[Maxx134]

I wanted to test these 8 resistors out before i swapped out all the other resistors out long with the additional cap changes and maybe the output wiring.

One thing is for sure though, Sonic improvements or not, the stock resistors are garbage.

Yes because they are most probably low cost low grade resistors, but only the resistors around tubes (grid,anode,cathode) were worth changing, not all the resistors..

 

[bloodhawk]

Don't forget that the quality of the Mills wirewound resistor is superior to a film or carbon or other composition resistor as the the electrons don't have to pass thru different materials with their inherent impurities.

It was mentioned in the other thread one user used carbon resistors when upgrading the power stage and noticed some different added aspect to the sound.

I myself tried another older wirewound and noticed an odd thinness as they were Inductive.

The Mills are non-inductive so they are the best choice in the ouput powertube stage.

Takmans or other film resistors are just fine in the driver stage the resistor upgrade was negligible.
It was the bias change that made the biggest massive difference there.

They are probably good but also probably carbon type and will not be best choice. Mills MR5 best choice for that output stage.

Do not decieve yourself,
You will ALWAYS hear all and any changes when you are using your own system in your own home that you are familiar with.
Familiarity and knowing ones own system is the key to distinguish things.
This is why "having it at home" is the best way to evaluate a product.

Perhaps. I swapped in the Mills 330Ohm MRA5 again today and have been listening to music for about 3 hours now. Still sounds exactly the same as the Kiwame 5W Carbon Film resistors. The background is pretty much blank with both Kiwame and Mills. One thing i noticed with all the Kiwames is that they read much better on my Fluke Multimeter. The Mills seem to be 1-1.5% tolerance. Kiwame's are under 0.75% tolerance and mostly around 0.5%, going as low as 0.4%.

End of the day im happy with either, so will probably with Kiwame's , since i like the color haha. Mills will be used in another high end project.

Also when i mentioned i will be replacing all the resistors, i meant all the resistors that were mentioned by you guys!

 

[SonicTrance]

Im sure the Kiwames are fine. The nr 1 reason for changing the anode and cathode resistors is to get rid of the crap stock resistors that burn under normal use.

 

[baronbeehive]

The amp sounds much better with the higher output 6SL7 (or 12AX7 in MK8).
I find with the change done to increase the bias point really sovled that issue and also used the higher gain setting,
But for MK6 the higher gain is stated by others to not sound as good.

Maye the gain settings combined with other tweaks is another area to look at, I think Sonic tested his gain settings prior to doing the mods, I could test it after the mods to see if it still sounds crap with the different driver bias setting for example. The amp might benefit from less feedback if we could overcome the problems that might cause somehow as you seem to have found.

Plus, I got carried away and re-tinned every solder point on the circuit board(!) with Cardas Solder...

Hey, I thought I was the perfectionist around here, but I didn't go to those extreme lengths!

Have you found the problem with one of your bias meters dropping after warmup?

It was mentioned in the other thread one user used carbon resistors when upgrading the power stage and noticed some different added aspect to the sound.

They are probably good but also probably carbon type and will not be best choice. Mills MR5 best choice for that output stage.

Yes I remember someone saying that. I had them fitted before the mods too and I didn't notice anything bad about them.

 

[baronbeehive]

Yes, I use a psu like this, with a 4-pin molex plug:


And then I connect it to this:


Remote controlled fans is the only way to go

That's for a dimmer switch isn't it?

Is the WiFi receiver in the 3 plugs or do you need anything else with this kit?

I was looking into this because I'm still waiting on my replacement speed controller from Hong Kong. I think the SMPS must have failed with my existing one, I can't see how the rheostat could have failed because it's sealed in.

That's why I was looking into the remote. The other option would be to have a power supply with a fixed 5v output rather than the 1-12v. I think 5 volt would be enough for a speed somewhere in the mid range of operation.

Failing that I could always knock up one myself I suppose.

 

[SonicTrance]

That's for a dimmer switch isn't it?

No, just power on or off. There's a reciever in the plug. You dont need anything else. The remote can control three plugs. Ive had an entire power strip on one plug. Very handy if you want to power on/off a lot of devices at once.

Some fans wont run at 5V. Check specs before buying.

 

[baronbeehive]

No, just power on or off. There's a reciever in the plug. You dont need anything else. The remote can control three plugs. Ive had an entire power strip on one plug. Very handy if you want to power on/off a lot of devices at once.

How do you control your fan speed then?

These fans are so powerful that max speed is too much, and the noise of the air flow replaces the bearing noise we had with the old fans.

 

[Maxx134]

End of the day im happy with either, so will probably with Kiwame's , since i like the color haha. Mills will be used in another high end project.

That is great.
Go with what you like.
Most performance issues are only at extremes.
Carbon Resistors do generate thermal noise.
Our amp gets hot, but with fan and
The 5W rating ,
It Most likely make it negligible as you have noted.
Still for impulse and other heavy duty characteristics, wirewound is Superior.

Maye the gain settings combined with other tweaks is another area to look at,

Yes, so I encourage bias testing for both high and low gain (NFB) settings,
as it should be slightly different sweet spots depending also on which tubes used.

