Nonono, just one 12AX7 tube, no resistors or capacitors. 😱 That would color the sound.
I actually knew a guy that always said "Resistors are bull. Why not just let the current flow?" I tried to explain basics like destructive current etc and it was still "bull."
Remember, these people know everything about electronics, but have never successfully designed anything.
Ya, that line preamp has a lot of visual symmetry in the schematic that does not translate to performance. Reminds me of the bad ideas I had in the 1970's. For example, R509 and R511 match R507 and R505 but are worse than useless, ie they create an unnecessary pole in the feedback path. The LTP current sources make almost zero difference compared to a resistor. 50 years later the same resources are better spent on a current mirror instead of R507 and R505. And the total symmetry, dual LTP, may reduce even harmonic THD but that is not a great idea. Not that it matters here, but its usually better to degenerate the LTP and not the VAS, for the sake of TIM. The VAS degeneration is related to the dual complementary LTP, ie the VAS current is dependent on the LTP current; not good. So, it's just old, lot of enthusiasm but a bit naive in 2020. To it's credit, the output followers are cross coupled, an important modern idea.
Last edited:
To use an op-amp in a power amp global feedback, you have to decide what you are going to do with the op-amp dominant pole. If it remains the dominant pole then the op-amp output has to slew as much as possible and the added stages have to be fast.
If it is not the dominant pole, then you want the op-amp output to slew as little as possible and you may want to add a zero to cancel it.
Then there is the output voltage saturation losses. Generally you will want some gain in the added power stage to get closer to the rails, and, push the amp above +/- 15VDC. A good idea is to boot-strap the op-amp rails, but, beware of the input common mode limitations, which, can be addressed with positive feedback. But this all gets complicated and leads to other stability issues like stray capacitance coupling. Note that many SPICE models for op-amps do not model the supply pins correctly.
Using an op-amp in the global feedback violates the idea of keeping your feedback loops short and fast, although there are plenty of discrete amps that are just as bad.
Using an op-amp as a DC servo or as a preamp is a better plan.
If it is not the dominant pole, then you want the op-amp output to slew as little as possible and you may want to add a zero to cancel it.
Then there is the output voltage saturation losses. Generally you will want some gain in the added power stage to get closer to the rails, and, push the amp above +/- 15VDC. A good idea is to boot-strap the op-amp rails, but, beware of the input common mode limitations, which, can be addressed with positive feedback. But this all gets complicated and leads to other stability issues like stray capacitance coupling. Note that many SPICE models for op-amps do not model the supply pins correctly.
Using an op-amp in the global feedback violates the idea of keeping your feedback loops short and fast, although there are plenty of discrete amps that are just as bad.
Using an op-amp as a DC servo or as a preamp is a better plan.
Hi ! thanks for the important advice. Actually i have seen some preamps from the late 70s showing this kind of symmetry. Simmetry is a characteristic of beauty So i was thinking maybe nice look provides nice sound ... 😱
Thank you very much for your deep technical analysis. I only understand that is not a design masterpiece. Still the Bryston pre very similar to this one has very low distortion. Maybe PSRR is not that great because from what i have gathered the quality of the psu has an impact on the sound.
For now i am only interested in line stages.
But i understand better that the idea of improving some commercial units is silly. Clearly they have already optimized the unit.
They spend hours and hours doing that.
Then the design fix the max obtainable from a unit. There is no way to fix a wrong design. I was curious about this Rotel preamp because it was considered almost on par with the TOTL preamps from Rotel in the same years, the Michi line. I am pretty sure that also the Michi preamp has a very similar design. I saw the schematic
Thanks again, gino
Last edited:
I've never had such a powerfull amplifier, 20W was the maximum and it sounded awful.
Replacing current sources (eg from current mirror configurations) by a simple resistor (or RCR network) reduces possible noise sources and other stablility issues. Tube amplifiers are 'elementary' by nature, and it cannot be a surprise if solid state have its own balance in complexity. But the challance is to find a suitable solution given the set of parameters which determine the needs, goal and result. What, which and how many parameter-ingredients are in the soup depends on the preferences of the cook.
Also, on this niche-market audio platform another opinion is in fashion compared with the pro-audio business. Both technical and commercial optimisations are possible and allowed.
Thanks for the advice about resistors They simplify the schematic.
To be honest without any technical basis i was thinking that this was a really great design. Then i have been explained that is not that great (what wer they doing at Rotel ?).
Also because when a design is right i guess it can be scaled-up to get more power if needed ?
Just to explain i am attaching another schema i like very much for its simplicity .... aura va-50
A friend of mine whose ears i trust more than mine ... told that it sound really good. Only issue weak in current delivery. Then why not replaced the output devices with something Gamut audio like ? big and chunky power mosfets ?
What puzzles me is why taking the hard way of complexity when an optimized simplicity can provide satisfaction ? i do not see the reason
Unless the more complex design superiority can be demonstrated
i understand. For me the look of a unit is not that important. For instance i like to listen in the dark possibly not distracted by any lights. Often i read darker background ... Bruce was dancing in the dark ... i like to listen in the dark ... like at a show i mean
But again i will try in the next years to start some simulation on working designs and see the effects of changing parts.
A very nice think would be if manufecturers provide models for their active devices. I need more time. I tried to simulate something in the past ... a Sziklai pair i am also attaching It provided interesting results
Attachments
Last edited:
Horses**t... My monoblocs make 110W RMS @ 30Hz. The only tube amp that has poor bass is one that was designed poorly.
