The driver transistors sit in a emitter follower configuration (ie. no amplification) and so it could be argued that they thus have less influence on audio quality than the previous stages. And indeed, in my experience you really have difficulties in hearing differences in A/B tests between good quality drivers.
But as 'audio quality' is inevitably subjective, the best advice is probably to simply try swapping sets of candidates - and have a good listen to each.
Per
PS.
The only reason that I can see that Rotel used the 600/631 as VAS transistors is because they set the VAS current to about 14mA, which with ~45V across gives over 0.6W heat generation in each. You could reduce the VAS current to say, 6mA by decreasing the values of R607-10 and R615-18.
But as 'audio quality' is inevitably subjective, the best advice is probably to simply try swapping sets of candidates - and have a good listen to each.
Per
PS.
The only reason that I can see that Rotel used the 600/631 as VAS transistors is because they set the VAS current to about 14mA, which with ~45V across gives over 0.6W heat generation in each. You could reduce the VAS current to say, 6mA by decreasing the values of R607-10 and R615-18.
Last edited:
Hi it is me again I have found some graphs of THD+noise in the thread but about the ra971 model
Looking at the various service manuals it is possible to see how the topology has been maintaned somewhat similar at least for the integrated amps
There must be a reason
They have been very conservative considering that the first amp goes back to 50 years ago ?
I do not want to sound trivial but also noise should be addressed For instance i have seen from the lab report some power supply noise clearly passing through
I wonder what an upgrade of the power supply could bring to the table My feeling is that the capacitance in the power supply is too small
uFs re never enough ...
Almost always an increase of uF in the primary power supply caps quality and quantity brings the noise down in a measurable way
The evident peaks before the 1kHz test signal are clearly coming from the power supply and are not suppressed by the amp circuit (low PSRR?)
Imho is impressive what a very low noise amp can do to the musical experience
When the power is not very high i would even consider regulated voltage rails I know this is an added complexity but noise is a really bad beast
The less the noise the better
Could you please direct me to any post containing lab measurements ?
I'm a bit like Saint Thomas. I have to see to believe. And then when I've seen I become a fervent believer
I also propose to make a link in the original post to those pages for those skeptics
I am sure that showing the improvements will bring more people to this very excellent thread
Thank you sincerely and have a nice day
gino
Looking at the various service manuals it is possible to see how the topology has been maintaned somewhat similar at least for the integrated amps
There must be a reason
They have been very conservative considering that the first amp goes back to 50 years ago ?
I do not want to sound trivial but also noise should be addressed For instance i have seen from the lab report some power supply noise clearly passing through
I wonder what an upgrade of the power supply could bring to the table My feeling is that the capacitance in the power supply is too small
uFs re never enough ...
Almost always an increase of uF in the primary power supply caps quality and quantity brings the noise down in a measurable way
The evident peaks before the 1kHz test signal are clearly coming from the power supply and are not suppressed by the amp circuit (low PSRR?)
Imho is impressive what a very low noise amp can do to the musical experience
When the power is not very high i would even consider regulated voltage rails I know this is an added complexity but noise is a really bad beast
The less the noise the better
Could you please direct me to any post containing lab measurements ?
I'm a bit like Saint Thomas. I have to see to believe. And then when I've seen I become a fervent believer
I also propose to make a link in the original post to those pages for those skeptics
I am sure that showing the improvements will bring more people to this very excellent thread
Thank you sincerely and have a nice day
gino
Last edited:
Hi Gino,
It is all too easy to get concerned (read: panic) about things and start changing stuff - when the problem actually lies somewhere else.
I have more than once torn my hair out (and I don't have that much left) over some inexplicable measurement signal (or lack thereof) - only to find a broken test lead, wrong instrument setting or noisy nearby SMPS device to be the problem.
And I, of all people should really know better - my very first commercial project as a newly minted R&D engineer involved an expensive super low noise metal can opamp - which I couldn't get to do anything. Nada. Whatever I tried and checked. Oh no
- had I somehow stupidly already blown this $$$ opamp?
So this young engineer of course started doubting his entire education and was ready to hand back the diploma - when as a last resort I tried to pop in a discarded old uA741 that someone had left on the bench. And it worked!
I called the supplier who admitted that B&O had just returned a shipment of the expensive opamp. They had simply cut the can open and found that the manufacturer had 'forgotten' the chip-to-pin 6 bonding wire.
Good advice: Have a regular quality control (double) check of the equipment on the bench before starting a project.
(I probably should start to listen to that.)
Anyway, back to your questions.
1. Yes, I have tried to measure THD+N, my problem is that my Xonar7 measurement setup seems to be super sensitive to hum fields from any mains connection.
I think that I covered that in earlier posts on this thread, but I get 50Hz RTA peaks even when both the amp and the mains wall outlet switch are off. (!)
So neither the power supply or other circuits in the amp are causing this RTA signal.
Pull the plug out of the wall and it goes away, but there is of course not much music coming out of the amp.
You would think that my Fluke or Keithley DMM could measure this 50Hz ac RTA signal, but there is nothing. And definitely no audible hum.
So I have pragmatically decided to live with the RTA trace hum field artifacts for the time being.
Yes, to see is to believe - but you shouldn't believe all you see.
2. The late Bob Peace once said in one of his "What's All This ... Stuff, Anyhow?" article series:
There are lies, damned lies ... and Spice. Also: My favorite programming language is ... solder.
Ok, I think that Spice is a valuable tool for initial evaluation of circuit ideas and cabirio did quite some excellent Spice work on the 820AX circuit upgrades (see around pg.10).
Hope this waffling long reply answers your questions.
Best,
Per
It is all too easy to get concerned (read: panic) about things and start changing stuff - when the problem actually lies somewhere else.
