What do you guys think about this transformer/pot hybrid volume control scheme? (Scroll down the page to the bright yellow box and description.)
How would something like this compare to the pure-transformer solutions like the S&B tx102? Respective strengths and weaknesses? Aside from voltage gain, what does the transformer bring this otherwise simple shunt system? If you need <unity gain, would a commercial 1:1 600ohm line transformer work ok/better, or would the whole point of the transformer then be moot?
I'd like to do a 6-channel volume control for triamping, and I can't afford 6 tx102 units. Is there anything to recommend the above system over a ladder-resistor attenuator?
How would something like this compare to the pure-transformer solutions like the S&B tx102? Respective strengths and weaknesses? Aside from voltage gain, what does the transformer bring this otherwise simple shunt system? If you need <unity gain, would a commercial 1:1 600ohm line transformer work ok/better, or would the whole point of the transformer then be moot?
I'd like to do a 6-channel volume control for triamping, and I can't afford 6 tx102 units. Is there anything to recommend the above system over a ladder-resistor attenuator?
Well, if we're both talking about that yellow box, then I'd say it really has nothing at all to do with a TVC. That doesn't mean it isn't a good idea though; using a line input transformer can lower the common mode noise of even an unbalanced connection by as much as 60-80dB at 60Hz. I think a 150-10k stepup transformer is a poor choice though, which results in that particular device being so finicky.
Jensen Transformers has an excellant whitepaper on unbalanced input line transformers.
Jensen Transformers has an excellant whitepaper on unbalanced input line transformers.
I tried to gather a list of transformers with statements about quality and cost. I also have some measurements on the second page of this thread.
http://www.diyaudio.com/forums/showthread.php?s=&threadid=35347
I would say the Edcor's are usable (barely). If you are really going for high end I would either take a gamble with Cinemag or get a Jensen, but for anything other than subwoofer bass the distortion is probably lower than that of the speakers themselves.
Talk to me if you go the Cinemag route, I might be interested in piggybacking if we'd have enough for a discount.
http://www.diyaudio.com/forums/showthread.php?s=&threadid=35347
I would say the Edcor's are usable (barely). If you are really going for high end I would either take a gamble with Cinemag or get a Jensen, but for anything other than subwoofer bass the distortion is probably lower than that of the speakers themselves.
Talk to me if you go the Cinemag route, I might be interested in piggybacking if we'd have enough for a discount.
Konnichiwa,
Well, I think it is time to give soem TVC background....
Some time in the dark days last millenium rumours of a Product that was hideously expensive and hideously looking but sounded great reached the Joenet. Not much was information was available except the thing was a transformer and called Silver Rock (but cost it's weight in 22 Carat Gold!), in fact they had resurrected a level control approach first shown by WE in the 1920's.
This originally got me thinking and the benefits of using this principle for a passive line controller where immediatly obvious. No specifications where available on the SR, so I had to think of my own.
I first approached Sowter in the UK about arranging a one-off production run of such devices for a group order of the Joe-Net and the London Live DIY Circle. I decided to specify the nominal impedance of the line controller transformer as 10K, based on a number of considerations (I'll come back to them below).
For a number of reasons (mainly Sowters refusal to make what we wanted and our refusal to buy what Sowter wanted to make). The group buy blew up. At any extent, a year later or so Sowter made a Volume Control transformer available, with the exact specifications I had originally asked for. This one was soon being "improved" by Sowter, based on listening to the "improvements" they seemed more related to manufacturing economics and measured performance than good sound.
Once I got in contact with Stevens & Billington Transformers I got them to make a "no compromise" design to the same specs I had given Sowter and the TX-102 was born.
Now to the specs.
I first of all considered (and still do) the following basic truisms:
1) The ideal linestage has an infinite Input impedance
2) The ideal linestage has a zero output impedance
The aim of the TVC's which I specified was to approach this ideal as much as possible, especially in the midrange.
If you use any significant level of attenuation on the typhical TVC your load on the source will become very high. If for arhguments sake you use a TX-102 into a 10K load and you apply 12db attenuation, your input impedance of the TVC will become around 160KOhm in the midband rising further as attenuation increases.
