DC adjustable double-Sziklai Power Diamond Buffer

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I have designed another 0 dB Diamond Buffer Power Amp.

I use this Power Buffer with my SRPP+Cathode Follower TUBE E-80
pre-amp to make a pretty good sound Non-Global-Feedback Amp.

With some changes from my ULTIMA-SDDB small Class-A power Amp and headphone amp ,
see my last thread at sziklai-darlington diamond buffer (SDDB) for power output stages ,
I desided to design this 0 dB Power Amp with double-Sziklai output stage.

The INPUT stage is similar to LH0033.PDF fast buffer, but with cascode
stage for JFET high voltage protect. So the output DC level can be
adjustable by tunning the CCS current.

About 40W with my tube pre-amp the output power can be achieved. ( not yet totally tested )
The bottleneck is the 2SK246BL's limited breakdown voltage only
50V, if use LS844 it will be better 60V.

An externally hosted image should be here but it was not working when we last tested it.
My Taiwan made E-80 tube amp.
 

Attachments

  • ultima-double-sziklai-power-db-v1.0.pdf
    54.5 KB · Views: 3,333
hi, Greg,

I think MOS is not suitable for the input FET here.

DMOS FETs are Enhancement type FETs, the Vgs will be positive
voltage from gate to source.
But JFETs are depletion type, the Vgs will be NEGTIVE voltage
from gate to source.

The source voltage of the Input JFET Q8 will higher than the input
(gate) voltage, if we set it to be Vbe, then we can get pure DC
output Buffer without using capacitors or DC-servo ckt.


hi, skychu,

I have a DZ T270 27-0-27 power transformer in my hand, so I use
this to get a +/-38V DC.
But I lower the voltage of the cascode Q7 (Q12) voltage to about
25V, and providing the Q14 (Q15) CCS sufficient voltage.

The voltage of the output tage can be lower to +/-25V,
only I have no other AC output voltage.
 
Hi Greg,

I think my amplifiers no matter with amplitude or not, basically are
designed with simple structure , though look like complex circuit.

Since my first My Ultima DB-01 Dual JFET+DB Head-Amp announced
on Head-Fi at 10-18 2003, I began to design with the same structure
using the same cascode JFET differential input stage, plus single-
ended JFET+CCS voltage amplitude stage, I want to design a high voltage gain
and large current output, so they can be pre-amp headphone amp
as well as a power amp ---- my 150W ULTIMA-1 .


And I love the diamond buffer and JFET's sound .

But they are all Global-Feedback designed, since I bought my SRPP
tube pre-amp, I think my power amp maybe can eliminate the
differential input stage and voltage gain stage, just use a SIMPLE
common drain JFET stage as input buffer plus a SIMPLE Power
Diamond Buffer
to drive not only the headphone but also the
speakers.


True, it came to be the same complex as my first amp, but I use
only one pair SANKEN 2SC2922/2SA1216, to have large cuurent
gain. And some circuit are for remote stand-by control optional.
LEDs are for CCS and power on indicators.

All BJTs are very cheap small signal except the driver and output
stage, total cost is not very high.

NGF 0dB Amp without the no-useful or no-benefit voltage ampltude
stage maybe can achieve better noise, THD, phase-shift, ...

I think it can sound more elegantly with the tube, but without
output transformer. :)
 
hi Kanwar,

Thank you for your appreciation.

I only know very few of my headphone amp have been sold or
made out of Taiwan, and also not many local diyers have finished
the whole set, because I can not support the parts especially the JFETs.
And one of your friend had done it , it is great! :)

You can see one of my friend alin even use a paper box as his headphone
amp's box here.
alin's DB-01 headphone amp.
 
Hi mtlin,

Yeah, in India jfets are hard to get, but my friend salvage the jfets from old japanese preamplifier.......
Taiwan has bigger semiconductor and electronic industry access than India and you guys have much greater variety of components to choose from and that a good fact....We have to import semiconductors from other countries to fulfill our requirements...

regards,
Kanwar
 
Hi Kanwar,

I hold an ASIC design house for 11 years, and had been merged
by another bigger local design company since 2000.

In Taiwan, because of UMC and TSMC, they are all CMOS everywhere,
some may have DMOS or even BJT process, but no JFETs.

We have to buy them from Japan or U.S.,too. I bought the
TOSHIBA 2SK246BL even from U.S. for NOS not from Japan.

