Hi,
there are various methods of coupling driver tubes to the output, which of the 3 methods is preffered due to decreased signal degredation?
1) resistor and coupling cap, or just resistor.
2) Interstage transformer as used by Andrea Ciuffolli.
3) plate choke as in drd circuits
I've done a bit of web searching, and each has it's merits in terms of cost and performance etc, has anyone been able to compare these methods of coupling for their sonic benefits?
Thanks
Raja
there are various methods of coupling driver tubes to the output, which of the 3 methods is preffered due to decreased signal degredation?
1) resistor and coupling cap, or just resistor.
2) Interstage transformer as used by Andrea Ciuffolli.
3) plate choke as in drd circuits
I've done a bit of web searching, and each has it's merits in terms of cost and performance etc, has anyone been able to compare these methods of coupling for their sonic benefits?
Thanks
Raja
They all sound very different. And it's not easy to make absolute claims as you may be dealing with different caps, chokes, interstages. Some people are absolutely in love with the interstage sound, no matter how crappy the iron. This i do not understand. Unless there is lots of nickel involved i'd stay away from interstage transformers. Best advice: ignore all opinions and find out for yourself - yes it costs some money but there simply is no other way.
Hi,
You're right, it would be expensive to try all three..........
My intrigue of the interstage trannies came due to reports that Transformer volume controls sound superior to resistive attenuators, so I was wondering if this phenomena was also true with i/s trannies.... especially now as it's possible to get amo' cores.
Well I'm no closer to making a choice between DRD with plate choke or i/s transfo a ciuffolli circuit..........
Thanks
Raja
You're right, it would be expensive to try all three..........
My intrigue of the interstage trannies came due to reports that Transformer volume controls sound superior to resistive attenuators, so I was wondering if this phenomena was also true with i/s trannies.... especially now as it's possible to get amo' cores.
Well I'm no closer to making a choice between DRD with plate choke or i/s transfo a ciuffolli circuit..........
Thanks
Raja
Yes, they do. In many respects but not all. If focus and air in the high frequencies are important for you TVCs and interstage transformers are not that great. It is in the overall presentation they really excell, not in the details. Choke coupling does not seem to have the same problems, at least not to my ears. If you read subjective opinions carefully you'll notice the transformer aficionados always talk of musicality and other elusive qualities but seldom mention incredible bandwidth. I like and use a TVC but in all honesty have already forgotten what a really great crashing cymbal sounds like.
Just to muddy the water....
Raja,
You can ignore opinions of what makes for musicality, though IMO it's best to keep your ears open.
What you cannot ignore is the engineering. CR coupling and inductive coupling are not equivalent nor are they interchangeable.
As you will find by studying examples of existing designs, CR coupling is frequently employed in amplifiers that use medium to high amounts of negative feedback. It is suitable because it gives little phase shift with frequency, and so a good margin of stability is possible.
CR coupling most common in later designs, but has 2 disadvantages: 1- when used to drive a low to medium mu triode output valve, the supply voltage needed for the driver stage is much greater than that for the output stage. This makes for complexity and inefficiency. 2- The behaviour under, and following overload is poor. This "blocking" can be quite noticeable. This is really unfortunate as soft overloading is one of the good features of valve amps.
LC coupling (choke in anode) is more costly, but overcomes one of the disadvantges of CR coupling. The available drive (P-P) at the anode of the driver valve is nearly twice the supply voltage, so a common, or lower supply is practical. Blocking can still occur. The phase stability margin means that large amounts if negative feedback are not practical.
DRC is direct coupled, so blocking may not occur, but it will depend on the detail of the biasing circuitry.
Transformer coupling overcomes the supply voltage problem and the blocking problem. It is the most costly. It does allow for a coupling ratio different from 1:1 to be used, but for most low mu output valves 1:1 is close to best. As with the LC coupling, this scheme is best for zero or low negative feedback designs.
It is my opinion that a zero global feedback amp running at low level will sound identical with each of these schemes properly implemented. It is at higher levels that the changes occur. With sources of a wide dynamic range, an amplifier can be driven into overload without realising it.
Raja,
You can ignore opinions of what makes for musicality, though IMO it's best to keep your ears open.
What you cannot ignore is the engineering. CR coupling and inductive coupling are not equivalent nor are they interchangeable.
As you will find by studying examples of existing designs, CR coupling is frequently employed in amplifiers that use medium to high amounts of negative feedback. It is suitable because it gives little phase shift with frequency, and so a good margin of stability is possible.
