~7000. So they are not exactly the same there.

And it also means that although the resistors that I added for the impedance mod are correct for (most) 6080/6AS7 tubes, they may not be of the correct value for 6N5P - as the values for the added resistors were calculated with 6AS7 values. So as I now have two Tung-Sol 6080WAs and two Shuguang 6N5Ps for power tubes, the value may be wrong for the latter two.

As for the correct resistor values for the impedance mod, the Cavalli formula again is Ra = (1/Gm*(1/Ka*(1+Ka)/(1+Kr)*(Rp+Rl)/Rp+Rl/Rp)) .

For the 6N5Ps in the mkVi+,

Gm = 0,00445
Rp = 450
RI = 18 (for the M1060Cs)
Ka = 0,7
Kr = 0,75

So the equation is Ra = (1/0,00445*(1/0,7*(1+0,7)/(1+0,75)*(450+18)/450+18/450))

Which results in ~333,3180 so it is roughly 333 ohms.

But... Isn't that almost exactly the default value for the power resistors here?
Maybe I could try removing my previously added 680 ohm resistors from the 6N5P's part of the board to see if that makes a difference or not. Or could there perhaps be some reason why this wouldn't be a good idea, if I'd still have the added resistors for the TS 6080s side?

And now it looks like the 6N5Ps actually need less power than the 6080/6AS7 tubes for 18 ohm impedance headphones... but they do still prefer it to be quick power?

With the impedance mod I now have in place that was calculated with the 6AS7 values (resulting in 220 ohms), the 6N5Ps now seem to be also getting more power than they need. Wonder if that is harmful in some way.

 

[CopperFox]

For the WCF to perform optimally, the tubes should be if the same model number. I never mixed tubes with different model numbers, as I didn't think possible unless similar.
Also there was another mod to switch bias for other power tubes.

Check your anode resistors in case of excess burn from unbalanced tubes, but if your replacement resistors are wirewound resistors, you should be ok.

Hmm, I had been mixing the power tubes (different 6080s and 6AS7Gs) before I begun with the mods. The people in the other mkVI thread had been doing this too.
It is actually one of the reasons I like this amp. With a combination of two different power tubes on each channel it is possible to get characteristics from two different power tubes at the same time and imo the combinations also sometimes do things that the respective single tubes can't do on their own as a set of four.

I think for now I would prefer to have this ability to over optimum WCF performance, if it works ok.

They had the best transparency, with lively and balanced sound, while also being good at top and bottom FR extremes. I just used adapters for them.

Sounds good. The 6C8G's values for amplification factor, plate current and plate resistance seem to be somewhere between 6SN7 and 6SL7. Could try them in my 6SN7 amp as well. I think I'll get a couple with adapters as I'm going to be getting some new tubes next week. Also going to order some 6SL7s from the newest Chinese brand Linlai who have released three different models of them in recent weeks. Possibly the JJ Electronic 6SL7 as well.

 

[baronbeehive]

So the equation is Ra = (1/0,00445*(1/0,7*(1+0,7)/(1+0,75)*(450+18)/450+18/450))

Which results in ~333,3180 so it is roughly 333 ohms.

But... Isn't that almost exactly the default value for the power resistors here?
Maybe I could try removing my previously added 680 ohm resistors from the 6N5P's part of the board to see if that makes a difference or not. Or could there perhaps be some reason why this wouldn't be a good idea, if I'd still have the added resistors for the TS 6080s side?

And now it looks like the 6N5Ps actually need less power than the 6080/6AS7 tubes for 18 ohm impedance headphones... but they do still prefer it to be quick power?

With the impedance mod I now have in place that was calculated with the 6AS7 values (resulting in 220 ohms), the 6N5Ps now seem to be also getting more power than they need. Wonder if that is harmful in some way.

.

As someone who hasn't done this to any extent I hadn't thought about this, so don't know about that .

So as I understand it you want to remove just one of the 2 resistors making up one of the channels. It sounds a weird thing, each channel constantly fluctuating in output impedances during the WCF loop. As the WCF function is the basis for the amps functioning I doubt this would be a good thing and thinking about it it certainly doesn't sound a good idea*. I think this is a case of customizing gone mad, because you have to start thinking about the effects of one thing you did every time you start changing the mix.

