Hi folks, I have a stainless chassis, some Chinese iron, and some Russian and Chinese tubes that I'd like to turn into a decent amplifier. The schematic that comes with these kits varies widely, and the one that came with this kit is a mess. It is attached. It shows the two halves of the 6H8C/6SN7 drivers wired in parallel, which I have been told is generally a bad idea and also not necessary. So my idea is to simplify and just use one half of a single 6H8C/6SN7 driver tube for each channel.
I also got rid of the tube rectification in the power supply discussion here: https://www.diyaudio.com/community/threads/5c4s-replacement-with-solid-state.405726/
I am at the point where I am trying to put together a schematic that will not just work but work well in case I decide to keep this amplifier. It started out as a learning tool and an experiment where I didn't care much about the final result, but now that I am this deep into it, I'd like to do at least a half way decent job.
I have redrawn the schematic to show the solid state power supply and 1/2 6H8C/6SN7 per channel.
I do not have an oscilloscope. I thought about buying one of these but I would have exactly no idea how to use it or how to begin. https://www.amazon.com/Hantek-Digital-Storage-Oscilloscope-Analyzer/dp/B015XVPPJ6
Proposed schematic updated 3/30/2025:
I also got rid of the tube rectification in the power supply discussion here: https://www.diyaudio.com/community/threads/5c4s-replacement-with-solid-state.405726/
I am at the point where I am trying to put together a schematic that will not just work but work well in case I decide to keep this amplifier. It started out as a learning tool and an experiment where I didn't care much about the final result, but now that I am this deep into it, I'd like to do at least a half way decent job.
I have redrawn the schematic to show the solid state power supply and 1/2 6H8C/6SN7 per channel.
- I have no idea how to determine the proper values for the negative feedback resistor and any parallel NFB capacitor. I'd like to move the negative feedback to the 4 ohm output. I will not ever use the 8 ohm outputs, and although I know in theory it should be fine to attach the negative feedback to the 8 ohm output anyway, to satisfy my curiosity I'd like to put it on the 4 ohm output.
- Is there any reason for me to try 4k (or higher) output transformers instead of the 3.5k specified? I'll be replacing the tiny ones that came in my version of this kit and I can get any value needed - update 3/30/2025 - larger replacements arrived today.
I do not have an oscilloscope. I thought about buying one of these but I would have exactly no idea how to use it or how to begin. https://www.amazon.com/Hantek-Digital-Storage-Oscilloscope-Analyzer/dp/B015XVPPJ6
Proposed schematic updated 3/30/2025:
Last edited:
Running the preamp tube in parallel halves the Plate resistance, increases the current. Is a good thing. But you could also wire the tube SRPP, or Mu Follower, Cathode Follower, etc. and see what it sounds like. Check out Designing High Fidelity Valve Preamps by Blencowe, there's a good section on the different ways you can "totem" the two gain stages in dual triodes.
Feedback is meant to be taken from the last tap on the OPT, you might not like the results of connecting it to an inner tap.
you could get a 50k stereo pot, run the FB wires thru that, adjust while listening. keep the pot or measure it to get the resistor values to keep it at your sweet spot.
Sure you can add a pentode/UL switch. I recommend DPDT ON-ON with a healthy amp rating, $10 or up, and don't switch it with the amp on. It'll wreck your tweeters and your ears,
play with This Calculator to see how the different parameters affect the tube's operation. It also has a Universal calculator that you can work out the resistor values for your preamp stage. 30k on the plate and 322 on the cathode seems low, OG drawing had 82k + 1150...
Feedback is meant to be taken from the last tap on the OPT, you might not like the results of connecting it to an inner tap.
you could get a 50k stereo pot, run the FB wires thru that, adjust while listening. keep the pot or measure it to get the resistor values to keep it at your sweet spot.
Sure you can add a pentode/UL switch. I recommend DPDT ON-ON with a healthy amp rating, $10 or up, and don't switch it with the amp on. It'll wreck your tweeters and your ears,
play with This Calculator to see how the different parameters affect the tube's operation. It also has a Universal calculator that you can work out the resistor values for your preamp stage. 30k on the plate and 322 on the cathode seems low, OG drawing had 82k + 1150...
Looking at the original schematic I would swear it was made for 6SL7, not 6SN7. The plate resistor is quite high, and a 6SL7 would make more sense using gNF. The open loop gain when using a 6SN7 is pretty low. I will simulate it to see what happens.
SIM:
The original schematic uses a 6SN7 (6N8S) indeed, DC values match pretty well. Distortion is pretty high at 5W output though, which makes sense, there is not enough open loop gain for the gNF to correct things. But the 10K resistor is wrong, it is too big, you can't get 285V from 310V by using a 10K.
SIM:
The original schematic uses a 6SN7 (6N8S) indeed, DC values match pretty well. Distortion is pretty high at 5W output though, which makes sense, there is not enough open loop gain for the gNF to correct things. But the 10K resistor is wrong, it is too big, you can't get 285V from 310V by using a 10K.
