Very interesting, thanks. You are of course aware that the low input impedance for MC has it's own staunch supporters.
Yes, a cap at the input is probably the best approach. Tell this to the purists camp
There's a placeholder for that on the V4 board.
Its because they throw away noise along with HF and they don't know it. So the cart fleshes out subjectively. Then they call it ''gray'' also. Many perceive the Denon 304 as thin for instance. Its because it is 0.18mV but 40R Rs. The phono stages need to cut en so they use 100RL to tame it, alas they lose 38% signal level. Back to square 1. Hence they sell it for silly prices on ebay bcs unpopular. Get one for the 4.0 phono of yours. Only such a phono is going to allow it play as it must. You gonna have a myth buster on the cheap.
Its because they throw away noise along with HF and they don't know it. So the cart fleshes out subjectively. Then they call it ''gray'' also. Many perceive the Denon 304 as thin for instance. Its because it is 0.18mV but 40R Rs. The phono stages need to cut en so they use 100RL to tame it, alas they lose 38% signal level. Back to square 1. Hence they sell it for silly prices on ebay bcs unpopular. Get one for the 4.0 phono of yours. Only such a phono is going to allow it play as it must. You gonna have a myth buster on the cheap.
At 40 ohm, you don't need V4 or V3.1 That's already 0.8nV/rtHz!
I am not aware of that cartridge, but it seems like it has a horrible calculated S/N: only 57dB!
Thanks, interesting approach, though I'm not sure it's worth a boatload of low noise MAT02/MAT03 monolithic pairs (at $25 a pop) and caps (in the signal path). Do you have a full spec for this pre?
You can load it down at 28.5R with 100 Ohm RL they use but keep the noise near its new en when having the benefit of using much gain without much penalty with the 4.0, that is my trick suggestion. Just because its an oddball as I said.
You can see some measurements i did on my "MPP" thread under "Audio Line Level"
I can se no caps in the signal path except the bypass caps for the current mirrors, the RIAA caps and the cap for the servo. The bypass caps improve 2nd harmonic by 10dB and 3rd by 4dB. They can be left out, then the noise is more like 0.7nVqHz.
I am still working on the optimisation of the circuit so parts values are only default.
Yes i will not use MAT02/03 in the finished version. It happened that i had them around and some people atribute magical qualities to them so i was curios. They are absolutely not neccesarry for good results.
I can se no caps in the signal path except the bypass caps for the current mirrors, the RIAA caps and the cap for the servo. The bypass caps improve 2nd harmonic by 10dB and 3rd by 4dB. They can be left out, then the noise is more like 0.7nVqHz.
I am still working on the optimisation of the circuit so parts values are only default.
Yes i will not use MAT02/03 in the finished version. It happened that i had them around and some people atribute magical qualities to them so i was curios. They are absolutely not neccesarry for good results.
Distortions, that's about what you can get open loop with a symmetrical complementary design with matched devices. I don't think there's room for any major improvements, but for some degrading when you'll give up the MAT monolithic pairs. You'll then get some 2nd harmonic as well
Those are the caps I was thinking of. It's ok, you can keep them; I was looking with a purist eye at your design
Glad you followed the advice and posted over here
Thank you very much for that i can keep the capacitors in.
I whould not call your designs purist either but that should not prevent them from sounding good. I can design for extremely low distortion and noise without capacitors in the signal path and do so in my commercial products. When you visit the MPP thread you can see some measurements of the Quantummusic Goldstandart Headamp.
The Goldstandart has only resistors and precision OP´s in the signal chain. There is even no servo and because it`s balanced it does not dump much current into the cardridge. It is DC coupled throughout.
I did the topology of the MPP out of other reasons i explain in the thread.
Actually performance is not that bad with 2nd at -90dB and 3rd at -100dB at 0dB/U
(1V) for the whole shebang including the following RIAA stages.
But you make me think. I should do a second version with voltage feedback and then compare the results. I also think the circuit has potential for improovement for example by bootstrapping the parallel cascode stage. Anyway what counts is how it sounds and this circuit sounds different (better ?) then the usual single ended, high impedance, high feedback designs. There is a certain liquidity in the mids and sweetness in the highs that is quite nice. Maybe it is " the sound of distortion" as Douglas Self whould say but i think distortion in the MPP is low enough to be not audible per se. Maybe it gives just a little spice to please the ear.
