No CCS?
Regards
M. Gregg
No, no CCS. If I did one it would be the SY DN2540 design, but for the space, component count and fiddliness I prefer just to throw another tube in and have an SRPP instead.
Would you say you are a purist? <<genuine question!
Is this just a preference or would you use SS if it out performed a tube in a particular instance?
Or redesign to avoid SS?
Regards
M. Gregg
Would you say you are a purist? <<genuine question!
Is this just a preference or would you use SS if it out performed a tube in a particular instance?
Or redesign to avoid SS?
Regards
M. Gregg
Not sure what a purist is TBH
I find I can get the sound I like with just tubes, maybe when I have more time I'll audition the CCS and see if it wins over the resistor or the SRPP etc, but for now it's too much like hard work!
I built many SS amps starting at my days at National Semi. class AB and pure class A.. Thought that Krell was the best amp made until I heard a little room warmer tube amp in TRIODE mode.....
It had the combination of tubes and SS CCS for the LPT and SS for zeners and power supply, and power supply delay ramp-up for B+..
The little $500 tube amp kicked the $2,500 Krells sonic butt. I'm more of a pragmatist than a purist and see a practical balance of both in the same amp.....
It had the combination of tubes and SS CCS for the LPT and SS for zeners and power supply, and power supply delay ramp-up for B+..
The little $500 tube amp kicked the $2,500 Krells sonic butt. I'm more of a pragmatist than a purist and see a practical balance of both in the same amp.....
I design circuits in fancy CMOS and BiCMOS processes all day at work. I'm having issues with devices blowing up at 3.3 V because of the fancy processes not allowing much voltage across the dielectrics. But I find that given enough time, I can design just about any circuit using sand technology.
At home, I'm working on the opposite end of the spectrum. Tube amps. It started one day when I was at work waiting for a simulation run to finish. I decided to look into vacuum tubes for amplification. It's become a bit of an escape or outlet if you will... I've gotten into the directly heated triode world of 300B tubes. I really like the open sound stage I get with my SET amps. It's miles beyond what I have been able to achieve with solid state designs.
Technically, though, the tube amp measures worse than the sand amp. Getting 0.001 % THD+N with sand is not exactly trivial, but it's not exactly hard either if one knows what to look for. My LME49811-based amp delivers that performance. Yet, I prefer the sound of my 6LU8 Spud - and especially 300B SET designs that have THD+N around 0.1~1 % at typical listening levels.
I do use a lot of sand in my tube amps, though. LED bias. CCS bias. Switchmode supplies for the filaments, etc. There's a place for everything... It doesn't have to be either or. It can be both.
~Tom
At home, I'm working on the opposite end of the spectrum. Tube amps. It started one day when I was at work waiting for a simulation run to finish. I decided to look into vacuum tubes for amplification. It's become a bit of an escape or outlet if you will... I've gotten into the directly heated triode world of 300B tubes. I really like the open sound stage I get with my SET amps. It's miles beyond what I have been able to achieve with solid state designs.
Technically, though, the tube amp measures worse than the sand amp. Getting 0.001 % THD+N with sand is not exactly trivial, but it's not exactly hard either if one knows what to look for. My LME49811-based amp delivers that performance. Yet, I prefer the sound of my 6LU8 Spud - and especially 300B SET designs that have THD+N around 0.1~1 % at typical listening levels.
I do use a lot of sand in my tube amps, though. LED bias. CCS bias. Switchmode supplies for the filaments, etc. There's a place for everything... It doesn't have to be either or. It can be both.
~Tom
I disagree.
Dominance of even or odd harmonics has to do with the circuit design, not tube versus transistor. Below are measurements of my amps. 2 are tube amps and 1 is transistor amp. Can you guess which (white, blue or red) is which?An externally hosted image should be here but it was not working when we last tested it.
Surely FETs, BJTs, MOSFETs, Triodes and Pentodes all have differing amount of various harmonics?
The white one has >1% second harmonics, so that most likely is a tube amp. I know valve amps with these kinds of distortion numbers that still sound OK, whereas I have never encountered such phenomenon with SS. With the other two, it would be mere guessing. Which I suppose is exactly what these pictures are meant to illustrate.
Tinitus makes an interesting point about high frequency detail; I was once comparing an expensive SS mic pre with one of my diy tube preamps at a recording session. The sound engineer,who is a friend of mine preferred the tube amp for what he described as its "naturalness" but thought that there was additional HF detail with the SS amp. I,for my part,on this occasion concurred,but I was not able to convince myself that this additional HF was actually real musical detail.
If the difference is little, it's basically meaningless in terms of audible attributes such as between blue and yellow (I labeled it red by mistake) graphs I posted. There may be something else other than harmonic signature that may attribute to audible differences but that's not what I'm discussing.
I left out the circuit info for now just to show CBS240 that it's not the tube or transistor that determines the characteristics of harmonics. I'll reveal the answers later.
We will never settle this one completely I am afraid. Some think that tube amps sound different because of what they add. Other think that it is because of what they don't add or don't subtract. In the end, other than the technical plusses and minuses that are objectively measured there will be little agreement on the issue. So I says run whatcha like and enjoy.
Yes, they do. There are differences between types as well. Take a look at the specs for the 807: THD figures as low as 1.8% (Class AB1, PP)
Vpk= 365Vdc
Vsgsg= 270Vdc
Vgk= -22.5Vdc (fixed bias)
Po= 26.5W
Looks pretty good, doesn't it? These figures are indeed accuarte, and well in line with the open loop performance I measured, even with using common Hammond OPTs, and not the premium iron.
However, 807s sound positively gawdawful open loop. What these figures aren't telling you is that a lot of that THD is h5 and higher. This becomes quite obvious during the Twin-T test: the residual has a period of 3F1, but is very distorted.
This type needs extra help, and this was noted by the developer of the type: O. Schade. His "white paper" cum advert recommends local feedback of 0.1Vop. This recommend is also spot-on, and after connecting the local NFB, the sonics improved very much, and all that was left to do was add some 7.0db of gNFB to take off the residual "edginess".
OTOH, these TV HD pents that had no mention of any use as audio finals mentioned in the spec sheet (and they could never be used as SE audio finals as the most linear part of the characteristic came in at Ip= ~100mA: red plate territory, so you have no choice but to use them PP). When operated as audio finals, the THD comes in at ~3.0%. Should sound worse? Well, they didn't, as almost all of that THD was h3. Open loop, these tended to sound overly "aggressive", but that pentode nastiness was never heard unless you cranked them almost to the point of audible clipping.
All that design needed was some gNFB to improve damping and to take off the "edginess". I never finalize a design without spending at least a week listening while running open loop before deciding what corrections need be applied.
Same thing crops up with SS design. Just because something measures better, doesn't mean it sounds better.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.