Ultra-Hi-Res headamp

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi all!
I design amplifiers high resolution with great depth of feedback. This is one of them.

The input stage VT1VT8 performed in a differential mode. VT3 VT7 is the source followers, and at the same time - stage with a common gate. This gives input cascade ability to work with almost no distortion from a signal source with high internal resistance, for example, directly from the volume control with a large output resistance.
Not all amplifiers can do this without introducing distortion. The amplifier is tested typically by a signal generator with low output resistance, and shows good results, and how it works from sources with high output resistance remains unknown to the developer.

The input stage thanks to a high-impedance load VT6-VT10 has a gain of 80 dB at frequencies from 20 Hz to 1 kHz, and 50 dB at the frequency 20 kHz. Practically, it is a gain limit for these stages.

The feature of this cascade is that VT1VT8 can be either JFET or BJT without additional schema changes.

The second stage built on VТ14 under the scheme with a common base and dynamic load VТ16. VТ10-VТ11 drive this cascade through the resistor R16, and its dynamic load - through VТ12 and current mirror VТ13-VТ16.

Why is it so?
Cascade with CB, due to the lack of connection input and output using the Ccb the most stable against the generation of a cascade of voltage gain. A nice bonus is that the amplifiers with such cascades it is easy to provide frequency compensation.

The second cascade up to a frequency of 20 kHz gives a 60 dB gain. Both cascade together, the first and second, have a gain up to 140 dB at 1 kHz and 110 dB at the frequency 20 kHz. Therefore, the depth of the NFB - 120 dB from 20 Hz to 1 kHz and 95 dB at 20 kHz. Such a large depth feedback allows to obtain a very high linearity of the amplifier. The output stage is built on the good old scheme of three-stage emitter follower and operates in class A.

OpAmp DA1 is used in the service node for stabilization regimes of the amplifier. This OpAmp can not install.





The first versions of the amplifier

You can order the PCB for this amplifier here: aleanu@gmail.com.
 

Attachments

  • UNIKUM-2 JFET.GIF
    UNIKUM-2 JFET.GIF
    85.4 KB · Views: 1,455
  • UNIKUM-2 bjt.GIF
    UNIKUM-2 bjt.GIF
    85.6 KB · Views: 925
  • UNIKUM-2 Loop gain.JPG
    UNIKUM-2 Loop gain.JPG
    83 KB · Views: 686
  • UNUKUM-2 test 19-20 kHz.JPG
    UNUKUM-2 test 19-20 kHz.JPG
    71.1 KB · Views: 387
  • 16300d624a113b9a3dfb8bffcebebc57.jpg
    16300d624a113b9a3dfb8bffcebebc57.jpg
    277.6 KB · Views: 369
  • Спектр ИМИ в тес&#.jpg
    Спектр ИМИ в тес&#.jpg
    277.6 KB · Views: 294
Last edited:
Hi, Sergey Bandera!
I'm glad you're looking for knowledge wherever you can!
We have enough high-tech to present it to the world. See here: Several schemes.
Off: No one can accuse us of not inhospitable: our prisons are always open to fascists, who shot down the Malaysian "Boeing" over the Donbas and shoot from the tank on the children.
 
Last edited:
The loop gain looks kind of weird. Are you sure it works?
Hi, Nickolay!
Yes, of course, this amp works. It is built in many exemplars, his scheme published in our journal, which is fairly competent editor Nikolai Sukhov. Not so hard to model it and work to make sure the advantages of such circuitry. A graph of loop gain you think it's strange because the amplifier is compensated for by the stability criteria of the Harry Nyquist.
 

Attachments

  • UNIKUM-1.GIF
    UNIKUM-1.GIF
    129.6 KB · Views: 466
Last edited:
Glad to see that nothing has changed and you are as insane as you always have been
Before us is a crazy idea. The question is whether it is crazy enough to work. Niels Bohr.
I'm glad you haven't changed, old friend. You're the same clown, and joking and laughing as Yorick from Hamlet. :)
Oh, I see you found a decent auditory over there :)
Yes, there is a good audience.
 
Last edited:
Just providing an email address to order your boards is pretty vague.

People would want to know the cost of the boards and whether they would need 2 boards or one board for 2 channels.

What's the deal with C20? Is it 22pF or 22uF and is it used in parallel with R23 or not? I see a dotted line connecting it to R23 in your schematic above.

What is the gain of the circuit and which 2 resistors set the gain?

Thanks...
 
Hi, ammel68!
Just providing an email address to order your boards is pretty vague. People would want to know the cost of the boards and whether they would need 2 boards or one board for 2 channels.
I told him, he registered on this forum and will respond personally.
Two channels of the amp need 2 of the Board + protection from DC (I always protect the headphones, which can be much more expensive, than the amplifier).

What's the deal with C20? Is it 22pF or 22uF and is it used in parallel with R23 or not? I see a dotted line connecting it to R23 in your schematic above.
This question relates to which of the diagram? According to my diagrams at the beginning of the topic, C20 - 1000,0.

