I've got two Counterpoint SA-220 FET output-Tube hybrid power amplifiers to fool around with.
I'd like to mod the PS (first job!) for the tube section. It is a Tube rectified PS, done with a 6CA4 (aka EZ81) Rectifier tube.
According to one engineer I was speaking to, this particualr tube in this particular amplifier is quite overloaded as far a current draw goes. I was about to upgrade the PS caps that are in there now, but he says the unit has too much capacitance already. The heavy draw makes the tube nosier, not quieter. The 'slap' of the rectifier turning on and off is far worse than it should be, due to the heavy current draw in this design. After he explained to me what he meant, it made sense.
He also said, that the unit had a 'motorboat' issue, early on, as the windings for the tube PS are on the same trasformer as the PS for the FET output section. Go figure, eh? 😉
The heater supply is under driven. I was toying with the idea of removing the mini 9-pin socket and retro-fitting a octal tube socket, and seeing what a bigger rectifier might do, but I need abit of advise first. 1 amp for the 6CA4, the 5AR4 uses 1.9 amps. I think it can do it, as the (heater rectification) rectifier is not hot under it's current stressing. Perhaps I can add a simple heater trasformer into the mix on the unit. I also have mounting space for Inductors, for a bit of filtering. There are no inductors right now, in the stock design. Shame.
Should I just go solid state here, and forget about it?
I'm not looking to design a car, here. Or send men to the moon. I'm looking for what is ultimately a simple, elegant, and photogenic fix ot the unit. No heavy mods and **** hanging all over the place. KISS. I will have to re-sell the units some day, maybe sooner than I think. So keeping the mods clean and simple is the way to go. Increase the value of the unit, not decrase it by scaring the potential customers away. Which, in my opinion, something as simple as a larger rectifier, the good ole' 5AR4, GZ34, or the ..uh..memory here..5U4GB?maybe that one too.
Any quick fixes? like an octal, in your opinion?
I'd like to mod the PS (first job!) for the tube section. It is a Tube rectified PS, done with a 6CA4 (aka EZ81) Rectifier tube.
According to one engineer I was speaking to, this particualr tube in this particular amplifier is quite overloaded as far a current draw goes. I was about to upgrade the PS caps that are in there now, but he says the unit has too much capacitance already. The heavy draw makes the tube nosier, not quieter. The 'slap' of the rectifier turning on and off is far worse than it should be, due to the heavy current draw in this design. After he explained to me what he meant, it made sense.
He also said, that the unit had a 'motorboat' issue, early on, as the windings for the tube PS are on the same trasformer as the PS for the FET output section. Go figure, eh? 😉
The heater supply is under driven. I was toying with the idea of removing the mini 9-pin socket and retro-fitting a octal tube socket, and seeing what a bigger rectifier might do, but I need abit of advise first. 1 amp for the 6CA4, the 5AR4 uses 1.9 amps. I think it can do it, as the (heater rectification) rectifier is not hot under it's current stressing. Perhaps I can add a simple heater trasformer into the mix on the unit. I also have mounting space for Inductors, for a bit of filtering. There are no inductors right now, in the stock design. Shame.
Should I just go solid state here, and forget about it?
I'm not looking to design a car, here. Or send men to the moon. I'm looking for what is ultimately a simple, elegant, and photogenic fix ot the unit. No heavy mods and **** hanging all over the place. KISS. I will have to re-sell the units some day, maybe sooner than I think. So keeping the mods clean and simple is the way to go. Increase the value of the unit, not decrase it by scaring the potential customers away. Which, in my opinion, something as simple as a larger rectifier, the good ole' 5AR4, GZ34, or the ..uh..memory here..5U4GB?maybe that one too.
Any quick fixes? like an octal, in your opinion?
well, I'm gonna anyway. Not that concerned about how much they sell for, if I do end up selling them, that is...: things are going farily well, yes. Actually, i may never sell them.
I'm going pro on the loudspeaker design thing, and will have to have two racks (or more) of amplifiers sitting around that I can swap out to check the passive crossover designs. Or, just tons of different sounding ampliers, period. I'm up to about 7 amps or so right now, and I'm a bit short on the tube units. I'm a bit tired of the electrocutions.....(see short comment above for the joke part)
"Electrocutions? Good.. Down the hall, first B+ rail on the left...."
