PSA 150 S (pure sound amplifier)---help needed

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hello DIYaudio, I'm new on this forum (only 4 months, but have learned alot from all of you) so i need your help.
Here is a schematic that my colleagues have had developed from alot of other schematics---only the good things from each one, on " Forumul Electronistilor" wich is just like diy audio (only in romanian language); can somebody trace its origins and can tell me if it can deliver good sound quality (ex. some simulations, personal opinions)?
I want to say that we made some simulations but, we (other colleagues) didn't have all of the transistor models in LtSpice, so the results where inaccurate.
We have already initiated a group buy for the components, and i am hoping that all of you could help me to tweak this interesting design for better results and sound quality.
I have attached the BOM and the schematic.
Please excuse my rough english.
Regards

Sergiu
 

Attachments

  • BOM PSA-150S.pdf
    123.5 KB · Views: 411
  • SCH PSA-150S COLOR.pdf
    83.7 KB · Views: 506
High collector voltage on input transistors. Noise ? Reliability ?

Q109 really needed ? Have you compared with/without ?
Purpose of JP101 ?

R128 R129 cause ground pollution. Need proper cautious wiring.
R144 R145 cause ground pollution. Need proper cautious wiring.

C111 (bias filter) 1µF optimal value ? Have you tried more ?

Is the amp stable when suppressing C116 C117 C118 C119,
shorting R150,
suppressing C122 C123 C124 C125,
shorting R151,
and placing a single V+ to V- decoupling cap like 220µF ultralow ESR ?

C109 C110 : allow mutiple choices, more than just WIMA and Nichicon.
Is there a version with FET input and no capacitor in the feedback path ?

DZ101 DZ102 : better without ?
C101 C102 C103 : allow mutiple choices, more than just WIMA and Rohm.

RC compensation network may be needed in parallel with R120 R121.
R = 6.6K
C = 1.8pF

No lowpass capacitor after L101, the output coil ?

F101 F102 rating ?
 
for JP 101 is something like a controller, that is under construcion, but in our case its not needed..
C111 (bias filter) 1µF optimal value ? Have you tried more ?---i didnt tried, i think that walue is optimal..
i dont know the value for the L101, but it will be winding on the R162, i will try to talk to the creators tomorrow..
F101-102 ---5.2 amps---just to be sure
 
Last edited:
details

hello,
F101,102 5.2 amps for the 150 w/4 ohm version
it work without JP101,102 wich are for a controller that is under construction..
that's what i was hoping to get help for: to find a solution that doesnt includes capacitors in the signal path
C111 (bias filter) 1µF optimal value ? Have you tried more ?---i didnt tried, i think that walue is within 0.1 and X uF, but it wouldn't have an influence in the sound ..

regards
sergiu
 
Last edited:
I cannot recommend to choose this mirror symmetric circuit for a homemade resp. diy amplifier.
In the schematic from post #1 and in the theory the negative half is a mirror image from the positive half. This means, that all even-numbered harmonic distortions goes to zero.
This isn't so in the real world, because PNP BjT haven't a mirror image character from NPN counterparts. Even the most professional engineers don't respect this. Additional there is the risc of unwanted oscillation through parasitic effects causes through spread capacity and inductitity of the PCB wires.
Most commercial products that I know, where the topology from post #1 is inside, don't sounds really better than products with an easier topoloy like that one from Mr. Nelson Pass.
For this reason I hate such topologies.

For perfect transmission you need a circlotron like "Sumo 9" - go to
http://www.diyaudio.com/forums/solid-state/23888-sumo-power-amp.html
or a SuSy topology (also much easier than that one from post #1).

Nearly the same audible results you will get with the german diy amp "Black devil" (very easy to build and very fine results by the use of a good power supply and high quality caps, eg from ftcap, Husum) go to post #13 about
http://www.diyaudio.com/forums/solid-state/154493-your-opinion-about-schematics-4.html
The whole paper you will find here:
http://www.eserviceinfo.com/downloadsm/56816/Elrad Black Devil_Black Devil Poweramp.html
Some years ago about
http://www.experience-electronics.de/englisch/index.htm
this amp was available as kit.

An additional solution for you are the "Circlophone" amplifier or the "Circlomos" from the member "ELVEE".
If you want to send me an PM, use my e-mail address and not the PM aera here on diyaudio:
kirschner-hifi@tiefbasswiedergabe.de
 
Last edited:
hello tiefbassuebertr,

why oh why discouraging people from trying all-symetrical amps like the Borbely-Lender types ? Look, it's not a Stochino arrangement. The VAS current actually gets limited by the long tail pair current. It's still on the safe side.

