Hi Chris,
the Cyrus Mono X looks very interesting.
I have not found any measurements and the declared distortion is not good as the acclaimed class D amp in the land of the Lord of measurements.
I assume you trust the design philosophy and the people behind the project.
"excellent amplifier design, and through incremental listening tests" IMHO is the right way to achieve the best results.
While of course I don't ask you to disclosure the schematic I would have a few questions about the design.
They call it ‘zero-feedback’ circuit topology, I assume there is no global feedback but maybe local.
Just to clarify I don't consider an emitter resistor (or cathode resistor) as feedback, for me negative feedback means something taken from the output and returned to the input.
Is there any feedback correction in the various stages?
Another interesting point is how they have got the voltage gain, it looks like a sort of active step up transformer.
Can you tell more about their approach?
Can you tell me more about the current amplifier?
I assume it's a class AB amp, EF CFP or what else?
Andrea
Not claiming that. It is already well known that IMD tends to be more easily audible than HD.
Not claiming that. Do claim it is possible to identify very small changes in low level distortion in A/B tests, allowing both that there has been sufficient listener practice, and that the system does not otherwise excessively mask the sound of the changes.
That was the experiment; the only claim is the specific result of the experiment.
Andrea,
You talked about 98db loudspeakers previously. Is it from horn, direct radiators or full rangers?
I ask because i wonder (if you have a fullranger) if you implemented the method described in the De Lima's article Indra gave a link to?
How would you characterise the effect of a cathode resistor if not feedback then? ( i'm more confortable with tube than transistors if you need a technical description in answer).
You talked about 98db loudspeakers previously. Is it from horn, direct radiators or full rangers?
I ask because i wonder (if you have a fullranger) if you implemented the method described in the De Lima's article Indra gave a link to?
How would you characterise the effect of a cathode resistor if not feedback then? ( i'm more confortable with tube than transistors if you need a technical description in answer).
IMD and HD are not audible because they are measurement numbers, not sounds or signals.
What you hear is a nonlinearity, and the exact same linearity will produce IMD or HD if you measure it with a single frequency signal or multiple frequency signal.
So, perhaps what you mean is that a nonlinearity that produces XX% IMD sounds worse than a nonlinearity that produces the same % THD?
Or that a complex signal with lots of frequency (orchestra, massed strings, etc) will sound worse than a simple signal on a circuit with the same nonlinearity?
You may have experienced what Jakob2 talked about in this message:
https://www.diyaudio.com/forums/the...quality-vs-measurements-1924.html#post6658929
There is a pro studio german company called SPL which use tube ( and passive LC eq on specific range) for some of their 'enhancers' and eq and on some settings it can give this kind of feelings of displacement along the elevation plane ( at least to me).
I wonder if the 'holographic' effect ( it is sometimes mentioned about some tube microphones -so mono!- by audio engineers in pro circles too) doesn't come from some kind of tubes artefacts ( microphonic, materials used,...) which enhance some specific range and harmonic contents?
More or less. As mentioned before, vocal harmonies are an example of where it can be heard. Human brains appear to be quite attuned to subtle changes in vocal sounds. More than two voices at once may be better.
Last edited:
^ this vocal intelligibility range ( 300hz -or 150hz depends on who talk about it- to 6khz) is the one we are the most sensible.
Evolution outcome ( we have a peak circa 3/4,5khz because of specie survival: baby crying in this range).
And we are relatively good at lateral localization not so much for vertical ( less chances to have predators coming from above).
This happen to be the range of interest if you have to focus on one only and you want sound quality.
Evolution outcome ( we have a peak circa 3/4,5khz because of specie survival: baby crying in this range).
And we are relatively good at lateral localization not so much for vertical ( less chances to have predators coming from above).
This happen to be the range of interest if you have to focus on one only and you want sound quality.
Regarding Cyrus Mono X, the way they describe getting voltage gain is basically the same method used in Pass Labs HPA-1 (by way of mirroring).
Mono X – Cyrus Audio
Pass HPA-1, what do we know?
Mono X – Cyrus Audio
Pass HPA-1, what do we know?
Last edited:
Hi Andrea,
Cyrus does not release service information and I have given my word to them I will not either. The design team at Cyrus are pretty sharp and use good, solid design principles. They also maintain an up to date lab, and a good listening area where the principles also have to listen and approve designs.
