Hoffman Amplifiers Tube Amplifier Forum
Amp Stuff => Tube Amp Building - Tweaks - Repairs => Topic started by: 12AX7 on October 18, 2013, 11:22:04 pm
-
I have a question concerning 500k vs 1M gain pots. Looking at a lot of cascaded preamp type amps like jcm 800's, soldonos, etc, I see both 1M and 500k for the gain pot after the first stage. I don't see any common denominator between the ones that use 500k or between the ones that use 1M. The question is what is it those that use one or the other are trying to accomplish? or in other words, what is the difference tonally? I HAVE tried this by paralleling a 1M resistor on my 1M gain pot and i have a opinion on what it seems to do, but i wanna hear from others and see what you all say. Just a gain changer or something more?
-
My amp has an interesting take on this issue. It's a Sovtek Mig60 that was modified by Splawn Amplifiers (with further tweaks by me). Since you mentioned the JCM 800 we know that the first stage has a 2K7 cathode R and a 100K plate R. Off the plate we have a 470K resistor and a 1M gain pot. So if we max the gain pot we get roughly 2/3 of the voltage that is generated across that voltage divider feeding to the second stage's grid.
In my Mig60 the first stage has a 1K1 cathode R (the stock value) and a 220K plate R (stock value was 120K). This increases the gain of this stage measurably over the JCM 800 first stage with both a lower value cathode resistor and a higher value plate resistor. But then Splawn used a 1M resistor and a 100K gain pot where the JCM 800 uses 470k/1M. So he gained up the first stage quite a bit but is only sampling 1/11 of the signal that is generated across the voltage divider if the gain pot is dimed (I never turn it past 12 o'clock). It's not so much about the level of the signal going to the second stage as it is the tone of that signal.
So you see, looking at the value of the gain pot itself is not necessarily very telling. You have to examine all of the surrounding circuitry to really see what is happening.
-
I have a question concerning 500k vs 1M gain pots. .... The question is what is it those that use one or the other are trying to accomplish? ...
Maybe something, maybe nothing. Alerich alludes to it, but you might have to point to 2 specific amps you want to compare/contrast.
Knowing nothing else, using a 1M pot after a gain stage instead of a 500k pot (and no other parts between the coupling cap and ground) is a lighter load on the previous stage, increasing that earlier stage's gain somewhat.
See the table I posted here (http://www.el34world.com/Forum/index.php?topic=16116.0.msg157296#msg157296).
-
Look at them as variable grid leak resistors...just controlling gain between stages ...you "could" have one between every stage (but where would you put all of those pots?)
Dumble puts them on the front panel as Trim/Drive/Level
I put them on the board as trimpots, so I can tweak the gain at each stage,,during the experimental process
I just made an amp for a guy,,,who I KNOW won't be able to just leave it as is.....so I used trim pots in 2 stages to give him something to monkey with (see pic)
To reword what Alerich said in my own simple-minded way.....
If your going to have one stage "smoking", you're going to need a way to drop some signal before you hit the grid of the next stage
The JCM800 first stage is "cool" enough to be able to "afford" a 1M into the 2nd stage
This Bogner preamp (pictured) is pushing the 1st stage a little harder, so it's a good idea to use a 500K before the next stage
-
All very good explanations above as you must consider the application for signal gain, consider the loading effect it presents on the prior stage (which also affects gain), and effectively varies the grid stopper amount into the succeeding stage. After summing these things up, there's other considerations that affects my decisions:
1. To get the most useable or usefulness range out of the control. Depending on design, if it's mostly used between 0 - 40% or 40% - 60% or say 60% - 100% then the useful range is not really all that useful?! Something can be adjusted to better correct this. Heck if the control's not used much or though a minimal range then a control pot should be replaced w/ a fixed resistor.
2. The second thing I consider is that if I can use the lowest value that still gives me plenty of gain (due to the loading effect by being in parallel w/ the prior stage's plate resistance) then my dropping resistor + pot (voltage divider) can be much less in resistance or better yet to not even need the dropping resistor and just have a single pot there if possible? This lessens or eliminates the need for bypass or treble bleed caps for a nice even signal and/or control response. Bottom line is I ALWAYS try to minimize the resistance in the signal path to retain the signal as purely and naturally as possible. The need for treble bleeds and bypass caps make up for what I believe are often errors in circuit design.
So the "ideal" or preferred way to go about designing between stages would be to have no voltage divider or dropping resistor in the signal path and to have the lowest effective grid stopper directly before the next stages' input and use a single load resistor to set your gain amount (and also serve as a grid return for succeeding stage for bias purposes). BUT you want to control this gain - so use the same value pot as the load resistor to keep the same loading effect on prior stage (and you shouldn't have to use a dropping resistor prior to the control pot?). The grid stopper then may not be (necessarily) needed because part of the pot's resistance can then effectively serve as the grid stopper also - unless this control is going to be often used up all the way (and depending on where this pot is in the circuit)? Then the grid stopper would still be useful - but it could be much lower than it otherwise would've been.
Lastly, if you're planning ahead properly and know that you're going to need to drop your signal by a certain amount, a better method would be to utilize a split load arrangement for that gain section to again eliminate or minimize the need for any resistance in your signal path and keep your original tone intact.