I'm going pro on the loudspeaker design thing, and will have to have two racks (or more) of amplifiers sitting around that I can swap out to check the passive crossover designs. Or, just tons of different sounding ampliers, period. I'm up to about 7 amps or so right now, and I'm a bit short on the tube units. I'm a bit tired of the electrocutions.....(see short comment above for the joke part)
"Electrocutions? Good.. Down the hall, first B+ rail on the left...."
Hi,
What exactly is the 6CA4 feeding?
A pair of 6DJ8s?
No idea how much capacitance they stuck right of the rectifier but if it can take for a few days it can usually take it for ages too....
Oh, what do you mean by underdriven? Too low voltage on the heaters of the tubes?
If so, how much do you measure?
As for the rest, I'm with SY here.
Cheers, 😉
What exactly is the 6CA4 feeding?
A pair of 6DJ8s?
No idea how much capacitance they stuck right of the rectifier but if it can take for a few days it can usually take it for ages too....
Oh, what do you mean by underdriven? Too low voltage on the heaters of the tubes?
If so, how much do you measure?
As for the rest, I'm with SY here.
Cheers, 😉
The 6CA4 is feeding 5 6DJ8's. The unit, in stock form, has about 800uf of capacitance. B+ rail is running at about 365 VDC or so. I haven't investigated much furthur than that, as of yet. I'm waiting to get a schematic.
The 800 uf is distributed as 8x100uf 450 volt snap mount UCC caps.
what I mean, is the SS 25 amp rectifier for the DC heaters is definitely underdriven. Don't know about the winding in the tranny, though! I wouldn't want to overtax that. That is what I meant about possibly running a separate heater transformer for the heavier heater on the larger rectifier. The 6CA4 is 1 amp,and the 5AR4 is 1.9 amps on the heater. that's substantial increase overall (5x 6DJ8 + 1 Rectifier tube, at 1 amp or 1.9amp) in heater current draw.
The 800 uf is distributed as 8x100uf 450 volt snap mount UCC caps.
what I mean, is the SS 25 amp rectifier for the DC heaters is definitely underdriven. Don't know about the winding in the tranny, though! I wouldn't want to overtax that. That is what I meant about possibly running a separate heater transformer for the heavier heater on the larger rectifier. The 6CA4 is 1 amp,and the 5AR4 is 1.9 amps on the heater. that's substantial increase overall (5x 6DJ8 + 1 Rectifier tube, at 1 amp or 1.9amp) in heater current draw.
800mF is an awful lot of capacitance for any valve amp, especially when it's only the input section, and valve rectified.
The 5AR4/GZ34 or 5U4 have 5V heaters/filaments, so they will need a different heater/filament voltage from the 6.3V 6DJ8s. Also, the 5U4 is directly heated, while the 5AR4/GZ34 is indirectly heated, but the cathode is internally connected to the heater. So you basically need a separate 5V winding, which will be at the potential of your B+.
The 5AR4/GZ34 or 5U4 have 5V heaters/filaments, so they will need a different heater/filament voltage from the 6.3V 6DJ8s. Also, the 5U4 is directly heated, while the 5AR4/GZ34 is indirectly heated, but the cathode is internally connected to the heater. So you basically need a separate 5V winding, which will be at the potential of your B+.
If you're going to do it, do it right. A Maida-style regulator will give you a very, very solid supply rail.
SY said:
And to quote the Director of the hospital, In the film "The meaning of Life", when he is told that the woman is about to give 'birth', ...."What sort of thing is that?"
Sounds like some sort of tube hybrid regualtor? maybe a SS regualtion of tube? Something simple, I hope?
Thanks for your effort!
PS: I have both the RC-30 RCA book, and I believe what is one of the last renditions of the Sylvania tube reference manual. I'm thinking of find in a way to get them on the company server, since I paid for the bloody thing and it's huge bandwith use. Got about 160 gig doing nothing...
It's a machine that goes, "piiiiiing." Most expensive one we've got.
The Maida regulator can be found in National Semiconductor's regulator application notes. It's basically an LM317 with some guard circuitry around it. Very simple, very effective, 98% as good as a super-complex ultra-high-performance discrete regulator (overkill for your application).
