as I am in love in the triplet of the transistors that peufeu has discovered, please try this single ended design.
To minimise "memory distortion" (or as do we wish to call them?) the transistors work at constant power.
The gain is slightly less than 10 times.
The sound is very smooth and detailed.
Please extremely care to use good quaality capacitors.
You can use positive rail up to 30V and negative only -5V as well.
This is the schematic of working circuit.
I have built the 3 bjt discrete buffer this thread refers to (from Decibel Dungeon). I am very happy with it and have made my impression known to DD.
Carlmart says these buffers can easily be configured for some gain. Can he, or anyone else, please explain to me how to do this?
Or maybe it would be better to build a preceding gain stage, or even to build a new pre, configured for gain from the start?
Thanks for showing interest. I was referring to the 3 bjt buffer found on Decibel Dungeon's website as referred to by Carlmart. (Found at this address:-http://myweb.tiscali.co.uk/nuukspot/decdun/gainclone2.html)
I have built it with 22uF Black Gate output caps and a different selection of transistors (by trial and listening) and find it simply outstanding.
I would also be very interested indeed to try your design - in fact it looks like the perfect solution for me - but I cannot read the drawing clearly, and also I am not experienced enough to be able to calculate correct values for use with my +/- 16.8V power supply. If you could post a clearer drawing, that would be much appreciated, and any help in how to work out best values would be most welcome.
Right, I built it. It was a bit of a disappointment. Not as much detail as I had with just the buffer, not as much bass control, although there was an improvement in dynamics. The big problem however is that it starts to clip at about 2/3 volume (not very loud, BTW) and distortion sets in from before that point (probably why the detail is muddied). I tried feeding my buffer from it with sonically pleasing results, but the clipping and the distortion were not in any way improved. I tried bigger values of input resistor, but of course all that meant was that I had to open my attenuator up more to get back to the same output volume and clipping set in at about the same loudness as before.
I suspect that the input impedance of this stage is not high enough for my DAC to drive comfortably. I don't know how to alter this, apart from building a completely different design (possibly something with a J-fet input instead of a BJT.)
In the meantime, I have built an opamp input/gain stage (OPA637BP) and it is feeding my buffer. After some tweaking, it is sounding pretty good, but I would still like to build a solid state input/gain stage. Any suggestions anyone?
I am feeding from a stepped attenuator (50k I think - can't remember - so long since I built it. I don't think there is any problem there, since the volume is correct and also I have tried a number of pre-amps in exactly the same way. From the pre, the output goes to my active crossover.
I don't think the output impedence is a factor, since I am connecting it to a buffer (with output cap to block DC). I can't see any mistakes in my build, but then perhaps there is something I didn't pick up in the (rather poor quality) picture. I would be very happy if it is something like that. I will draw it for you exactly as I built it, but a bit later. (Right now I got to go do some work.)
thanx for the support.