Where are the distortion figures of the HPS4.0 at 1V output (RIAA in place). Did I overlook something?
FWIW: Here is another type of balanced phono preamp - "RTP3C" - completely open loop, including passive RIAA filter. Ok, it has caps in the signal path. This version has been on the WEB for quite a while but in case someone has not seen it: Schematics
Distortion may be a bit worse than in the MPP, but still pretty good for a zero-feedback design; At 1 Volt output as stated in the "tube preamp cookbook": K2 = 0.004%, K3 = 0.006%, K4 & K5 = 0.001%.
Measuring low distortions in such high gain preamps is not easy. Although I have all the equipment one could wish, I was never able to measure distortions at 1V output, with a gain of 60dB, and that's because of the noise and artifacts. One day I'll build some 1:10 shielded attenuators with the Maxim precision resistive networks that I have, then try again. All I can tell, THD @1KHz is under 0.001%.
I had to reliable measure the distortions at an insane 17V output, and they were around 0.01%.
Those schematics are Allen's Wright designs, as far as I know. If I would be a tube fan, I would certainly consider them and further experiment with.
Pretty hard to follow your schematic, I'll take the time and do some redrawing to make it (to me) more legible. But at the first glance, it looks good. The 3dB penalty in noise is unavoidable in the differential/balanced configurations. As usual for feedback configurations, the loop needs careful compensation to provide unconditional stability.
One thing I am concerned in your approach is matching. It seems to me (again, at the first glance) that everything, from distortions to PSRR is very tightly related to component matching. Of course, using AD or THAT pairs provides what's required, but this may or may not be economincally feasible, in particulat when N/P matching has to be provided as well (THAT304 does NOT provide close N/P matching, but only close (thermal) tracking).
Again, I'll take a closer look ASAP.
P.S. Greet Vienna on my behalf
You can print out the pdf`s and paiste them by hand. Sorry but i had no better possibilities here in my hotel room and i put the circuit together in a hurry.
For people that do not like the balanced nature : in can be build single ended too.
Maybe i post a circuit tonight that has that change.
You are right about the matching. That is one reason i do not plan to make a commercial version until somebody pays for the lot of work. I rather post it hear so is does not get lost in time and space.
For people that do not like the balanced nature : in can be build single ended too.
Maybe i post a circuit tonight that has that change.
You are right about the matching. That is one reason i do not plan to make a commercial version until somebody pays for the lot of work. I rather post it hear so is does not get lost in time and space.
P.S: as i was talking about THAT they can make transistors to spec, also with hfe and u/be matching if so desired. i was also surprised that the MPP has more 2nd then 3rd harmonic so the imbalance in hfe is an advantage if you buy the philosophy that distortion should exponentially fall with rising order for best subjective result (eg. Hiraga and Hannes Mayer of Stereoplay a German Audio Magazine)
The nature of the circuit should cancel 2nd harmonic when it is totally symmetric.
The nature of the circuit should cancel 2nd harmonic when it is totally symmetric.
Here is the bare bones single ended MPP with "all in one go" shunt RIAA JLH style.
The capacitor at the output can be omitted with a servo but that would certainly be less
"elegant". If you do not like the inverting nature the leads of the cartridge can simply be reversed in case that the minus pole is not connected to ground.
I left out the bypass capacitor and simplified the current mirrors. In that state the noise is the same then the bypassed balanced version. The noise can be improved by 3dB with the capacitors in place and with the input transistors doubled i am in the noise ballpark of Syn08`s creations. Have fun ! Everybody should be able to build that circuit very easy.
P.S: NFB can be added too to satisfy the feedback camp.
The capacitor at the output can be omitted with a servo but that would certainly be less
"elegant". If you do not like the inverting nature the leads of the cartridge can simply be reversed in case that the minus pole is not connected to ground.
I left out the bypass capacitor and simplified the current mirrors. In that state the noise is the same then the bypassed balanced version. The noise can be improved by 3dB with the capacitors in place and with the input transistors doubled i am in the noise ballpark of Syn08`s creations. Have fun ! Everybody should be able to build that circuit very easy.
P.S: NFB can be added too to satisfy the feedback camp.
May be a silly question but should the distortion spectrum not change with the load impedance (collector load I mean)? For comparison, pentodes (common cathode of course) have varying levels of 2nd and 3rd harmonics depending on the anode load impedance, with K2 having a clear minimum at an optimal load point and rising towards lower and higher load, and K3 monotonously rising with higher load. So for a balanced pentode circuit a low load impedance would minimise K3 at the cost of higher K2, which would cancel out to an extend. I wonder if this was feasible with BJTs too?