At 20Hz the inductive reactance limits your input impedance to around 50KOhm, but even that is quite high by anyones accounts. However, importantly the Midband where the ear is most sensitive sees very little load on the source and thus low distortion.
Now let us look at the output side of things. If we have a source with 3k Output Impedance (not ususual for many Valve Circuits) and we attenuate 12db, what will our output impedance be? Around 200 Ohm, this will drop further as attenuation goes up.
So what we have created is in effect a "gearbox" that attempts to load the source as little as possible while making ALL the power drawn from the source (minus a small amount of loss) to the load. The low load on the source minimises distortion there and the low output impedance allows a TVC to "drive" their loads and cables.
Much of what makes typhical TVC's sound the way they do is (IMHO) a result of this principle.
Now let us look what EP does. Note that I cannot comment how their approach sounds, I have not heard any of their stuff, I can only make a comparison on electrical grounds and based on the discussed specs....
The EP "Passive Preamp" takjes a transformer with a 1:8 Stepup and apply the result to a 10K Pot. Therefore, on the output side of things the whole thing looks like a classic "passive preamp" with a 10K Pot.
The output impedance can be low but may reach 2.5KOhm if the source has zero ohm source impedance, higher if there is a significant output impedance. We come back to this.
This 10K Pot loads the transformer which transformers the impedance by a factor of 64 (square of the ratio). In other words the 10K will appear to the source driving the EP arrangement as a 150 Ohm Load. You may wish to peruse the measurements in many stereophile reviews to see how the output stages of much equipment react to being loaded with a 600 Ohm load, which is in line level circuits considered a severe ,oad, 150 is four times as severe!
So, in electrical terms at least the EP arrangement achieves the exact and polar opposite of what the classic TVC achieves. Instead of loading down the source minally while keeping output impedance low as the classic TVC does the EP arrangement maximally loads down the source and creates a fairly high output impedance.
Sayonara
Bill F. said:
Well, I think it is time to give soem TVC background....
Some time in the dark days last millenium rumours of a Product that was hideously expensive and hideously looking but sounded great reached the Joenet. Not much was information was available except the thing was a transformer and called Silver Rock (but cost it's weight in 22 Carat Gold!), in fact they had resurrected a level control approach first shown by WE in the 1920's.
This originally got me thinking and the benefits of using this principle for a passive line controller where immediatly obvious. No specifications where available on the SR, so I had to think of my own.
I first approached Sowter in the UK about arranging a one-off production run of such devices for a group order of the Joe-Net and the London Live DIY Circle. I decided to specify the nominal impedance of the line controller transformer as 10K, based on a number of considerations (I'll come back to them below).
For a number of reasons (mainly Sowters refusal to make what we wanted and our refusal to buy what Sowter wanted to make). The group buy blew up. At any extent, a year later or so Sowter made a Volume Control transformer available, with the exact specifications I had originally asked for. This one was soon being "improved" by Sowter, based on listening to the "improvements" they seemed more related to manufacturing economics and measured performance than good sound.
Once I got in contact with Stevens & Billington Transformers I got them to make a "no compromise" design to the same specs I had given Sowter and the TX-102 was born.
Now to the specs.
I first of all considered (and still do) the following basic truisms:
1) The ideal linestage has an infinite Input impedance
2) The ideal linestage has a zero output impedance
The aim of the TVC's which I specified was to approach this ideal as much as possible, especially in the midrange.
If you use any significant level of attenuation on the typhical TVC your load on the source will become very high. If for arhguments sake you use a TX-102 into a 10K load and you apply 12db attenuation, your input impedance of the TVC will become around 160KOhm in the midband rising further as attenuation increases.
At 20Hz the inductive reactance limits your input impedance to around 50KOhm, but even that is quite high by anyones accounts. However, importantly the Midband where the ear is most sensitive sees very little load on the source and thus low distortion.
Now let us look at the output side of things. If we have a source with 3k Output Impedance (not ususual for many Valve Circuits) and we attenuate 12db, what will our output impedance be? Around 200 Ohm, this will drop further as attenuation goes up.
So what we have created is in effect a "gearbox" that attempts to load the source as little as possible while making ALL the power drawn from the source (minus a small amount of loss) to the load. The low load on the source minimises distortion there and the low output impedance allows a TVC to "drive" their loads and cables.
