SONDEKNZ,
I can not answer for other hard-core audio guys, but here are my reasons:
My tube amplifiers are built on old chassis. Most of them have completely different parts and circuits than the original amplifiers.
The original chassis and original power transformers are all that remains; every other part is different.
And each of those amplifiers I designed have had more than one circuit topology on the same chassis.
The only steel chassis [yes, it is magnetic], had a PCB for the input and driver tube. The PCB got traces cut in order to do new design modifications,
and eventually the PCB went into the trash (good riddance).
I am getting ready to use a brand new chassis, and brand new power transformer. Just have not decided on a circuit topology, tube types, etc.
Now my vision and unsteady hand will not allow me to build solid state with super fine lead spacing, etc. I had to use a Loupe even back then.
I only like to design and build vacuum tube amplifiers, with very few solid state devices, my preference.
The only solid state devices I use now are silicon B+ rectifiers . . . a practical part for choke input B+ filters.
I sometimes draw out the schematic of my design, sometimes do not (it is already in my head). Then I start wiring the amplifier.
Most of my circuits are relatively simple, and I can connect the parts and points with only a tube 'pin-out diagram', just in case I forget which pin is which. Of course, I check my wiring, and do some Ohmmeter tests, before I power up the amplifier.
Once I wired in a 7591, but I mistakenly/forgetfully used 6L6 tube pin-outs. At first power up . . . Ouch! Fortunately, I turned the amplifier off before any damage was done.
Have fun learning, designing, building and listening!
I can not answer for other hard-core audio guys, but here are my reasons:
My tube amplifiers are built on old chassis. Most of them have completely different parts and circuits than the original amplifiers.
The original chassis and original power transformers are all that remains; every other part is different.
And each of those amplifiers I designed have had more than one circuit topology on the same chassis.
The only steel chassis [yes, it is magnetic], had a PCB for the input and driver tube. The PCB got traces cut in order to do new design modifications,
and eventually the PCB went into the trash (good riddance).
I am getting ready to use a brand new chassis, and brand new power transformer. Just have not decided on a circuit topology, tube types, etc.
Now my vision and unsteady hand will not allow me to build solid state with super fine lead spacing, etc. I had to use a Loupe even back then.
I only like to design and build vacuum tube amplifiers, with very few solid state devices, my preference.
The only solid state devices I use now are silicon B+ rectifiers . . . a practical part for choke input B+ filters.
I sometimes draw out the schematic of my design, sometimes do not (it is already in my head). Then I start wiring the amplifier.
Most of my circuits are relatively simple, and I can connect the parts and points with only a tube 'pin-out diagram', just in case I forget which pin is which. Of course, I check my wiring, and do some Ohmmeter tests, before I power up the amplifier.
Once I wired in a 7591, but I mistakenly/forgetfully used 6L6 tube pin-outs. At first power up . . . Ouch! Fortunately, I turned the amplifier off before any damage was done.
Have fun learning, designing, building and listening!
@6A3sUMMER
Thanks for these thoughts.
So, in your case, point-to-point wiring provides you a "blank canvas" enabling you to create your own entirely new circuits.
Very cool!
Thanks for these thoughts.
So, in your case, point-to-point wiring provides you a "blank canvas" enabling you to create your own entirely new circuits.
Very cool!
Heading back to the main topic of this thread: the awesome BEWITCH 6550 tube amp from China...
I have spent the last few years trying to improve this amp - with good success.
I saw the interesting posts by saltonm73 and posted some questions about his mods a few posts ago. But it seems saltonm73 has quit this forum - no reply to my questions and no finding him in membership.
So, I would like to ask other learned members here...
(I have included the schematic provided by saltonm73 and circled the areas of interest...)
Why add a .68uF capacitor between C3/4 and C5? What does it do and how does it impact the sound of the amp?
Why add a 1Meg resistor across the volume potentiometer? Is this simply an attempt to gain more usable adjustment of the pot?
Why run the 6SN7 in the first tube position - instead of the (original) 6SL7? I fear that this will mess with the overall amplifier Gain and NFB structure.
I look forward to all expert thoughts. Appreciated.
I have spent the last few years trying to improve this amp - with good success.
