High Output Current Buffers/Line Drivers

[ninjadave7]

Hey all,
I've been researching buffers and came across a forum conversation about "output buffers", being different from input buffers in that they had high output current to drive long cable lengths (30'+). I guess you could call them line drivers instead of buffers? I did a little more research and found this in a couple other tech articles from companies like Rane, who suggested using ICs like the 5534 to line drive because they have high output current (~38ma), compared to a TL072 (17ma). There are many buffers of all kinds on the guitar market, but all except a couple only advertise buffering in the traditional sense, making a high impedance signal into a low one and stabilizing it in the process. But the conversation I referred to earlier was about these not being good for pushing a signal more than a few feet without starting to lose tone because of cable capacitance.

My questions are: What is the thought around here? And I have a CAE/MXR MC401 "Line Driver" with an LF442 in it, but on the spec sheet didn't see output current and I'm not sure how to calculate or measure it. Is this suitable? Would a buffer built with a 5534 just work? Or is there some other design aspect?

Thanks, hopefully this will be an interesting discussion for people.

[mozwell]

LF442 is 1/10 the current draw of LM1458, it is a FET input op amp.
Input noise is relatively high at 35nV/sqrt(Hz), output current capability is really only 4mA, TL072 is much less noisy & has better driving capability.
As far as i know LF442 is an "old part", and there are much better low supply current, low nosie devices these days.

If you want to drive a low impedance, 5534 is the best, as it is specified to run into 600 ohm loads

LM833 is low noise 4.5nV/sqrt(Hz) & can drive 2k loads
LM4558 can drive into 600 ohm loads
MC33077 is low noise 4.5nV/sqrt(Hz) & can drive 2k loads
MC33178 is low noise 7.5nV/sqrt(Hz) & can drive 600 ohm loads
TL072 is low noise 18nV/sqrt(Hz) & can drive 2k ohm loads
OPA2227 is low noise 3nV/sqrt(Hz) & can drive 600 ohm loads
OPA1662 is low noise 3.3nV/sqrt(Hz) & can drive 2k ohm loads
OPA1652 is low noise 4.5nV/sqrt(Hz) & can drive 2k ohm loads
OPA2134 is low noise 8nV/sqrt(Hz) & can drive 600 ohm loads


Any of these will be ok to drive a stomp box or amp input.
If you really have a long cable run & are worried about driving impedance & cable capacitance, then any of the 600 ohm capable drivers will be ok.
The old blues guys used long cables on purpose, so the capacitance woudl take the harsh bite out of their signal.....

[Steve Conner]

For guitar systems, the distinction between "buffer" and "output buffer" really isn't worth worrying about. Any op-amp based buffer can drive more cable than you'll ever find in even the most ambitious pedalboard.

Worked example: RG58 coax cable has 28.5pF of capacitance per foot. (Guitar cable should be less.) So, a 2k load is 140ft of RG58 at 20kHz, a 600 ohm load is 465ft.

If building your own buffer, remember that you need a 50 or 100 ohm resistor between the op-amp output and the cable, otherwise the cable capacitance will make the op-amp oscillate. The resistor won't hurt the drive capability, since it's small compared to 600 ohms.

The old blues trick works because an electric guitar has a very high output impedance and can hardly drive any cable at all.