Mpp - Page 152 - diyAudio
Go Back   Home > Forums > Source & Line > Analogue Source

Analogue Source Turntables, Tonearms, Cartridges, Phono Stages, Tuners, Tape Recorders, etc.

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 6th August 2010, 10:57 AM   #1511
diyAudio Member
 
Joachim Gerhard's Avatar
 
Join Date: Mar 2009
Sam, your experiment with PSU´s is important. As i have said several times here, i am not good at designing PSU´s. I rather try to make my circuits imune to PSU issus as good as i can. Nevertheless your shootout could spearhead the path i should follow ones you found the best sounding solution. I will then try to learn why that particular topology works best and then we have another thread about PSUs for Audio.
I still have not build the Meaner Shunts that Salas has deshned for me, but i have now all the parts i need and i asume i can make it work in the next two weeks.
  Reply With Quote
Old 6th August 2010, 11:55 AM   #1512
diyAudio Member
 
Onvinyl's Avatar
 
Join Date: Aug 2002
Location: Germany
Joachim, yes, but the helper transistors (current mirrors) nail the thing DC-wise.

Rüdiger
__________________
"I can feel what's going on inside a piece of electronic equipment. I have a sense that I know what's going on inside the transistors." Robert Moog
  Reply With Quote
Old 6th August 2010, 12:26 PM   #1513
diyAudio Member
 
Joachim Gerhard's Avatar
 
Join Date: Mar 2009
Yes, it´s a great way to do it. Send me your e-mail adress by PM and i show you something that could interest you.
  Reply With Quote
Old 7th August 2010, 03:37 PM   #1514
diyAudio Member
 
Joachim Gerhard's Avatar
 
Join Date: Mar 2009
On Holger Barskes Website is a small listening test of my Nobrainer Phonostage. What i found is that a typical DIY´s likes simple circuits. Look at the success that Salas has for example. So i thought the Nobrainer is ideal for DIY, even for beginners. On Holgers site are also comments about my stage and one guy shows a circuit from a Japanese designer that uses the NE5534 for MC duty. I think this is not such a good idea and when i expressed my doupth a small discussion stated about the merits of the NE5534. I know that OPamp very well. The first design i made in 1983 with the NE5534 was a preamp. At that time the NE5534 was revolutionary. It was the first OP i really liked.
It has some problems with common mode distortion and offset. Voltage noise is 3.5nVqHz and current noise is quite low for a bipolar so it is low noise with MM cartridges. For MC it is too noisy but still some companies like LFD use it in more affordble stages. Common mode distortion can be avoided by using shunt feedback and voltage noise can be reduced with massive paralleling. I thought when a caveman can build an amplifier then a German Neanderthal man can build a phonstage so i designed a stage that should work and is extremely simple. I use 8 NE5534A in parallel with a one stage shunt feedback RIAA. Gain of that stage with a Titan i that has 6 Ohm impedance including cable loss is 60dB. Gain at 20kHz is around 40dB with RIAA and the NE5534 has a gain of 5000 @ 20kHz leaving a feedback factor of 34dB @ 20kHz. This is not much for an Opamp but still i think it is sufficient. This is not a state of the art Phonostage in any way but everybody has some NE5534 around or can buy them for less the 50 cents a piece. This stage can be build for less then 20 € if a semy decent PSU is available. The NE has good PSU rejection and even a simple Pi filter or some batteries buffered with elcaps of say 2000 uF will do. I thouht about how i will call it and no, i do not call it the Neanderthalman but it will be known as the Phonoman.
P.S. noise of that circuit is around 1.3nVqHz and offset shuld be low with a low impedance cart like mine. The plus and minus input is at near zero potential in that design and when you use a cart like the DL130 with 40 Ohm the resistors in the RIAA have to be increased by a factor of 40/6 and the values of the caps lowered with the same factor. 320 Ohm resistors have now to be fiited from Pin2 to ground in each Opamp to balance the offset voltage. Maybe i will ad a coupling cap at the output for savety.
Attached Files
File Type: pdf Phonoman.TSC - TINA.pdf (50.6 KB, 176 views)
  Reply With Quote
Old 8th August 2010, 01:06 AM   #1515
diyAudio Member
 