I saw the interesting posts by saltonm73 and posted some questions about his mods a few posts ago. But it seems saltonm73 has quit this forum - no reply to my questions and no finding him in membership.
So, I would like to ask other learned members here...
(I have included the schematic provided by saltonm73 and circled the areas of interest...)
Why add a .68uF capacitor between C3/4 and C5? What does it do and how does it impact the sound of the amp?
Why add a 1Meg resistor across the volume potentiometer? Is this simply an attempt to gain more usable adjustment of the pot?
Why run the 6SN7 in the first tube position - instead of the (original) 6SL7? I fear that this will mess with the overall amplifier Gain and NFB structure.
I look forward to all expert thoughts. Appreciated.
hf bypass for electros, its use is contested...discussed here to death, search for the posts of member Eva....
to prevent problem of possible tube saturation when the pot failed as an open circuit, a good practice imho..
the lower mu tube will result in lower open loop gain and is not good for stability, it is possible to get rid of gnfb this way, just use it in triode mode or do a OSchade feedback from plate of 6550 to plate of driver tubes, 120k 2 watt is okay try...
Many thanks @TonyTecson for your prompt answers to my questions above.
I understand the principles of capacitor bypass and have in fact bypassed my own C3, C4 and C5 on my BEWITCH 6550.
But the position that saltonm73 has installed his .68U (circled in red above) suggested to me that perhaps more was happening - than simply bypassing the C3 and C4 caps.
Regarding swapping the first position preamp tube from 6SL7 to 6SN7 - the lower gain 6SN7 generally considered to provide superior sound quality - is this something I could simply try with my triode only BEWITCH with its present stock circuit values? Or will the 6SL7 circuit need to be changed to support the incoming 6SN7?
Presently, the "OSchade feedback" you mentioned above is an unknown to me, so I'm keeping it simple...
For reference, I have added a schematic below, of my current BEWITCH 6550 triode only circuit, with many values showing what the factory had installed - as opposed to the original 2016 schematic. My mods are also shown - being slightly different from saltonm73.
I understand the principles of capacitor bypass and have in fact bypassed my own C3, C4 and C5 on my BEWITCH 6550.
But the position that saltonm73 has installed his .68U (circled in red above) suggested to me that perhaps more was happening - than simply bypassing the C3 and C4 caps.
Regarding swapping the first position preamp tube from 6SL7 to 6SN7 - the lower gain 6SN7 generally considered to provide superior sound quality - is this something I could simply try with my triode only BEWITCH with its present stock circuit values? Or will the 6SL7 circuit need to be changed to support the incoming 6SN7?
Presently, the "OSchade feedback" you mentioned above is an unknown to me, so I'm keeping it simple...
For reference, I have added a schematic below, of my current BEWITCH 6550 triode only circuit, with many values showing what the factory had installed - as opposed to the original 2016 schematic. My mods are also shown - being slightly different from saltonm73.
Last edited:
@TonyTecson
The triode switch is already long gone and my BEWITCH is run purely in triode.
Thanks for the Schade Feedback link. I'll do some reading.
As for removing the NFB resistor, this would be a very big move as it appears to set the entire Gain structure for the amp. I imagine it also has a lot to do with how the amp sounds - which is very nice!
That said, I did experiment with larger (and larger) NFB resistors in the early days - effectively reducing the level of NFB - and the sound of the amp deteriorated. I can't imagine that the BEWITCH would sound better with no NFB.
I don't know... Perhaps I have somehow missed your point?
The triode switch is already long gone and my BEWITCH is run purely in triode.
Thanks for the Schade Feedback link. I'll do some reading.
As for removing the NFB resistor, this would be a very big move as it appears to set the entire Gain structure for the amp. I imagine it also has a lot to do with how the amp sounds - which is very nice!
That said, I did experiment with larger (and larger) NFB resistors in the early days - effectively reducing the level of NFB - and the sound of the amp deteriorated. I can't imagine that the BEWITCH would sound better with no NFB.
I don't know... Perhaps I have somehow missed your point?
Last edited:
Hello SONDERKNZ
Sorry that I'm only getting in touch here now, in Germany Sunday and Monday were public holidays and we were out with the family.
that bridges with the additional .68 uF MKP (if you look closely at the assembly on the pictures) the two capacitors C10-C11, which are also missing on the original Bewitch circuit diagram but are present on the board!