We denote microfarad with a comma for brevity: 1,0 = 1 µf, picofarad not designated: 220 = 220 pF, 5600 = 5600 pF. Similarly, the megohms indicated by the comma: 1,0 = 1 MOhms, and ohms is not indicated: 47 = 47 Ohms.
What is the gain of the circuit and which 2 resistors set the gain?
Gain amplifier - 2 times. This gain is set by resistors R13 and R11. If for high impedance headphone will need more voltage output and need more gain, reduce the resistance of R11 unchanged in the circuit.
Thanks...
Please. I will be glad to help.
 
Last edited:
Thіs thrеad bеlongs to vеndor's bazаar.

Вut i hаve tо аdmit, kіds nоwadays dо hаve а lоt оf things tо plаy with - thеse simulаtions , CAD softwаre, FFTs. Whеn i wаs of thеir аge, аll і hаd wаs a pеncil аnd sоldering irоn.
Thumbs uр fоr thе kids еxploring thе wоndrous wоrld оf еlectronics!


BTW, whу dо уou prеfer blаckman over hаnn?
 
Last edited:
Hi, s3tup!
Thank you for very wise words. You were out here in may 2015. Where have you been, old friend? They say you very sick. You have recovered, or still sick?
How do you like this amplifier?
Thank you for spending your precious time on my topic. I will be glad to see your always so competent tips to improve my amps.
 
Last edited:
Hеllо mу dеаr,
I'm glаd уоu'vе kеpt trасk оn mу аppаrеnt dіsаppеаrаnсе. Dоn't lеt thеіr wоrds tо mіsguіdе уоu, і hаvеn't bееn sісk fоr а whіlе. Bу thе wау, whо аrе thеу whо sаіd suсh thіngs tо уоu?
I аm hаppу tо sее уоu соmіng bасk frоm wаr wіth pоlіtеnеss аnd pеасеfulnеss, sееms уоu hаvе fоund оut hеr truе fасе.
As fаr аs і саn sее (аnd thіs іs nоt tоо fаr, gіvеn thе sіzе оf thе pісturеs уоu hаvе pоstеd іn thе stаrtіng mеssаgе hеrе), nоthіng hаs сhаngеd іn thе lауоut dеpаrtmеnt, whеrе thе mоst prесіоus sсhеmаtісs fасе thе rеаlіtу оf bеіng stuсk іntо 2-dіmеnsіоnаl PCB wіth іmpеrfесt соmpоnеnts аnd еnоrmоus lооps оf trасеs. Lеt аlоnе thеsе hugе rеsіstоrs уоu usе. Yоu'd bеttеr lау thеm оut уоursеlf tо bеttеr knоw thе іnсоnvеnіеnсе оf сrоss-соnnесtеd sсhеmаtісs, аs wеll аs bulkу сhаіns оf thіngs оvеr thе rаіls. Thіs mіght gіvе уоu thе rіght іnsіght іntо асtuаl wоrld whеrе thе sсhеmаtісs оught tо bе еmplоуеdת in vеrу sаmе mаnnеr аs уоu hаvе fасеd thе rеаl humаn spесіеs іn thе tіmе оf wаr.

Sincerely yours,
s3t.
 
T117, thank you for answering my questions. I did receive a PM from Alexandr about the boards.

Looking at your diagram in your first post, what is the purpose of L1 and what value is it?
Is it a through-hole inductor you can purchase or does it have to be wound from insulated wire?

The 2SC2785/A1175 aren't easy to find here in the U.S.. I see some on eBay but have no idea if they are real or fakes.
Are these transistors common in your country?
 
T117, thank you for answering my questions.
Please.
Looking at your diagram in your first post, what is the purpose of L1 and what value is it?
I was interested in how much gain can give two cascades VAS, and how deep the NFB can be obtained. As can be seen from the graph, the audio frequency range in such a simple amplifier can be quite easy to obtain the depth NFB 120...95 dB.
But the band steps NFB is limited to 10 MHz. If I would use standard Miller compensation with a decline rate of 20 dB/decade, from frequencies 1 kHz to 10 MHz only 4 decades, and would have to sacrifice the depth of NFB, displacing the first pole lower in frequency. Therefore, it was necessary to use three-pole frequency compensation with a plot twice more rapid decline in frequency response from 100 kHz to 1 MHz.
The inductor L1 100 uH together with R4 and R8 forms the third pole of the frequency response at the frequency of 1.5 MHz, because these elements create a local NFB in the input stage.
Is it a through-hole inductor you can purchase or does it have to be wound from insulated wire?
I bought this inductor. It is better to use inductors with a closed magnetic system to avoid electromagnetic interference: toroids or pot-core type.
The 2SC2785/A1175 aren't easy to find here in the U.S.. Are these transistors common in your country?
Just found here are matching transistors. Model in LTspice IV contain 2N3904 and 2N3906, so you can build an amplifier with them.
I see some on eBay but have no idea if they are real or fakes.
We are in exactly the same situation with these Chinese. :)
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.