Although I've spent a lot of time designing high performance tube amps, my time listening to my designs is limited. Money is always a limiting factor as to what I can build and when and I usually tear one project down in order to build another. I've yet to get around to building something resembling a true SET amp although It's reasonably high on my priority list.
I specifically avoid designing anything with a signal transformer or a capacitor on the output.
I've been thinking, I noticed people talk about the "magic" of SET amplifiers. I've heard magic from amps I've designed before in the form of euphonics but I think SET sound is a bit different from the way I see it described.
My question is, is SET sound the product of the output transformer or can SET sound be achieved in a SET OTL as well? No I don't mean paralleling tubes. Ignore the power output for the purpose of this question.
I specifically avoid designing anything with a signal transformer or a capacitor on the output.
I've been thinking, I noticed people talk about the "magic" of SET amplifiers. I've heard magic from amps I've designed before in the form of euphonics but I think SET sound is a bit different from the way I see it described.
My question is, is SET sound the product of the output transformer or can SET sound be achieved in a SET OTL as well? No I don't mean paralleling tubes. Ignore the power output for the purpose of this question.
Last edited:
The PMillet Nutube amp is effectively a low-end hybrid amp. Not really a SET. It opamp buffers the output signal. I think any non push-pull thing with a tube in it is going to have SET distortion patterns.
I'm hoping this is not the case, it challenges the parameters I use for my designs. Namely avoiding overly imperfect reproduction methods, such as signal transformers. What a can of worms it would be if I had to start designing signal transformers.
Can you elaborate on your experience?
Last edited:
Just do a search here at the forum, there are probably hundreds of posts on the topic, not to mention the many articles written about the OPT in the popular press during the heyday of tube amplifier, e.g., Crowhurst's articles in Audio, Radio-Electronics, etc.
Can you tell us what kind of amps you are designing and building? Are these all OTL? So is your question how to build a SET OTL amp?
Anything as obvious as this would have been done by now if it were relatively easy to design. And then, why would you need to design a transformer when there are plenty of good designs already made and in existence? The difficulty you'd have is that they will cost money, which doesn't seem to be in your calculations. But building really good SET transformers is a bit of a black art.
This is all a question of trade-offs. To get the SET sound you use an output transformer. If you go OTL you get different outcomes - you gain and you lose.
Anything as obvious as this would have been done by now if it were relatively easy to design. And then, why would you need to design a transformer when there are plenty of good designs already made and in existence? The difficulty you'd have is that they will cost money, which doesn't seem to be in your calculations. But building really good SET transformers is a bit of a black art.
This is all a question of trade-offs. To get the SET sound you use an output transformer. If you go OTL you get different outcomes - you gain and you lose.
Last edited:
Yes I only design directly coupled amps. I already know how to build a SET OTL, that's not the issue. My question is, will I get SET sound without the transformer?
If not, it's my understanding that a transformer will be of higher quality with a lower turns ratio. So if I use a 1:1 turns ratio signal transformer I should get a pretty close to ideal output transformer right? But then, how is an ideal output transformer any different than any other high performance method of power transfer in terms of sound?
Is imperfection what I seek for SET sound? I'm trying to figure out what parameters cause the SET sound.
If not, it's my understanding that a transformer will be of higher quality with a lower turns ratio. So if I use a 1:1 turns ratio signal transformer I should get a pretty close to ideal output transformer right? But then, how is an ideal output transformer any different than any other high performance method of power transfer in terms of sound?
Is imperfection what I seek for SET sound? I'm trying to figure out what parameters cause the SET sound.
Last edited:
The main problem of OT SE trafo is the L at low frequencies where the dc current has a big (negative) influence and parasitic at high frequency.
Same issues for the PP , of course but there isn't the dc current
There is a test can show us in which way a output stage is working. It is a THD vs. freq.
Normally with a poor stuff we can see an increment of THD fromm 100 Hz to 20 Hz and from 8-10 kHz to 20kHz and more.
This means that at low frequency, due to the poor inductance (where the dc current ha a big influence), the Z seen from tube going lower and the THD, of course, increase due to the increment of the current that the tube must deliver.
Same effect are in the high frequency due the parasitic that are capacitive (dominant).
This is the major issue about SE trafo and these aspects heavely "colour" the sound .
On lab test about THD vs frequency we have an idea in which way the amp is working in a best condition; in real world, with a complex load of the speaker, the job done by the tube is more heavy.
This as general aspect.
Walter
Same issues for the PP , of course but there isn't the dc current
There is a test can show us in which way a output stage is working. It is a THD vs. freq.
Normally with a poor stuff we can see an increment of THD fromm 100 Hz to 20 Hz and from 8-10 kHz to 20kHz and more.
This means that at low frequency, due to the poor inductance (where the dc current ha a big influence), the Z seen from tube going lower and the THD, of course, increase due to the increment of the current that the tube must deliver.