Also I did notice a difference in sensitivity with change of power tubes, as I am not using stock tube type in output stage either, (I using old cheap Russian 6h6 at higher bias),
So I must agree what MrCurwen aludes to the input grid sensitivity of the power stage will affect your settings on the driver output.
So you should use the new bias points as a place to start, and adjust the Anode lower untill you find a good spot.

Then setup a CCS after you find your best bias points.

Yet I remember both me and Sonic found massive improvement with higher bias so you can't go wrong.

For my MK8,
I use tubes that worked well on high gain setting (less nfb), and then adjusted bias.
In general, both stages did better with higher bias settings.

I remember Sonic changed his output tube bias right?
Not the impedance mod.
We need find write good bias settings for output tubes of mk6

Have you found the problem with one of your bias meters dropping after warmup?

It was not meter , it was a "cold solder" point in the driver stage socket.

 

[SonicTrance]

How do you control your fan speed then?

I use fan speed reducer adapters. They have a dropping resistor on the +12V wire. That's easiest for me as I really don't need to change the speed. I just want it to be quiet and at the same time cool the components.

I remember Sonic changed his output tube bias right?

That's the 5998 bias mod you're thinking about. That one actually made it to the first page! Ha!

Might just end up replacing the each and every wire with 22 AWG and 18 AWG single core silver plated copper wiring.

If you replace the wiring from the transistors on top side of board (recommended as they're going to break loose at some point) you need to use a flexible wire. If you use single core wire there it will also be prone to break from solder connection when you put the board back in. I used 20 awg stranded silver plated PTFE wire. Pretty stiff but OK to use, have no issues.

 

[bloodhawk]

If you replace the wiring from the transistors on top side of board (recommended as they're going to break loose at some point) you need to use a flexible wire. If you use single core wire there it will also be prone to break from solder connection when you put the board back in. I used 20 awg stranded silver plated PTFE wire. Pretty stiff but OK to use, have no issues.

Gotcha, yeah i realized for a lot of the wiring at the back that doesnt need to be routed around using flexible wiring would be better. I might still end up using single core wire routed around the board for the output. But ill know more once i take the whole board out.

Thanks again for the helpful replies guys!

 

[Maxx134]

That one actually made it to the first page! Ha!

Good!
Let me know if you wanna put anything else first page,
If Ridge78 not available (as he been MIA here),
I can squeeze in my old posts.

Basically only need some topics baronbeehive mentioned.

 

[baronbeehive]

Yes, so I encourage bias testing for both high and low gain (NFB) settings,
as it should be slightly different sweet spots depending also on which tubes used.

Also I did notice a difference in sensitivity with change of power tubes, as I am not using stock tube type in output stage either, (I using old cheap Russian 6h6 at higher bias),
So I must agree what MrCurwen aludes to the input grid sensitivity of the power stage will affect your settings on the driver output.
So you should use the new bias points as a place to start, and adjust the Anode lower untill you find a good spot.

Then setup a CCS after you find your best bias points.

MrCurwen did suggest removing NFB from the APPJ and went on to suggest other tweaks to mitigate that, that's why I thought about trying to optimize the high gain setting on the MKVI.

I've lowered the driver bias to the Sonic settings near enough with the parallel resistor add on. So I would start with this new setting and maybe try another setting intermediate between this and the stock 220K ohms. I'm not sure how less feedback affects other component settings, generally I would think that some components need to be higher and some lower, sorry if that's pretty vague.

Then on to the CCS mod. This should have a big effect on the amp. As you know I have already said that one area, the only area, that the MKVI is slightly down on comparisons with the APPJ is on articulation, especially transients, at least with my tube setup, and this could change a lot with the CCS mod. Also a different tube type, power or driver would help, the GE's that I got recently are much sharper in sound so should be better in this respect.

You are ready to do this mod right?

For my MK8,
I use tubes that worked well on high gain setting (less nfb), and then adjusted bias.
In general, both stages did better with higher bias settings.

I remember Sonic changed his output tube bias right?
Not the impedance mod.
We need find write good bias settings for output tubes of mk6

Yes Sonic has optimized power tube bias for the 6AS7G's/421A's, and I have also done this mod. I don't plan on using any other power tube types so power tube bias is already done for me. The optimum point on the tube load line does not change with differing NFB does it?

If so all I need to do is adjust the driver bias again with the high gain setting

As I said I'm going to continue using my existing power tube types so I don't know if I can find a more optimum type to work well on high gain as you have done. How would I do this anyway?

ATM I will continue with burn in when I get my speed controller from HK probably next week, only after burn in can I do the sound tests when the system has settled down. But I don't plan to do any more mods until I get the last 421A because then I can fix the resistor value for these tubes, and go on then to complete the mods.

 

[bloodhawk]

Anyone ever try the RIFA Class X1 caps ? The PME-271E.

From the information i could find they are the 300V variants, and have higher voltage tolerances. I haven't been able to find a solid way to find the proper capacitance tolerance though. Since the Class X2 PME271M' have both a 10% and a 20% variant.

 

[MrCurwen]

Are there any capacitance critical spots in these amps? Coupling cap values are not at all important within 20% ranges.

 

[bloodhawk]

Are there any capacitance critical spots in these amps? Coupling cap values are not at all important within 20% ranges.

I honestly do not know the answer to that. But from what i do know, these are being used to bypass the 220uF Cathode bypass Caps in parallel.