Sonic implemented a switch for his impedance mod but that was to change it with a change of headphones. I suppose it would be possible to have each channel output a different impedance so long as it wasn't constantly fluctuating but you wouldn't want to do that for the sake of your ears .

What do your meters read for the 6N5P's?

* Disclaimer: Of course I may be totally wrong LOL

Edit: I think I know someone who would have tried this if he could though, naming no names..., that doesn't mean you should try it necessarily .

 

[baronbeehive]

We're getting into strange territory here because as I understand it each half of each channel operates in a phase/antiphase way so that each identical half cancels each other out regarding noise artifacts so what happens when you unbalance the channels like this.. I don't know is the short answer .

Edit: Perhaps if we both fall down that rabbit hole and meet on the other side we might see what goes on in this peculiar world where up is down and vica versa, and we can sit at the table and discuss it with the Mad Hatter who is sitting at the head of table under the grinning cheshire cat .

 

[CopperFox]

.

As someone who hasn't done this to any extent I hadn't thought about this, so don't know about that .

So as I understand it you want to remove just one of the 2 resistors making up one of the channels. It sounds a weird thing, each channel constantly fluctuating in output impedances during the WCF loop. As the WCF function is the basis for the amps functioning I doubt this would be a good thing and thinking about it it certainly doesn't sound a good idea*. I think this is a case of customizing gone mad, because you have to start thinking about the effects of one thing you did every time you start changing the mix.

Sonic implemented a switch for his impedance mod but that was to change it with a change of headphones. I suppose it would be possible to have each channel output a different impedance so long as it wasn't constantly fluctuating but you wouldn't want to do that for the sake of your ears .

What I meant is remove the impedance mod (which I have installed based on calculations for 6AS7 values) from the two power rails that feed the 6N5P tubes, as the default resistor has the correct value for them and leave the mod on for the power rails that feed the 6080 tubes, so those two will have the added resistors delivering more power.

Here the idea would be to reduce the imbalance between the 6080 and 6N5P on the same channel. It is also an imbalance that I've had for most of the time for the last six months as I've been using the combination of the Tung-Sol 6080WA and the Shuguang 6N5P with the same power resistor values, be it with or without the impedance mod.

As for the possible switches, the amplifier already does have individual switches (for low/high gain) leading to each of the four power tubes. Maybe it wouldn't be exceedingly difficult to also have individual 6080/6N5P/5998 switches for them as well

What do your meters read for the 6N5P's?

Since starting with the new PCB in spring, the meters have been steadily at ~60 ma values on all kinds of tubes, only hopping upwards on bass hits or having some imbalance when the impedance matcher is attached. The impedance mod had no effect on the values.

We're getting into strange territory here because as I understand it each half of each channel operates in a phase/antiphase way so that each identical half cancels each other out regarding noise artifacts so what happens when you unbalance the channels like this.. I don't know is the short answer .

Edit: Perhaps if we both fall down that rabbit hole and meet on the other side we might see what goes on in this peculiar world where up is down and vica versa, and we can sit at the table and discuss it with the Mad Hatter who is sitting at the head of table under the grinning cheshire cat .

Well, as stated above, I have already had that kind of imbalance for the last six months. Now that I think of it, there has been this effect which could occur if I left the amp running for a longer time - above 90 minutes - where it sounded like the sound characteristics of one of the two mixed power tube types would somehow become more distant or hollow sounding. I think I had thought it is normal for tube amps to have their sound vary over time with tubes warming up and capacitors charging up, but it could perhaps possibly be better attributed to phase or impedance fluctuations in the WCF. Rebooting the amp would change it to varying degrees, not always fixing it. But putting four 6N5Ps on instead of the mix always did.

Anyways, I now have the amp back up running after removing the impedance mod resistors from the 6N5P's part of the circuit. I'll keep an eye on if there is some change on that temporal stability thing over a few weeks.