Oh, I agree with your assessment. it is a tough read, especially if you're like me. I haven't done any "mods" besides what came with the "DX" kit. I have kept the rolling of tubes to a minimum. Mainly am wanting to mod mine to where I can use a less microphonic tube besides the 12au7.....
Was trying to find more info in the thread on the mod to upgrade the driver current so I could use 12by7, since I have a HK Citation II as well. But I can't find it in that thread, since it goes everywhere.....
Was trying to find more info in the thread on the mod to upgrade the driver current so I could use 12by7, since I have a HK Citation II as well. But I can't find it in that thread, since it goes everywhere.....
The 12AU7 should not be microphonic. Power tubes tend to be more microphonic than driver tubes in a power amp anyway.
If you apply all the modifications that I recommended in this thread then you will be able to use 12BH7, which is what you presumably meant. 12BY7 is a pentode (beam tetrode actually). On the other hand, 12BH7(A) does not really fit the TU-8200.
You also sent these questions in PM, which is what I replied to.
Conclusion is that the spreadsheet is correct, the new C34 and C35 are OS-CONs with the same value like the cathode bypass capacitors in the DX upgrade kit.
Driver tubes with 600mA heater current can be used after these mods as long as you use them with a 6L6GC variant.
Modifications to Driver Section
You mentioned in one of your posts that you were satisfied now that you had completed you updates to the Driver Section. I assume that those changes were above and beyond the changes in the spreadsheet you posted. Could you tell us what mods you made to the Driver Section.
Thanks,
Sandy
You mentioned in one of your posts that you were satisfied now that you had completed you updates to the Driver Section. I assume that those changes were above and beyond the changes in the spreadsheet you posted. Could you tell us what mods you made to the Driver Section.
Thanks,
Sandy
Yes, the updates in my spreadsheet are recommended even if you do not want to make any other change to the amp except for tube rolling.
Almost 30 years ago I bought 4 Tungsram E80CC tubes based on Jean Hiraga's praising comments in l'Audiophile. These tubes had already been discontinued back then so that he recommended 12BH7A for the driver in his design as an acceptable compromise.
I wanted to design a build a tube amp but getting OPTs was very difficult back then especially here in Hungary, so that I designed and built a solid state amp instead. These 4 tubes, along with some Russian triodes that I stole from spare parts stock or simply pulled out of disassembled units during my military service, had remained in the closet for a few decades unused.
I thought I would give them try after the power supply upgrades and I was deeply impressed. I took 10 seconds to recognize the huge step forward via headphones, which resolved all my complaints about the overly warm and laid back character of the stock amp. The highs cleared up, the stage opened up into multiple layers, and the dynamics finally arrived. Mr. Hiraga was right...
Later, when playing music with higher volumes especially via speakers, I noticed some harshness in the treble. The reason is obvious. Even though the E80CC has similar bias current to the ECC82/12AU7 in the TU-8200, it has about 50% higher gain that affects both stages. This amp has a global negative feedback, and the resulting higher loop gain may cause transient intermodulation distortion and maybe parasitic oscillations.
The solution is simple - desolder C5 and C6 cathode bypass caps, at least if you like the sound otherwise. This will restore the loop gain and fix the harshness problem without sacrificing anything of the benefits of the tube replacement. No cap is the best cap anyway, so if you can live with a slightly less gain, you can get rid of C1 and C2 as well. In that case the negative feedback to the cathodes of the input tube needs to be restored by increasing R3 and R4 from 12 to 16-20 Ohms. (I use 20 Ohms and it works well.) The amount of the GNFB applied has significant impact on the sound, I even tried it without global feedback and it sounded flat and boring so "no feedback" is not always the best option as many believe...
Note that due to the higher heater current requirements of the driver tubes, you can use them with 6L6GC and its variants only. KT66 is on the edge but is probably OK. (I have a pair and have had no issues but preferred the RCA 6L6GC sonic character overall, not to mention the much neater look with this amp compared to the oversized KT66.) No EL34 or KT88 can be used as the total heater current may damage the power transformer.
All brands are great, right now I use Valvo/Philips variants mainly because I am not emotionally connected to them
You can still put the 12AU7 back any time if you want to use these big power tubes but as the circuit has changed, it may not sound the same as the stock model. (To my ears, there isn't that much difference though.)Thanks for that. I’m almost finished building the tu-8200 with the mods you suggested. I’m still waiting on the DB207 from China for the D7 position. I’m getting very anxious to try it. You’re posts are greatly appreciated. I don’t have the design background for electronics. I chose mechanical engineering.
Ready for this......
I have a Bachelors in Electronic Engineering, but my job is Mechanical Design Engineer.
I had to step in for a fired ME at an old job to get out projects done on time, found out I am good at it too....and stayed with it.
Its always fun to tell people I am an ME with an EE....lol
Yes, I meant the 12BH7, I tend to get the nomenclature confused at times....
