Willsenton R8: Push Pull Global Negative Feedback Tuning.
Вставка
- Опубліковано 6 вер 2024
- In this final mod segment, I explore adjusting the Global Negative Feedback, or GNFB for short, on this vacuum tube amplifier. Using a scope + the Analog Discovery 2 with the audio analyzer suite software, I work to get a clean square wave, avoid instability/oscillations that can kill your speakers/tweaters and at the same time not kill off power.
This is a video series about the popular Willsenton R8 KT88 EL34 Push Pull tube audio amplifier. We will go through the repair this amp needs, testing and reviews of the amp as delivered, and then work on modifications to improve the sound quality and durability.
A great book for DIY amp building info!
Morgan Jones: Building Valve Amplifiers.
amzn.to/3X2hKfk
You can buy this amplifier here
amzn.to/3Q3Ynjr
If you would like to join my Patreon to help support the channel:
www.patreon.co...
Or donate on my website:
www.skunkiedes...
Howdy.
I have seen schematics using 220 pF in the global neg. feedback. I should think it is spot on.
Global neg. feedback has a tendency to produce kink spikes and ringing. I would think the reason is the phase shifting of the OPT.
For myself I never use GNF in my designs. Another reason is I like the extra electrical safety that keeping the OPT secondary galvanically separated provides.
I may suggest taking the feedback from the output tubes' anodes. That is what I do. There is practically no phase shifting.
Best regards.
Thanks to this video, I was finally able to fix my ringing problem. Even though the frequency response isn't flat to 30KHz, it's only 1.2dB different from 20Hz to 20KHz.
Woohoo!
This series really made me happy, and brought you a new subscriber! Thanks so much! I just might buy the amp and mod as you have…looks like a heap of fun and very worthwhile for sound and practical education in valve tuning.
Welcome aboard!
I have watched this video couple times to understand everything single thing you said in here. it helped a lot that I can apply to other amps. Please have more video of NFB for other amps if there is. Thank you so much.
Glad it was helpful!
I love that you found a way to adjust the tone if needed. Couldn't wait for the wrap up video. Please include a total list of part. Thank you so much Stephe
That is the final plan, a "wrap up video" at the same time I post a BOM.
PLEASE DON'T BE IMPATIENT! Just as most of you know who are asking me to do this for you: looking up all these parts, with part numbers, making up a spread sheet and double checking it takes HOURS to compile.
@@SkunkieDesignsElectronics trust me Im patiently waiting haha. Anyway, thank you Stephe, really. For making this all available to everyone.
Excellent analysis Skunkie. I agree with you regarding the resistors. When I did that A12 modification I used all 2w resistors for the same reason you said. I think you found the feedback sweetspot. Great video, great R8 project!!!!!!
Love ya' work, Stephe 👍
Thank you!
Another great video Skunkie.
thanks for the suggestions
Good demonstration of the perverse impact of exess NFB................what one can also do and try is to keep the original value of the Fb resistor and istalling another complementary value before it .........then only try values of Capacitance shunting onlt the original rersistor , values mostly between 100-470 PF....I like to use good quality silver mica capacitors for the job....
Honestly, the variations are almost endless :)
This was a fabulous series on the fabulous Willsenton R8. I own the R8, the Willsenton R800i, as well as the Cayin Soul 170i. If you do buy your own R8 I would recommend the MELZ 6H9C in the 6SL7 position. The Sylvania 6SN7GTA straight-plate (not the angle-plate), MELZ 6H8C or the MELZ 1578 in the 6SN7 position, and Mullard 6CA7 xf2 in the power tube section. In the power supply filtering position, I usually use a good high mu well balanced 6SN7. The amp, with these tubes, sounds amazing, and I can only imagine how much better it might sound with your recommended mods.
Thanks great video series
Hey Skunkie, if you add 110pf parallel to 220pf you get 330pf
I had a pair of 110pf and a pair of 330pf. I wanted to try a 220pf :)
Hi A good way of determining the proper amount of inverse feedback is to divide the open loop gain ( feedback resistor disconnected ) by the closed loop gain this will give a number which (NFB) you then apply to the formula 20(log) NFB this will give you the amount of feedback in DB's The amount of feedback should not exceed 15 DB as the soundstage begins to collapse I limit the amount of NFB in amps I build to 10 DB and I find this gives me a stable amp with a good open sound Usually a small capacitor like 10 to 15 PF takes care of any ringing which I find is rare using this method.
