Quote
" have been producing world class pre-preamps for the last 35 years. Most here are rank amateurs, without experience, understanding, or even having done the computations necessary to determine what they are stating. I just can't cope with this disparity in understanding and experience. It is pointless. Also, 50 ohms is usually too low, 100-200 ohms is usually better with modern cartridges. Capacitance is optional and should point out the silliness of worrying about the intrinsic capacitance of the input devices."
You should find the Tylenol in the medicine cabinet in the bathroom. Take two of those, make yourself a nice cup of hot chocolate and go to bed.
" have been producing world class pre-preamps for the last 35 years. Most here are rank amateurs, without experience, understanding, or even having done the computations necessary to determine what they are stating. I just can't cope with this disparity in understanding and experience. It is pointless. Also, 50 ohms is usually too low, 100-200 ohms is usually better with modern cartridges. Capacitance is optional and should point out the silliness of worrying about the intrinsic capacitance of the input devices."
You should find the Tylenol in the medicine cabinet in the bathroom. Take two of those, make yourself a nice cup of hot chocolate and go to bed.
Charles re your post Post #111.
I think the overload question has to consider the equalization topology to have any real meaning. I would expect that passive equalization - where the firs t stage is purely amplification - would have less 20KHz headroom than an active one in which GNF rolls the HF resonse off (ignoring for the minute any concerns about other non-linearity mechanisms relating to feedback).
This is where the trade-off for noise vs overload capability has to be made (Walt Jung's paper covers this quite well in my view).
I agree discussing general overload capability with respect to pre-amp outputs that can deliver 7VRMS when the power amp is already clipping at 2VRMS or less is pointless (a '******* contest'?). My comments earlier were concerned with 20Khz overload capability, and more specifically wrt passive equalization. I'll try to find the JA review tonight.
I think the overload question has to consider the equalization topology to have any real meaning. I would expect that passive equalization - where the firs t stage is purely amplification - would have less 20KHz headroom than an active one in which GNF rolls the HF resonse off (ignoring for the minute any concerns about other non-linearity mechanisms relating to feedback).
This is where the trade-off for noise vs overload capability has to be made (Walt Jung's paper covers this quite well in my view).
I agree discussing general overload capability with respect to pre-amp outputs that can deliver 7VRMS when the power amp is already clipping at 2VRMS or less is pointless (a '******* contest'?). My comments earlier were concerned with 20Khz overload capability, and more specifically wrt passive equalization. I'll try to find the JA review tonight.
Hi Tino.
I lost my password for diyhifi.org and can't see the circuits there anymore.
No, I did not build it yet, for three reasons: I'm examining another topo in my sparse spare time, it is (too) complex for an input stage, and, offest and drift might indeed very well be to big. But just maybe I have an solution for this, but it needs to be tested in reality.
Rüdiger
I lost my password for diyhifi.org and can't see the circuits there anymore.
No, I did not build it yet, for three reasons: I'm examining another topo in my sparse spare time, it is (too) complex for an input stage, and, offest and drift might indeed very well be to big. But just maybe I have an solution for this, but it needs to be tested in reality.
Rüdiger
Hi Tino,
thanks for the schematic.
What I have build and currently using it a variation of this threads post #27
It's along these lines of thinking. The input offset is under 1mV now, but may well be higher when using unmatched components.
The sound of the corresponding pre is detailed and powerful, but still a work in progress.
Rüdiger
thanks for the schematic.
What I have build and currently using it a variation of this threads post #27
It's along these lines of thinking. The input offset is under 1mV now, but may well be higher when using unmatched components.
The sound of the corresponding pre is detailed and powerful, but still a work in progress.
Rüdiger
it's this article?
http://stereophile.com/phonopreamps/307sim/index2.html
In my experience, headroom is very important in a phono stage, because some phono stages can overdrive with the pops and clicks on the record.
Some time ago, we compared a Benz phono stage to one of my design's, and me and the other listeners were surprised that my phono had much less noise replying old records than the Benz.
I think this is a important step of a design.
Rick
http://stereophile.com/phonopreamps/307sim/index2.html
In my experience, headroom is very important in a phono stage, because some phono stages can overdrive with the pops and clicks on the record.
Some time ago, we compared a Benz phono stage to one of my design's, and me and the other listeners were surprised that my phono had much less noise replying old records than the Benz.
I think this is a important step of a design.
Rick
Bonsai said:
Again, the question of being able to drive the EQ network is a simple one. If the circuit designer can't get past that step, perhaps he should be in a different line of work.
Regarding your comments that a passive EQ would tend to have less headroom than an active EQ, that is absolutely a wrong generalization.
In a previous post I gave an example where a Dyna PAS-3 (with active EQ) could run into trouble at high frequencies. On the other hand, look at post #145. Here is a schematic with a passive EQ driven by a transconductance stage. There will be absolutely no problem with "overload margin" varying with frequency in this design.
Again, these are elementary problems to solve.
I am also surprised by the designs that have been suggested so far. Some have a large electrolytic coupling capacitor at the input. No matter how good that design might sound, an equivalent design without the capacitor is clearly going to sound better. Why bother to make a DIY circuit that has such in-built performance limitations?
