MPP

This is an even more purist version i call the Malkolm Hawsksford Homage because it has some resemblance to the Essex Battery Phonostage that was made by LFD . It has no voltage feedback at all does amplification and RIIA in one stage and uses no integrated circuits.
It has an interesting gain structure : I-U, U-I, I-U. U-I
expect more distortion from that stage but it could sound good.
 

Attachments

  • MPP Basic Structure single ended Malkolm Hawksford Homag.pdf
    57.9 KB · Views: 859
Administrator
Joined 2004
Paid Member
Hi Joachim,
Thank you for sharing these!

I do know that the output buffer in the Malkolm Hawsksford Homage sounds excellent. It's almost exactly what I do except that I use red LEDs as the voltage reference in the CCS sections. So nothing negative to blame on that output buffer design.

Your input stages are a bit less familiar to me, so I'll have to get my head around that. Interesting designs.

-Chris
 
Do you mean Complimentary-Longtail-pair ?
I think James Bonjiorno and Borbely where the first.
The first folded cascode i have seen was around 1983 in L`Audiophile by a guy named Rambeyrolles. I do not know if i spell his name correctly. He certainly inspired me.
The advantage of the folded cascode is quite high gain with very little topological complication. As you see in the MPP the signal is only flowing trough two transistors and has the resistors in parallel so you could say they are not in the direct signal path.
Nevertheless distortion of that stage is quite low. I have access to an Audio Precision
SYS-2522 here in Vienna and my engineer friend was quite surprised that distortion remained under the noise floor at 1V output without voltage feedback in the input stage.
Some integrated circuits use the folded cascode gain stage albeit not balanced as i do and with bootstrapping.
The AD797 and LT1469 are examples. Their gain structure can be considered as single stage and still they belong to the lowest distortion op amps on the marked. I like both very much if i use op amps and they are recommended for the instrumentation amp stage in the MPP. Used in the correct topology they have a very neutral and true "matter of fact" sound that i like to compare my more holistic tube or discreet designs too.
 
AP measurements

Here are some measurement on the AP analyser of the MPP.
Distortion is under the noise floor at 1 kHz and 1V output.
The noise shows a lot of 50Hz harmonics. We had a hum problem we could not quickly remove. Nevertheless you see that signal noise ratio is -80dB in the midrange improving to -90dB in the treble. Not bad, i am happy.
Also the transfer linearity is shown. Overload margin is ca.20dB at 1kHz for a 0.5mV input improving to ca.40dB at 20 kHz due to the way i implement the 75 usec time constant. This is no world record but the improved margin in the treble will make shure that there is no overload due to scratches and dirt on the record that are the major cause of overload in designs that have no satisfactory overload margin in the treble.
 

Attachments

  • MPP FFT.pdf
    54.3 KB · Views: 538
  • MPP Linearit_t dbv.pdf
    51 KB · Views: 452
diyAudio Chief Moderator
Joined 2002
Paid Member
Hello Mr. Gerhard

I am interested in your experience, objective and subjective, on the matter of MC cartridge correct loading value. Please expand if you feel like it at a point.
I am following your current circuit investigations and findings with much interest.
Thanks for contributing.

Regards
 
Hello Salas !
Nice to have you around.
Tell me your cartridge and maybe i have an idea what to say.
I watched your simple riaa thread and that attracted much more attention then what i am doing hear. great that you are coming up with something that people actually need and will build. i have the parts that you suggest and could build it too but i am too occupied with what i do. sorry.
to anatech : could you post your buffer ? i would be very interested
buffers can be extremely use full for a lot of circuits and can improve the sound (i think especially dynamics ) a lot
 
the cartridge and the input impedance constitute a low pass filter.
so i need the equivalent circuit as we say in germany
a passive filter can be modeled with equivalent capacitors, resistors, losses (hysteresis) and transformers. the active element is harder to grasp because noise and distortion are creeping in from the circuit, the layout and the power supply. today there is a lot of concern about electromagnetic radiation too. passive parts are not ideal and i could come up with ever so complicated "equivalent" circuits. the lumped parameters of the poor capacitor contain resistors, capacitors and inductors in spades the higher you look in frequency. it simply is not that simple. a friend of mine has a DENON DL103R and that is shurely on the agenda.
 
Administrator
Joined 2004
Paid Member
Hi Joachim,
My scanner just stopped working properly. I'll see what I can do if there are electronic copies around still.

The buffers I tend to use are very simple, basic. However, there are some members here that have done a wonderful job with their buffer designs. Let me see if I can find their names for you. PMA comes to mind for one.

Best, Chris
 
diyAudio Chief Moderator
Joined 2002
Paid Member
the lumped parameters...

Since there is discussion about minimal frequency response deviations for different enough MC loading, hence contradiction between the amplitude domain and the too commonly reported substantial subjective difference perceived, almost a traditional ''fact'' between TT owners, I was leaning towards the idea that it can be possibly researched as a matter of correct termination of a complex conduit of interacting parameters standing between the cartridge output pins with its generator characteristics, and chosen load. I.e. cabling, phono input charactristics. So, maybe proper loading could be about terminating the transmission well, avoiding reflections. Is that too wild a guess in your experience?
 
well, why not. i have seen newest research on transmission lines and the conclusion was that reflexions ARE relevant in audio frequencies i look if can find the literature
what amases me is that nobody is talking about terminating an mc cartridge with a resistor
AND a capacitor to compensate the substantial rise in the treble that they all have to a certain degree with circuits like the MPP i had success with a 10nF in parallel
 
diyAudio Chief Moderator
Joined 2002
Paid Member
I have used a trap circuit in the past so to catch tip mass on vinyl resonance in the last octave. While the bode flattened out, mid to upper range dynamics suffered. Did you tame the last octave fully with 10nF? Here is what I have with DL103R and about 450pF (cable and one fet input Miller), res 1/48 oct. Telarc OmniDisc pink noise L+R band 13/side 2. 150R load resistor. Look to that Lyra test too. Werner has some FFTs and some interesting opinion above the first graph he shows. I think you mentioned having familiarity with Lyra models.
 

Attachments

  • DL103R.gif
    DL103R.gif
    14.9 KB · Views: 1,575
the combination of the 5.5 ohm 0.0005mH titan and 10nF with the 25 Ohm input impedance of the MPP is only 3dB down at 3.5 Meg Hz my simulation shows nevertheless the sound got more "natural" without any spittyness to be honest i just listened to various combinations and it sounds crazy good for my taste a lot of dynamic slamm and big-focussed soundstage with tremendoes height and projection people are quite shocked that hear the system because it is so emotional
maybe i also have to measure inductance of my wire but it should not be much
 
diyAudio Chief Moderator
Joined 2002
Paid Member
I thought you meant response manipulation when I read ''to compensate the substantial rise''. Sorry. So you don't talk about ironing out its 10-20kHz on its bode plot on test disc FFT, but about softening the subjectively perceived HF edginess? Hmm, hints to the time domain. Impulse response.