Thanks for your response!
I have some hesitation about the value of the feedback resistor when the cathode bypass capacitor is removed.
Theoreticaly the gain of the E80CC stage without the cathode capacitors equals the gain of the stage with 12AU7 and bypass capacitors installed. So the total ammount of feedback should be equal..
By removng the bypass capacitors the cathode 1k resistor comes into play for the audio signal and messes up the ammount of feedback set by the original 12 ohm resistor. Raising it to 18 ohm is in my opinion inadequate. Soms basic calculation points at a value of 100- 120 ohm for the feedback resistors.
Am I on the right track here or do I err somewhere?
Theoreticaly the gain of the E80CC stage without the cathode capacitors equals the gain of the stage with 12AU7 and bypass capacitors installed. So the total ammount of feedback should be equal..
By removng the bypass capacitors the cathode 1k resistor comes into play for the audio signal and messes up the ammount of feedback set by the original 12 ohm resistor. Raising it to 18 ohm is in my opinion inadequate. Soms basic calculation points at a value of 100- 120 ohm for the feedback resistors.
Am I on the right track here or do I err somewhere?
This assumption is correct. (The gain of the modified stage is about 10% less actually.)
How did you get to 100 Ohms?
Both the input and the driver stages are grounded cathode circuits. The global feedback goes from the secondary winding of the OPT to the first stage. The first stage works as a grounded grid amp from the feedback perspective.
Grounded grid circuits do not invert the phase, which is why the feedback is negative despite the input and feedback signals being in the same phase.
The cathode of the E80CC represents a 1.3k Ohm load impedance in the given working conditions. If you remove the bypass cap then the 1k cathode resistor will be connected in series with the cathode and about 43% of the feedback voltage falls on it. This needs to be compensated with the higher feedback voltage. Note that similar part of the input signal also falls on the cathode resistor.
(btw, I initially miscalculated the value and used 30 Ohm resistors. It sounded harsh at higher volumes, exactly like the direct driver tube replacement. This led me to revise the calculations and found the error. 18-20 Ohms are the correct values for both theoretical and subjective reasons.)
Thank you very much for your fast, clear and eleborate reply!
I am an autodidact on tube electronics and the most difficulties I find with the subject of feedback and phase shift. I have no oscilloscope to check wat I am doing, so I try to get a theoretical grip on the different subjects and check potential changes on existing circuits (Most to counter the loads of ignorance I seem to find on the internet...).
I tend to look at the feedback circuit as a voltage divider, that is how I arrived at the mentioned value, but you demonstrated this to be wrong. thank you for that. I go back to the books....
I am an autodidact on tube electronics and the most difficulties I find with the subject of feedback and phase shift. I have no oscilloscope to check wat I am doing, so I try to get a theoretical grip on the different subjects and check potential changes on existing circuits (Most to counter the loads of ignorance I seem to find on the internet...).
I tend to look at the feedback circuit as a voltage divider, that is how I arrived at the mentioned value, but you demonstrated this to be wrong. thank you for that. I go back to the books....
I just finished installing the component changes from the spreadsheet.
Along with swapping out the volume control with a 50 K , 21 Position
stepped attenuator (from eBay) and i was very pleasantly surprised
with the results. It sounds much better than I was expecting.
Recently I built a ST-120 amplifier for my living room I hooked it up to
the same speakers that my TU-8200 is on for testing and it sounded
much better than the TU-8200 did. At that point I strongly considered
replacing the TU-8200. Then I found this thread and figured It was worth
trying the new components. and since my volume control was crackly
I had to open it up to repair it anyhow.
The component changes listed in the spreadsheet made quite a difference
and is really what the amplifier should have sounded like in the first place.
The low end is much fuller (It was always lacking before) and over all it
just sound better (crisper perhaps).
I would strongly suggest these component changes.
It really has breathed new life into this little amplifier.
Thank you OP for your work in coming up with these changes.
Along with swapping out the volume control with a 50 K , 21 Position
stepped attenuator (from eBay) and i was very pleasantly surprised
with the results. It sounds much better than I was expecting.
Recently I built a ST-120 amplifier for my living room I hooked it up to
the same speakers that my TU-8200 is on for testing and it sounded
much better than the TU-8200 did. At that point I strongly considered
replacing the TU-8200. Then I found this thread and figured It was worth
trying the new components. and since my volume control was crackly
I had to open it up to repair it anyhow.
The component changes listed in the spreadsheet made quite a difference
and is really what the amplifier should have sounded like in the first place.
The low end is much fuller (It was always lacking before) and over all it
just sound better (crisper perhaps).
I would strongly suggest these component changes.
It really has breathed new life into this little amplifier.
Thank you OP for your work in coming up with these changes.
Thank you. I had the same impression after making these changes, which is why I started this thread.
