Which output buffer?

[ninjaaron]

Hey, I want to put an output buffer after an effect I'm tinkering with. It's a JFET faux tube thing (not emulating any particular amp, just trying to make some good sounds). I want a buffer so I will have low impedance output regardless of where the volume is set. I don't want any distortion from the buffer, so I suppose I will have to positive bias the input voltage regardless of what kind of buffer I use.

Right now I'm planing on using a TS buffer with a 2N3409. However, I also have some TL071s, and that is an option as well... or should I rather just use a JFET (MPF102 or J201), regardless of the higher ouptut impedance?

I have some other BJTs as well, 2N2222, and 2N4401. I doubt they would make much difference opposed to the 2N3409, but if you know something I don't, I welcome any advice.

[Ben N]

I don't know why people feel compelled to use jfet buffers with jfet circuits. Seems to me you ought to do fine with the TS-style buffer, with any of the transistors. If you are going to go with an opamp, a 5534 would be better than a TL071.

[Gus]

Why would a 5534 be a better buffer than a TL071?

   In this case and other cases it depends.  A TL071 will do fine as a buffer at 9VDC levels and  driving a load that is current source and sink output is not exceeded.  A 5534 has a lower input resistance than a FET input TL071.

[ninjaaron]

I don't know why people feel compelled to use jfet buffers with jfet circuits.

I don't really understand it either, but I'm too much of a n00b to know if there is a good reason to use it on the output (where as I am not so much a n00b that I don't know why it might be better to use it on the input). I'll take your word for it and continue under my previous assumption that there is no good reason to use a JFET here.

Seems to me you ought to do fine with the TS-style buffer, with any of the transistors. If you are going to go with an opamp, a 5534 would be better than a TL071.

Ehm... is there any reason to prefer opamp to a transistor in this case?

Anyway, the only single opamps I have right now are TL071 and LM741 (which I presume is an inferior due to its lo-fi reputation). I also have a few 4559 opamps that could work well, I guess (I just bought them because I saw that it's what Paul C. is using in the Timmy, so I figured it must be pretty legit.).

To Gus:
I don't know why a 5534 would be better either (never used one for anything), but I can't say that I see how a FET input presents a great advantage at the end of the signal path. 


I don't really know too much about the differences among different kinds of opamps, and I also don't really know how to find out other than reading lots of data sheets. n00bdom is tough.

[Ben N]

To respond to Gus, the reasons I suggested a 5534 was (a) coming after the jfet-based circuit, I assumed that its low input impedance would not be an issue and (b) it would therefore be lower noise, and (c) it has the lowest output z of any of the usual suspects, so I assume it would best serve the goal of predictable behavior with any following circuit, assuming (a) & (b). Probably makes very little difference, though, and anytime you here from Gus & me on the same issue, just assume I'm babbling--go with Gus.

As for whether an opamp buffer is preferable to a transistor here, within my limited knowledge the answer is probably not. Most of the circuits I've seen that use the opamp (at least at the output) are using half of a dual (like those 4559s of yours) that happens to be left over in their design--might as well use it. I suppose the only way to know for sure is to try them both. But with unity gain, none of these devices is likely to alter your signal very much.

[alanlan]

That's true.  In commercial designs, every component must justify its existence and it would be crazy to use a discrete emitter follower if there is a spare op-amp stage handy.  You will see quite often that where they have used all the op-amp stages a discrete follower will be used instead.

[ninjaaron]

Thanks guys.

I tried the TS buffer (emitter follower) and it worked... eventually. I didn't have any 10k resistors (which is what the TS voltage divider uses, as well as on the shunt to ground), so I used 15k for everything. At first there was some distortion and then sound died altogether. Then I tried another transistor (of the same kind), and it worked perfectly. must have had a bad device.