Sorry that I'm only getting in touch here now, in Germany Sunday and Monday were public holidays and we were out with the family.
that bridges with the additional .68 uF MKP (if you look closely at the assembly on the pictures) the two capacitors C10-C11, which are also missing on the original Bewitch circuit diagram but are present on the board!
with the 6sl7 in the firsts hole, i will not remove global negative feedback, you may even try the 6sn7 in first hole and 6sl7 in the second hole, see if you like what you hear...
if using the 6sn7 in the first hole and 6sn7 in the next one, removing the feedback resistor will give you some surprises, you may even like it...
running the kt88 in triode lowered its "plate resistance" significantly so that you may get away with global negative feedback....i will not hesitate to try it...
@saltonm73
Many thanks for this reply.
Sorry that I gave you so little time to respond. Actually, previously I tried searching your username with "@" but it did not bring you up as an option. Today it did bring you up. Strange.
Personally, I am not qualified to comment on whether your decision to install an additional .68U bypass cap is an unnecessary "fad". That said, I have enjoyed good sonic improvements with every cap I have bypassed in my BEWITCH 6550 amp, so fair enough giving this one a go.
In any event, I'm sure that you would agree with me, that with a little bit of time and effort invested in a few tweaks, the BEWITCH is a superb sounding amp.
For listening comparisons, our other tube amp is a stock standard LEBEN CS-300F - a very costly little brute - and the BEWITCH certainly gives it a run for its money. Both amps have a very different sonic signature: the LEBEN with a bit more detail, sparkle and articulation; and the BEWITCH, a weightier presentation with a bit more "meat on the bone". Both amps deliver a quality sound that is entirely valid, in its own way.
Hard to choose and spoilt for choice!
Frankly, my ongoing tweaks and experiments on the BEWITCH are a constant attempt to bring a bit more of the LEBEN-type sparkle and articulation to the sound, but I'm not interested in a "bright" amp - of course.
Thanks for these thoughts, Tony. Appreciated.
From what you have suggested above, it sounds like I could happily place a 6SN7 tube in the "first hole" tube position, with the existing circuit left as it is - WITHOUT inflicting any damage on circuit or tube(s).
On this basis, I will give it a go and revert with my findings.
Also, if I am removing the 6SL7 from the amp completely - and operating with all 4 x 6SN7 tubes in the preamp stages, I may take the opportunity to experiment with lower/no Gain.
Firstly, I will try the 4 x 6SN7 and see how it sounds.
As an aside, I don't run KT88 tubes. The amp was built for 6550 output tubes, so I have happily stuck with those. They sound very good - but will ned to be repalced at some stage soon, I suspect.
I completed a bit of testing today, with interesting results.
My "normal" configuration for this amp is 2 x 6SL7GT tubes in the first hole and 2 x 6SN7GT tubes in the second hole. The output tubes are 4 x 6550.
TEST 1: I simply "swapped" the 6SN7GT and 6SL7GT tube positions and had a good listen.
The sound was less good and sounded rather homogenised. Not terrible, just less good and less enjoyable.
What this test did show me though, was that either the 6SN7GT or the 6SL7GT could operate (somewhat) without injury in either hole 1 or hole 2. No too brightly burning tubes, no offensive distortion and no problems. Probably not optimised, but no injury.
TEST 2: I removed the 6SL7GT tubes and installed 6SN7GT tubes in all 4 preamp positions, including two new NOS tubes.
Again, there were no too-brightly burning tubes, no offensive distortion and no problems.
The sound change was quite dramatic. A greater sense of space and detail than any tube configuration I have yet conjured with the BEWITCH. The "sparkle" I have been chasing in the tops seems to have arrived.
But on the downside, dynamics are down, versus my "normal" configuration and there is a bit of unwelcome hardness in the tops. It is not as easy to enjoy the music as my normal configuration, but I feel that this 4 x 6SN7GT configuration holds great promise in terms of extracting a bit more detail from the BEWITCH.
I have left TEST 2 configuration running for the moment, understanding that the new (NOS) tubes may need a bit of time to settle down and make friends.
I suspect that the support circuitry previously optimised for the 6SL7GT tubes now needs to be revised for the 6SN7GT tubes.
I do not have the knowledge or experience to determine the correct values for the resistors in question - R1, R2, R3, R4 and R5.
Also running 4 x 6SN7GT at the front end of his BEWITCH, I see that saltonm73 has simply halved the values of R2, R3, R4 and R5.
On this basis, I will try the same approach - unless others have alternative suggestions.
I look forward to your thoughts.
I have attached below, an updated schematic of how my BEWITCH with 4 x 6SN7GT now exists. Please note that the switched have been removed and the amp ONLY runs in triode.
My "normal" configuration for this amp is 2 x 6SL7GT tubes in the first hole and 2 x 6SN7GT tubes in the second hole. The output tubes are 4 x 6550.
TEST 1: I simply "swapped" the 6SN7GT and 6SL7GT tube positions and had a good listen.
The sound was less good and sounded rather homogenised. Not terrible, just less good and less enjoyable.
What this test did show me though, was that either the 6SN7GT or the 6SL7GT could operate (somewhat) without injury in either hole 1 or hole 2. No too brightly burning tubes, no offensive distortion and no problems. Probably not optimised, but no injury.
TEST 2: I removed the 6SL7GT tubes and installed 6SN7GT tubes in all 4 preamp positions, including two new NOS tubes.
Again, there were no too-brightly burning tubes, no offensive distortion and no problems.
The sound change was quite dramatic. A greater sense of space and detail than any tube configuration I have yet conjured with the BEWITCH. The "sparkle" I have been chasing in the tops seems to have arrived.
But on the downside, dynamics are down, versus my "normal" configuration and there is a bit of unwelcome hardness in the tops. It is not as easy to enjoy the music as my normal configuration, but I feel that this 4 x 6SN7GT configuration holds great promise in terms of extracting a bit more detail from the BEWITCH.
I have left TEST 2 configuration running for the moment, understanding that the new (NOS) tubes may need a bit of time to settle down and make friends.
I suspect that the support circuitry previously optimised for the 6SL7GT tubes now needs to be revised for the 6SN7GT tubes.
I do not have the knowledge or experience to determine the correct values for the resistors in question - R1, R2, R3, R4 and R5.
Also running 4 x 6SN7GT at the front end of his BEWITCH, I see that saltonm73 has simply halved the values of R2, R3, R4 and R5.
On this basis, I will try the same approach - unless others have alternative suggestions.
I look forward to your thoughts.
I have attached below, an updated schematic of how my BEWITCH with 4 x 6SN7GT now exists. Please note that the switched have been removed and the amp ONLY runs in triode.
I often see on other circuit diagrams that the capacitors are bridged with an additional foil cap!
So do you think that at this point bridging the capacitors (c10-c11) with a 0.68uF cap is superfluous and would not help?
@saltonm73
Nice job there!
I particularly like the way you have brought the volume pot to the input, keeping the input trace as short as possible. Smart move, but not always easy to achieve, mechanically.
That said, I guess your "The Goblin" input feeder still needs to stretch across the connect with the PCB.
In any event, it looks like progress.
Like everything else on my BEWITCH, the (factory) input connectors are like the Rock of Gibraltar - absolutely rock solid. I'm not sure why yours were so loose and flimsy.
Question: Given that the BEWITCH has a simple passive preamp - and the fact that you have lowered the gain by opting for 6SN7 tubes all round - which reduce the original level of Gain - do you run your BEWITCH with an active preamp?
Nice job there!
I particularly like the way you have brought the volume pot to the input, keeping the input trace as short as possible. Smart move, but not always easy to achieve, mechanically.
That said, I guess your "The Goblin" input feeder still needs to stretch across the connect with the PCB.
In any event, it looks like progress.
Like everything else on my BEWITCH, the (factory) input connectors are like the Rock of Gibraltar - absolutely rock solid. I'm not sure why yours were so loose and flimsy.
Question: Given that the BEWITCH has a simple passive preamp - and the fact that you have lowered the gain by opting for 6SN7 tubes all round - which reduce the original level of Gain - do you run your BEWITCH with an active preamp?
saltonm73,
A 0.68 uF capacitor across an electrolytic: Either it Improves the sound, Or it does Not Improve the sound.
