Simple Symetrical Amplifier

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

http://www.diyaudio.com/forums/soli...x-gain-mosfet-output-stage-6.html#post2659204

In this stage with minor modification it performa very well. In fact it has top and midrange is better than a burmester integrated. The bottom lacks a bit compared to burmester. But it is only a 50watt design with a single 300va for both channels compared to a 100watt with a larger powersupply, also the bottom has the same signature as Many mosfet based amps.

The top and midrange is very detailed and in the same time not harsh. I Think mainly due to the 400ma bias current.

This amp is the mark for comparison. This AMP is also being updated, and will get better.

my own designs Will follow the path from the mirand A1 to get this design more purified. The AMP i am doing for another person is made on the basis from the AMP in this thread, not my mirand AMP.

I Will try to compare Them but it Will maybe be in favor of my own designs. Such comparison should not be done vy the designer himself.
 
diyAudio Chief Moderator
Joined 2002
Paid Member
I will put 1k trimmer for the output stage bias resistor in the one with the CCS patches I am preparing to test because I smell there will be a few Iq favorites to listen to (if the rest will go well, first things first). 100mA looks interesting in LT SPICE at least.
 
Could you explain the subjective sound when you achieved 1000V / uS slew.
I would also strive for high slew rate but in practice I have found that slower can sound more pleasant and realistic in some ways.
It may seem curious, but the more noticeable improvement is in the basses. Specially kick drums. More solid, less "one note", more fast and detailed. Separation is better and treble less aggressive (you can feel you have lost some of them). Of course all depend of the source. You can discover some saturation and clipping in CDs that you never noticed before, and an aggressive recording will sound more aggressive, while a good one softer. By contrast.
You can feel you have lost some power too (less distortion at high level ?).
Of course, there is not a huge improvement in my amp, as it was quite fast before (~200V), and the incoming signals where yet kept slower than his capability. And it is difficult to say what is the part of the speed and the one of the reduced distortion SSA had brought. Thanks to Andrej's talent
Last, i would like to precise that my loudspeakers slows down after 16KHz, i do not like to reproduce trebles in excess, often added in mixes, and i'm more aware about dynamic and energy (they are JBL 1"drivers+spherical horn in wood, cut at 1000Hz). And i do not have my young ears anymore too. But I believe there is not so much interest to increase slewrate after something like 500V/µs, while it cannot be bad and it is fun.

Now, if you like to reduce the bandwidth, and i know some CD with imperfect filters can be better that way, a low pass filter before the input will do it in a linear way, with coherency between phase, bandwidth and global slewrate, and will reduce further the chances of any TIM.

What you did to Crescendo is just simply amazing, I am impressed.
I'm impressed too by the results. But the idea comes from you you. Did-you amaze yourself ;-)
 
Last edited:
Thanks Christophe for your comprehensive reply to my question.

From your description I suspect that as well the increase in slew rate, the other improvements you made significantly contributed to the increase in the sound quality and it sounds like you have a very nice amp - thx for sharing :)

mike
 
I have packed the prototype today.

I had it on my system shortly -

MiniPC + itunes + M2Tech Young DAC - Amplifier (DUT) - Audiovector M1 Super.
The speakers needs to be upgraded but they are not bad bad...

I tested it on different types of music for 30minutes before packing it down.

The amp is not bad. You will very quickly notice if it dull to listen to.
It is musical no doubt about that.

There is less edge on the transient than on my CFB + EXICON. The detail level is good.
It has more details in the bottom, but i have made a fault on my CFB + EXICON.

The SSA and the CFB + EXICON has nearly the same bandwith of approx. 600KHz.

The servo in my amp has a minus 3dB point of 3.3Hz. On the Mirand AMP it is set too 0.3Hz (And this was stated to be bottomless on the hififorum.nu)

This this SSA it is set to 1Hz. It is in the Middle.... It the bottom detail level is also in the middle between those two.

The Mirand A1 AMP is more stable regarding DC offset than the SSA without SERVO, there the SERVO is only used for removing DC from a Source.

Regarding the SSA it is also there to control the "wild" swings. But at present time with DC servo the out floats below 1mV... So thats is okay. I could maybe lower DC invoke frequency, but that will time tell..

The Facit:

For the sound it is worth building - And as allways it is matter of taste... But i think most people would be very satisfied.

My opinion - There is still a lot of work to do on it! Many things need to be solved....., but the sound is worth it...

I will comment more on other amps, and in comparison. I have some sets of amps.. Mirand A1 V1.1, SITO Audio CFB + EXICON V1.1, FETZILLA and SITO/MIRAND A2 V1.1 (Next weeks)..

If the SSA was AC coupled with lower rail voltage it would make a PREAMP. There is no HISS...
 