I hate ask a silly question but I have done most of these mods and watched all the videos (I think) and I can’t seem to find where you actually show where and how to adjust the global negative feedback. Thank you!
@Skunkie Designs Electronics At or around 21:30, you said you had a 110pF and a 330pF but no 220pF. If you had another 110pF, you could put two 110pF capacitors in parallel to get the 220pF that you wanted 😊 I hope that helps 😉
Also, thanks for posting your amp designs on your website 👍 I am going to try some of them in the not too distant future (hopefully). Thanks for sharing all this with us!
You are correct, but I didn't have another pair of 110pf caps. And being totally honest, I'm not saying this GNFB is the be all end all. I only had a few hours to try to come up with something decent that was better than what it comes with.
That was the topic in the late seventies by Matti Otala when he exposed excess NFB as one of the causes of what he called TIM ( transient intermodualtion distortion with what you experience on yhe scop and saturation of the input......)
Interesting to look up his work........one of the first amplifiers build along his study was the glorious Two Channel Stereo Amplifier from Electrocompaniet .....a class A amplifier witch set a new standard in music reproduction at that time...
The idea that "If some is good, more is better" almost never applies to anything.
Great series Stephe
@Skunkie Designs Electronics So, without a picture of your feedback network, I believe it's a 24kohm resistor and a capacitor (110pF, 330 pF, etc.) in parallel. You might try a 24kohm resistor in parallel with a capacitor (110pF, 330 pF, etc.) in series with a second resistor (like 1/4 to 1/2 of the other resistor, so like 6kohm to 12kowm). Or perhaps double the second resistor value from the first resistor, so like 28kohm for the resistor in series with the capacitor. I think that proposed feedback network won't remove so many of the high frequencies, but you should still see some ringing (overshoot). Anyway, it could be fodder for another video 😉
Hi.. It would have been nice to see if you tried it triode-wired and without (or very low) NFB.. But I’m glad the owner is now enjoying the amp..
No, and honestly I simply ran out of time to test the GNFB in triode mode.
Thanks, Stephe. An excellent and enlightening video. I was wondering what advice would you have for R8 owners looking to enhance the TR part of the circuit with the GNFB resistor value, and possibly bypassing it completely. Thanks!
If I was serious about listening to this amp only in triode mode, I would hardwire it with a 100ohm resistor between the plate and the screen (after removing the existing resistors) and give it a try with zero GNFB. Yeah it will have higher THD, but there is more to how a tube amp sounds than THD measurements.
@@SkunkieDesignsElectronics I did it (100R and no GNFB), just after ALL you mods. Big thank you for all videos!!! Longest soldering session in my life.
dear stephe bless you for amazing work! just for make sure the steps you do , all mods fit 100% to users that use in 230v electricity ,yes?
Again anything AFTER the secondary of the power transformer on any tube amplifier is 100% the same. The 115/230V stuff is ONLY between the power socket and the power transformer. It's why they can make this a switchable 115/230V as the outlet power has ZERO to do with anything after the power transformer.
Skunkie, what are your favorite test tracks for your audio projects? We need to know what tunes you jam to determine what’s the budder and what’s bunk!
One of my fav test track is Elton John Madman Across the Water. That is an easy song for an amp/speaker combo to butcher. My Jam is the smoother side of Jazz, think like Dexter Gordon, Oliver Nelson + some rock like Heart, Eagles even some B52's.
@@SkunkieDesignsElectronics I love the entire Madman Across the Water album, and there is a much grungier version of the title track song. It is the last track on "Tumbleweed Connection". It may not be a great reference, but it is still a great song. I have thoroughly enjoyed this series of videos. I have been building guitar amps for twenty years, and you have got me into a new hobby, On another note, my wife likes your hair. She watches your channel with me on occasion........"Have a great day".
Skunkyie: Thomas stereo is selling his modified /upgraded Doge Chinese amp with variable negative feedback.interesting.
I thought we were going to see a before and after of how it looked on that machine. Were all these mods worth the effort do you think? It sounds like you weren't extremely happy with the results.