I was over there touring the factory as the Canadian Service Manager for the distributor at the time. I was developing the same type of amplifier at then and was able to understand and ask questions on the uncompleted design. The lead engineer recognized I was intimately familiar with the design and we struck up a long discussion about it. We finished up and I simply stated that I want a pair when they were complete. I have the very first pair that hit Canada.
The design is basically various configurations of a diamond buffer. There is no overall feedback, or any feedback between stages at all. Like a CFP output stage, no transformers either.
I see I haven't tested mine yet. I'll have to do that. They have the British speaker connectors and I will be replacing them with normal banana jacks at some point. Once that is done, I will test them. Like I said, the first in Canada, production models for Canada used Banana connectors.
They sound great, and are very reliable as I have used them for years, often turn up on my 4 ohm speakers.
-Chris
Cyrus does not release service information and I have given my word to them I will not either. The design team at Cyrus are pretty sharp and use good, solid design principles. They also maintain an up to date lab, and a good listening area where the principles also have to listen and approve designs.
I was over there touring the factory as the Canadian Service Manager for the distributor at the time. I was developing the same type of amplifier at then and was able to understand and ask questions on the uncompleted design. The lead engineer recognized I was intimately familiar with the design and we struck up a long discussion about it. We finished up and I simply stated that I want a pair when they were complete. I have the very first pair that hit Canada.
The design is basically various configurations of a diamond buffer. There is no overall feedback, or any feedback between stages at all. Like a CFP output stage, no transformers either.
I see I haven't tested mine yet. I'll have to do that. They have the British speaker connectors and I will be replacing them with normal banana jacks at some point. Once that is done, I will test them. Like I said, the first in Canada, production models for Canada used Banana connectors.
They sound great, and are very reliable as I have used them for years, often turn up on my 4 ohm speakers.
-Chris
N101N,
Well yes ( we agree on benefits) except that in my understanding once you take a stepback and introduce the psu in the schem it is part of a feedback loop if your psu is unregulated.
I asked because if it is not feedback what about miller effect which is feedback too or the heater which is too part of a feedback loop - if unregulated(i'm not into transitors i talk about valves).
I found tube fun about that: you've got many loops around the component. Transistors are more sneaky to me: less interaction but some parameters are modulated ( dynamic behavior of some parameters) which i've got difficulties to deal with ( i'm somewhat limited to understand some things).
Well yes ( we agree on benefits) except that in my understanding once you take a stepback and introduce the psu in the schem it is part of a feedback loop if your psu is unregulated.
I asked because if it is not feedback what about miller effect which is feedback too or the heater which is too part of a feedback loop - if unregulated(i'm not into transitors i talk about valves).
I found tube fun about that: you've got many loops around the component. Transistors are more sneaky to me: less interaction but some parameters are modulated ( dynamic behavior of some parameters) which i've got difficulties to deal with ( i'm somewhat limited to understand some things).
Last edited:
Now I have only got the drivers.
The speakers project is still far.
It will take at least a year (optimistically speaking) to complete the other projects like the digital front end, the DAC and the amplifier.
We wouldn't care about the digital source because the FIFO should perfectly isolates it, since the source and the DAC will operate in totally different time domain.
It will be a 3 way OB with a fullrange (Supravox 165 GMF) refined by the ESS Great Heil and a pair of AE LO 15.
The crossover will be as simpler as possible.
Nothing for the AE drivers, I will implement a mechanical low pass filter.
Then a simple capacitor for the Supravox and the Heil drivers.
XOVER frequencies will be around 120-160 Hz and 7-8kHz.
To understand if a mechanical low pass filter will also be necessary to cure the break up of the Supravox.
This will be the starting point.
I don't have a definition for the emitter or cathode resistor, but since there is not a loop from the output to the input I would not consider it as a feedback.
This too is an eternal question with different replies.
Technically speaking, an emitter resistor is a form of local negative feedback. However, it may not sound the same to some people as less localized forms of feedback, so they may prefer to classify it on that basis. A connection does exist between the output and an input, since an emitter can function as both. This is true even though the connection might be considered an infinitely small loop 
Ever wonder why a tunnel diode can act as an amplifier or oscillator given that it only has two terminals?
Ever wonder why a tunnel diode can act as an amplifier or oscillator given that it only has two terminals?
Thank you for answers N101N and Andrea.