Joachim Gerhard's Avatar
 
Join Date: Mar 2009
I build the Phonoman and got it working after some small difficulties at the beginning.
My last post has shown a DC coupled version and with the NE5534A it does not work that way. I measured around 28mV to cartridge in both channels ones i connected a 5 ohm dummy resistor. That made the output jump to over 10V with a +-12V supply. I knew that the NE5534 is not famous for low offset voltage and the OP27 is much better in that regard becuse it has a bias cancelation circuit. Still i think even the OP27 is not good enough for DC coupling at the input. Bias cancelation also brings a new source of noise. That is the reason Douglas Self does not like the OP27 or the LT1115 or LT1028.
I can not see that noise in the spec sheet so i have mixed feelings if Self is right. Comparing the AD797 he likes to the LT1028 i can not hear much difference in noise. Quite the contrary because the LT1028 seems to be well to spec all the time whereas i have some samples of the AD797, say one in twenty that have to be sorted out when you use them at the input of an MC phonostage. I solved the issue by simply inserting a 10.000uF 6.3V Pannasonic FM i keep handy for that pupose. Just to feel better i bypassed with a 1uF Wima MKT and a 10nF Rifa SSM. Actually i think the cap does more good then evil at that place. All DC coupled Phonoclone like stages i see here send considerable DC into the cartridge. After a hot debate on this thread if this is audible i build a transimpedance input stage some time ago where i could adjust the DC
at will. I found that 100uV is just audibe in my system and non of the phonoclones i see here match that criterion. If you are concerned about the distotion that this cap brings i can asure you that it is very little. An elcap develops distortion ones a high voltage develops over it. That happns in the bass in poweramps but not at the mV levels we talk about here. Burkhard Vogel had the same problem and in his typical rationate way developped a circuit where he could listen to Elcaps in the chain at the levels that are involved over headphones. He could not hear a difference with the cap in or out.
I do not know how good Burkhards ears are but i will publish that circuit tomorrow and you can make your own test. the second chage i made is that i used only 4 stacked NEs because i am soldering them directly over each other and when the circuit does not work i loose 16 chips that are hard to desolder. That buys me a worcening of Signal-Noise of 3dB and i thought i can live with that. Actually the circuit is still some 5dB better then a single OP27 that i see here in very alaborated phonos. It surprises me that stil at the noise level of the OP27 the users are extremely happy and pleased with the result. I can only asume that they listen at much lower volume then me on speakers with under 90dB efficiency. On the other hand "When the stone hits the plastic" as our chief moderator Salas whould say a hunt for extremely low levels of noise is rather academic. Well, i enyoy the spooky silence that my new Transconductance stages have the the Phonoman AC as i will call it with a stack of 4 is no match when i cranc the volume full. The noise is not particular agressive because the shuntfeedback RIAA nicely shapes it but there is a culmination of noise audible, i whould say i the upper midrange just under the presence region so it does not sound forward but slightly back. Its hard to describe but it does not go on the nerves like flicker or popcorn noise does. It´s like the sound that the wind makes when Boris Becker plays tennis ( was i the one that wrote this down ? ). At my normal listening level it is well surpressed and just audible, faintly in the backround from the midle of the speakers. That is much better performane then a 20€ stgae like that has any right too.Well i have not counted the time it needs to make it but i whould say, the next time i build it, i need 2 hours max. So how does it sound ? The first think that was obvious and i knew that from the discrete INA where i use stackes of NEs in shunt feedback is that typical robust bass. I had to turn down my sub a bit after i switched from the lates Transconductance stage. Wel actually i got a lot of noise in the beginning on one channel. After substituting the 4 Nes everything was fine so i thought they, or one of them was faulty. Actually i had taken the NEs from the considarable waiste that was left over from the development of the Discrete INA so that was a logical explanation.
I was wrong. I had forgotten to solder one row of contact but i found that out much later until i looked again at the object to photograph it. See the atatchement.
In the midrange i whould say it sounded unspectacular and neutral. I can live with that because most cheep stages do not get the midrange right at all so particular clasical music can sound dreadfull. Strangly agressive and foggy the same time. Not in this case but also not extremely tansparent. If this is good or bad depends on your system also because the Phonoman AC has a particular way to treat the treble. I have seen numorous times that the NE series of Opamps gets acused for agressive and sterile treble that borders on the edgy. Nothing could be further from the truth in this particular application. The treble is silky smooth with a tendency of rounding and a rose tinted brownish tan. If this is the consequence of the topology or my use of components that have been soldered on and from one row down of my collection i can not say.
Actually i do not have really bad components so you can find some used antimagnetic Dales, two very expensive Welvyn Metalfim, Reichelt Styroflex, Rifa PPS and COG ceramics. The elcaps are Pannasonic FM and Sprague blue so nothing here that could explain what i hear this time. I have used the same components numorous times in circuits that sounded different. Maybe i have to wait until tomorrow and listen more.
The elcas may still be a little leaky and need some forming. That can need days or weeks. Next time i will preform at least the input cap for 24 Hours.
A real fun project that was. Making something out of nothing reminds me on the times i was a teenager and we had no money but wanted to hear music well. That brought me into this game and that is the reason i love DIY.
Attached Images
File Type: jpg NE5524 stack.jpg (63.9 KB, 397 views)
File Type: jpg NE5534 stack unsoldered in the midle.jpg (35.2 KB, 355 views)
Attached Files
File Type: pdf Phonoman AC.TSC - TINA.pdf (56.8 KB, 117 views)
  Reply With Quote
Old 8th August 2010, 01:23 AM   #1516
Salas is online now Salas  Greece
diyAudio Chief Moderator
 