Hi i see. I should have said i would never use tubes for bass below 100Hz.
I am convinced about solid state on bass since i heard a Krell ksa 100 power amp many years ago. That was a really great bass. Wall shaking with the right speakers of course. I would not ask anything more.
Ah I see what you mean. Most people use a class D amp for subwoofers now. I prefer the good old stereo approach myself 🙂
If you think that Krell is impressive, check this out 🙂 MCLAREN VAN FULL OF SPEAKERS AND SUBS AT TDH SHOW 2015 - YouTube
One time at the Canadian National Exhibition, there was a "speaker on wheels" made out of a Dodge Caravan. The body had been reinforced, not the doors though which flexed in and out like the speakers a good 2 inches without the windows!
Then there was the bass at Guvernment Night Club in Toronto before they knocked it down to build condos. So loud I have permanent hearing loss in my right ear. That's when I learned about attenuating hi-fi ear plugs...
I never got to experience the 96 KILOWATT Valve sound system though: Valve Sound System
If you think that Krell is impressive, check this out 🙂 MCLAREN VAN FULL OF SPEAKERS AND SUBS AT TDH SHOW 2015 - YouTube
One time at the Canadian National Exhibition, there was a "speaker on wheels" made out of a Dodge Caravan. The body had been reinforced, not the doors though which flexed in and out like the speakers a good 2 inches without the windows!
Then there was the bass at Guvernment Night Club in Toronto before they knocked it down to build condos. So loud I have permanent hearing loss in my right ear. That's when I learned about attenuating hi-fi ear plugs...
I never got to experience the 96 KILOWATT Valve sound system though: Valve Sound System
Last edited:
If you really want to learn and understand, there are science and technology. And, may I add, tertiary teaching institutions.
Using time listening and exploring opinions, is gossip and a waste of time, as I am doing right now. So, first and foremost, this applies to ME.
Last edited:
Hi ! now i see what You mean as well 😀
Disclaimer ... i have not been exposed to best tube amps out there. So i was really generalizing in a wrong way. Please excuse me.
But i liked that bass from that system ... i think the power amp was a ksa100 ...
I would more willing to use tubes in the voltage gain stages, upper in the chain
Anyway i have understood that the Zen approach to audio design is not that popular. There are cases ... but the majority of audio designers prefer more complex designs. I just do not understand what a more complex design can provide with respect to a well optimized more minimalist design.
The more amazing thing for me is that some experts can clearly spot design faults just by looking at a schematic and nevertheless these products perform quite well at the lab bench 😕
I do not know if the scale-up process can be used also in audio design.
I mean to design a great discrete op-amp schematic good, with change for a line stage and even an integrated by changing Vsupply, output pair and so on.
For instance ... taking the schematic attached moving the output pair upstream and place instead a pair of Sanken and increase the gain could provide a nice minimalist integrated ? if you see what i mean
To conclude ... tube amps are much more minimalistic in their design and do sound very good.
Can bjt amps be designed with the same approach ? if the answer is yes ... then why they do not do it ?
Attachments
Last edited:
Hi ! please try to understand my pain. 🙁 There is a design. It has very good lab measurements in terms of noise and distortion (can we call this scientific evidence?) And then i am told that this design is fundamentally wrong. 😕
How can a wrong design have very good measurements ?
i am quite willing to concede the contrary ... that designs with so so measurements can sound pleasant, enjoyable.
with the monkey being me of course. 😱
Better ... how can we say that a design is wrong or right ? what you designers look for during simulation ? i guess distortion basically
So if a design is let's say stable in normal working conditions and provide very low distortion how can it be wrong ?
In the review of which i have attached an extract they say like ... the design is very basic, even trivial ... and still provides levels of distortion almost not measurable. Well i would love to get a full basket of that basicality, triviality ...
what designers are looking for really ?
Last edited:
I prefer to keep things simple when I can. This is mainly why I build with tubes in unconventional ways. Also because they don't need thermal management.
The main difference is cost. 100W RMS from a tube amp costs 1000$+. 100W RMS from a class D eBay special is only like 25$ 🙂
The main difference is cost. 100W RMS from a tube amp costs 1000$+. 100W RMS from a class D eBay special is only like 25$ 🙂
I see and respect your design approach of course.
my problem is that i'm trying to say things i don't know in a language that isn't mine. As I said above, can a schematic/design be condemned even if the unit built according to the schematic performs very well at the measuring bench?
Better ... how a designer knows when the project is good ? by listening to it ? like i think they do for musical instruments ?
well i guess also the sound must be quite different. I mean it is not only a cost issue clearly.
The fact that class D is more popular for subwoofers make me think a little. Maybe is not that good for the rest of the audio spectrum ? i do not know.
Last edited:
To each his own, everyone learns differently.......In the end your way is probably less painful, but I’ve always been the masochistic type! 😀
ginetto, I fully sympathize with you......there are many things that are supposed to be ‘right’ but yet so wrong!
Last edited:
Hi you are very nice. Thank you ! Still i would like to get an idea of what is wrong and what is right. For instance ... if we agree that the measurements are dependable let's focus those. To start with saying ... this is a **** design with world class lab results ... what is this ? the hideous beauty concept ?
Some stakes must be fixed 🙄
if it sounds wrong but yet measures right, in the end it’s still wrong......why? Is a question best answered by yourself because that’s who matters.
In other words being told it’s right because it meets certain standards is not the end all.
In other words being told it’s right because it meets certain standards is not the end all.
Last edited: