Ww has lower noise floor compared to metal film. Pulse response in WW better than metal film. Parasitic inductance can be an issue but again an unknown value though a valid point. Everything is still in the planning stage.
Upgrading my Dared VP-300B Tube Amplifier
- Thread starter Hari Iyer
- Start date
Ww has lower noise floor compared to metal film. Pulse response in WW better than metal film. Parasitic inductance can be an issue but again an unknown value though a valid point. Everything is still in the planning stage.
@ Ravindra Desai
I further did some calculation for various values of resistors and their usual parasatic inductance and capacitance. The simulation shows that these values are usually in the uH and pF region and usually dont affect for audio frequencies. At the max the affect can be after 50KHz which is way above the audible range. My simulaton showed stability till 500KHz. If using wire wound allows lower noise figures and overall better stability then it should not be ignored imo. But only building and testing can finally confirm what is superior.
I may custom order some wire wound resistors for this experiment and unfortunately higher values are not available. so will stick with max 2K ohm/
I further did some calculation for various values of resistors and their usual parasatic inductance and capacitance. The simulation shows that these values are usually in the uH and pF region and usually dont affect for audio frequencies. At the max the affect can be after 50KHz which is way above the audible range. My simulaton showed stability till 500KHz. If using wire wound allows lower noise figures and overall better stability then it should not be ignored imo. But only building and testing can finally confirm what is superior.
I may custom order some wire wound resistors for this experiment and unfortunately higher values are not available. so will stick with max 2K ohm/
Yes, i agree about the SMD soldering. I will not be doing the soldering as i dont have the skills for this. I will be asking my elder brother to do this as he is an expert in smd soldering. I have seen him removing even a 64 pin soldered IC and resoldering them back with no issues. I have already asked him this and he has confirmed to me as OK.
I finally completed my final modifications in my 300B tube amp by replacing the smd metal film resistors with silicon coated wire wound resistors. The only resistor that i did not replace was the grid leakage resistor which was high value (470K) and WW is not available in small size for high values.
There was a significant improvement in the mids and highs with less compression and clutter. It was also very much relaxing and the lows were less masked by the mids and highs. I tried to find out why this is so and why changing just a resistor type from metal film to wire wound has such a great impact in sound stage. Searching in the net could not find any answer but only mention that WW sounds great / awesome.
IMO if the plate resistors are WW inductive (mine is around 200uH) then it allows for filtering of RF/EMI noise from the DC and prevents coupling of the same with the audio signal. Also i have added WW for the grid blocking resistor which prevents RF/EMI noise which cannot be handled by the tubes and make them run with less HF distortion. Adding WW to cathode also prevents bypassing of HF through the resistor but find the path through the bypass capacitors. Its very common to use a carbon composite resistor instead of WW in the grid blockage and there are many theories of amp oscillation etc. but my findings are completely different and it actually help to improve HF resolution. Another observation is since RF/EMI noise does not get into the tube, the output transformer runs much cooler than before and even after 2 to 3 hrs of playback, the output transformer is cold. Earlier it used to be slightly warm. So i think my 300Bs are also happy with the change.
IME WW is the way to go for tube amplifier, but the only catch is to get a good quality one of the shelf. In my case i got all resistors custom build.
Pending mods:
- Replace the internal PS wires from thin pvc stranded wire to 20 swg single core copper wire.
- Add a Ferrite bead to the cathode bypass resistor lead to block HF path to ground and use the bypass capacitor path.
There was a significant improvement in the mids and highs with less compression and clutter. It was also very much relaxing and the lows were less masked by the mids and highs. I tried to find out why this is so and why changing just a resistor type from metal film to wire wound has such a great impact in sound stage. Searching in the net could not find any answer but only mention that WW sounds great / awesome.
IMO if the plate resistors are WW inductive (mine is around 200uH) then it allows for filtering of RF/EMI noise from the DC and prevents coupling of the same with the audio signal. Also i have added WW for the grid blocking resistor which prevents RF/EMI noise which cannot be handled by the tubes and make them run with less HF distortion. Adding WW to cathode also prevents bypassing of HF through the resistor but find the path through the bypass capacitors. Its very common to use a carbon composite resistor instead of WW in the grid blockage and there are many theories of amp oscillation etc. but my findings are completely different and it actually help to improve HF resolution. Another observation is since RF/EMI noise does not get into the tube, the output transformer runs much cooler than before and even after 2 to 3 hrs of playback, the output transformer is cold. Earlier it used to be slightly warm. So i think my 300Bs are also happy with the change.
IME WW is the way to go for tube amplifier, but the only catch is to get a good quality one of the shelf. In my case i got all resistors custom build.
Pending mods:
- Replace the internal PS wires from thin pvc stranded wire to 20 swg single core copper wire.
- Add a Ferrite bead to the cathode bypass resistor lead to block HF path to ground and use the bypass capacitor path.
Last edited:
Output transformer was very very slightly warm earlier, now its fully cold. I think since the 300Bs are very close to the output transformer, the heat from the tube gets transferred to the outer case of the transformer though very very less. I will need to keep the amplifier on for over 3 to 4 hours to justify my finding on this though.
Did the following modification last weekend,
1. Changed the hum pot which is a 100 ohm, 1 watt potentiometer connected to the directly Heated cathode and which is used in the autobias circuit. The cathode bypass resistor is now a 25 watt boat resistor with aluminium heat sink and as per my calculation has a dissapation of around 5 watts continious. So the hum pot should ideally also be around 5 watts as this feed the cathode bypass. As per my guess when i removed the hum pot, and measured it it was very leaky and did not show the proper resistance and was broken from inside. I had ordered a 4 pairs of 100 ohm resistor and i connected them in parallel and series to get 100 ohm at the ends and 50 ohms at the center. this replaced the hum pot and had a total rating of 6 watts. This was a very important modification as this changed the sound stage to the next level with very tight presence and resolution.
When i touched this resistor after 30 minutes of playing the amplifier, it was quite warm and was sure the hum pot could not take this current.
2. Also i noticed the bridge rectifier module which supplies the 300B cathode voltage was quite warm / hot as this supplys the voltage and current to the 300bs. Hence am planning a heat sink for them. Hopefully the bridge rectifier should be happy with them.
3. The connectors that supply voltages from the power module to the power amps are of bad quality and will also need to be replaced and will be doing that too. Also will replace the wires at the same time of replacing the connectors.
4. Am also ordering some ferrite beads for the cathode bypass resistors.
1. Changed the hum pot which is a 100 ohm, 1 watt potentiometer connected to the directly Heated cathode and which is used in the autobias circuit. The cathode bypass resistor is now a 25 watt boat resistor with aluminium heat sink and as per my calculation has a dissapation of around 5 watts continious. So the hum pot should ideally also be around 5 watts as this feed the cathode bypass. As per my guess when i removed the hum pot, and measured it it was very leaky and did not show the proper resistance and was broken from inside. I had ordered a 4 pairs of 100 ohm resistor and i connected them in parallel and series to get 100 ohm at the ends and 50 ohms at the center. this replaced the hum pot and had a total rating of 6 watts. This was a very important modification as this changed the sound stage to the next level with very tight presence and resolution.
When i touched this resistor after 30 minutes of playing the amplifier, it was quite warm and was sure the hum pot could not take this current.
2. Also i noticed the bridge rectifier module which supplies the 300B cathode voltage was quite warm / hot as this supplys the voltage and current to the 300bs. Hence am planning a heat sink for them. Hopefully the bridge rectifier should be happy with them.
3. The connectors that supply voltages from the power module to the power amps are of bad quality and will also need to be replaced and will be doing that too. Also will replace the wires at the same time of replacing the connectors.
4. Am also ordering some ferrite beads for the cathode bypass resistors.
Finally i received the ferrite beads required for the last and final modification # 4 in the post #28 after a five weeks wait. I will be more than happy to implement this mod and close this project with astonishing success.
Thanks to everyone for their inputs and suggestions.
Thanks to everyone for their inputs and suggestions.
Contemplating a partial negative feedback loop from the output tube plate to its grid for increasing bandwidth, reducing distortion and lower output impedance. This will however reduce the output gain - a price to pay for better linearity and tone.
I will be using a 100K resistor and a 1uF capacitor from the plate to grid in the negative feedback loop and test today. Hope it allows greater bandwidth, lower distortion and lower output impedance. Voltage swing between plate grade and grid cathode will be almost similar to the original design. I to V conversion swing around 100volts for a 1mA input current / volt. The capacitor used will be Audyn, 600volts (voltage swings can be as high as 450volts) and resistor will be Vishay 1/2 watt metal film type.
Completed the above partial negative feedback loop today. Initially listening revealed the frequency response has now become more linear with the dominant mid-range tamed quite well. The vocals now do not have the shout nor do they announce their presence. The sound stage has become more relaxed and will now allow longer listening sessions. To my surprise the highs have become more rounded and much pure. I cannot see much difference in the bottom end but can find the mid-bass much better because of the tamed mids. The sound stage is now more liquid and organic and wants you to listen more. But all this have happened with the sacrifice of some gain. I have now to increase the volume 2 more notches up to get the same SPL. But i am not complaining due to the benefits realized. This mod is certainly for the keep.