SSA BIGBT High Performance

Going to build this one. :)
 

Attachments

  • SSA BIGBT HP.jpg
    SSA BIGBT HP.jpg
    721.4 KB · Views: 732
Excellent on 1600x1200 notebook

Why yours circuit diagrams are always so big?? Not easy to check on line, need to copy first and this is not practical.
dado
@ dadod,

:) They are excellently fit on the 1600x1200 screen res. before click on to left lover corner X to expand it to see it in a full resolution (Opera), maybe Your brovser automatically expand the img file to full resolution of 3.508 x 2.337 pixels @ 24 bits but then again perfectly fit on 1900x1200 screen res. :)

@ LC

:cheers: LC :cheers: for the HiQ of Yours SSA BiGBT version.
You perhaps already listen it for a while.
Could You pass some of Yours listening impressions for this HiRes HiPower SSA amp?
What speakers You drive with it or You maybe use a multi amp arrangement ?

- Would it'll be necessary to change much of the values for the IP and OP stage if the voltage
for the OP Stage would be + & - 50V DC (2x45V AC - CM500W trafo) and separate 60V DC for the IP stage ? :scratch1:

Kind regards
 
SSA BIGBT High Performance

OK guys this one has also on board remote on/off capability, so no output relay is needed. :)

I would really like to make this amp on a compact double-sided professional PCB not bigger than standard euro format. ;)

I hope the schematic is OK and now I am in a waiting mode for ... :wiz:
 

Attachments

  • SSA BIGBT HP.jpg
    SSA BIGBT HP.jpg
    762.7 KB · Views: 680
LC:cheers: LC :cheers: for the HiQ of Yours SSA BiGBT version.
You perhaps already listen it for a while.
Could You pass some of Yours listening impressions for this HiRes HiPower SSA amp?
What speakers You drive with it or You maybe use a multi amp arrangement ?

- Would it'll be necessary to change much of the values for the IP and OP stage if the voltage
for the OP Stage would be + & - 50V DC (2x45V AC - CM500W trafo) and separate 60V DC for the IP stage ? :scratch1:

Kind regards

Yes, until now I have a plenty of feedback info from my system and quite so from a fellow forum members. That is sufficient for me to start planning one PCB according to my wishes. It will be pain in ... to make it but at the end PCB will probably ended in a group-buy. Will see hehe :cool:

If there will be wishes like yours to adopt SSA BIGBT HP to different supply voltages, I will prepare all the necesary data, it is really nothing special, only a few resistors have to be changed accordingly. ;)

Cheers, Andrej :cheers:
 
PCB will probably ended in a group-buy. Will see hehe :cool:

If there will be wishes like yours to adopt SSA BIGBT HP to different supply voltages, I will prepare all the necesary data, it is really nothing special, only a few resistors have to be changed accordingly. ;)

Cheers, Andrej :cheers:

Excellent :) Andrej, Yes please to the 60/50 V DC - Balanced - :cheers:

Kind regards.
Andrew
 
Last edited:
diyAudio Chief Moderator
Joined 2002
Paid Member
CCS patching update

I made a second SSA Sziklai Latfet channel, this time with CCS current feeders to first stage as I was planning to test. Good news, the offset can remain between -25mV to +25mV in operation. I employed the attached schematic. Started with R8,R9 1K. Alas, the wanting 2SC2240/2SA970 models overestimated currents for .op LTSpice analysis, and I had to patch 1K additionally for the drivers to turn on. So 2K it is, as drawn. R14 is 1K trimmer. I1 is a 7mA IDSS BF245B, I2 is a 10mA IDSS BF245B with a 100R trimmer between its S&G pins.
Output stage current got easily set with R14 and stable. Tried 110mA and 170mA. J1 is 2SK30ATM(Y) Toshiba 2mA IDSS sample. It can even be a lower pinch off type non matched between channels when having R14 a trimmer. Low noise and high DC stability JFETs are needed.
The test circuit starts with a momentarily 200mV DCout shot when plugged to power, then goes low fast. It can be fixed at about +10 to +25mV by varying I2's 100R trimmer a bit. With sine signal it shows -25mV steady DCout. When the plug is pulled, the supplies are getting drained unequally and the DC can momentarily shoot again, but under 1V tops. The gain I measured is 27.6dB. -3dB is 1.2MHz. No instability. FFT on scope eyeballs about 0.02% THD at 1W/8R with H2 leading, but it can be the old function generator lending that profile. The bandwidth got 20% higher since the last experiment with resistors as current sources. This maybe explained by 8dB more feedback available in the analysis because of the current sources giving 5dB more OLG and ~3dB less gain used in this one. My beautiful square wave rising corner got a bit peaky now though, but this is secondary to tune when packaged because it may change then. Still some time before really listening to it as it needs modding the other channel to match, a dirt cheap Epay delay and DC sense relay guard to arrive, and some putting it together to resemble an amp somehow.
 

Attachments

  • SSAccs.gif
    SSAccs.gif
    24 KB · Views: 608
  • SSAperf.JPG
    SSAperf.JPG
    196.6 KB · Views: 548
  • SSAon.JPG
    SSAon.JPG
    150.7 KB · Views: 473
  • SSAoffsetNoSignal.JPG
    SSAoffsetNoSignal.JPG
    338 KB · Views: 283
  • SSAoffsetOnSignal.JPG
    SSAoffsetOnSignal.JPG
    341.2 KB · Views: 269
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.