E-H LPB-2 Question re: hiss

[gui_tarzan]

My cousin brought me an old booster pedal and asks if I'll look at that. It's an Electro-Harmonix LPB-2 that he somehow - well - I'm not even sure how to describe what he did to that but I rebuilt the circuit to original specs and layout. And yes, it was like the photo here which kind of shocked me since I'm not really into pedals and haven't looked inside too many. He told me when it worked it was LOUD when kicked on. I looked up the 2n5088 and said no wonder, it has a gain of 900!!!

So it's all back together with a new footswitch and my question is about the amount of hiss I hear when it's on running with my LP Jr. with a Duncan PAF style humbucker into my solid-state Fender Stage 112SE amp. In bypass it's quiet like it should be, it is just the amount of gain that this transistor has? It just surprised me that the hiss is so loud. Is there a way to reduce that hiss with a simple filter or is it just what it is?

[J M Fahey]

Electro-Harmonix LPB-2 .... it was LOUD when kicked on. I looked up the 2n5088 and said no wonder, it has a gain of 900!!!

Thatīs current gain.
That pedal has a voltage/signal gain around 20X, in any case loud indeed.

the amount of hiss I hear when it's on running with my LP Jr. with a Duncan PAF style humbucker into my solid-state Fender Stage 112SE amp. In bypass it's quiet like it should be, it is just the amount of gain that this transistor has? It just surprised me that the hiss is so loud. Is there a way to reduce that hiss with a simple filter or is it just what it is?

Itīs an OLD, very dated design, FWIW the very first EH product, go figure.

Not bad, but I guess you are using it in front of your already High Gain Stage 112.

You might use a lower noise transistor there, but it will still hiss some.

[gui_tarzan]

Thanks. I figured as much. The 900 came from a website describing it but the data sheet does say current gain. It actually is a very simple circuit, there are eight or nine parts total. I think I will build one with a 2n3904 and see how they compare.

[J M Fahey]

Check DIY Audio for that.
They have long and heated discussions over everything, including low noise transistors.

Check which names appear often, typically Japanese from the 80s.

Donīt mortgage the house for the unobtanium fairy dust ones, but Iīm certain there are a few 1 or 2 buck alternatives (which in my view must be as good as the $80 ones) and add a couple to your next Mouser/Digikey order.

Not worth paying $10/12 for $2 in parts either.

Then you tell us

[Mick Bailey]

I think I will build one with a 2n3904 and see how they compare.

That's what I would do first off and keep the original design.

Filtering out the hiss would take out some of the guitar's upper frequencies. The simplest way is a low-value cap across the output. You could experiment with values and a couple of clip leads just to see, but I'm thinking the results may not be too good. You could even have a noisy transistor.

EDIT; If you did want to filter the hiss, then a cap between collector and base would probably be a better method. Try 220pf or whatever is needed. This provides negative feedback for the higher frequencies and doesn't affect the tone so badly as a cap across the output.

[Helmholtz]

A capacitance between collector and base appears multiplied by the voltage gain between base and ground - due to the Miller effect.
Assuming a voltage gain of 30, e.g. a 220p cap between collector and base would result in a huge input capacitance of 6.6nF, most likely killing all PU treble.

Noise typically reduces with lower base current/higher beta (current gain), so a 2N5089 would be a better choice than a 2N3904.
As there is considerable variation between individual transistors regarding noise, it makes sense to try several transistors of the same type.

Generally BJTs in high gain/high impedance input stages are more noisy than Jfets or tubes.

[Mick Bailey]

I did say "or whatever is needed" for the cap value. Some experiment is needed to arrive at the best compromise between hiss reduction and treble loss. Another thought on noise is that it can be as result of ultrasonic oscillation due to layout or wiring proximity and even a small value cap of 22pf can be beneficial if this is the case.