For those who are techincally inclined here is the link of the theroy and inspiration behind this mod.
https://pearl-hifi.com/06_Lit_Archive/14_Books_Tech_Papers/Broskie_John/2001/01_02_Mar_2001.pdf
For those who are techincally inclined here is the link of the theroy and inspiration behind this mod.
https://pearl-hifi.com/06_Lit_Archive/14_Books_Tech_Papers/Broskie_John/2001/01_02_Mar_2001.pdf
To summarize more than a dozen mods done in the past 6 to 8 months the top 3 mods which had a huge impact in sound stage are,
- partial negative feedback
- Ultrapath
- wire changes with proper sizing.
The only mod that I could not subjectively quantify was adding the ferrite beads on the cathode resistor.
- partial negative feedback
- Ultrapath
- wire changes with proper sizing.
The only mod that I could not subjectively quantify was adding the ferrite beads on the cathode resistor.
Now, i am looking at replacing the 300B tubes. Have short listed a couple of them and will order soon. After this the only component that i would have not replaced will be the output tranformers, the power transformers, the volume potentiometer (Alps), some Nichicon capacitors. In the power supply there are a 5H choke, Aerovox capacitors (high voltage), schtoky diode rectifiers which wont need any replacement.
Original was zero negative feedback.
Without feedback: The sound has sparkle in the mids / highs and feels more open and has an extended openness with energy. The midbass sounds a bit weak due to this. I think most of the sound signature is of the tube and the output transformer quality. By changing either/both the amp without any feedback will also sound quite good imo. I am planning to change the 300B tubes shortly as the current one has gone very weak. I shall post my findings post that.
Withfeedback: The sparkle and the agressiveness of the mids / highs is lost and the sound appear more balanced. This will be good for some vinyls though but not for other media like CD & youtube. I actually now am liking without the feedback and have removed the negative feedback loop. I am now so much used to the sparkle of the 300Bs and cannot live without them now. The energy of this tube is fascinating and more lively and can related to real life music which gets absent with the negative feedback loop - again imo.
Since yesterday, my right channel 300B tubes has become so weak that it shorts within 30 minutes of operation. Hence after more than 6 months i connected my Denon AV receiver to replaced the tube amplifier. I played my usual music and noticed that i could not stand not even a single track with this amplifier. The Denon sounded very very laid back, dull, non-energetic, lifeless and all negative adjectives that you can attribute to an amplifier. Needless to say i turned them off in 3 minutes and watched TV instead.
Now my worries are can i ever go back to my old way of listening music - and i think probably not. Once you get used to something more refined then looking back is not a possibility for me in hifi. Please share your experience too.
I think the OPT is the one contributing to the distortion in the SE design. There are no cancellation of noise/ distortion in a SE and the quality of the OPT decides the final outcome of the amp. Also the tube used has a significant contribution to the final outcome in a SE design. For this reason, tube rolling makes more sense in a SE amp, In a push-pull amp tube rolling does not have that much impact.
Iam interested to know how cables can make a difference.Will not any quality cable do the job.? iam not doubting any of the findings .
Is my understanding correct?
The circuit has more influence than the individual components on the final tone.The tube by itself has inherently low gain compared to solid state BJTs/mosfets.So it does not need much negative feedback.Secondly the circuit further reduces the gain that is almost independent of device characteristics.Then how do these components make such a big difference.
The 300B grid stopper resistor uses a 1 kohm wire wound resistor which has some self inductance. This can combine with the miller capacitance and transconductance of the 300 B and create a series resonant circuit. The resistor inductance and the miller capacitance will tune it at around 2.5 MHz. There could be possible oscillation at this frequency in my power tube grid which could influence SQ.
The best way to avoid this is to use carbon composite resistors, but these are hard to find and quite noise prone, and drift value over time as experienced by others. The carbon composite have extremely low parasitic inductance. Second choice will be to use metal glass glazed resistor which again is not commonly available. Trying with FM, yogibear - yet to get my hands on it.( I require just a pair).
Alternatively, I am thinking of trying ferrite beads at the grid stopper resistor leads. Ferrite beads suppress RF from 100 KHz to 10 MHz and work as a lossy inductor in that frequency. This also falls within my resonant frequency range.
Will post updates next week after implantation.
Thanks for looking.
The best way to avoid this is to use carbon composite resistors, but these are hard to find and quite noise prone, and drift value over time as experienced by others. The carbon composite have extremely low parasitic inductance. Second choice will be to use metal glass glazed resistor which again is not commonly available. Trying with FM, yogibear - yet to get my hands on it.( I require just a pair).
Alternatively, I am thinking of trying ferrite beads at the grid stopper resistor leads. Ferrite beads suppress RF from 100 KHz to 10 MHz and work as a lossy inductor in that frequency. This also falls within my resonant frequency range.
Will post updates next week after implantation.
Thanks for looking.
Get the Award Winning Diamond 12.3 Floorstanding Speakers on Special Offer
Similar threads