You are one emitter follower, one cascode and one current feedback loop away from what I am showing in this thread
The cascode will certainly help increasing the PSRR while the feedback will bring down the open loop distortions. Tight matching no longer required, for anything but low input bias current reasons.Now, one big question that still has to be responded is about the input impedance. I chosed to make it "large" (using followers) because I can then always bring it down. Not so easy to bring it up! Has anybody seen any credible work/study related to cartridge loading? I am not interested for in any subjective stuff, at least for the moment.
Lumba it is really hopeless. I come from another planet. In fact you get the least distortion when you put many circuits in series with less gain and more feedback. Scott Wurzer did a paper about that some time ago. It is the total opposite of what you claim. You need lot`s of gain to apply lot`s of linearising feedback. I do not claim that circuits done that way sound any better. I can not hear an improvement when the distortion is less then 0.1% TOTAL (all distortions from the recording, the electrocics and the speakers) of low order harmonics. You can test your ability on the www.klippel,de website. A few people did much better then me but that can simply be statistical coincidence.
There is a survey with various amounts of distortion added and it was very edjucating for me. I understand your concern for simplicity but even on that ground i can beat Hiragas recommendation of Le Pre-Pre and Kaneda Preamp. Look no further then my MPP thread where i posted an even more purist version of the MPP coined Malcolm Hawsford Homage. I could even reduce the component count by going to a single ended topology with only one input transistor and one cascode transistor. But what the heck, listening to music is very subjective and when you are happy with your system who am i to debate your emotions. I only wish you could open your mind and build one of Syn08`s designs and listen to it on your own system. I will certainly build the HPS 4.0 when the PCB`S and full documentation is ready. I think these designs deserve a fair subjective accessment. Maybe next time i will put forward my experience with cardridge loading. I was the Lyra importer for many years and have heard and setup many systems word wide.
There is a survey with various amounts of distortion added and it was very edjucating for me. I understand your concern for simplicity but even on that ground i can beat Hiragas recommendation of Le Pre-Pre and Kaneda Preamp. Look no further then my MPP thread where i posted an even more purist version of the MPP coined Malcolm Hawsford Homage. I could even reduce the component count by going to a single ended topology with only one input transistor and one cascode transistor. But what the heck, listening to music is very subjective and when you are happy with your system who am i to debate your emotions. I only wish you could open your mind and build one of Syn08`s designs and listen to it on your own system. I will certainly build the HPS 4.0 when the PCB`S and full documentation is ready. I think these designs deserve a fair subjective accessment. Maybe next time i will put forward my experience with cardridge loading. I was the Lyra importer for many years and have heard and setup many systems word wide.
One last thing . The MPP does not have an ugly distortion profile because at least with the MAT02/MAT03 combination there is an element of asymmetry not obvious for the eye.
2nd comes first out of the noise floor then 3rd. At higher levels 4th and 5th arrive in Hiraga approved exponentially decaying order. Se my MPP thread for measurements.
I decided that this is my last commend on this thread about the MPP. I would rather like to contribute some simulations and measurements concerning cardridge loading. The dust has not settled on that topic and loadings from 47kOhm to zero are a possible choice with very opinionated users in each camp. What i can do is measure the impedance curve of an MC cardridge without destroying it. I did that several times successfully and put the data into a simulation. When i find the time i will post some results here.
P.S. : Has anybody ever designed a phono MC input with negative input resistance ?
2nd comes first out of the noise floor then 3rd. At higher levels 4th and 5th arrive in Hiraga approved exponentially decaying order. Se my MPP thread for measurements.
I decided that this is my last commend on this thread about the MPP. I would rather like to contribute some simulations and measurements concerning cardridge loading. The dust has not settled on that topic and loadings from 47kOhm to zero are a possible choice with very opinionated users in each camp. What i can do is measure the impedance curve of an MC cardridge without destroying it. I did that several times successfully and put the data into a simulation. When i find the time i will post some results here.
P.S. : Has anybody ever designed a phono MC input with negative input resistance ?
Joachim,
And it can get even worse. I forgot to say say that I was talking about perceived, not measured distortion, these are not nearly the same thing. Do I start being obscure? Presenting objective evidence-based claims is highly desirable, but not always possible. Sound perception is an intricated phenomenon, for which we currently don`t have any reliable analysis method and won`t ever have. The discussion will continue eternally...In the meantime, don`t be fooled by LTSpice.
Ho Joachim,
I would be interested in the paper Scott Wurzer (not sure about which Scott you meant?) wrote, where can I find the paper, would be pleased if you could help giving a direction where to search for it, thanks!