Okay, within Borbely-Lender types, there are "easy" subjects like the Maplin amp, more elaborate subjects like the Leach amp, and infamous subjects like the CFP-equipped Giesberts (Elektor) amps. Given that our friends have chosen modern & fast & linear output devices, only two in series, I think that if they oversize a little bit the miller capacitors (say only 33V/µs slew rate on 8 ohm resistive), they should encounter no stability problems. Do you agree ?

Perhaps we may advise them to build their amp in an incremental way. They may simplify the first assembly in order to acquire some experience, and while doing so, do measurements. The first assembly may omit the buffer before the VAS, may omit the VAS cascode, and may omit the overcurrent protection. Need to recalculate the collector resistors, of course.

At this stage they should take the time that's needed for simulating their amp using LTspiceIV, using the generic BJTs that are in the default library. That's a basic requirement.

Then, after they manage to have something stable both in simulation and in real world, they may introduce the buffer and again make measurements. Finally, they may introduce the VAS cascode, reduce the miller capacitors, add some subtle compensation in the differential collectors, target a 100V/µs slewrate (8 ohm resistive), and see if the distorsion, bandwidth and square wave on capacitive load are the best.
And as a last move, introduce the overcurrent protection and determine if it is ruining the performance (usually caused by ground pollution).

I don't know if their amp will deliver good subjective results. From what I have read here and there, Borbely-Lender structures tend to deliver a deceptive, dull sound. True or false ? What is your opinion ?

Now if on top of this, our friends want to relace the BJTs at the input by FETs, that's a completely different story. They'll need to cascode the FETs, complexify the PCB, reconsider the frequency compensation, and certainly enter trouble. A plain & simple Kuroda arrangement using an ultralow offset FET opamp would be more effective. By the way, tiefbassuebertr, did you manage to get a Kuroda running troublefree those last days ?

If our friends really want to suppress the caps (input cap and feedback cap), they should not waste their time. In such scenario, the most practical solution is the Kuroda. They'll be able to tailor the acoustic signature by 1) chosing the opamp brand and model, 2) defining the closed loop opamp voltage gain and load and 3) the VAS gain and max current. With the real estate that's made free on the PCB, they shall implement a non-switching class AB like Kawanabe (working for Pioneer) filed in feb 1979 under number US 4.254.379. And check if it is not ruining the THD performance.

Regards
Steph
 

Attachments

  • US4254379 Pioneer - Kawanabe Non-Switching (Feb. 1979).pdf
    151.3 KB · Views: 109
  • Pioneer Non-Switching.jpg
    Pioneer Non-Switching.jpg
    259.8 KB · Views: 392
Last edited:

Attachments

  • Borbely-Lender True Symmetry.jpg
    Borbely-Lender True Symmetry.jpg
    194.4 KB · Views: 367
Last edited:
This PSA 150S amp is a modernized SMD stereo version of the well known Leach amp.
Check it out here : Leach Amp Plans - Part 1
There is a LTspiceIV simulation file provided.

Why haven't you told us since the beginning ?

Are you realizing what you are actually doing ?
Renaming and rebranding a well known design.
Asking help for "improving" somebody else work.
How would you call this ?
Especially, less than one year after the originator died.

I urge you to take contact with :
Georgia Institute of Technology
School of Electrical and Computer Engineering
Atlanta, Georgia 30332-0250

There is an email provided on the webpage Leach Amp Plans - Part 1
Maybe somebody is still in charge.

Ask them if you can base on their design for deriving a modernized SMD stereo version of the well known Leach amp.
Ask them if they grant you the permission to name your design and PCB : "Buca-Leach 2011".
Ask them if they will claim monies for that, once you will make money.

Steph
 
Last edited:
hello tiefbassuebertr,
why oh why discouraging people from trying all-symetrical amps like the Borbely-Lender types ? Look, it's not a Stochino arrangement. The VAS current actually gets limited by the long tail pair current. It's still on the safe side.

Okay, within Borbely-Lender types, there are "easy" subjects like the Maplin amp, more elaborate subjects like the Leach amp, and infamous subjects like the CFP-equipped Giesberts (Elektor) amps. Given that our friends have chosen modern & fast & linear output ...............
Regards
Steph
My english isn't so good. Thus I don't have understand all your comments exactly. But please note: member "sergiu2009" want to realize his first diy amplifier project, but he don't want to realize his own developement ideas.
Independend of this - a great amount of commercial amp products from small manufacturers (so called "High End Audio" manufacturers) with similar topology in their amp devices like Leach as to see about
The Leach amp 200W amplifier
have problems in one way or another (mostly based of unwanted oscillation).
And the few Amps, which haven't such problems and which sounds very very good (e.g. RESTEK E3 monos - go to
Meine alten helden-thorens restek e3 - magnus.de Forum )
don't sound clearly better than more easy concepts, even if there are only a single ended input instead a LTP and no Hfe enhancer between the input and VAS stage (of course by checking with the same power supply and transformer). Only the THD results and perhaps the square-wave behaviour at various complex resistorloads could be better by use of Borberly/Leach topologies.
But the audible enhancement by listening tests isn't comprehensible (usually only a small alteration in one or the other kind). I guess, the main reason therefore is the too low idle current through the output (usually 20-50mA, i.e. Class AB).
In this mode, a certain quality level is not to exceed, independend of the topology from the front-end (but I want to know a Class AB amp topology with small quiescent current and with a sonic transmission character like a good SE pure class-A, but such a circuit must yet to be invented).