Another design is having trouble getting the DC offset at the input below the millivolt level. This is surely an unacceptable circuit. Many MC cartridges have coils with a DCR in the order of a few ohms. So this would result in currents on the order of a milliamp flowing in the cartridge coils. I would *never* connect an expensive phono cartridge to such a design.
In my opinion, the only way to get past that is to use tubes or FETs as the input device. John Curl has demonstrated that FETs are capable of as low a noise level as any device, regardless of the driving source impedance. So that is my choice.
It could also be done with paralleled high Gm tubes, but that is not my preference. One of the problems is that most people are so unimaginative. How many times have you seen a phono stage with 12 AX7's or 6DJ8's as the input tubes? Surely it must be close to 99% of the time.
But if you are going to use tubes, why not do something original. For example, why not try a 6080 (6AS7) on the input? I haven't looked at the data sheet for decades, but I recall that they have a much higher Gm than those puny input tubes.
And then, of course, there is the "Steve Eddy" approach. Just throw a transformer on the input and be done with it....
rickpt said:
No that wasn't it. That one appears to be a case of poor circuit design. I distinctly remember the review that Bonsai mentioned. It was more like two years ago (or more???). I'm pretty sure that the overload margin was given as 6 dB at 20 kHz. It stuck in my mind, because as I read the article it was clear to me that the circuit was simply clipping the output stage due to the high gain configuration, and not a case of poor circuit design.
Charles Hansen said:
It seems that
this
phono stage has the common base input approach.....
No idea how they manage it to avoid these dangerous offsets...
Maybe Werner could comment?
I went away from that idea, and am thinking about something along this conceptual schematic.
Which would also be my contribution for the thread topic.
Tino
Attachments
Tino -
The circuit you have posted is broadly similar to what we have been using at Ayre for many years. Of course, the devil is in the details, and I will leave it to you to work those out. However, here are a few points:
a) There is no need for paralleled parts in the second stage. The gain of the first stage ensures that.
b) You are better off to have the adjustable gain in the first stage. Otherwise you are open to overload problems in the first stage with high output cartridges.
c) Your design is DC coupled. There is no way to have this work with an overall gain in the 60 dB to 80 dB range -- and that's at 1 kHz. At DC, the overall gain will be in the 80 dB to 100 dB range!! You need to either AC couple or use a servo, pick your poison.
And of course, the J74 is now officially discontinued. Luckily, we placed a large order for our lifetime buy purchase. I estimate a 50 year supply....
The circuit you have posted is broadly similar to what we have been using at Ayre for many years. Of course, the devil is in the details, and I will leave it to you to work those out. However, here are a few points:
a) There is no need for paralleled parts in the second stage. The gain of the first stage ensures that.
b) You are better off to have the adjustable gain in the first stage. Otherwise you are open to overload problems in the first stage with high output cartridges.
c) Your design is DC coupled. There is no way to have this work with an overall gain in the 60 dB to 80 dB range -- and that's at 1 kHz. At DC, the overall gain will be in the 80 dB to 100 dB range!! You need to either AC couple or use a servo, pick your poison.
And of course, the J74 is now officially discontinued. Luckily, we placed a large order for our lifetime buy purchase. I estimate a 50 year supply....
Kostya-M said:
Nice circuit Kostya
Using FETs for input is what I would do, too.
We shall not be too paranoic about trying to get a '100% noise free' input.
Because this is not practically possible, anyway.
We live our lives in a constant noisy environment.
Background noise we will never notice.
Those that have entered a room without noise, however, will get a spooky experience.
The absence of noise can almost drive a person mad
The total lack of noise is not = silence, because in natural silence there is a noise we are used to.
And think of all those well sounding Tube amplifiers using 'noisy' carbon resistors.
I rate harmonic distortion and lack of transient capacity (slewrate)
as far, far worse things in audio amplifiers.
Okay, let me get to my main message in this post.
It might be well worth it, to study and read a Pass Labs paper:
Pearl Phono.
This 'all fet' RIAA amplifier for DIY people, has been evaluated with different source cartridges.
1. Uses 4 paralllelled 2SK170BL at full max IDSS.
2. These are cascoded and
3. put into second stage of one 2SK389 parallelled pair.
4. The cascoded SK389 collector delivers the output.
The article is put together by one of Nelson Pass' chosen men, Wayne Colburn.
I think, it is not a bad guess to assume Colburn did his project
collaborating with The One and Only inside the Pass Audio laboratory.
Regards, Lineup
A few statements worth quoting:
An externally hosted image should be here but it was not working when we last tested it.
Pass D.I.Y. Preamps: Pearl Phono: part1
Lineup, you are right about noise. I think that noise level of amplifier must be on 4-10 dB lower of free disk groove. Therefore I not use parallel connection of transistors, noise level is good thus.
My correction method is interest in that distortion level is not rise with frequency because of rising of head output level. Schematic with linear amplification and passive correction has this distortion problem.
PS. My amp is for MM
My correction method is interest in that distortion level is not rise with frequency because of rising of head output level. Schematic with linear amplification and passive correction has this distortion problem.
PS. My amp is for MM
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.