Cheers Michael
Not sure exactly why exactly the "hopeless" Lumba, whose posts i greatly enjoy, won't build it, but here are my reasons. To quote Joshua_G from another thread: "I'm interested in checking out designs made only by audio designers (engineers or not) who can hear differences between cables. " Well, i am not so fussy about cables but i am only interested in phono designs made by designers who can hear differences between phono amps.
A design can attract my attention only if the designer has not only built it but has some reasonable base reference - either some recognised commercial or diy design and preferably several of each type. If the designer does not hear any difference between well-measuring designs than why on Earth would anyone be keen to replicate it? Possibly because there is some ultra-cool topology involved but emitter followers at the input hardly qualify as ultra-cool.
Lumba, i understand now better what you would like to put forward. I totally agree that measured performance does not tell you how good a product sounds unless it is designed totally incompetent. As i mentioned my threshold is at 0.1% low order. I think Earl Geddes is researching that topic since a long time and you can read some interesting results on his website. He is also posting here on diyaudio. Here is a ceveat : as far i i understand his work he is more interested to work with "normal people" with no particular sofisticated training. So he wants to find out what the majority of people can hear. This is a noble root but in the Klippel test there where certainly some people that did much better then me on the distortion test. For a minority of people small differences may be important. I compare it often with cooking and dining. I know some people that can not even stand a homeopathic dose of garlic in the food or are able to differentiate wine from various makers, vintage and grape. I pride myself to detect diffences between solder, resistors, wire and even OP´s that have less distortion then 0.001 % although i failed miserable on the Klippel test. For me there is an element of art involved but wishful thinking and the Placebo effect can lead to poor choices. That is one of the reasons i watch closely what people like syn08 are doing because he looks at the problem from the engineering perspective and is not much flattered by subjective accessment. By the way i like the sound of my own phonostages better then the Hiraga.
I worked closely with Hirage in the beginning of the 80th and have heard various incarnations of Le Pre-Pre over time. I hear more information over my designs. There is more sparkle on top and more speed and grunt in the deeper regions. I hear more micro information and better focus and 3-dimensionality. It simply sounds more "real" to me. High end has progressed in the last 25 years,that is at least my experience.
The Scott Wurcer i mean is the one that designed the AD797. Although nearly 20 years old is still one of the lowest distortion and noise Opamps on the marked.
Look at the Dennis Collin paper about a low distortion phono stage. You can download that for free from AudioXpress. In that paper he mentioned the work of Wurcer with reference where he published it. Another possibility is to chat with him directly. He is on this forum. I think he posted in the Blowtorch thread and others.
Concerning negative impedance there needs to be a sensor inserted from cardridge to input. The simple solution would be a resistor somewhere between 0.5 and 10 ohm. Feedback could be taken from there or is it feedforward ? When i find the time i will put some thought into it. By the way i am hopeless too being trapped into me own little world.
I worked closely with Hirage in the beginning of the 80th and have heard various incarnations of Le Pre-Pre over time. I hear more information over my designs. There is more sparkle on top and more speed and grunt in the deeper regions. I hear more micro information and better focus and 3-dimensionality. It simply sounds more "real" to me. High end has progressed in the last 25 years,that is at least my experience.
The Scott Wurcer i mean is the one that designed the AD797. Although nearly 20 years old is still one of the lowest distortion and noise Opamps on the marked.
Look at the Dennis Collin paper about a low distortion phono stage. You can download that for free from AudioXpress. In that paper he mentioned the work of Wurcer with reference where he published it. Another possibility is to chat with him directly. He is on this forum. I think he posted in the Blowtorch thread and others.
Concerning negative impedance there needs to be a sensor inserted from cardridge to input. The simple solution would be a resistor somewhere between 0.5 and 10 ohm. Feedback could be taken from there or is it feedforward ? When i find the time i will put some thought into it. By the way i am hopeless too being trapped into me own little world.
Excellent, you got a great team there (I can think of a few more names you could add to the list), also a great motto to lighten your way
Good luck!Do you have a little bit more information on the topology? Is it similar to yours?
Reinhard
It is not but he also uses a common base input stage. He did not disclose any details but the inductors most be of enormous values. I have seen the PCB on his screen and i ordered a sample. I his words the amplification devices are fets and he uses topologies similiar to his tube designs. Some time ago he designed a Tube-Fet Pre-Pre for me that used the Dr.Loesch topology. It is a fet, tube cascode and has around -100dB distortion 2nd harmonic at cruising level. Jürgen is certainly no advocate of NFB and uses mostly local current feedback. I my experience designs like that sound more "liquid" and high NFB designs sound more "clear".
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.