Read also Nelson Pass' articles carefully and you understand, what I mean.

An for me additional very interesting amp topology is the MJR amp about follow URLs:
Audio Amplifier Design
Renardson-Audio, Hi-Fi Audio Amplifier Designs.
and the topology from the Lavardin amplifiers
Memory Distortion Philosophies - Part 8 : More tests
Lavardin Technologies - distorsion de memoire

Hello,
First of all i want to thank you all for the fast responses and sugestions; i want to learn , and experiment the new/old, topologies/schematics from the diy audio, and i'm hoping that someone could help me to improve the sound quality and to eliminate the input capacitors from the design that i have posted..
I DON'T WANT TO SELL the amp, i want to make one (in stereo mode) for me (for some long night listening---:D); i have allready told you that it combines some of the topollogies from other schematics (only the good things) from the begining of the thread...
regards
sergiu
Which loudspeakers you have in use?
 
Last edited:
Hello,
First of all i want to thank you all for the fast responses and sugestions; i want to learn , and experiment the new/old, topologies/schematics from the diy audio, and i'm hoping that someone could help me to improve the sound quality and to eliminate the input capacitors from the design that i have posted..
I DON'T WANT TO SELL the amp, i want to make one (in stereo mode) for me (for some long night listening---:D); i have allready told you that it combines some of the topollogies from other schematics (only the good things) from the begining of the thread...
regards
sergiu
 
Terrific summary, Tief. Agree with everything you have said.
Do you have a 'Black Devil', which has attracted my attention also?
Hugh
My old device I have sold since a long time. It works with only 30 mA idle current and runs with 35 V supply voltage. The transformers for each channel was a version with 12VAC and 500VA (fully encapsulated, halogentransformer for 12V illumination stuff, two pieces/channel).
The cap unit consists of a Kendeil 10mF/63V, a choke (air coil, INTERTECHNIK) with 10mH and 1 ohms dc resistance and three additional Kendeil 10mF/63V.
Pics from the coil and caps you will find by this ZEN project there:
http://www.diyaudio.com/forums/pass...itional-op-amp-ultimate-sounding-phl1230.html
The supply parts by my old black devil are the same.
Only various Class-A amps had beat the 'Black Devil' in combination with this supply, despite caps in the input, the NFB-loop and the output in serial to the speaker (serial cap to the speaker is clearly preferable to a contact from a DC protect relais !!!)
For me this is a clearly evidence, that the audible influence from the circuit topology is much lower than generally assumed.

The German test magazine for recording studio equipment "Production Partner" (better than the most other for usually audio) also checks the internal resistance of the power supply for the amplifier test with tone bursts by loads from 16,8,4,2 and 1 ohm speaker impedance. Here the reader see a lot of effects, that isn't to see by the other magazines. Conclusion: Power supplies are mostly too small, even by various studio amplifier.
 
Last edited:
Good evening,
Mr steph, i hope you understand my intentions...i don't want to sabotage/improve/or modify the leach or other designs from whom where token the pieces from...
I want to make an amp that is compact and sounds beautiful, and i want to put it in the living room for long session listenings..
I want to remove the input cap because of what i have read from the diy forum, that it influences the sound, and, because here in Romania they are very expensive..
I have already ordered the components for this amp (because i want to know how it sounds), and believe that it can sound better, that it can be modified (the input) to be more stable, and to deliver "warmth" sounding, that's why i requested help from all our colleagues.. i want to modify it after the first listenings and benchmarks, so now i want to prepare myself for that..
please excuse my rough english
cheers
sergiu
 
Last edited:
At the core concepts of this project are more modern than Leach Amplifier which served as reference. For project development we inspired Leach Amplifier, Motorola Amplifier, Amplifier Crest, Dynacord, McIntosh Audio Bryston Audio and more ...
Books are written in books and not believe that someone should ask permission to develop an amplifier that has no "footprint".
Sorry for bad English, I translated with Google Translate.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.