Did you also modify the driver as I explained in a later post? That's what makes the most significant difference. Some of the component upgrades from the spreadsheet are prerequisite.
I have not upgraded the driver just the component
swap from the post with the spreadsheet.
I'll need to look at the driver portion as well.
But just the first part made a huge difference.
@harryY
I too have been looking at the ST-120 - and the ELEKIT TU-8200. Decisions, decisions...
Out of interest, did you opt for the SS rectification on the TS-120? Or tubed?
I know the ST-120 and the TU-8200 have vastly different power ratings, but now that you have completed the TU-8200 upgrades suggested by Ferenc, how do the two different amps compare? Have you got a clear favourite?
I would be very interested to hear any of your observations comparing the sound of the two amps and their respective build complexity. Appreciated.
This comment should be posted to a different thread, but I would never buy or build a class A/B amp with tube rectification.
The current draw from the power supply in these amps largely depends on the audio signal. Transients require few hundred mA immediately, but tube rectifiers are current limited. What makes the situation even worse is the high ripple voltage caused by the current limitations, which is then usually suppressed using a C-L-C filter with a large choke in the middle in order to keep the hum at acceptable levels.
Unfortunately the inductance of the choke does not know whether the current changes are coming from the rectifier's ripple voltage or from the amp's increased current requirements so that it will fight both and drop the B+ voltage to decrease current draw from the amp, which is the last thing we want.
You are practically disconnected from the power line from a transient perspective with this design.
The only component that can compensate the voltage drop is the capacitor after the (last) choke. This cap is actually fully included in the signal loop and its quality becomes very critical. Considering that you need high capacitance at high voltage, the only reasonable option is the traditional electrolytic cap.
Most tube amp owners who worry a lot about the quality of their coupling caps between stages and buy NOS rectifier diodes don't even suspect that they use components that have much more detrimental effect on the sound than even the cheapest polypropylene coupling cap.
The original Dynaco amp was not built with tube rectifier because it was that great, just because it was the only option back then.
I suspect the tube rectification became popular in the last decade mainly because the manufacturers can earn extra money on these "upgrades".
Between the ST-120 and the TU-8200 with mods I really don't have a
clear favorite. The TU-8200 is in my home office/den and the ST-120
is in my living room.
I hooked the TU-8200 up to my living room speakers while I was building
the ST-120 and while it did work better than expected I prefer the
additional headroom the ST-120 provides for my living room
That being said for my home office the TU-8200 is a good fit it has about
the right amount of power (the ST-120 would be a bit much for my den).
It really comes down to finding a good fit for your needs. In my
living room I tend to listen a bit on the louder side but In my home
office/den I sit much closer to the speakers so I run it much lower.
Last edited:
@HarryY
Thanks for these thoughts, Harry.
Aside from the massive power difference between these two amps, I had imagined that they might have entirely different sonic signatures.
I.E. One warm, rich and a bit more of an easy listen; the other snappy, dynamic and more extended at the frequency extremes.
It seems from your reply above that this is not the case - and that the power difference is the main consideration.
Interesting. If this is the case - as always - the main factor to consider is the room size and the specific speaker needing to be driven.
Truthfully, I've not heard either amp and I'm not likely to, short of building one or the other. That said, this project is going to be more projected than I had anticipated.
Returning to topic I like to share my experience with the modification of the driver part of the TU-8200. I removed the cathode bypass capacitors and changed the feedback resistors to18 ohms as advised to accomodate the E80CC.
Listening tot the result I experienced a sharp sibilance and roughness in the sound of selected tracks that seemed to hit instability at soms points. Using 12AU7 or 5814 did not make much of a difference on sibilance, but induced no instability Strangely the use of 5963 cleared things up and showed more detail in the sound.
By changing the feedback resistors back to 12 ohm the E80CC sounded much better again. So I like to think there may be soms benefit in lowering the value of this resistor even more.
I am puzzeled by this experience and like to hear if there is any comment on why my experience differs so much from the above mentioned good results with raising the value of the feedback resistors? Is it possible that I went wrong somewhere? I use the standard OPT. B+ is about 5% higher than max. In the feedback loop I used silver mica capacitors.
Listening tot the result I experienced a sharp sibilance and roughness in the sound of selected tracks that seemed to hit instability at soms points. Using 12AU7 or 5814 did not make much of a difference on sibilance, but induced no instability Strangely the use of 5963 cleared things up and showed more detail in the sound.
By changing the feedback resistors back to 12 ohm the E80CC sounded much better again. So I like to think there may be soms benefit in lowering the value of this resistor even more.
I am puzzeled by this experience and like to hear if there is any comment on why my experience differs so much from the above mentioned good results with raising the value of the feedback resistors? Is it possible that I went wrong somewhere? I use the standard OPT. B+ is about 5% higher than max. In the feedback loop I used silver mica capacitors.