EDIT: Michael Maida, National Semiconductor LB-47, from "Linear Applications Handbook." It's also reviewed in Morgan Jones "Valve Amplifiers."
The Maida regulator can be found in National Semiconductor's regulator application notes. It's basically an LM317 with some guard circuitry around it. Very simple, very effective, 98% as good as a super-complex ultra-high-performance discrete regulator (overkill for your application).
EDIT: Michael Maida, National Semiconductor LB-47, from "Linear Applications Handbook." It's also reviewed in Morgan Jones "Valve Amplifiers."
Thanks Sy.
BTW, a project for the folks here, might be to crash into Paul McGowans party (a-la 'Blazing saddles'...ie, the fight crashes through the wall into the cafeteria), and design a 'tracking' AC regulator. Something with a phase shift to the 60hz signal (cap-ind loading anyone?), to give the unit some'headroom' to work with. What I am speaking of is the PS Audio 're-generators' that Paul is selling. I figure even a Tranny that takes the AC, cranks it up by, let's say 25%, then, reg it back down to the desired 120AC. USe something that tracks with the 120AC, and regs what is above that 120.
Should be very doable, and efficient. Huge headroom in terms of thermal capacity and current throughput. Just spec a big-*** tranny....
Is that a stupid idea? It's been floating on the edge of thought for quite some time and PS Audio's $3500 toy may be nice, but is it really nessessary, in the end, for maximum fidelity? Can we do similar for considerably lower cost? Ie, for about $5-600 in parts, make one with three-four times the current capacity?
I realize that draw-phase angle may shoot this idea in the foot, but it may be doable. I haven't looked at it too closely.
BTW, a project for the folks here, might be to crash into Paul McGowans party (a-la 'Blazing saddles'...ie, the fight crashes through the wall into the cafeteria), and design a 'tracking' AC regulator. Something with a phase shift to the 60hz signal (cap-ind loading anyone?), to give the unit some'headroom' to work with. What I am speaking of is the PS Audio 're-generators' that Paul is selling. I figure even a Tranny that takes the AC, cranks it up by, let's say 25%, then, reg it back down to the desired 120AC. USe something that tracks with the 120AC, and regs what is above that 120.
Should be very doable, and efficient. Huge headroom in terms of thermal capacity and current throughput. Just spec a big-*** tranny....
Is that a stupid idea? It's been floating on the edge of thought for quite some time and PS Audio's $3500 toy may be nice, but is it really nessessary, in the end, for maximum fidelity? Can we do similar for considerably lower cost? Ie, for about $5-600 in parts, make one with three-four times the current capacity?
I realize that draw-phase angle may shoot this idea in the foot, but it may be doable. I haven't looked at it too closely.
My first thought is a big Variac.Most will go a decent percentage over the line voltage (I think both of mine go up to 140Vac output.Seen some that go to 150-160Vac output.)
I just picked up an older Best Power 660VA (400W) UPS for $5 at a surplus place,the batteries were dead,but I've replaced them with car batteries.I'm ready for another ice storm. 😀
This particular unit regulates the output (always right at 120vac out,and has a sinewave output. In brownout situations it will switch to batt power,and in overvoltage situations,I believe it will regulate the output for a bit(up to a certain voltage,maybe 130vac?),and then switch to batt power.
A good larger UPS (1KVA) will cost a fair bit,but you'll also have the ability to run some lights,refridgerator (gotta keep the beer cold!),etc. when the power goes out. 🙂
Gotta be cheaper than $3500,I'd think.
I just picked up an older Best Power 660VA (400W) UPS for $5 at a surplus place,the batteries were dead,but I've replaced them with car batteries.I'm ready for another ice storm. 😀
This particular unit regulates the output (always right at 120vac out,and has a sinewave output. In brownout situations it will switch to batt power,and in overvoltage situations,I believe it will regulate the output for a bit(up to a certain voltage,maybe 130vac?),and then switch to batt power.
A good larger UPS (1KVA) will cost a fair bit,but you'll also have the ability to run some lights,refridgerator (gotta keep the beer cold!),etc. when the power goes out. 🙂
Gotta be cheaper than $3500,I'd think.