Much of what makes typhical TVC's sound the way they do is (IMHO) a result of this principle.
Now let us look what EP does. Note that I cannot comment how their approach sounds, I have not heard any of their stuff, I can only make a comparison on electrical grounds and based on the discussed specs....
The EP "Passive Preamp" takjes a transformer with a 1:8 Stepup and apply the result to a 10K Pot. Therefore, on the output side of things the whole thing looks like a classic "passive preamp" with a 10K Pot.
The output impedance can be low but may reach 2.5KOhm if the source has zero ohm source impedance, higher if there is a significant output impedance. We come back to this.
This 10K Pot loads the transformer which transformers the impedance by a factor of 64 (square of the ratio). In other words the 10K will appear to the source driving the EP arrangement as a 150 Ohm Load. You may wish to peruse the measurements in many stereophile reviews to see how the output stages of much equipment react to being loaded with a 600 Ohm load, which is in line level circuits considered a severe ,oad, 150 is four times as severe!
So, in electrical terms at least the EP arrangement achieves the exact and polar opposite of what the classic TVC achieves. Instead of loading down the source minally while keeping output impedance low as the classic TVC does the EP arrangement maximally loads down the source and creates a fairly high output impedance.
Sayonara
Thanks KYW. I'm a transducer wonk, and my electronics understanding is weak, so I appreciate the explanation.
A couple more questions:
1) In terms of input/output impedance, etc., how do autoformer volume controls (AVCs?) compare to TVCs?
2) TVCs like the Silver Rock, S&B, etc. typically have a single primary winding and multiple secondaries, correct? Could even better loading and bandwidth characteristics be had from multiple primary and secondary windings connected by a ladder switch?
Bill
A couple more questions:
1) In terms of input/output impedance, etc., how do autoformer volume controls (AVCs?) compare to TVCs?
2) TVCs like the Silver Rock, S&B, etc. typically have a single primary winding and multiple secondaries, correct? Could even better loading and bandwidth characteristics be had from multiple primary and secondary windings connected by a ladder switch?
Bill
Konnichiwa,
Exactly the same, assuming the (primary) inductance and ratios are the same.
I cannot talk about others, but the current S&B TX-102 uses quite complex layering and sectioning for both primary and tapped secondary. The bandwidth of this transformer is as it is as a deliberate choice, t could have been made anything one likes.
Not neccesarily. There are many ways of doing things. There are winding structures that would allow a frequency response well into the MHz range. I do not feel however that using these offers benefits.
Sayonara
Bill F. said:
Exactly the same, assuming the (primary) inductance and ratios are the same.
Bill F. said:
I cannot talk about others, but the current S&B TX-102 uses quite complex layering and sectioning for both primary and tapped secondary. The bandwidth of this transformer is as it is as a deliberate choice, t could have been made anything one likes.
Bill F. said:
Not neccesarily. There are many ways of doing things. There are winding structures that would allow a frequency response well into the MHz range. I do not feel however that using these offers benefits.
Sayonara
Konnichiwa,
There are no real advantages either way if you aim at high sonic quality, the seperate input and output current loops of the true transformer makes it more versitale and can in some situations lead to better sound.
At some point in time a large customer of S&B was looking into having the "slagleformer" made by S&B (under licence and all as I understand). Calculations suggeste that the cost savings over a 102 would be minimal but the 102 is a much more generally compatible design.
If you want to save money it is easiest to replace the expensive large nickel permalloy transformer core with a smaller Steel core and to forgo potting the transformer in shielding cans (as some makers of commercial TVC's have done).
The price to pay is comparably poor sound and high sensitivity to external hum fields.
Sayonara
Bill F. said:
There are no real advantages either way if you aim at high sonic quality, the seperate input and output current loops of the true transformer makes it more versitale and can in some situations lead to better sound.
At some point in time a large customer of S&B was looking into having the "slagleformer" made by S&B (under licence and all as I understand). Calculations suggeste that the cost savings over a 102 would be minimal but the 102 is a much more generally compatible design.
If you want to save money it is easiest to replace the expensive large nickel permalloy transformer core with a smaller Steel core and to forgo potting the transformer in shielding cans (as some makers of commercial TVC's have done).