Last edited:
ya, Here's a Post from 2016 about the same kit, it has the 6H9C (6SL7) wired in parallel. A different kit has a similar topology but a 6H8C(6SN7) tube wired as two gain stages.
I did some work on a Chinese SE amp which used one 6SN7GT per channel in a cascode config.
https://www.diyaudio.com/community/threads/chinese-el34-6sn7gt-se-unusual-topology.401911
The seller claimed 10watts at 0.1%: it made barely 2watts at 10%
After the mods it did make 10watts at about 1.2%
I added a constant current anode loa to 6SN7, removed the capacitor across the EL34 cathod resistor, added a 2n2 cap across the feedback resistor and gave each channel a separate HT supply.
https://www.diyaudio.com/community/threads/chinese-el34-6sn7gt-se-unusual-topology.401911
The seller claimed 10watts at 0.1%: it made barely 2watts at 10%
After the mods it did make 10watts at about 1.2%
I added a constant current anode loa to 6SN7, removed the capacitor across the EL34 cathod resistor, added a 2n2 cap across the feedback resistor and gave each channel a separate HT supply.
Why would that be a bad idea? Each triode drives one EL34. I don't see what would be wrong with that.
No, it's both halves of a preamp tube wired in parallel to drive a single EL34. 4 tubes for a stereo amp.
There are two schematics. The original one uses two paralleled triodes on tbe inout stage.
Your schematics don't show the triodes in paralell. You have one half for each El34 and only one 6SN7.
Yes I didn't spot the double pin numbers BUT the photos only show one 6SN7 on the chassis so one half per EL34 as shown in the schematic marked up in red.
OP, how many watts are you expecting?
PC scopes are only 8bit so no use for measuring distortion. For that you need an audio interface. I have used a Behringer UAC222 (24b/48khz) and iRig 2 HD (24bit / 96kHz) however you'd need to make an attenuator for the UAC as it's input is about 1v RMS max and they cannot measure DC.
Last edited:
@Never Get Old If you got sensitive speakers, 1Vrms source, I would go for a triode-strapped EL34 and no negative feedback. Second harmonic rich, around 4.6W max before clipping, and the 6SN7 will have enough gain for the job.
I don't think a 6SN7 as a single or even paralleled triode is gonna drive a EL34 well, unless you use a pretty hot pre-amp signal to drive it and even then.. Swap this for a correctly biased 6SL7 (or a 12AX7) and it should work well. And while I haven't sat down and calculated it, the changes to the plate and cathode resistors on the 6SN7 doesn't seem right.
I'm also not a fan of just paralleling a triode when there is no reason to do it other than "well they are there". I know the internet/simulators love doing this, in practice I haven't seen it work out that great. The chance of having a real matched pair of triodes across the whole working range is slim to none, and while theoretically there are some advantages, people ignore the fact you are also doubling the Millar capacitance. I've rewired a few amps that did this, from parallel to single triode, and they all sounded better after the conversion.
"I suspect this amplifier will be used primarily in pentode mode." While they haven't stated, it doesn't sound like they are looking for a 2W SE amp. I haven't sat down and drawn load lines, but neither of these schematics look like a good plan to me. Just my experience, a 6SN7 as a triode struggles to have enough drive to actually sound good across the frequency range with these type output tubes in actual use. It might sim OK, go build one and listen to it. YMMV.
I have not built an SE with a 6SN7+EL34, but I built a PSE, 2x6P41S + 6N1P driver, sounds very good, freequency response was limited by the OPT, -3db/20Khz/9W. Not sure why the 6SN7 would do worse.
No problem at all. And sometimes required to drive a difficult Power Tube.
That might be when the power tube has a large grid & requires a grid resister at 50K or less.
In some cases that also occurs when driving paralleled output tubes.
In order to reduce interstage distortion the ratio of the driven grid resister value &
the resistance of the load resister on the driving tube needs to be as large as possible.
Say 12K driving 50K. One on the primary reasons for using interstage transformers.
A few years back I put together a page showing this, can't find it now. But easy to do,
another page is in the works,👍
That's a lot to read. I'm still pretty near the bottom of the learning curve, but not totally at the bottom, so I can understand most of it. 😉
It was. The schematic and driver tubes have changed across many variations of these kits over time, and several versions of the schematic show the Chinese 6N9P, which is 6SL7, not 6SN7. This version came with 6H8C, which is 6SN7. Of course, I do not have to use the tubes that came in the kit. Because I haven't committed to any design or schematic yet, I am still early enough that I can change things. Step 1 was the power supply, which required its own thread last month. I linked to that thread in post #1.