Hi Ken,
If your voltage regulation inside the unit is working, you do not need to regulate the incoming AC (in general). The SA-220 does not have any regulation, just filtering. It may benifit from AC regulation but it's simpler to regulate the HV for the tubes. These are your voltage amplification stage and are the most sensitive to noise and voltage changes. With the additional filtering used in this unit, you will not hear any difference between SS and tube regulation. The same holds true for the rectifier used. The 6CA4 is a good choice with reduced capacitance, some are using the 6BK4 (higher internal R), it needs the socket rewired.
Since these like to explode, and are a devil to repair to original configuration, mods are not that hard on the value. SY, you would not believe how close the darn fets have to be matched! I think the most reasonable first step would be to reduce the input capacitance and install a SS regulator for the tube B+. A new set of tubes (Electroharmonix is a reasonable choice) would be next. Stop there and have a listen.
-Chris
If your voltage regulation inside the unit is working, you do not need to regulate the incoming AC (in general). The SA-220 does not have any regulation, just filtering. It may benifit from AC regulation but it's simpler to regulate the HV for the tubes. These are your voltage amplification stage and are the most sensitive to noise and voltage changes. With the additional filtering used in this unit, you will not hear any difference between SS and tube regulation. The same holds true for the rectifier used. The 6CA4 is a good choice with reduced capacitance, some are using the 6BK4 (higher internal R), it needs the socket rewired.
Since these like to explode, and are a devil to repair to original configuration, mods are not that hard on the value. SY, you would not believe how close the darn fets have to be matched! I think the most reasonable first step would be to reduce the input capacitance and install a SS regulator for the tube B+. A new set of tubes (Electroharmonix is a reasonable choice) would be next. Stop there and have a listen.
-Chris
FET matching
I would believe it. I've never been a fan of the Counterpoint design.
What size source degeneration resistor do they use?
I would believe it. I've never been a fan of the Counterpoint design.
What size source degeneration resistor do they use?
Hi SY,
There are no degeneration resistors. "They are bad for the sound". Hmmm, I think that means that the already low damping factor gets worse still. Now the designer is on a "solid state is bad" kick. His current mods include pulling all semiconductors and using "magic" parts. He has gone to bipolar transistors as matching the fets is an all day affair. I know this from experience.
-Chris
There are no degeneration resistors. "They are bad for the sound". Hmmm, I think that means that the already low damping factor gets worse still. Now the designer is on a "solid state is bad" kick. His current mods include pulling all semiconductors and using "magic" parts. He has gone to bipolar transistors as matching the fets is an all day affair. I know this from experience.
-Chris
Oh, lord, that's BAD design. I mean REALLY bad. Trading off known fatal hazards against, ummm, "imaginative" problems....
My already low regard for Counterpoint has just gone down a few more notches.
My already low regard for Counterpoint has just gone down a few more notches.
That and they sound rough at best, as you can imagine. The SA-100 has far more serious faults plus the ones already mentioned.
However, a friend had an SA-100 that didn't sound good. So I started to fix the design errors. That increased the sound quality more than you can imagine. Now I've totally redesigned the vas (tubes) and most lately, the output section.
Now, with just over 7 dB feedback I have 0.06% THD up to my measurement limit of 20KHz (my oscillator is rated 0.05%, so there is error here). The damping factor is up to 45 ish. The heatsinks run cool and this amp now sounds good enough for me to own. It does compare with a Marantz 300DC in sound quality. The Marantz is still a better amp. Oh, and the Counterpoint (now something else) drives my PSB Stratus Golds with authority. This was unexpected.
-Chris
However, a friend had an SA-100 that didn't sound good. So I started to fix the design errors. That increased the sound quality more than you can imagine. Now I've totally redesigned the vas (tubes) and most lately, the output section.
Now, with just over 7 dB feedback I have 0.06% THD up to my measurement limit of 20KHz (my oscillator is rated 0.05%, so there is error here). The damping factor is up to 45 ish. The heatsinks run cool and this amp now sounds good enough for me to own. It does compare with a Marantz 300DC in sound quality. The Marantz is still a better amp. Oh, and the Counterpoint (now something else) drives my PSB Stratus Golds with authority. This was unexpected.
-Chris
That's significant. Do you have an idea of how the harmonics are distributed? Is the 7dB figure the global feedback?