Anyway, if i find a way to improve distortion i will do so.
I whould not call your designs purist either but that should not prevent them from sounding good. I can design for extremely low distortion and noise without capacitors in the signal path and do so in my commercial products. When you visit the MPP thread you can see some measurements of the Quantummusic Goldstandart Headamp.
The Goldstandart has only resistors and precision OP´s in the signal chain. There is even no servo and because it`s balanced it does not dump much current into the cardridge. It is DC coupled throughout.
I did the topology of the MPP out of other reasons i explain in the thread.
Actually performance is not that bad with 2nd at -90dB and 3rd at -100dB at 0dB/U
(1V) for the whole shebang including the following RIAA stages.
But you make me think. I should do a second version with voltage feedback and then compare the results. I also think the circuit has potential for improovement for example by bootstrapping the parallel cascode stage. Anyway what counts is how it sounds and this circuit sounds different (better ?) then the usual single ended, high impedance, high feedback designs. There is a certain liquidity in the mids and sweetness in the highs that is quite nice. Maybe it is " the sound of distortion" as Douglas Self whould say but i think distortion in the MPP is low enough to be not audible per se. Maybe it gives just a little spice to please the ear.
Anyway, if i find a way to improve distortion i will do so.
Finally... Matched out of the tube, and smiling with their silvery faces
Will test ASAP and be back...
So: this low noise Rohm pair has about the same Rbb as Hitachi's 2SC2547/2SA1085. They also seem to have the same amount of 1/f and a little GR exces noise, corner frequency is also slightly under 100Hz for both.
From all perspectives that I can look after, Rohm's 2SD786/2SB737 and Hitachi's 2SC2547/2SA1085 are equivalent. Unfortunately, I have the ROHM pairs in the low (Q) beta class only, so my favorite is still the Hitachi pair.
syn08,
Bonsai-san,
Right, my assumptions are not supported by the often cited advertising department.
The question is what the fragile audio signal, comprising a vanishingly small power, think about the brusque treatment by your contemporary, advanced, high-end circuitry. Be sure, it would gladly prefer the outdated Hiraga design, offering maximal signal energy transfer and minimal damage.
Bonsai-san,
Not being aware of anything better, I recommend the highly insightful Hiraga design. It will reproduce all the traditional Japanese percussion instruments with great precision. (Simulated distortion figures won`t look pretty though, which is always a good sign).
I just had a conversation with my cartridge and I was informed that the signals are on strike; sick of being bullied by this unmerciful in it's cruelty preamp, their are running in circles around the unbalanced outputs, carrying "We want Hiraga" signs.
Some sumptuous miniature velour pillows to support the cables from your turntable, and assuage the delicate sensibilities of the signals, might be in order
.
I bet "somebody" is sitting in a dungeon on the Siberian Tundra with only a net-book at his/her disposal spamming DIYAudio by citing random words and sentences from a copy of Horowitz and Hill The Art of Electronics, and Oxfords dictionary illegally downloaded by a Torrent program....
btw, Nice design Ovidiu!
Cheers Michael
Attachments
Last edited:
I bet "somebody" is sitting in a dungeon on the Siberian Tundra with only a net-book at his/her disposal spamming DIYAudio by citing random words and sentences from a copy of Horowitz and Hill The Art of Electronics, and Oxfords dictionary illegally downloaded by a Torrent program....
btw, Nice design Ovidiu!
Cheers Michael
Michael
Good post
I like the SpamBot quote
And as you say it's a nice design.
stinius
Syn08, your insistence on low distortion really makes me think, So voila from my hotel room in the town of music Vianna comes the low distortion MPP with voltage feedback. I coined it Sans Pareil to differentiate it from the transimpedance MPP. The circuit i show has more parts then neccesarry because it opens up a plethora of fascinating options. I start to like it and will build it to compare it with the as you say "open loop" version. I learned at the university that this is current feedback because without any kind of steering method no circuit works. Anyway thanks for challenging me. I really enjoy your dry humor and intelligent circuits.
So here is some explanation: C1 and C2 can be inserted for Miller compensation in case of instability. The feedback OP´s need to have ample current drive so the one you recommend is a good choice here too. I will not use as extremely low resistor values as you do because i go for more gain (36dB) and will use 2 Ohm resistors instead of your 1 Ohms. I can not beat you in the noise department anyway because you have a 6dB noise advantage as you so eloquently explain to even the deaf. I can not use the very linear shunt feedback you use ether because my feedback path has to be non inverting.
