Discrete buffers

[Ted205]

Originally posted by Nuuk


Faulty transistor(s)?

i changed the output cap from a 2.2uf to a 4.7uf and that has stop. So i have removed the LP filter.

All i'm getting now is a small pop/thump when i turn the amp on/off.

Would it be worth including a filter in your opinion?

[Nuuk]

Would it be worth including a filter in your opinion?

More important than my opinion is did you hear any difference when using the LP filter and taking it out?

[Christer]

Timpert, yes you do have a point. But on the other hand, is the small benefit of a factor 2 in Cgs and Early effect worth the extra transistors? And do we need more headroom than we get with the bootstrapped version? We could always use even higher rail voltage to overcome this problem. I already use +/- 18 V for JFETs alone, with no "cascodes" at all.

[timpert]

I think so. Not because the benefits are substantial (they are sublte, if audible at all), but because transistors are so cheap that it doesn't hurt to include them. On the other hand, I think that the headroom is not the main issue here, but well behaved clipping is.

[juma]

The 3 transistor buffer published on Nuuk's site is very simple circuit - an emiter-follower with 5 mA CCS. I experimented a lot with this buffer - different transistors, different CCSs (JFET, LED, 2 diodes...) and it sounded different (nuances) but it worked great every time. One really has to be a master of disaster to mess it up - I guess nothing is foolproof
There are better sounding buffer circuits, but this one can not be beaten for its simplicity/sound quality ratio. Only thing it may need in addition is an input cap (small input DC offset can make some pots crackle).
Nuuk, thanks for great circuit and all patient help you've been providing to less experienced DIYers

[Nuuk]

Thanks Juma, and a reminder again that we owe thanks for the circuit to Andrew Rothwell. As you say, it's a great circuit for trying out different transistors and learning how they sound.

I know we often want to remove any caps in the signal path, but I remember another buffer circuit that AR designed for me. That used DC blockers on the input and output and sounded superb so I don't think we should become too obsessed with avoiding caps when they are needed. In the case of the three-transistor buffer, one on the input will remove and noise caused by the preceding attenuator/pot!

[justblair]

Originally posted by juma
One really has to be a master of disaster to mess it up - I guess nothing is foolproof

Touche

And yes another big thankyou to Nick at DD for what is to my mind, the very best 'Clone' resource for beginners.

[Shoog]

My last remaining Gainclone which I use on my computer has this buffer on it. The reason its the last remaining gainclone I have is because it sounds really nice.
I had issues when building it. It kept on self destructing until I used some hefty TO220 power transistors. I have it on a seperate power supply. When I discovered +4V on the input I decided it needed an input blocking cap. This is the big down side with this amp as I don't like the fact that it has two caps in the signal path.

This amp is very easy listening with a smooth none fatiguing presentation. I attribute this at least in part to the sound of this buffer.

Shoog