I have more than once torn my hair out (and I don't have that much left) over some inexplicable measurement signal (or lack thereof) - only to find a broken test lead, wrong instrument setting or noisy nearby SMPS device to be the problem.
And I, of all people should really know better - my very first commercial project as a newly minted R&D engineer involved an expensive super low noise metal can opamp - which I couldn't get to do anything. Nada. Whatever I tried and checked. Oh no
- had I somehow stupidly already blown this $$$ opamp?So this young engineer of course started doubting his entire education and was ready to hand back the diploma - when as a last resort I tried to pop in a discarded old uA741 that someone had left on the bench. And it worked!
I called the supplier who admitted that B&O had just returned a shipment of the expensive opamp. They had simply cut the can open and found that the manufacturer had 'forgotten' the chip-to-pin 6 bonding wire.
Good advice: Have a regular quality control (double) check of the equipment on the bench before starting a project.
(I probably should start to listen to that.)
Anyway, back to your questions.
1. Yes, I have tried to measure THD+N, my problem is that my Xonar7 measurement setup seems to be super sensitive to hum fields from any mains connection.
I think that I covered that in earlier posts on this thread, but I get 50Hz RTA peaks even when both the amp and the mains wall outlet switch are off. (!)
So neither the power supply or other circuits in the amp are causing this RTA signal.
Pull the plug out of the wall and it goes away, but there is of course not much music coming out of the amp.
You would think that my Fluke or Keithley DMM could measure this 50Hz ac RTA signal, but there is nothing. And definitely no audible hum.
So I have pragmatically decided to live with the RTA trace hum field artifacts for the time being.
Yes, to see is to believe - but you shouldn't believe all you see.
2. The late Bob Peace once said in one of his "What's All This ... Stuff, Anyhow?" article series:
There are lies, damned lies ... and Spice. Also: My favorite programming language is ... solder.
Ok, I think that Spice is a valuable tool for initial evaluation of circuit ideas and cabirio did quite some excellent Spice work on the 820AX circuit upgrades (see around pg.10).
Hope this waffling long reply answers your questions.
Best,
Per
Hi ! first of all sincere congratulations for your project and this thread. It is very educational for me that i am just a poorly skilled beginner
I have been impressed by your story It is quite surprising to realize that a manufacturer of such advanced device has skipped on QC
I had one and it seemed to work
https://www.ebay.it/itm/322543092641
Or you can do it with a little od DIY You just need a usb 2.0 type -b extension cord Open it in the middle and leave only the data wires connected and use the power wires with an external +5VDC low noise power supply Depending on the sound card consumption maybe even a battery could work providing a complete separation from the pc power supply noise
I am using a Behringer U-Phoria UMC202HD and it seems ok But i am focusing mostly line buffer and low voltages Up to 2-3Vrms the level that can bring a power amp to max power nothing more
Recently i have started some use of LTSpice and i am very fascinated Sometimes trying to change values i get results that look even ridiculous
But the blame is entirely on me of course
Your answers have been completely satisfactory and very kind.
Thank you very much again and kind regards,
gino
I have been impressed by your story It is quite surprising to realize that a manufacturer of such advanced device has skipped on QC
I see the issue You could try one of those Y usb cable that allows for the use of an external power supply instead of the pc power1. Yes, I have tried to measure THD+N, my problem is that my Xonar7 measurement setup seems to be super sensitive to hum fields from any mains connection.
I think that I covered that in earlier posts on this thread, but I get 50Hz RTA peaks even when both the amp and the mains wall outlet switch are off. (!)
So neither the power supply or other circuits in the amp are causing this RTA signal.
Pull the plug out of the wall and it goes away, but there is of course not much music coming out of the amp.
I had one and it seemed to work
https://www.ebay.it/itm/322543092641
Or you can do it with a little od DIY You just need a usb 2.0 type -b extension cord Open it in the middle and leave only the data wires connected and use the power wires with an external +5VDC low noise power supply Depending on the sound card consumption maybe even a battery could work providing a complete separation from the pc power supply noise
I am using a Behringer U-Phoria UMC202HD and it seems ok But i am focusing mostly line buffer and low voltages Up to 2-3Vrms the level that can bring a power amp to max power nothing more
yes but speaking of design the calculation step is not avoidable I have no skills in electronic design but i think that simulation SWs are very convenient
Recently i have started some use of LTSpice and i am very fascinated Sometimes trying to change values i get results that look even ridiculous
But the blame is entirely on me of course
i agree absolutely and thank you for the suggestion I will look for those sims immediately
Your answers have been completely satisfactory and very kind.
Thank you very much again and kind regards,
gino
I am quite sure that the 50Hz does not come through the usb. I put in an usb isolator, tried to run it from a laptop on battery power, etc.
And as I said, the hum peak disappears when I pull the mains plug from the wall.
It is most probably an electromagnetic field pickup from the amp mains side. As suggested by sgrossklass, the remedy could be to put an instrumentation or differential opamp before the Xonar7, making its input balanced.
Done right, eg. INA217 can give an extraordinary high CMRR that will effectively cancel any common mode signals on the inputs. I have actually made a protoboard for this, but haven't had the time to try it out yet. I will get round to it someday....
Chiao,
Per
And as I said, the hum peak disappears when I pull the mains plug from the wall.
It is most probably an electromagnetic field pickup from the amp mains side. As suggested by sgrossklass, the remedy could be to put an instrumentation or differential opamp before the Xonar7, making its input balanced.
Done right, eg. INA217 can give an extraordinary high CMRR that will effectively cancel any common mode signals on the inputs. I have actually made a protoboard for this, but haven't had the time to try it out yet. I will get round to it someday....
Chiao,
Per