I COULD beat you in the distortion department if i use two inverting stages in series with nested feedback loops but i have no desire to do so. I am not on this forum for competition but for enjoyment and i enjoy myself if i see your elegant work.
RIAA can be implemented in many ways. For example i could do the 75usec active in the first stage with R9, C7, R21, C10 with rather high value capacitors. The response would of cause flatten out to a gain of 1 at high frequencies and for perfection that could be equalised passive with R22, C9, R23, C13. Another option is to do that active, passive in the last stage. See the circuit around OP5. The 75usec could also be done passive with R22, C9, R23, C13. In that case C7 and C10 could be used to optimise the pulse response. You really inspire me!
To Lumba Ogir : that Hiraga circuit is really old and tired. It has tremendous distortion and very bad power supply rejection. I have seen varieties with such big supplies that you can build a whole decent system for the money. Anyway it can not be beaten in simplicity and is "philosophically" interesting. I can show you a circuit with 2 more transistors that has two orders of magnitude better distortion without adjustment but i will not do that. I can not rescue you anyway because you seam to be in heaven already.
"Make things as simple as possible but not simpler" Albert Einstein
P.S. Maybe i win the record for lowest noise in a BALANCED circuit. As i see on various threads people here seem not to be very sensitive for noise. For example a circuit with the INA163 is 6dB noisier then mine. I have even seen the OPA1632 that gets a lot of attention and is noisier still. The OP37 and the still much more noisy OPA637 (good for MM but not for MC) makes the people happy too. As i tried to put forward in my MPP thread i think that extremely low noise per se is not the key to great sound. I hate to much noise and distortion too but i also know that each part has it`s own sound so i try to avoid incredible complication. So i am somewhere between the objectivist and subjectivist camp. A person that tells me that he knows everything better then me is often not a good teacher.
So here is some explanation: C1 and C2 can be inserted for Miller compensation in case of instability. The feedback OP´s need to have ample current drive so the one you recommend is a good choice here too. I will not use as extremely low resistor values as you do because i go for more gain (36dB) and will use 2 Ohm resistors instead of your 1 Ohms. I can not beat you in the noise department anyway because you have a 6dB noise advantage as you so eloquently explain to even the deaf. I can not use the very linear shunt feedback you use ether because my feedback path has to be non inverting.
I COULD beat you in the distortion department if i use two inverting stages in series with nested feedback loops but i have no desire to do so. I am not on this forum for competition but for enjoyment and i enjoy myself if i see your elegant work.
RIAA can be implemented in many ways. For example i could do the 75usec active in the first stage with R9, C7, R21, C10 with rather high value capacitors. The response would of cause flatten out to a gain of 1 at high frequencies and for perfection that could be equalised passive with R22, C9, R23, C13. Another option is to do that active, passive in the last stage. See the circuit around OP5. The 75usec could also be done passive with R22, C9, R23, C13. In that case C7 and C10 could be used to optimise the pulse response. You really inspire me!
To Lumba Ogir : that Hiraga circuit is really old and tired. It has tremendous distortion and very bad power supply rejection. I have seen varieties with such big supplies that you can build a whole decent system for the money. Anyway it can not be beaten in simplicity and is "philosophically" interesting. I can show you a circuit with 2 more transistors that has two orders of magnitude better distortion without adjustment but i will not do that. I can not rescue you anyway because you seam to be in heaven already.
"Make things as simple as possible but not simpler" Albert Einstein
P.S. Maybe i win the record for lowest noise in a BALANCED circuit. As i see on various threads people here seem not to be very sensitive for noise. For example a circuit with the INA163 is 6dB noisier then mine. I have even seen the OPA1632 that gets a lot of attention and is noisier still. The OP37 and the still much more noisy OPA637 (good for MM but not for MC) makes the people happy too. As i tried to put forward in my MPP thread i think that extremely low noise per se is not the key to great sound. I hate to much noise and distortion too but i also know that each part has it`s own sound so i try to avoid incredible complication. So i am somewhere between the objectivist and subjectivist camp. A person that tells me that he knows everything better then me is often not a good teacher.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.