Hi SY,
Yeah, I'm really happy about the way it turned out. The 7 dB figure is total global feedback. The distortion is mostly 2nd harmonic from the tube voltage amp.
I did very little to that section. Mostly, the second stage was current starved (corrected) and the current source for the output follower was tied to the first stage cathode resistor. Talk about non-linear feedback! (corrected by separating). The other thing that was done was the addition of a nice HV regulator.
-Chris
Yeah, I'm really happy about the way it turned out. The 7 dB figure is total global feedback. The distortion is mostly 2nd harmonic from the tube voltage amp.
I did very little to that section. Mostly, the second stage was current starved (corrected) and the current source for the output follower was tied to the first stage cathode resistor. Talk about non-linear feedback! (corrected by separating). The other thing that was done was the addition of a nice HV regulator.
-Chris
yeah. It blew my mind when I got both working and then settled down and read the owners manual. A damping factor of 6? WTF? And I bought these to run in MONO????
And..now you tell me they are dangerous? ****.
On the Other hand... if they sound good, and like music, instead of an engineering exersize.. I'll take that first. I don't have an opinion of them yet.. haven't had them long enough. Id love to have you have a go at them Chris, but I suspect it is much like our loudspeaker endeavor. If you want a pair, at this time, you have to sign some -very- binding NDA's. I'll sign your NDA, if you are of that sort of mind on the subject.
Don't take this as a slight guys, but.. I've (More specifically WE!)been building speakers over here that do very seriously challenge the state of the art in any field of acoustics (be that military, pro, industrial or even black ops), or more specifically, damping solutions. What I'm saying is, I've got a window to look through that you haven't ever heard the likes of. Chris has a small bit of an understanding of what I speak of. A minimum specification is the entire noise floor in all dimensions is 50-70db down from the signal, which puts it ..2-5 times better than anything else out there at this time. that's one component of what is going on. The best drivers in the world become the weak link, at that point.
Hopefully Goo Systems soon gives us the cash to get these loudspeakers into the marketplace. What I'm saying is, and hopefully politely.. I hope you guys aren't confusing false detail over real detail. Or music, for that matter. Not that I think that this amplifier needs defending or that I am defending it. I've just never head an all solid state amplifier sound that good. Which is whY I bought the counterpoint amps.
When I go to audio shops to hear anything, I invariably leave disgusted with the state of what people seem to think that music sounds like. $200-400k systems that are so boring and bad that I want to pee in the $3000 leather listening chair just to cause some excitement. And I'm praying to the gods that you guys aren't headed in that direction.
In essence? To hell with good engineering if it does disservice to music.
BTW, Goo Systems sole purpose in my mind when I created it, was to commit an act of brigandage on the screen industry to get the cash to get into the loudspeaker business.
I bought bunch of basic Sovtek 6922's, and matched them on my B&K dyna 707. For whatever that's worth. I hadn't seen the schematic at the time, so I put the best match/gain in the phase splitter (for I was looking to go mono), the worst matched in the gain socket (an I/O tube I figured), and the second best matched in the driver (presumably ouput fet driving..so....) sockets. I bought the standard sovtek,as I also bought 8 pairs of 7591A tubes..a bunch of 12ax7LPS sovteks,..and some 7199's..and... some 396A's for the shanling SACD players.
At that point, my wallet was lying on the floor...crying, shaking...desparately trying to cover its wounds..etc... I showed mercy and stopped.
So no electro-harmonix 6922EH at this time. Next time maybe.
PS. dange eet Sigh, thes plac neds a spel checkr.
And..now you tell me they are dangerous? ****.
On the Other hand... if they sound good, and like music, instead of an engineering exersize.. I'll take that first. I don't have an opinion of them yet.. haven't had them long enough. Id love to have you have a go at them Chris, but I suspect it is much like our loudspeaker endeavor. If you want a pair, at this time, you have to sign some -very- binding NDA's. I'll sign your NDA, if you are of that sort of mind on the subject.