The only way to know for sure is to do a very carefully controlled Double Blindfold test setup, using lots of different kinds of music recordings.
Any other test method is suspect at best.
I have done a few different setups of double blindfold testing.
One test setup used un-marked double switches, with an umbilical cord.
The 4 combinations of switch settings (up/down for each switch), connected the amplifier in 4 different ways (up up; down down; up down; down up). With no markings on the 2 switches, and No discussion of what the switches Did, or Did Not do. The listener(s) were completely in the dark.
They did not know what they were listening for, except to listen for any difference.
Even the up down were not marked on the switches; one listener might grab the switches differently than another listener (one listener's 'up' might be another listeners 'down').
A lot of thought went into designing this completely fair test.
This is key, neither the test administrator, nor the listener(s) knew what the different switch settings were, as they listened for sound differences.
For all anybody knew, only one combination of 4 did anything different and the other 3 settings were all the same connections;
Or, there were 2 different connection combinations (2 of one connection, and 2 the other connection),
Or, all 4 setting combinations were totally 4 different connections.
The different combinations of switch settings did Not sound different to one listener, so in frustration . . . He literally threw the two switches and umbilical cord across the room.
By the way, this same amplifier and switch setup was repeated at 3 different venues; with 3 different sets of listeners; each venue had a different signal source (3 different signal sources); and each venue had a different set of loudspeakers (3 different manufacturers and models).
Of course, there were 3 different listening rooms.
Just my opinions and experience.
Note: Measurements are one thing. Double Blindfold Listening tests are another thing.
They may agree, or they may not agree.
Have fun setting up a true double blind test.
A 0.68 uF capacitor across an electrolytic: Either it Improves the sound, Or it does Not Improve the sound.
The only way to know for sure is to do a very carefully controlled Double Blindfold test setup, using lots of different kinds of music recordings.
Any other test method is suspect at best.
I have done a few different setups of double blindfold testing.
One test setup used un-marked double switches, with an umbilical cord.
The 4 combinations of switch settings (up/down for each switch), connected the amplifier in 4 different ways (up up; down down; up down; down up). With no markings on the 2 switches, and No discussion of what the switches Did, or Did Not do. The listener(s) were completely in the dark.
They did not know what they were listening for, except to listen for any difference.
Even the up down were not marked on the switches; one listener might grab the switches differently than another listener (one listener's 'up' might be another listeners 'down').
A lot of thought went into designing this completely fair test.
This is key, neither the test administrator, nor the listener(s) knew what the different switch settings were, as they listened for sound differences.
For all anybody knew, only one combination of 4 did anything different and the other 3 settings were all the same connections;
Or, there were 2 different connection combinations (2 of one connection, and 2 the other connection),
Or, all 4 setting combinations were totally 4 different connections.
The different combinations of switch settings did Not sound different to one listener, so in frustration . . . He literally threw the two switches and umbilical cord across the room.
By the way, this same amplifier and switch setup was repeated at 3 different venues; with 3 different sets of listeners; each venue had a different signal source (3 different signal sources); and each venue had a different set of loudspeakers (3 different manufacturers and models).
Of course, there were 3 different listening rooms.
Just my opinions and experience.
Note: Measurements are one thing. Double Blindfold Listening tests are another thing.
They may agree, or they may not agree.
Have fun setting up a true double blind test.
Last edited:
As I mentioned (see a few pages back) I looked at several modifications from various sources on the net and considered which ones would work well for my Bewitch 6550, my changes to the circuit are based on Lampizator modifications
http://lampizator.eu/amplifiers/china/bewitch 6550/bewitch.html
and adopted even if the gain appears to be slightly lower than the first and second level with Matched
Russian 6SN7 tubes with identical values per side.
my Bewitch is either operated alone, or amplifies the incoming signal via an Adcom preamplifier.
The factory-installed RCA jacks were actually a disadvantage that did not have a reasonable contact, this problem also existed with some RCA cables of different price quality, so I exchanged them without further ado, maybe you have other RCA jacks from other manufacturers installed in your Bewitch than mine!
Hello 6A3sUMMER
Yes, that's true with your statement about the test method, so it certainly brings more clarity with the test whether it's audible or not!
This will certainly be a bit difficult for me to implement, but if there is a chance I will try.