CR coupling most common in later designs, but has 2 disadvantages: 1- when used to drive a low to medium mu triode output valve, the supply voltage needed for the driver stage is much greater than that for the output stage. This makes for complexity and inefficiency. 2- The behaviour under, and following overload is poor. This "blocking" can be quite noticeable. This is really unfortunate as soft overloading is one of the good features of valve amps.
LC coupling (choke in anode) is more costly, but overcomes one of the disadvantges of CR coupling. The available drive (P-P) at the anode of the driver valve is nearly twice the supply voltage, so a common, or lower supply is practical. Blocking can still occur. The phase stability margin means that large amounts if negative feedback are not practical.
DRC is direct coupled, so blocking may not occur, but it will depend on the detail of the biasing circuitry.
Transformer coupling overcomes the supply voltage problem and the blocking problem. It is the most costly. It does allow for a coupling ratio different from 1:1 to be used, but for most low mu output valves 1:1 is close to best. As with the LC coupling, this scheme is best for zero or low negative feedback designs.
It is my opinion that a zero global feedback amp running at low level will sound identical with each of these schemes properly implemented. It is at higher levels that the changes occur. With sources of a wide dynamic range, an amplifier can be driven into overload without realising it.
Heh, heh, heh. Overload behaviour, my idee fixee. Every time we've heard differences in two properly designed amps, we've been able to trace it to just that cause. And when we take rigorous care to prevent clipping or overload, we seem to lose our magic ability to distinguish amp A from amp B.
There was an interesting article in Audio Amateur about this, perhaps twenty years ago, Krueger and Carlstrom were the authors, IIRC.
There was an interesting article in Audio Amateur about this, perhaps twenty years ago, Krueger and Carlstrom were the authors, IIRC.
SY there is only one kind of coupling possible in the amps that you like 
Actually I had a pair well built Mullard 5-20's on the bench the other day. With their distributed load and push-pull distortion cancellation, they measured beautifully right up to when clipping suddenly happened. Very much like a SS amp.
Unfortunately they sounded harsh and metalic on anything but the lowest volume. The source mattered. CD was the worst.
At the same time I has a pair of 2A3 SET's in. These had rather small parafeed transformers. The LF end was nothing like the thunder out of the 5-20's, nor was the total noise level (3W vs 20W). But even with my inefficient speakers, I know which I would rather live with.
You might argue that the SET overloaded from the word go, but is was in a competely musical way.
That's why we use valves, isn't it?
Actually I had a pair well built Mullard 5-20's on the bench the other day. With their distributed load and push-pull distortion cancellation, they measured beautifully right up to when clipping suddenly happened. Very much like a SS amp.
Unfortunately they sounded harsh and metalic on anything but the lowest volume. The source mattered. CD was the worst.
At the same time I has a pair of 2A3 SET's in. These had rather small parafeed transformers. The LF end was nothing like the thunder out of the 5-20's, nor was the total noise level (3W vs 20W). But even with my inefficient speakers, I know which I would rather live with.
You might argue that the SET overloaded from the word go, but is was in a competely musical way.
That's why we use valves, isn't it?
Peter,
Not all at the same time, I must admit.
I've done CR vs transformer though. These are as far apart as any in terms of topopogy.
Does anything ever sound identical? Well allowing for the occasional cosmic particle passing through my brain an affecting my perception, yes. But of course that entire sentence brings into doubt my absolute judgement
But then we were talking coupling, weren't we....
Not all at the same time, I must admit.
I've done CR vs transformer though. These are as far apart as any in terms of topopogy.
Does anything ever sound identical? Well allowing for the occasional cosmic particle passing through my brain an affecting my perception, yes. But of course that entire sentence brings into doubt my absolute judgement
But then we were talking coupling, weren't we....
Frank, the article in question was Carlstrom, Krueger, and Greenhill, Audio Amateur 3/1982, p 30, "Some Amplifiers Do Sound Different." The amps they compared included an Audio Research D120, all running an Acoustat electrostat (5 ohms impedance, pretty flat). The amps measured fine on the bench, but in double-blind comparisons, they could identify the AR amp to better than a 0.05 significance.
They determined that the amps were being driven into clipping about 1% of the time, and that there was an overload recovery issue with the AR. It was most audible on a bass drum from an Eagles album, where they heard a distinct buzz. They couldn't distinguish the amps with symphonic material (unspecified what that album was).
They determined that the amps were being driven into clipping about 1% of the time, and that there was an overload recovery issue with the AR. It was most audible on a bass drum from an Eagles album, where they heard a distinct buzz. They couldn't distinguish the amps with symphonic material (unspecified what that album was).
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.