reducing noise, where to start?

[onusx]

hi, i have a couple of 70s devices i'd like to quiet down so they're useful to record with; one is an SS gibson reverb unit, other is a PAIA synthspin; both are noisy when on or off, consistent hum/hiss that's very noticeable when miked, though not totally unusable.

just wondering where to start with eliminating this. both units have two prong cords, and i'm not sure how to install a grounded plug. is it best to just go ahead and switch caps out, is that likely to be the problem? i know they might be inherently a little noisy, but what could i do to improve their performance?

sorry this is probably a redundant question on this forum, this is just where i'm at right now. i'm very much a beginner at electronics work.

thanks,

brian

[Paul Perry (Frostwave)]

The first thing is to distinguish between 'noise' and 'hum', the latter you might improve with bigger filter caps on the power rails (with the synthespin, you could try temporarily running it on batteries to be sure).
As for the intrinsic electronic component noise, if it just sounds like a continuous white noise, then it is probably a design problem. The synthespin trransistors could well be noisy, compared to today's. But, I don't think they can be swapped out willy-nilly, one is quite crucial to teh operation.
The best thing is to besure that the input signal is as high as possible, without overloading anything. And maybe a noise gate after it, or (better) a downward-expanding noise leveller.

[petemoore]

Start at the beginning...the guitar itself, then the cable to the first box.
 The guitar probably makes noise if it's like my factory guitars, I don't know how to get rid of it but by shielding. I Double shielded my Clone Yer OWN and it's super quiet in comparison.
 At the input, and near the beginning of the chain is where a small amout of noise sounds HUGE...the same amout of noise injected later in the path, after some gain stages...is lesser by the amout of the amping even later in the chain TIMES the earliest gain stages, so even a small improvement near the source can seem a HUGE difference...because it gets amplified by the first gain stage...AND Every stages after, [thats a multiplication equation..].
 I hope you're confused but get the Gist, because it looks confusing the way it's worded...'at or near source'...hope this helps !!!

[Paul Perry (Frostwave)]

And a problem with spring reverb units is that the signal is so small when it is picked up at the end of the spring, you have to be careful with earthing and shielding. Plus I suspect the magnetic pickup can get hum into it :x Think of the reverb pickup like a guitar pickup but wiht a signal MUCH smaller.

[Mark Hammer]

If I remember correctly, the Synthespin uses a quad Norton amp (LM3900) which may not be the quietest thing on earth.  There was an article in Polyphony in 1978 about reducing noise in the PAiA Phlanger by replacing a few select Norton stages with 4739 op-amps.  There are thankfully many more low-noise op-amp choices these days than the 4739, but conceivably, the few stages in the audio path (some of the Norton stages are used for LFO and CV handling) that employ portion of the LM3900 can be similarly replaced with an op-amp that has much better noise performance.

Personally, I often find that more bandwidth is built into many pedals than is really needed, likely because the manufacturer has no idea what you will be using their product with and provides for worst-case scenario.  What that means is that many devices provide the opportunity for hiss to be heard, rather than filtering it out.  I often find it worthwhile to tack a feedback cap in an output stage to tame the 10khz and up content.  My tastes, though, and needn't be yours.  After all, I also don't know what you're using it with.

All reverbs suffer from the same problem: reverb pans are a particularly inefficient and problematic medium for transmitting sound energy.  If you drive them too hard, they go splat.  If you don't drive them hard enough, then recovering a signal at the other end which is of sufficiently useful amplitude involves adding a lot of gain.  And boosting a wimpy signal too much almost always involves acquiring hiss.

There are a few bits of good news, though.  First, the bandwidth of even the best reverb pans is severely limited, so if you filtered out content above 3khz (and sometimes even lower) in the recovery circuit, you probably wouldn't miss much except the hiss.  Second, there are better and worse methods of both driving a spring and recovering a signal.  I'm unfamiliar with this specific unit, but it is entirely possible that some sort of interstage transformer (if there isn't one already) at one or both ends might improve matters some.  If the springs are driven harder (but not too hard), then what gets recovered at the other end will be of a decent amplitude, relative to any hiss in the recovery stage.  If the recovered signal is of a decent level, then the gain in the recovery stage may be reduced and further eliminate and unwanted high end.  As well, it is also possible that the unit is really designed for line levels, so your S/N ratio problem might just disappear if the unit is fed a hotter signal to begin with.