Don't take this as a slight guys, but.. I've (More specifically WE!)been building speakers over here that do very seriously challenge the state of the art in any field of acoustics (be that military, pro, industrial or even black ops), or more specifically, damping solutions. What I'm saying is, I've got a window to look through that you haven't ever heard the likes of. Chris has a small bit of an understanding of what I speak of. A minimum specification is the entire noise floor in all dimensions is 50-70db down from the signal, which puts it ..2-5 times better than anything else out there at this time. that's one component of what is going on. The best drivers in the world become the weak link, at that point.
Hopefully Goo Systems soon gives us the cash to get these loudspeakers into the marketplace. What I'm saying is, and hopefully politely.. I hope you guys aren't confusing false detail over real detail. Or music, for that matter. Not that I think that this amplifier needs defending or that I am defending it. I've just never head an all solid state amplifier sound that good. Which is whY I bought the counterpoint amps.
When I go to audio shops to hear anything, I invariably leave disgusted with the state of what people seem to think that music sounds like. $200-400k systems that are so boring and bad that I want to pee in the $3000 leather listening chair just to cause some excitement. And I'm praying to the gods that you guys aren't headed in that direction.
In essence? To hell with good engineering if it does disservice to music.
BTW, Goo Systems sole purpose in my mind when I created it, was to commit an act of brigandage on the screen industry to get the cash to get into the loudspeaker business.
I bought bunch of basic Sovtek 6922's, and matched them on my B&K dyna 707. For whatever that's worth. I hadn't seen the schematic at the time, so I put the best match/gain in the phase splitter (for I was looking to go mono), the worst matched in the gain socket (an I/O tube I figured), and the second best matched in the driver (presumably ouput fet driving..so....) sockets. I bought the standard sovtek,as I also bought 8 pairs of 7591A tubes..a bunch of 12ax7LPS sovteks,..and some 7199's..and... some 396A's for the shanling SACD players.
At that point, my wallet was lying on the floor...crying, shaking...desparately trying to cover its wounds..etc... I showed mercy and stopped.
So no electro-harmonix 6922EH at this time. Next time maybe.
PS. dange eet Sigh, thes plac neds a spel checkr.
SY,
I should mention that most of my testing was done at one watt output. The distortion figure goes up to 0.21% 1KHz at 32W. This is due to the vas. I didn't want to burn out my zobels at high frequency / high power. I was driving both channels into 8 ohm dummy loads for testing. One waveform I used was a differentiated squarewave. The phase shift occurs mostly in the vas as you would expect, the phase shift wasn't very high. I didn't bother to measure it. I was more interested in the magnitude of the phase shift in general and how much distortion the output stage was contributing (not much). The vas is loaded by 100K || 1.5 Meg resistive for the most part.
I am limited in that I'm using a Leader LAG-120B generator, Leader LDM-171 and HP 331A analysers. The amp is running on a metered variac for tests. The dummy loads are Dale 250W 8 ohm models mounted on a big heatsink.
Listening tests are done with a Denon DCD-S10 connected directly to the amp. with a different selections (mostly jazz and female vocalists).
I did find that as the feedback ratio went up, the dynamics disappeared. The feedback tap off point includes the output fuse within the loop.
Ken,
You can design a good amp with all the dynamics and transients of a low feedback tube amp. using solid state parts.
-Chris
I should mention that most of my testing was done at one watt output. The distortion figure goes up to 0.21% 1KHz at 32W. This is due to the vas. I didn't want to burn out my zobels at high frequency / high power. I was driving both channels into 8 ohm dummy loads for testing. One waveform I used was a differentiated squarewave. The phase shift occurs mostly in the vas as you would expect, the phase shift wasn't very high. I didn't bother to measure it. I was more interested in the magnitude of the phase shift in general and how much distortion the output stage was contributing (not much). The vas is loaded by 100K || 1.5 Meg resistive for the most part.
I am limited in that I'm using a Leader LAG-120B generator, Leader LDM-171 and HP 331A analysers. The amp is running on a metered variac for tests. The dummy loads are Dale 250W 8 ohm models mounted on a big heatsink.
Listening tests are done with a Denon DCD-S10 connected directly to the amp. with a different selections (mostly jazz and female vocalists).
I did find that as the feedback ratio went up, the dynamics disappeared. The feedback tap off point includes the output fuse within the loop.
Ken,
You can design a good amp with all the dynamics and transients of a low feedback tube amp. using solid state parts.
-Chris
- Status
- Not open for further replies.