DIY Audio Projects Forum

"The Beast" 6336 Dual SET Amplifier
Page 5 of 7

Author:  Suncalc [ 03 Feb 2019, 19:32 ]
Post subject:  Re: "The Beast" 6336 Dual SET Amplifier

First the Miller capacitance question. I calculated the C-M of the big triode at about 92 µµf at this bias point. Using the 14.9kΩ output impedance of the driver and the (470kΩ grid resistor) with my original 10kΩ grid stopper, this puts the high frequency rolloff at about 71kHz. This should not be affecting the high frequency response at 20kHz. Your 3.3kΩ stopper puts the high frequency rolloff at ≈97kHz. So the power stage grid stopper is not the issue. The Miller capacitance of the driver stages is only ≈20µµf bypassed and ≈13µµf unbypassed. With the 5kΩ grid stoppers these roll offs should be well above 400kHz. So this shouldn't be an issue.

The first thing I would recommend is to replace the 12BH7 with a 12AU7. Then see what the amp sounds like.

Author:  ToddD [ 03 Feb 2019, 22:12 ]
Post subject:  Re: "The Beast" 6336 Dual SET Amplifier

I replaced the 12BH7 tubes with the 12AU7 tubes and there was no change.

I am running this amp straight from a CD player with 2 Vrms signal. I have the 250K (audio) pot at about 9 o'clock for casual listening and at most 12 o'clock for loud listening. Is the resistance of the volume pot causing a problem? If I change to a 100K pot would that help? Could I add a voltage divider between the CD player and amp?

feedback appreciated.


Author:  Suncalc [ 04 Feb 2019, 14:11 ]
Post subject:  Re: "The Beast" 6336 Dual SET Amplifier

The 250kΩ volume pot peaks it presented impedance to the amplifier at 62.5kΩ+Rsource/4. The CD player probably has a very low output impedance (≈32Ω?). So the high frequency rolloff due to the volume pot would still be ≈127kHz. Switching to a 100kΩ pot shouldn't matter.

I'm kind of at a loss for where to go from here. Can you measure the passband of the amp from like 10Hz to 30kHz and see what it looks like?

Author:  ToddD [ 04 Feb 2019, 17:31 ]
Post subject:  Re: "The Beast" 6336 Dual SET Amplifier

When time permits, I will do a thorough check of the circuit. Every thing else I have checked so far (voltages, current, and gain) has been right in line. This problem appears to be in both channels and it has had the hf attenuation from the initial turn on but I had temporary speakers and I thought it might improve as it broke in.

Hopefully I will find the fault or at least have more information to work with.

Thanks for your help.

Author:  ToddD [ 10 Feb 2019, 20:17 ]
Post subject:  Re: "The Beast" 6336 Dual SET Amplifier

Ever try to do too many things at once?
I think between my wife's comments that it sounded "muffled" and me looking at the frequency response chart in the build, I jumped to conclusions that I may have Miller effect.

I took some time to play some pink noise and measure with a frequency response app on my phone. Not the best, but it does give me an indication. The readings indicate no noticable high frequency roll off. So I took so more time to listen to some different music and see if that told me anything. What I am hearing and probably what my wife is picking up on is a reduced volume at a frequency where some of the vocals fall. As a result on many recordings the vocals sound pushed back in the mix instead of up front. As a result they could be called "muffled".

Current setup is CD player (with 600 ohm output impedance) into the Beast and then to the speakers. Prior to the Beast I was using a preamp with tone controls, but I had to borrow tubes so I took out the preamp. I will get the pre/tone back in line and see if that is the difference I am hearing.

Otherwise, I matched the schematic and values except I used 0.1uF coupling caps instead of the 0.068uF caps. Not sure if my problem could be amp layout, component choices, or the sound of the output tube. I only have one so I do not have another to try out.

As for the lack of dynamics, my other amps have CCS loaded anodes in the driver stage. Would that be an option on the Beast? or choke loaded anodes?

But this amp does some things well. When I listen to Zeppelin, Page's guitar cuts through and Bonzo's drums have great weight. As I have said this amp has grunt.

Author:  ToddD [ 11 Mar 2019, 18:33 ]
Post subject:  Re: "The Beast" 6336 Dual SET Amplifier

I have spent the last month making changes and trying things and trying not to make too many changes at once.

In order as I remember them.

Initially I built this amp with all carbon comp and carbon film resistors. It is what I had and I guess I wanted to see if I would get that "euphonic sound" that I have seen attributed to carbon comp resistors. However, I think I ended up with a higher noise floor. Then I went all metal film and wire wound resistors. Wire wound on the cathode resistors and metal film everywhere else. This did lower the noise floor, but at the same time it also sounded metallic, no pun intended. Finally, I replaced the metal film grid stoppers with carbon comp. I think this is a good combination of low noise floor and euphonic sound. It was nice to make the changes and hear the subtle differences.