Only DACs. I have a large collection of CDs since I bought my first one in 1984, and that's all I ever listen to. I won't say what my first CD was as it's now embarassing! 🙄
Correct. I'd like to try to squeeze as much power out of it as possible. I have no problem with pentode operation or driving the EL34 very hard. I am not after the SET sound right now. SEP would be fine. Although my speakers are roughly 92dB efficient, experimentation over the years has shown that I will not be happy with 2 watts. Yes, I am aware of 10x the power needed vs. dB output, and on and on that discussion of how much power do you "need" goes. That discussion will go on forever. I want wattage, especially at 25Hz. Yes, I also am quite aware of the "there is no need for that low frequency" arguments. My speakers are quite capable, and if it's not there, I won't be able to tolerate the amplifier. I have always enjoyed my Dynaco ST-70 series ii (Panor corp. circa 1992). It has "enough" power down in that low range for me to be satisfied with it for 30 years. It has quite impressive bass power actually. I also have the much smaller pentode "Huaji Audio" push-pull EL84 amp that I built in another thread. It also is capable, but I have to say that there are times when I want more power out of it. XrayTonyB's tests on Youtube got about 12 watts/channel out of it at the bottom end. I love that little amp though, and I'm going to be building another one soon.
How bad a match is it and what problems would/might this cause? Is it bad enough for me to change driver tubes?
That might work, but I suspect I will want more power, which is why I was hoping to add a pentode-triode switch. That was done by someone who built one of these in another thread years ago so he could compare.
Correct. I have a third one that I have not posted because it will confuse things further. It is a version that tries to address the issues related to paralleling the triode sections in two 6SN7 drivers. If for some reason we need to go that route, I can post that third schematic later.
Re: two 6SN7 tubes with the two halves of each tube in parallel:
Is it necessary and/or advisable in this amplifier? If so, I was told that it might work if each half triode is treated separately, as in this post and the Glass Audio article mentioned here:
So, if using two 6SN7 tubes with the two halves of each in parallel, that would mean individual bias resistors for each half and individual grid stoppers for each half.
Based on that thread, and a lengthy one-on-one discussion with another member, we did try it on paper with two 6SN7s with the halves of each one in parallel, just to see what might happen. I have that schematic, but based on the info from @6A3sUMMER above and absent some real need for it, it seems like an unnecesary and possibly inadvisable use of the 6SN7. So, instead of using just one half of two 6SN7 tubes, I came up with the idea of just using one driver tube. I love the sheer simplicity of that, but if two driver tubes are needed to do the job correctly, parallel triode or not, I can do whatever is required.
It was. The schematic and driver tubes have changed across many variations of these kits over time, and several versions of the schematic show the Chinese 6N9P, which is 6SL7, not 6SN7. This version came with 6H8C, which is 6SN7. Of course, I do not have to use the tubes that came in the kit. Because I haven't committed to any design or schematic yet, I am still early enough that I can change things. Step 1 was the power supply, which required its own thread last month. I linked to that thread in post #1.
Only DACs. I have a large collection of CDs since I bought my first one in 1984, and that's all I ever listen to. I won't say what my first CD was as it's now embarassing! 🙄
Correct. I'd like to try to squeeze as much power out of it as possible. I have no problem with pentode operation or driving the EL34 very hard. I am not after the SET sound right now. SEP would be fine. Although my speakers are roughly 92dB efficient, experimentation over the years has shown that I will not be happy with 2 watts. Yes, I am aware of 10x the power needed vs. dB output, and on and on that discussion of how much power do you "need" goes. That discussion will go on forever. I want wattage, especially at 25Hz. Yes, I also am quite aware of the "there is no need for that low frequency" arguments. My speakers are quite capable, and if it's not there, I won't be able to tolerate the amplifier. I have always enjoyed my Dynaco ST-70 series ii (Panor corp. circa 1992). It has "enough" power down in that low range for me to be satisfied with it for 30 years. It has quite impressive bass power actually. I also have the much smaller pentode "Huaji Audio" push-pull EL84 amp that I built in another thread. It also is capable, but I have to say that there are times when I want more power out of it. XrayTonyB's tests on Youtube got about 12 watts/channel out of it at the bottom end. I love that little amp though, and I'm going to be building another one soon.
How bad a match is it and what problems would/might this cause? Is it bad enough for me to change driver tubes?
That might work, but I suspect I will want more power, which is why I was hoping to add a pentode-triode switch. That was done by someone who built one of these in another thread years ago so he could compare.
Correct. I have a third one that I have not posted because it will confuse things further. It is a version that tries to address the issues related to paralleling the triode sections in two 6SN7 drivers. If for some reason we need to go that route, I can post that third schematic later.
Re: two 6SN7 tubes with the two halves of each tube in parallel:
Is it necessary and/or advisable in this amplifier? If so, I was told that it might work if each half triode is treated separately, as in this post and the Glass Audio article mentioned here:
So, if using two 6SN7 tubes with the two halves of each in parallel, that would mean individual bias resistors for each half and individual grid stoppers for each half.
Based on that thread, and a lengthy one-on-one discussion with another member, we did try it on paper with two 6SN7s with the halves of each one in parallel, just to see what might happen. I have that schematic, but based on the info from @6A3sUMMER above and absent some real need for it, it seems like an unnecesary and possibly inadvisable use of the 6SN7. So, instead of using just one half of two 6SN7 tubes, I came up with the idea of just using one driver tube. I love the sheer simplicity of that, but if two driver tubes are needed to do the job correctly, parallel triode or not, I can do whatever is required.
Last edited: