Yet another "bright idea...."

[CurtZHP]

Yes, after reviewing the previous discussion, and running the numbers (thank God for internet HPF calculators!), I now see the error of my ways as far as the cap values are concerned.  I'll have to experiment with that a bit, but I think I can find a "sweet spot" that will work for most cases.

I think I found a better pot, so I can ditch the scratchy one.  That was half the problem.

(If you're ever tempted to strip an old First Act battery-powered practice amp for parts, don't .)

 

[CurtZHP]

Ah, nuts!  The new pot I found is a linear pot.  Tried it anyway, but it's not giving me the "resolution" I want, so I gave the crappy reverse-log pot a good cleaning with some nasty chemicals.  It's a little better now.

So, I tried a few different cap values for the HPF and this is where it gets interesting.  I have to be missing something....

Our previous discussion (up-thread) suggested a 100uF cap as a good starting point, but when I use that, it appears to have no effect whatsoever.  I'm trying several low frequencies, starting up around 250Hz-300Hz, and looking at it on a scope as well as listening.  But, then, running the numbers on the HPF calculator, I'd have to be at max gain, or very close to it, to see any difference.

So, I tried smaller caps.  I had to put a 1uF cap in there before anything looked or sounded like it was HPFing!

Weird!

Outside of that, the thing works great....

 

[Bo Deadly]

But, then, running the numbers on the HPF calculator, I'd have to be at max gain, or very close to it, to see any difference.

Right. That's what I just said in my last post. The only solution is to use a small gain range. In my circuit I used 45-67dB. That's it. Then Rg turns out to be small enough that the low-cut works over a usable range.

Now that I think about it, even a 1K is much too large. You would need a reverse log pot of more like 30 ohms which does not exist.

You would have to pretty much make it stepped and you probably really need two switched caps. Meaning just like my circuit in Reply #1.

 

[CurtZHP]

Coulda swore I had a 1K laying around here somewhere.

At least I know I'm not as crazy as my wife thinks I am.

 

[JohnRoberts]

Right. That's what I just said in my last post. The only solution is to use a small gain range. In my circuit I used 45-67dB. That's it. Then Rg turns out to be small enough that the low-cut works over a usable range.

Now that I think about it, even a 1K is much too large. You would need a reverse log pot of more like 30 ohms which does not exist.

This is an old problem. At Peavey last century we tooled up a custom gain pot that had nice adjustability at the low single digit resistance (high gain) end, while simultaneously delivering high resistance at the low gain end (something like 17k ohm), for better headroom. These pots involved something like 3 or 4 separate screened resist layers, so were slightly more expensive than standard tapers but when buying hundreds of thousands a year, the cost increase was manageable.

These pots may not be in DIY friendly footprint configurations but are perhaps available from Peavey service dept, as repair parts. Perhaps check specific sku schematics before yanking the service dept's chain as they may not understand what you are asking about. 

You would have to pretty much make it stepped and you probably really need two switched caps. Meaning just like my circuit in Reply #1.

I haven't shared this other old story lately, but back in the day one of our competitors spent piles of money claiming how quiet their mic preamps are, while intentionally undersizing the cap in series with the gain pot. This had the intended effect of rolling off the very low frequency response at high gain settings, making the 1/F noise appear lower than reality.  Of course their published frequency response was measured at lower gain where the LF response is still relatively flat.  :

JR

 

[CurtZHP]

I haven't shared this other old story lately, but back in the day one of our competitors spent piles of money claiming how quiet their mic preamps are, while intentionally undersizing the cap in series with the gain pot. This had the intended effect of rolling off the very low frequency response at high gain settings, making the 1/F noise appear lower than reality.  Of course their published frequency response was measured at lower gain where the LF response is still relatively flat.  :

JR

There's lies, damn lies, and spec sheets.