C1 in the power supply
I continued to try different first capacitor in the power supply. 22uF Solen then 10uF Solen then 7.5uF motor run. Of course each smaller capacitor brought down my B+, but this kept giving me improvement so I kept going lower. The change from the 10uF Solen to the 7.5uF motor run was considerable in sound. Better clarity and bounce. The music had more rhythm now. This was the best the amp had sounded. The amp had swing, so of course I changed it. but I made it such that I could go right back if I did not like the next change. A detour so to speak.

Choke loading on the plate for the first driver stage
In an effort for even more dynamics I replaced the 100K resistor on the driver plate with a choke. I took the idea from the La Luciernaga Amp on the website. A great mix of information and humour by the way. Like him I used a Hammond 156C 150H choke. Only on the left channel first stage and I kept the right channel unchanged. After letting it settle in, it was more dynamic, but I lost my low frequency grunt. As he points out in his write up, the voltage gain at the lower frequencies is less. Once it hits ~1280hz it has a gain of 17 with a 12bh7 tube, but below ~80 hz it is noticabley less. I used a 1000hz signal to check gain and on the left channel and I had a gain of 16 for the first stage and 7.4 on the second stage. A total gain of 118. By comparison the right channel was 12.76 on the first and 7.3 on the second for a total gain of 93.
All that to say I took out the choke loading on the plate for the first stage and went back to the 100K resistor. This gave the amp back its grunt and it still had its bounce and swing.

More work on the power supply
I used the PSUD2 program on Duncan's Amps page to see what other changes I should consider to streamline my changes. Two changes looked promising and worth trying based on parts I had on hand. I reduced the first cap, C1, to a 5uF motor run cap. This was another step in the right direction. More bounce and swing. Then next I added 30uF to the second cap. I had a 60uF already, so I paralleled a 30uF motor run for a total of 90uF. These were both a step in the right direction. Per the program both together gave quicker rise time and quicker settle time with less overshoot. I would say the amp has better damping. Also I added 0.1uF bypass caps on C1 and C2. I did this at the same time as the C2 change, so I can't say what effect if any these had, but the overall sound is nice, so they will stay in.

I have a few more changes I can make, but they be should be small I think. I can add bypass caps on the remaining caps in the power supply and I have some russion PIO K40y-9 in both 0.1uF and 0.068uF to try as coupling caps. Right now I have 0.1uF Obligatto Gold coupling caps and they sound good so I am not in a hurry to change those. I can also change out my cathode caps with motor runs that I have. and now my B+ is low enough that I can change between the 5U4GB and GZ34 rectifier tubes. I am currently using the 5U4GB. I think I can also change my cathode resistors on the 6336 back to the 820ohm.
All things to consider.

I will post a marked up schematic when I stop making changes, but it stays overall pretty close to Suncalc's.

This has been a fun built and I want to thank Suncalc again for posting this well documented build.

Author:  ILoveHiFi [ 11 Mar 2019, 20:48 ]
Post subject:  Re: "The Beast" 6336 Dual SET Amplifier

Wirewound resistors have inductance which may kill sound quality compared to standard ones

Author:  ToddD [ 12 Mar 2019, 09:27 ]
Post subject:  Re: "The Beast" 6336 Dual SET Amplifier

How much impact would that have when they are used as cathode resistors?

Author:  ILoveHiFi [ 12 Mar 2019, 15:30 ]
Post subject:  Re: "The Beast" 6336 Dual SET Amplifier

I don't have much expirence with wire wound cathode resistors as the tube amps I built don't require more than 2w.
Theoritcally the diffrence would be too small for a diffrence and negletibal. But since you can hear diffrence between carbon and metal film resistor then you might hear a diffrence here.

Author:  Suncalc [ 12 Mar 2019, 18:26 ]
Post subject:  Re: "The Beast" 6336 Dual SET Amplifier

ToddD wrote:
How much impact would that have when they are used as cathode resistors?
None. The self inductance is too small to have any effect at audio frequencies.

However, if you're really paranoid (even though you shouldn't be) use wire wound resistors with Ayrton-Perry windings like the Vishay-Dale NS series (e.g. NS02B910R0FE12). These have shockingly low self inductance numbers; far too low to worry about below about 100MHz.

Page 5 of 7 All times are UTC - 6 hours [ DST ]
Powered by phpBB® Forum Software © phpBB Group