Same effect are in the high frequency due the parasitic that are capacitive (dominant).
This is the major issue about SE trafo and these aspects heavely "colour" the sound .
On lab test about THD vs frequency we have an idea in which way the amp is working in a best condition; in real world, with a complex load of the speaker, the job done by the tube is more heavy.
This as general aspect.
Walter
So your saying it's due to distortion from increased load at low and high frequencies? I thought SET sound is subjectively best in the mirange though? How does SET sound produce superior mid range euphonics to higher performance circuits if the performance through a signal transformer is optimized within the midband?
Last edited:
The results of the sound is the sum first of OT then the other components.
There are circuits where from the first stage are designed poor.
So we can have the best OT but if the swing at the primary of the trafo is not very good.............
In the midrange, where we can consider the OT ideal (this just to play) where the value of Z is high and the power tube can work in the linear region, we can see only the products of the IInd harmonic ; this is the mainly reason about euphonics.
In addition we have also to consider the Zout of the entire power circuit because with SE it is not so low (low dumping factor) and the response, when we have the speaker connected, at low frequency it is modified.
On my circuits, in SE, I halways use a single secondary fixed at 5 or 6 ohm and the tranformer ratio ( related with the type of tube) is relatively high, so we can get a reasonable dumping factor; I lost just little power but is not important.
Walter
There are circuits where from the first stage are designed poor.
So we can have the best OT but if the swing at the primary of the trafo is not very good.............
In the midrange, where we can consider the OT ideal (this just to play) where the value of Z is high and the power tube can work in the linear region, we can see only the products of the IInd harmonic ; this is the mainly reason about euphonics.
In addition we have also to consider the Zout of the entire power circuit because with SE it is not so low (low dumping factor) and the response, when we have the speaker connected, at low frequency it is modified.
On my circuits, in SE, I halways use a single secondary fixed at 5 or 6 ohm and the tranformer ratio ( related with the type of tube) is relatively high, so we can get a reasonable dumping factor; I lost just little power but is not important.
Walter
I'm not sure what low-end means in this context, but the signal processing is through a SET. I know an SET when I hear one, Nutube sounds like one, especially easy to tell in solo instrumental music (cello, violin, duets, trios, etc), and with <198kbps MP3s.
>It opamp buffers the output signal
That's also one way to impedance match the output (at low voltages) without trafos. By using a very high quality opamp, it is basically ensured that everything the Nututbe produces gets amplified with less than 0.1% distortion.
> I think any non push-pull thing with a tube in it is going to have SET distortion patterns.
I only measured push-pull, SET and the Nutube. SET and Nutube were very similar. The Nutube works as a signal processor, just as much as a 300B does.
Last edited:
The overall design / setting of the amplifier is most important. Of course OT needs to be quality but not crucial for good sound.
Example:
Last week we had HiFi-show here in my city. Some manufacturers have exhibited their hi-fi systems, and DIYeri's also theirs.
Two DIY systems have attracted special attention:
- Loftin-White 2A3 monoblocks (6SL7-Sylvania in to RCA 2A3) in combination with Fostex full-ranger in Lowther box.
- 6C33C monoblocks with separate PSU (half 6H6 in to half 6H6 in to 6C33C with fixed bias). In combination with Vienna-Acoustic "Mozart" speakers.
Output transformer in both amplifiers from the same manufacturer, winding on double C-cores.
In short, 6C33C monoblocks sound much much better
Example:
Last week we had HiFi-show here in my city. Some manufacturers have exhibited their hi-fi systems, and DIYeri's also theirs.
Two DIY systems have attracted special attention:
- Loftin-White 2A3 monoblocks (6SL7-Sylvania in to RCA 2A3) in combination with Fostex full-ranger in Lowther box.
- 6C33C monoblocks with separate PSU (half 6H6 in to half 6H6 in to 6C33C with fixed bias). In combination with Vienna-Acoustic "Mozart" speakers.
Output transformer in both amplifiers from the same manufacturer, winding on double C-cores.
In short, 6C33C monoblocks sound much much better
Attachments
The chances that this would properly work are rather slim.
No wonder... Would have been interesting to be able to swap these two speaker sets, and then compare again.
Of course, but I was not the controller
I just tried to say how important is the overall setting for good sound from the SET amp
Last edited:
Just a reminder to please use proper attribution when quoting.No I've never used an OPT, and yeah I always use regulated b+.
So far I've always used some form of power buffer circuit as an output stage.
I try to avoid interstage caps when I can but I've found that decent caps don't appear to have a sound if the next stage is easy to drive so it's probably not worth the extra effort.
So far I've always used some form of power buffer circuit as an output stage.
I try to avoid interstage caps when I can but I've found that decent caps don't appear to have a sound if the next stage is easy to drive so it's probably not worth the extra effort.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.