The price to pay is comparably poor sound and high sensitivity to external hum fields.
Sayonara
Konnichiwa,
Well, here a final one bit to note....
My favourite "traditional" linestage is the line part of the Arthur Loesch Preamp.... Look here to see what it looks like: 404 Error - Page Not Found
Sayonara
Well, here a final one bit to note....
My favourite "traditional" linestage is the line part of the Arthur Loesch Preamp.... Look here to see what it looks like: 404 Error - Page Not Found
Sayonara
Kuei,
I was planning to build a TVC with a cheaper transformer than TX102 but if it looks like a passive preamp I don't like to give it a try at all.
Aren't out there any other (cheaper) transformers with multiple secondary taps?
So only with S&B we don't have something that looks like a passive preamp?
I was planning to build a TVC with a cheaper transformer than TX102 but if it looks like a passive preamp I don't like to give it a try at all.
Aren't out there any other (cheaper) transformers with multiple secondary taps?
Konnichiwa,
There are others (Audio Consulting & Sowter come to mind) that make similar transformers.
I have not tried Audio Consultings Transformers but even their "cheap" DIY Model is very expensive, I am familar with Sowter but due to my peripheral involvement with S&B should perhaps not comment too publically on the differences, so I'll leave it to someone more independent:
http://www.diyaudio.com/forums/showthread.php?postid=557174#post557174
I might add that I agree.
Sayonara
resident said:
There are others (Audio Consulting & Sowter come to mind) that make similar transformers.
I have not tried Audio Consultings Transformers but even their "cheap" DIY Model is very expensive, I am familar with Sowter but due to my peripheral involvement with S&B should perhaps not comment too publically on the differences, so I'll leave it to someone more independent:
http://www.diyaudio.com/forums/showthread.php?postid=557174#post557174
I might add that I agree.
Sayonara
Konnichiwa,
Some general notes here...
First, using a "Transformer & Pot" approach - IMHO this reliably combines the negative sides of either approach, with few of the benefits (except the ability to break groundloops).
A possible exception is in a case where you have WAY TOO MUCH gain in the system and always attenuate by a large amount such as 6 - 12db or more, then placing the "bulk" attenuation in the Transformer and using a low impedance volume control becomes feasible and even benficial. That said, you may still be better off sonically restructuring your systems gainstructure.
A typical case for the above might be using digital pro-audio gear where the full scale levels are usually around 12db higher than in consumer gear and them possibly a use with a high gain Amplifier. Then using a 4:1 Stepdown transformer followed by a 1K attenuator might be a good idea.
Secondly, multitapped transformers, these are subject to the usual drawbacks and design compromises of transformers only more-so. You can of course make your own.
If you have a very low source impedance and you use very few turns you can quite easily wind your own with good results, be it an autoformer or a true transformer (place the tapped secondary between two layers of 1/2 primary each), however the nominal impedance will be low, usually in the region of a few 100 Ohm and level handeling tends to be shitty.
As you increase turns first of all your primary inductance goes up. This can be translated into a high input impedance and it leads (with a given core) also to better level handling. The downside is that now your parasitic capacitance goes up (and thus the input impedance at high frequencies gets worse) and your leakage inductance too rises.
This means any resonances are higher Q and lower in frequency, as a TVC leaves parts of the secondary unterminated these resonances can become really troblesome, the problem rises exponentially with the nominal primary impedance. Autoformers can be a little better on that issue, but it only goes so far.
This is also a pointer as to why certain TVC's sound better than others and what core specifications to look at.
The lower the primary inductance the lower the effective input impedance and the easier it is to make the device, but the less compatible it will be with higher than average output impedance sources. The better the HF response the better the management of supersonic resonances.
As a reference, the S&B TX-102 has around 400H small signal Inductance (IIRC), Sowters nominally also 10K Transformers have 85H and the DIY Transformer from Audio Consulting has 50H (I suspect their "silver rock" is at similar levels of inductance).
Now any of these will do fine when driven from a brute force solid state output stage with 10 Ohm Zout and a few 100mA current on tap. But the lower the primary inductance the more problems will appear at low frequencies if the source impedance creeps up.
