[guzzler]

Has anyone any experience with this amp??

I've done a search of Headwize and here and not really turned up much about the sound of it... i was going to modify the circuit slightly to place the opamp into a multiloop mainly. I've already designed a digitally controlled power supply stage, and was perhaps thinking about going for a digital pot at the input, but i'd need to get hold of a microprocessor programming device and learn it, and i suppose i like the feel of a decent pot as well...

perhaps a compromise and get a motorised ALPS pot, for remote control later...

but my original query is about the SDS topology as a whole!

g

 

[Tomo]

Hey,

I think Mr. Stokes is around. He is probably working now, but he still likes fiddling with his DAC. Perhaps do a google search with his name.

Note that SDS amp is identical to PRR's Tori amp with constant current source loading a source follower. Which means that feedback is extremely important to DC bias at the output.

The fundamentally power mosfets and mosfets used in opamps and buffers are quite different. It is probably good idea to compare non-multiloop version and multi-loop version. According to what I have heard ... long ago, before META craze ... , SDS amp sound pretty warm and nice.

Lastly, (food for thoughts) SDS amp's mosfets have considerably higher gate capacitance (input capacitance) than BUF634. This means that the amp will be slower. This means that high frequency effect due to extreme gain reduction with strong feedback could already be reduced.

Tomo

P.S. Strong Feedback Effect can be favored. In fact, some people do it on purpose. It's all about your taste. Me? I don't use it.

 

[peranders]

Check also my QRV-02 headphone amp, the same idea. Works like clockwork.

 

[Tomo]

PerAnders:

I don't think so. Your amp is very different.

T

P.S. PerAnders - Have you tried Mr. Jorgensen's amps?

 

[aos]

SDS amp can sound pretty good. It's class A all the way - you can and should load the MOSFETS at 50 to 100mA idle current to put them in more linear region. This amp should be able to drive pretty much anything including small speakers. I don't see the point of multiloop since the MOSFETs are in fact your buffers. Also I cannot remember if this schematics already load the opamp into class A as well, otherwise it would probably benefit from adding that.

Choice of components is EXTREMELY important for sound. I found that 2SJ76 and 2SK213 complementary pair (there are 3 more variations) sound best here, and this is the first amp where I managed to hear the difference between resistors. Normally I can't but upon inserting Holcos instead of some generics, I got better sound.

 

[peranders]

Quote:

Originally posted by Tomo PerAnders: I don't think so. Your amp is very different.

Not VERY and if you change the BD139/BD140 to MOSFET's and increase the Vbe multiplier. Not that I have current sources instead of resistors for biasing the transistors. The pcb has also heatsinks so you can "kräma på" lot's of current (100-200 mA) in class A.
Quote:

Originally posted by Tomo PerAnders: P.S. PerAnders - Have you tried Mr. Jorgensen's amps?

No, but the idea is interesting though, original.

BTW: Have you checked out my real monster, QRV-01? The looks are very similar to Sheldon's amp but my is monster. Noone has dared to build it.....

I have tested with IRFD120 and IRFD9120, very nice performance. I have also as option room for PNP/NPN/N-Channel/P-Channel TO-126/TO-220 devices. I have tested with BD139/BD140, works allright also but I like MOSFET's more.

 

[Tomo]

PerAnders -

But one of the complementary pair of the SDS amp is DC coupled and the bottom mosfet is biased by 4 LEDs.

Your monster is little too much for me. I am still interested in Q-3 all the more after I heard about the Korean site.

But then, I am kinda feeling it is time for doing some weird stuff. I am thinking inverted push-pull ground driving amp. I have simulation. I just need time and cash. Waiting to settle down ... at least temporarily.

T

 

[peranders]

Quote:

Originally posted by Tomo But one of the complementary pair of the SDS amp is DC coupled and the bottom mosfet is biased by 4 LEDs.

If you really take a close look you will see the biasing scheme is more or less the same, 4 diodes instead of a Vbe multiplier.

If you want to build the QRV-03 you have to wait or make it by yourself, not so hard as you can see. I have made my prototype on a Vero-board, work good but a real pcb is better.

The component values are fairly OK, also according to our Korean friend. The only thing you have to take care of is the trimming of the current mirrors. I was planning to have a DC-servo for that.

 

[guzzler]

hi, thanks for all the replies...

I noticed that Intersil now make the RFP15N05 N-channel and the RFP15P05 MOSFETs, so I might just go for those as RS is really my only source of slightly esoteric components...

i'm going to redo the circuit layout to cut out the original power supply section, and get it all properly made up by Olimex PCB prototyping, really cheap: $21 for SS 160x100mm, $26 for DS 160x100mm and then only $5 for shipping, and they do silkscreen, solder mask, and tinning and accept directly from Eagle which is always nice cos I can't afford any Gerber tools

http://run.to/pcb

g

 

[rickcr42]

Sheldons new wb address

http://www.quadesl.com/index.shtml

 

[peranders]

Quote:

Originally posted by rickcr42 Sheldons new wb address http://www.quadesl.com/index.shtml

The address isn't very new but Sheldon is linked in many places and few of them has his new address.

It's a shame that noone can update links here but Jason & Co is going to fix this. It's a hard job to update link lists. I supplied Steve Ekblad with a file of incorrect links. 25% of his links are faulty.