Where did I say that? Most of the changes are subjective and you can watch the first test video and then this one. Clearly just the changes in this video improved the accuracy of this amp. As I mentioned, the grounding upgrades solved a bunch of weird hum and ground loop issues. I'm already hearing from people how much these mods improve this amp. Again I don't get where you think I'm not happy with the results. Don't confuse the fact I like SE amplifiers more than Push-pull that this wasn't a win.
Another great video. Thanks for this!
One thing I'm curious about: Does the ideal capacitor value change if the resistor value is changed? As in, if I went with a 220pf cap with the 24k resistor, would I require a different cap for the same tuning if I changed to say a 16k or 28k resistor?
Yes, especially if the resistor value is changed a lot. Down to 16K, likely. Up to 28K, probably not so much.
@@SkunkieDesignsElectronics Thanks for the answer.
Do you have any thoughts on an adjustable feedback setup? Possibly either a variable resistor in series with a fixed resistor (for a minimum) or a rotary switch? Are 2 pole variable resistors accurate enough, or would it be best to go with a 2 pole rotary switch where you could match the resistors yourself for each channel? Would a switch pop when switching?
I probably missed it, but what results did you achieve (by rerouting the ground scheme and the heater supply wires) in reducing hum and noise? I recently bought an R8 and it does have a tiny bit of hum into my ZU DWII speakers (fairly efficient). I've been looking for ground loops without success.
I very much appreciate this series and plan to install many of your mods, especially the two large power supply caps and the resistors on the bias boards.
Yes, all the hum and especially with weird interactions between my tube phono stage and this amp disappeared when I redid all the grounding and heater wiring.
@@SkunkieDesignsElectronics Thank you for your reply. I have downloaded your schematics and am watching for when you post your BOM. Really good content. I appreciate your thoroughness and all the work you have done on this project. I really like my R8 but I know I will need to do these mods.
Please Tell me which oscilloscope is good for the works of tube amp..? My Iwatsu 40 MHZ is dead and I wish to buy a new one. But the CRT based good quality ones are not available now days.
I honestly have never used one of the modern ones.
Isn't zero feedback preferable? The Prima Luna amps for example have no feedback.
If you were only ever running the amp in triode mode, you can get away with zero feedback. With UL or pentode, you are gonna need some. That's true for every amp I have worked on. On some SE UL amps I have used plate to plate or "Schade" feedback instead of GNFB and I am a fan of that.
Cool! Which cap was this exactly? Diagram?
I'll get a schematic up before the next video, it's a cap that doesn't exist in the original amp.
@@SkunkieDesignsElectronics you rock!
Do you recommend to use this feedback 24K/220pf with triode aswell?
No, and honestly I simply ran out of time to test the GNFB in triode mode. I hope someone who owns one of these amps, and also has a scope, has the time to play around with the endless combinations. That said it is not a good idea to just throw parts in, without a square wave generator and a scope to see the results.
@@SkunkieDesignsElectronics so guessing leaving it as is then?
I prefer zero negative feedback if the amplifier is well designed and not prone to oscillations, because said design gives the most realistic input signal production.
No NFB works well for triode mode, especially single ended. UL and pentode need a little negative feedback to sound good IMHO.
@@SkunkieDesignsElectronics Please explain the role of global negative feedback in the amplifier. I have a Cayin A55T push pull KT88 amplifier. I did a modification on it so that I put a switch in the negative feedback line. It is designed that from the output transformer I have a return to the cathode resistor of the first stage, which is 1.2k, via a 180k resistor. The feedback ratio is very weak 1.2/(1.2 + 180) ≈ 0.007 but I prefer an amplifier sound without NFB. The amplifier does not oscillate without the NFB, so I am wondering if there are any negative consequences for the operation of the amplifier if I turn off the negative feedback connection. I really do not care if my THD without NFB is greater than with NFB but for my ears i preffer no feedback sound. Thanks for the answer 😀
Hi there! Would it be possible you took a closer look on the these days so popular Willsenton R300 ? As an affordable and thus to many attractive 300B SET, it may disappoint some purists by being a global negative feedback design. How would it react to modifications of the NFB ? Could it be modded for adjustable NFB ? How does it hold its own stock ? A rough comparison to the Skunkie Modded A50 soundwise ?
One is being shipped to me ask I write this.
@@SkunkieDesignsElectronics Oh, terrific! Thanks for swift response. BR Magnus