Salas's Avatar
 
Join Date: Oct 2002
Location: Athens-Greece
Hey, not bad at all for a bunch of creepy bugs on a pack of Luckies. Nice write up too.
  Reply With Quote
Old 8th August 2010, 02:19 AM   #1517
diyAudio Member
 
Joachim Gerhard's Avatar
 
Join Date: Mar 2009
Thanks, tomorrow i post the "cap listener" and i am working on a transformer coupled, inductive RIAA with 4 Josi Opamps. An expensive beast, so to say the total opposite of the Phonoman. A friend of mine has an original Audio Note IO with 1Ohm impedance and only 0.025mV output and then it gets tough, even for my best electronic stages. I already have some nice Lundahls and the good coils that Brian Sowter made me in house. I wait for Sigurd Ruschkovski to finish the tests on the Josi and then i try an all out asault.
Will use your meaner shunts this time, i promiss.
  Reply With Quote
Old 8th August 2010, 07:50 AM   #1518
Werner is offline Werner  Europe
diyAudio Member
 
Join Date: Jan 2002
Location: Belgium
Quote:
Originally Posted by Onvinyl View Post
I build both versions of your phono
...
I had to make the load resistor for the input fet trimmable to get the best working conditions for it.
Which versions exactly?

I have been working on and off on a seriously big version of a similar phonostage. It would be two full-width chassis, with LC regulation, pretty high voltages for the front-end etc.

But there wasn't much progress towards a commercial-grade build of this, and I wanted something to listen to quickly. Further, we were buying a house for renovation (more work!), so I decided to make a tiny version of the circuit in the few months left to me before I would re-invent myself as carpenter/mason/... So this brought some constraints:
-PCB 100x160 mm max
-only one 2SK369 per channel (don't want to deplete the stock!)
-100nF cap for 75us (wanted to use either MKV or MCap)
-approx. 9nF cap for the rest
-60dB gain
-18V supply for the front-end
-accuracy better than +/-0.1dB

This ended up being like a Chinese puzzle. I succeeded, but only with some effort and a lot of pre-characterisation of active element and measurements on partial and full circuits.


Now back to calculating heat losses and isolation layer thicknesses...
__________________
bring back dynamic range
  Reply With Quote
Old 8th August 2010, 08:25 AM   #1519
diyAudio Member
 
Join Date: Jun 2005
Location: East of Uddersfield
Hello Joachim.

How much gain does your schematic in post 1488 have?,

Also, does the gain come from the cascode transistors, or do the current mirrors provide additional gain?

Thanks
  Reply With Quote
Old 8th August 2010, 01:13 PM   #1520
diyAudio Member
 
Joachim Gerhard's Avatar
 
Join Date: Mar 2009
The innitial gain of the input stage is defined at 2 x Gm x RL. RL is the output impedance of the cascoded current mirror. The cascoding raises the output impedance of the mirror
by the beta of the cascode transistors. In the case of the BC550C-BC560C beta ( or hfe )
is ca.500. The factor of 2 in the formular comes from the fact that the gain in a parallel symmetrical stage is double then the gain in a single ended stage. So you see, the mirrors do not add gain but raise the gain in the input stage by raising RL. Actually the bipolar cascodes shield the Fets so the gain comes from the bipolar cascode transistors. In that case the gain equation is (( beta x RL ) : ( Rbe + Rbv )) x 2. RL is the output impedance of the uncascoded mirror times the beta of the cascode transistor If we asume that the output impedance of the uncascoded mirror is 1kOhm then RL is 1000 x 500 = 500.000 Ohm. Rbv is zero in case of that circuit because i shunt the base current AC wise to ground with a capacitor. Rbe is ca. 35 Ohm with the cascode transistors i am using so we get 500.000 : 17,5 = 31.847. Bandwidth of the input stage is 1Mhz -3dB so gain-bandwidth product is 1M x 31.847 = 30GHz. Now you know what a beast that is.
Well i think 30Ghz is rather optimistic because i did not take stray effects into account but still in praxis we end up in the several GHz range. That gain now is shaped by the RIAA filter to ground. I have to calculate the input impedance of that filter at 1Kz and then you have to total gain of 500.000 : Ohm @ 1kHz.
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off



New To Site? Need Help?

All times are GMT. The time now is 11:03 PM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright ©1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2