I have finished my 0dB power DB for both channels but no case now.
The dynamic range is HUGE, my tube pre's VR is set at 3
o'clock (maxium 5). :bigeyes:

The picture of the system is here.
 

Attachments

  • kif_0003-s.jpg
    kif_0003-s.jpg
    79.7 KB · Views: 1,635

An externally hosted image should be here but it was not working when we last tested it.


0-dB Diamond Buffer for headphone using DT-01 PCB


An externally hosted image should be here but it was not working when we last tested it.


Compare with modified normal DB-01 headphone Amp ( RIGHT )
0-dB Diamond Buffer has no 1st differential input stage and 2nd VAS,
If add a pair of Power BJT like SANKEN 2SC2922/2SA1216 then it
will be a 0dB-Power-DB for speakers or DB-01 to be a Power Amp.


An externally hosted image should be here but it was not working when we last tested it.


Plugging this DT-01 0dB DB PCB into the DT-02 Power BJT & protector-ckt PCB then


An externally hosted image should be here but it was not working when we last tested it.


it will become one channel of my ULTIMA-0dB-Power-DB ! :D
 
An externally hosted image should be here but it was not working when we last tested it.

[url]http://www.tubecity.com.tw/[/URL]




An externally hosted image should be here but it was not working when we last tested it.


My E-80 SRPP+Cathode Follower Pre-Amp can supply very high
voltage output. E-80 use 4*Mullar EC90 and 2*MiniWatt E80CC.



An externally hosted image should be here but it was not working when we last tested it.


I like not only the amp's structure, but also its all tubes regulator at
B+ 280V. The tubes are EZ81¡B6RA2¡B6AU6¡B5651.


An externally hosted image should be here but it was not working when we last tested it.


These are my audio equipments, the E-80 pre-amp, the marantz
DV-8300 SACD player, the B&W P6 speakers, and the 0dB POWER AMP.


An externally hosted image should be here but it was not working when we last tested it.


I have done myself other things like NIXIE TUBE CLOCK, DZ CD-PRO2,
and my 150W*2 ULTIMA-POWER AMP.

:)
 

An externally hosted image should be here but it was not working when we last tested it.


0-dB Diamond Buffer at 2K Hz square wave 10V p-p RL=8 £[.


An externally hosted image should be here but it was not working when we last tested it.


0-dB Diamond Buffer at 20K Hz square wave 10V p-p RL=8 £[.



An externally hosted image should be here but it was not working when we last tested it.


No matter 20K Hz or 200K Hz, there is almost no phase shift.


An externally hosted image should be here but it was not working when we last tested it.


0-dB Diamond Buffer at 100K Hz sine wave 8V p-p RL=8 £[.


An externally hosted image should be here but it was not working when we last tested it.


0-dB Diamond Buffer at 1M Hz sine wave RL=8 £[, you can
see there is a little voltage gain drop, and a not-so-smooth
sine wave.


ULTIMA-0dB is truly a FAST SPEED POWER BUFFER! :)
 
Workhorse said:
Good SineWaves! , no phase shift , too good for a buffer, yet excellent....

I know it is only a buffer, but it needs to be a very good power buffer.

Because I want to join with a TUBE amplifier to be a good hybrid
amp. One of my friends , Dr. SteveSun said in Chinese, "In theory,
tube can be a very linear voltage amplifier, and BJT can be a very
linear current amplifier, its F.R. can be very wideband, and DB is a
very good structure, putting two tigether would be a nice choice. ..."

But he prefers differential mu-follower structure, rather than SRPP,
I know very little about tibes or differential mu-follower ,
but I truly believe that tube+Power-DB can be a very good sounding
amplifier, like PATHOS INPOL one of the Italian Amp maker's patent says,
"But it's a hybrid!", you say. Yes, apparently. The big difference is the fact that does not introduce any manipulation of the audio signal. In order to express the idea better, INPOL
works like a pantograph, copying the signal absolutely faithfully.
 
Hi,

I forgot to measure the output impedance of this DB.

I use my two 4 ohm DALE power resistors as load, and a function
G. with a multimeter, and roughly test it at 1KHz 1V output. Then
after calculating, it is about 0.5 OHM.

Is it low enough for a normal 8 OHM speaker?

Should I lower the Re from 0.22 to 0.1 ohm?

thanks,

mtlin12
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.