I noticed that Dick Olsher in his review found it not to be a good match to pair the Silver Rock TVC with the Lamm LP-2 Phonostage, which has an above average level of output impedance (around 3k).
Sayonara
resident said:
Some general notes here...
First, using a "Transformer & Pot" approach - IMHO this reliably combines the negative sides of either approach, with few of the benefits (except the ability to break groundloops).
A possible exception is in a case where you have WAY TOO MUCH gain in the system and always attenuate by a large amount such as 6 - 12db or more, then placing the "bulk" attenuation in the Transformer and using a low impedance volume control becomes feasible and even benficial. That said, you may still be better off sonically restructuring your systems gainstructure.
A typical case for the above might be using digital pro-audio gear where the full scale levels are usually around 12db higher than in consumer gear and them possibly a use with a high gain Amplifier. Then using a 4:1 Stepdown transformer followed by a 1K attenuator might be a good idea.
Secondly, multitapped transformers, these are subject to the usual drawbacks and design compromises of transformers only more-so. You can of course make your own.
If you have a very low source impedance and you use very few turns you can quite easily wind your own with good results, be it an autoformer or a true transformer (place the tapped secondary between two layers of 1/2 primary each), however the nominal impedance will be low, usually in the region of a few 100 Ohm and level handeling tends to be shitty.
As you increase turns first of all your primary inductance goes up. This can be translated into a high input impedance and it leads (with a given core) also to better level handling. The downside is that now your parasitic capacitance goes up (and thus the input impedance at high frequencies gets worse) and your leakage inductance too rises.
This means any resonances are higher Q and lower in frequency, as a TVC leaves parts of the secondary unterminated these resonances can become really troblesome, the problem rises exponentially with the nominal primary impedance. Autoformers can be a little better on that issue, but it only goes so far.
This is also a pointer as to why certain TVC's sound better than others and what core specifications to look at.
The lower the primary inductance the lower the effective input impedance and the easier it is to make the device, but the less compatible it will be with higher than average output impedance sources. The better the HF response the better the management of supersonic resonances.
As a reference, the S&B TX-102 has around 400H small signal Inductance (IIRC), Sowters nominally also 10K Transformers have 85H and the DIY Transformer from Audio Consulting has 50H (I suspect their "silver rock" is at similar levels of inductance).
Now any of these will do fine when driven from a brute force solid state output stage with 10 Ohm Zout and a few 100mA current on tap. But the lower the primary inductance the more problems will appear at low frequencies if the source impedance creeps up.
I noticed that Dick Olsher in his review found it not to be a good match to pair the Silver Rock TVC with the Lamm LP-2 Phonostage, which has an above average level of output impedance (around 3k).
Sayonara
Thanks Kuei,
I think I'll buy S&B TX102
I lost the group buy ... no money when the group buy exists...
Sowter are not so less expensive to buy them.Around 90-100GBP (as I remember) So it worth it to give some more and go with TX102.
As Sigurd wrote.....
But anyway thanks again for your help.
I think I'll buy S&B TX102
I lost the group buy
... no money when the group buy exists...Sowter are not so less expensive to buy them.Around 90-100GBP (as I remember) So it worth it to give some more and go with TX102.
As Sigurd wrote.....
I also don't like Sowter ... one more reason for this is your story with TVC...
But anyway thanks again for your help.
From me ... don't worry about this.
Yes,I also see it in forums over here (diyaudio) ,the only disadvantage is cost.
But if it has difficult windings,interleaving,etc I understand it.
Today I recieved an e-mail from Jonathan.He gave me an offer for the TX102.Not so bad if you think that Sowter is around 140euro.
May I say to you....
Sayonara
Interesting! Your enumeration of the difficulties of engineering a full-bandwidth TVC makes what I want to do seem easier. Wouldn't it be far less of an engineering challenge to put together three TVC/AVC pairs for tri-amping?
Since the bandwidth required of each would be relatively narrow, you could send your bass through a large-cored transformer with many turns and lots of inductance--let the parasitic capacitance and leakage inductance fall where it may. The mid transformer would require only modestly high inductance/low capacitance. The treble transformer could be far fewer turns of thinner wire since low capacitance would be far more important than inductance.
Am I correct in thinking this would be much easier to DIY for high performance than a TX-102 knock-off?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.