N101N, well you may be right, i'm self taught in electronic and i confess i have 'holes' in my knowledge but i was still able to work on different audio processors with relative success: microphones, preamps, compressors ( this one was very interesting and not easy), passive eq with tube makeup gain stage...
From my experience with tube microphone circuits you can't take the psu out of the equation about feedback loops. Those are deceptively simple circuits (if you take out the fringe to use very high input impedance source with tubes which shouldn't accept this in theory, undervoltage of heater psu, etc,etc,...) very touchy about any change you make.
And from my experience with Neumann U47 tube inspired circuit i can tell you there is feedback at play ( once you use a regulated psu with a shunt regulator and ccs the circuit sound very different. Same for the heater supply - but i've not tried shunt psu there because of noise issues ( fixed bias derived from main ht supply in the original circuit) as well as microphonics involved in the sound of the original mic ( i was lucky to have access to a well maintened pair at work for comparison).
So maybe i'm wrong but in practice this is not what i observed. I'm not a theorist though so no offense taken if i need to be corrected ( about my conclusion).
But from someone denying ohms law, i will wait other pov. ( I'm teasing you but i'm open to discussion ).
I don't get the increasing with freq though: you talk about real world part behavior (inductance of non perfect components)?
Andrea , nice project.
Supravox can sound great. Not my kind of rendering but it sound nice.
Yes i know there is different pov about it hence why i ask: i'm always open to new way ( different than the one i have i should say) to see things. As long you can discuss the point and not go too much into esoterical things that is food for thoughts ( and i'm hungry!).
About global fb. Well i understand the academic concerns but i must confess i've heard enough circuits implementing them to know that in my own case i don't care about the academic concerns most audiophile have with the principle. Tbh one of my most beloved mic ref ( Neumann U67) use it to great effect and almost all tube compressors are done around this principle. And for me they sound wonderfull.
That doesn't stop me to do the other way around ( my tube preamp is global feedback free, but i use heavily local feedback on it ( except for an srpp stage i don't use bypass capacitor on cathode- but i can disable it if i want).
I've got no issues with cathode followers too, white cf and slcf ( i love this circuit!).
Thank you Myleftear. Will take a look at it.
N101N, well you may be right, i'm self taught in electronic and i confess i have 'holes' in my knowledge but i was still able to work on different audio processors with relative success: microphones, preamps, compressors ( this one was very interesting and not easy), passive eq with tube makeup gain stage...
From my experience with tube microphone circuits you can't take the psu out of the equation about feedback loops. Those are deceptively simple circuits (if you take out the fringe to use very high input impedance source with tubes which shouldn't accept this in theory, undervoltage of heater psu, etc,etc,...) very touchy about any change you make.
And from my experience with Neumann U47 tube inspired circuit i can tell you there is feedback at play ( once you use a regulated psu with a shunt regulator and ccs the circuit sound very different. Same for the heater supply - but i've not tried shunt psu there because of noise issues ( fixed bias derived from main ht supply in the original circuit) as well as microphonics involved in the sound of the original mic ( i was lucky to have access to a well maintened pair at work for comparison).
So maybe i'm wrong but in practice this is not what i observed. I'm not a theorist though so no offense taken if i need to be corrected ( about my conclusion).
But from someone denying ohms law, i will wait other pov. ( I'm teasing you but i'm open to discussion
). I don't get the increasing with freq though: you talk about real world part behavior (inductance of non perfect components)?
Andrea , nice project.
Supravox can sound great. Not my kind of rendering but it sound nice.
Yes i know there is different pov about it hence why i ask: i'm always open to new way ( different than the one i have i should say) to see things. As long you can discuss the point and not go too much into esoterical things that is food for thoughts ( and i'm hungry!).
About global fb. Well i understand the academic concerns but i must confess i've heard enough circuits implementing them to know that in my own case i don't care about the academic concerns most audiophile have with the principle. Tbh one of my most beloved mic ref ( Neumann U67) use it to great effect and almost all tube compressors are done around this principle. And for me they sound wonderfull.
That doesn't stop me to do the other way around ( my tube preamp is global feedback free, but i use heavily local feedback on it ( except for an srpp stage i don't use bypass capacitor on cathode- but i can disable it if i want).
I've got no issues with cathode followers too, white cf and slcf ( i love this circuit!).
Thank you Myleftear. Will take a look at it.
Last edited:
- Status
- Not open for further replies.