[CopperFox]

Cathode1.5k's circled in red, anode 220k's circled in yellow below:

Thanks! I assume the recommended values (from the previous graph) are 46K for the both two yellow-circled position resistors (on each side) and 390R for the one circled in red on each side.

Edit: What is the "Last R" 3K3 resistor/value?

 

[baronbeehive]

Thanks! I assume the recommended values (from the previous graph) are 46K for the both two yellow-circled position resistors (on each side) and 390R for the one circled in red on each side.

The best values I found were 69K and 380R. Maxx also preferred 69K but he went with 1K for cathode. I had 500R before I put in the CCS, so anything between 380R and 1K would suffice.

When you have done this you should find that the sound and liveliness is pushed to end game, and the impact and textural detail is fantastic, violins scratch, drums crash, piano notes come out forcibly, guitar picks hit you palpably, and vocals are stunning, all have that detail that says the sound is real and "in your room"!

 

[baronbeehive]

Re: decoupling caps, we felt the need to add them to the output tubes as well as the driver tubes.

This is a quote from page 1.. always a handy reference LOL! "To improve this PSU, we have decided to add to the original design a "decoupling" cap near each power tube. The current feeding the headphones will be provided by the PSU and/or the decoupling cap, depending on the respective impedence at this time."

 

[baronbeehive]

Edit: What is the "Last R" 3K3 resistor/value?

We went down to 1K for the last resistor in the CRC power supply filter. This boosted B+ to around 185V.

 

[CopperFox]

We went down to 1K for the last resistor in the CRC power supply filter. This boosted B+ to around 185V.

Should I replace that too now, or should the default resistor stay there if it's part of another mod?
Oh, and where is this resistor located

 

[baronbeehive]

Should I replace that too now, or should the default resistor stay there if it's part of another mod?
Oh, and where is this resistor located

You can replace them now, it won't interfere with anything else.

So how far do you want to go with the mods? It might be an idea to figure that out beforehand so that, for example you can do everything you want if you have to remove the PCB again. LD obviously didn't want anyone to interfere with the insides of their amp because they fitted such flimsy wires that they were bound to break the moment you opened it up, unless you were extremely careful, they have no sense of fun! Make sure you check these before you put the board back, or better still replace some of the wires perhaps. We have all torn our hair out many times over this!

The further you go with the mods the more you need to think about uprating other components eventually, for example you should uprate the resistor wattage from stock values anyway. Being slightly off the wall, we went the extra mile and uprated much of the insides, just for something to do, especially around the PSU area where the stock garbage resistors are likely to cook eventually.

One other thing switch over to high gain, with less negative feedback the sound is better.

Good luck, and happy modding!

 

[Maxx134]

There has been allot of changes and variations to the mods later in this thread, and so I see that the first page needs a bit of updating, so I will try to update things next week.

I do not believe @Redge78 is around to do it.

 

[CopperFox]

Okay, so I replaced the power and anode/cathode resistors, added decoupling caps for the driver tubes and before those did some cathode bypass cap rolling.
Also re-soldered the Miflexes with 3% silver solder.

With the cathode bypass caps I tried Elna Silmic RFS 1000uf 50v and Nichicon KZ 1000uf 50v. Both with the same Kemet/Rifa 0.1uf 220v bypass added.
Both of the 1000uf caps sounded worse than the 470uf Nichicon KZ. The 470uf had clearly more bass slam and dynamics than the larger capacities. The Elna had better mids but less bass slam than the 1000uf KZ. Did about one week of burn in on both of the larger caps and there was almost no change in sound in them. The 470uf KZ had had a clearly perceivable gradual change during burn in. So I got tired and put them back in.

For the power resistors used the Mills 330ohm 5Ws. Most of the brands/products that were recommended on the first page for the anode/cathode positions were not available so I used Kiwame resistors for those since they seemed to be both highly rated and affordable and could not find any descriptions of negative experiences with them. used 68K for anode (69K was not availabe) and got three different values to try for the cathode position, which were 1k, 470 and 390. The 1k sounded very flat and the 390 had very low amount of bass, but the 470 brought a clear improvement in dynamics and detail. Would be interesting to try the next two values, which would be 510 and 430.

EDIT: Saw some old posts that stated that the impedance of headphones used would have an effect on the correct value here. Mine are modded Monolith M1060Cs whose rated impedance is 18 ohm.


The old power resistors were in good condition. There were some other resistors which had discoloration though:



Those four stacked resistors at the top which are near the power capacitors. Are there any recommendations/mods as to how those should be replaced?

The driver tube decoupling capacitors, where I used the 100uf 500v Wima MKP4s, did have effects on sound. Less noise/distortion, more definition, more nuance. Also the sound seems to have less difference at different volumes. Installed those just two days ago though so they're probably still burning in. Will need to see what happens and also to do more tube rolling. Will try the decoupling caps on power tubes later as well. But first I need to figure out where to put them so that they stay in one place and don't get too hot. I'll probably add one or two more fans to the bottom plate as the rear middle part of the amp can get pretty hot.

The various mods have also changed the way the different tubes sound. Looks like the larger bottle-shaped power tubes have improved the most (Svetlana 6N13S and Shuguang 6N5PJ).

Also tried some 6SN7 tubes after the latest changes and they sounded ok at least. Made me wonder if there is some way to "optimize" the amp for 6SN7's by changing the resistor values.

 

[CopperFox]

So how far do you want to go with the mods? It might be an idea to figure that out beforehand so that, for example you can do everything you want if you have to remove the PCB again. LD obviously didn't want anyone to interfere with the insides of their amp because they fitted such flimsy wires that they were bound to break the moment you opened it up, unless you were extremely careful, they have no sense of fun! Make sure you check these before you put the board back, or better still replace some of the wires perhaps. We have all torn our hair out many times over this!

(...)

One other thing switch over to high gain, with less negative feedback the sound is better.
Good luck, and happy modding!

Well that's a tough question as I'm not sure if I'm aware of all the mods that can be done.
Replacing all the wires with spc sounds too tedious at the moment.

The stock wires haven't felt brittle or come loose yet. They're thicker and and have more solder on them than DAP wirings for an example.

By high gain do you mean the high gain mode that the manual only recommends for use with the AKG K1000s?

 

[baronbeehive]

With the cathode bypass caps I tried Elna Silmic RFS 1000uf 50v and Nichicon KZ 1000uf 50v. Both with the same Kemet/Rifa 0.1uf 220v bypass added.
Both of the 1000uf caps sounded worse than the 470uf Nichicon KZ. The 470uf had clearly more bass slam and dynamics than the larger capacities. The Elna had better mids but less bass slam than the 1000uf KZ. Did about one week of burn in on both of the larger caps and there was almost no change in sound in them. The 470uf KZ had had a clearly perceivable gradual change during burn in. So I got tired and put them back in.

Really good work CopperFox!

That is very valuable observation and backs up what I said about the tradeoff between capacity and speed. You said that the 470uF had more dynamics and this is very similar to what I found with another amp. In that amp the high capacity caps sounded dull and flat, but were relieved by putting on the bypass cap. It may be that a lower capacity bypass cap on the cathode bypass would work well here and bring out the higher frequencies more to make for a more lively sound. This is something that we haven't experimented with atm.

 

[baronbeehive]

For the power resistors used the Mills 330ohm 5Ws. Most of the brands/products that were recommended on the first page for the anode/cathode positions were not available so I used Kiwame resistors for those since they seemed to be both highly rated and affordable and could not find any descriptions of negative experiences with them. used 68K for anode (69K was not availabe) and got three different values to try for the cathode position, which were 1k, 470 and 390. The 1k sounded very flat and the 390 had very low amount of bass, but the 470 brought a clear improvement in dynamics and detail. Would be interesting to try the next two values, which would be 510 and 430.

That's another great observation! You say the 1K cathode was flat. I suspect there is an interaction with headphones and personal preference. Maxx has a 1K cathode and used to use Sennheiser HD800's. He likes the 1K, I don't know if he has tried lower like I did. I couldn't tell much difference between 390 and 500R, and there shouldn't be a massive difference because it wouldn't change the operating point that much. I felt that a cathode somewhere around those values produced the ideal amount of impact and texture.The point on the graph where the 6SL7 operating point is is clear of distortion but has less voltage swing. Again this is something we haven't one into in detail. It may be an idea to trade some linearity for some more voltage swing, although you don't need much swing with headphones.

I will post again with the graph to illustrate this.

 

[baronbeehive]

The drive tube decoupling capacitors, where I used the 100uf 500v Wima MKP4s, did have effects on sound. Less noise/distortion, more definition, more nuance. Also the sound seems to have less difference at different volumes. Installed those just two days ago though so they're probably still burning in. Will need to see what happens and also to do more tube rolling. Will try the decoupling caps on power tubes later as well. But first I need to figure out where to put them so that they stay in one place and don't get too hot. I'll probably add one or two more fans to the bottom plate as the rear middle part of the amp can get pretty hot.

More interesting observations LOL!

I think we were of the opinion that driver tube decoupling woun't have much effect because the signal is so small, and therefore decoupling would have more of an effect on the power tubes.

Did you install the larger fans or the same size as stock?

 

[baronbeehive]

We put our white decoupling caps here without any issues:

...they are very close to the hot 330R power resistors, but so long as you have good air flow.

Also you could put in a switch and resistor combo to change from 6SL7's to 6SN7's like here with a switch to change from 6AS7 to 421A power tubes.

We didn't optimise for 6SN7's ecause the higher gain of the 6SL7's is a better match for this amp, which is relatively low in power.

Edit: Sonic's amp is the most beautiful.

 

[baronbeehive]

Well that's a tough question as I'm not sure if I'm aware of all the mods that can be done.
Replacing all the wires with spc sounds too tedious at the moment.

I thought the same but for the few that are most important, like output wires and any that look fragile, it actually is quite a quick operation!

By high gain do you mean the high gain mode that the manual only recommends for use with the AKG K1000s?

Yes. It means less negative feedback in the loop so better transparency in theory, provided there is no distortion.

 

[SonicTrance]

We put our white decoupling caps here without any issues:

...they are very close to the hot 330R power resistors, but so long as you have good air flow.

Also you could put in a switch and resistor combo to change from 6SL7's to 6SN7's like here with a switch to change from 6AS7 to 421A power tubes.

I recognize this amp! It seems like forever ago now. Brings back memories =) I see that I hadn't completed all the mods yet in this pic. Here's a pic with the Ra mod and CCS tail completed as well. I think the Ra mod is what @CopperFox is talking about here:

EDIT: Saw some old posts that stated that the impedance of headphones used would have an effect on the correct value here. Mine are modded Monolith M1060Cs whose rated impedance is 18 ohm.

That is where we optimized the white cathode follower output stage for a specific load, using the Cavalli formula Ra = (1/Gm*(1/Ka*(1+Ka)/(1+Kr)*(Rp+Rl)/Rp+Rl/Rp))
I even went so far as to test my own tubes for actual gm, using my heathkit tester, and then calculate the different values to optimize the WCF for different tubes and loads.
I found this chart, this is for "stock" gm though: