|
|
Post by Blackdawg on Jan 17, 2019 10:33:20 GMT -6
Brain storming here..
Is it possible to add at 24 or 47 stepped pot such as an Elma or smaller like a grayhill to any circuit?
Thinking like lets say the Threshold and Gain on a SSL style bus compressor. Both are 10k potentiometers. Does this mean when in the final "step" the total resistance has to be around 10k?
How would one calculate the steps?
Thanks!
|
|
|
|
Post by matt@IAA on Jan 17, 2019 10:50:52 GMT -6
A potentiometer is a voltage divider with a variable ratio between the "two" resistors. You need to figure out how the pot is configured in the circuit and what taper is required. For example, if you're using it for gain, it almost certainly needs to be log based...EQ will likely be linear.
Once you do that, then you need to know what the resistor is affecting. Again with the gain example, looking at something really simple like a 312 circuit or something the gain in dB is a formula based on the feedback resistor - 20*log(1+Rf/R) where R is your pot. Flipping that back to solve for R, you get R = Rf / [10^(dBgain/20)-1]
Make a table of the gain points you want, then calculate the resistor values required for each step, starting with the smallest resistor value (maximum gain, in this case). So if we have an Rf of 20k and we want 40 dB out of the op amp, you get a calculated resistor value of 202, or 200 for nearest nominal. If you need 35 dB, you get a resistor of 362. But typically what you do is make a resistor ladder, so you add more resistance at each point on the switch. Based on that, we already have 200 so we need an additional 162. That's a nominal value, so life is good. For 30, you need a total of 653, or an additional 291 - use 287 or 294. And so on.
A linear circuit is much easier because you can just subdivide the total range into the steps you want, provided there is a nominal value available.
Just as a side note, something like a 1073 is different, because that uses a combination of voltage dividers and feedback resistors - and even switching various gain blocks on and off at different points. Something like a Focusrite Red or the newer RND stuff is halfway - voltage divider plus changing the feedback on an op amp.
So the short answer is yes. Long answer is it might be a pain, or not, depending on the circuit.
|
|
|
|
Post by matt@IAA on Jan 17, 2019 10:58:33 GMT -6
Sorry - for your specific example, in a VCA compressor the sidechain is usually varied in the linear domain because the output of the RMS detector is the log of the input signal level - it's already in dB. So you change the voltage in the sidechain linearly in order to get a linear dB change on the compressor. Replacing those pots with a switch should be as simple as subdividing 10k by the steps you want.
Threshold and gain in a VCA sidechain are usually a set voltage divider between the +V and -V rails - or some high and low reference voltage. As you vary the threshold you're changing the bias point in the threshold detector's input, which turns on or off at the reference voltage. For gain, you're doing the same thing except you're adding a DC offset to the final signal going into the VCA control voltage.
|
|
|
|
Post by Blackdawg on Jan 17, 2019 11:18:34 GMT -6
Sorry - for your specific example, in a VCA compressor the sidechain is usually varied in the linear domain because the output of the RMS detector is the log of the input signal level - it's already in dB. So you change the voltage in the sidechain linearly in order to get a linear dB change on the compressor. Replacing those pots with a switch should be as simple as subdividing 10k by the steps you want. Threshold and gain in a VCA sidechain are usually a set voltage divider between the +V and -V rails - or some high and low reference voltage. As you vary the threshold you're changing the bias point in the threshold detector's input, which turns on or off at the reference voltage. For gain, you're doing the same thing except you're adding a DC offset to the final signal going into the VCA control voltage. Thanks! Man I haven't done math like it a loooonnngggg time haha That is a very helpful response thanks man. I knew this could and probably would be complex. But I want to know more about the circuit design and how it all works. So this is a cool place to start. Okay so let run with the VCA compressor idea since that is the first thing I will try this on and thankfully it should be simpler. So if I used a 47 step pot(something like this guy: www.don-audio.com/Elma-A47-High-End-Audio-Attenuator-rotary-switch?gclid=Cj0KCQiA7IDiBRCLARIsABIPohg1mupIHDpt0Ei5nEMpdtu1zDcMsG3LCdrf1HRHO-lI_fNQEPj1-4AaAojxEALw_wcB) and its for the gain and threshold. I basically just need to divide 10k into 47 stepped in an additive way? So first step is 0ohms or whatever the minimum value is normally in the original pot. Then go up by about 213 ohms a step which amazingly is a nominal value. So each step is a 213 ohm resistor. That would work for the threshold for sure but the gain Im not sure if that is a feed back setup like how your where describing. It a make up gain circuit so this might be different yes? Am I thinking of that right? |
|
|
|
Post by Blackdawg on Jan 17, 2019 11:23:32 GMT -6
FYI
Id be attempting this with DIY-Racked MB2 VCA compressor as it has the ratios I like and built in Parallel processing which can also be stepped but they have done the leg work there. Thinking of making this a fully stepped guy for mastering.
|
|
|
|
Post by Blackdawg on Jan 17, 2019 11:36:13 GMT -6
|
|
|
|
Post by Blackdawg on Jan 17, 2019 12:15:02 GMT -6
Huh. I think I can just get some stepped 10k Pots. Thats would make it simple. The V47 comes in a 10k ohm with SMDs: www.goldpt.com/compare.htmlOnly $248 each...ouch. CAPI sells a 2x24 for 140 each..ouch. Doubles the price of this compressor LOL Edit: Im an idiot. I don't need them in 2 pole. So that makes the price way more reasonable at 130 a switch with SMDs for a 10k. Bit high still. Can get them with TH PCBs for 81 bucks. Seems worth it to me. |
|
|
|
Post by matt@IAA on Jan 17, 2019 14:18:33 GMT -6
Yes, the first step will be zero as a pot is shorted to one side a the extreme. And yes, that is correct for both threshold and gain. The gain is done by offsetting the VCA voltage on a linear basis, so it's identical to the threshold. You don't have to buy a high end premade attenuator - you can make your own using a rotary switch and resistors. Maybe this would work - an Electroswitch C7D0123S would have 23 positions - not sure if that's enough or not. Put a resistor ladder daisy chaining using 453 ohm resistors (like Xicon 1% - 271-453-RC) between each contact. At position 12 you have a middle position for a zero gain (4983 on each side). Same pot would work for the thresholds. Total is $80 plus shipping. For the dual pot, the C7D0223S is the same switch with two gangs. What is the dual / parallel pot doing? It's also linear. |
|
|
|
Post by Blackdawg on Jan 17, 2019 14:25:38 GMT -6
Yes, the first step will be zero as a pot is shorted to one side a the extreme. And yes, that is correct for both threshold and gain. The gain is done by offsetting the VCA voltage on a linear basis, so it's identical to the threshold. You don't have to buy a high end premade attenuator - you can make your own using a rotary switch and resistors. Maybe this would work - an Electroswitch C7D0123S would have 23 positions - not sure if that's enough or not. Put a resistor ladder daisy chaining using 453 ohm resistors at each contact. At position 12 you have a nice middle position for a zero gain (4983 on each side). Same pot would work for the thresholds. Oh cool. That would be much more affordable. Are there any disadvantages to using that switch vs the one I was looking at? I was looking at doing one of these  Its a through hole design still. But much easier to build thats for sure. And I can get it in 47 steps, which I like the idea of for mastering. |
|
|
|
Post by matt@IAA on Jan 17, 2019 14:36:00 GMT -6
|
|
|
|
Post by Blackdawg on Jan 17, 2019 14:44:24 GMT -6
damn can't see the photos without being a member. Well since its a lot simplier than I thought it would be. I think i'll go for the 47 position one. Might even try it on my SB4000 im building right now for drum bus duties. Edit: Won't work with the SB400 due to the PCB design. Oh well. I'll get the MB2 soon. |
|
|
|
Post by matt@IAA on Jan 17, 2019 15:35:01 GMT -6
Why won't it work on the SB400? No room or what?
|
|
|
|
Post by Blackdawg on Jan 17, 2019 15:57:21 GMT -6
Why won't it work on the SB400? No room or what? Yeah no room. The PCB for that one goes all the way across the front. The switches for threshold and Gain are PCB mounted an the LED meter is next to the gain. I'll do more measurements tonight at the house. But pretty sure a stepped switch will be too deep. |
|
|
|
Post by drbill on Jan 17, 2019 18:16:32 GMT -6
We use the 41 detent pots on some stuff. Then work great. But these days I'm leaning more towards preferring the 21 detents. (I think those are what is on the EQP5 IIRC). I like that they have more deleniated "clunks" and are much easier to recall accurately. That is if you can deal with the increased space between steps, I think they deliver a better "feel" and ease of recall.
And even if Eric doesn't like em, I offset his dislike with my like!!! <thumbsup>
Bravo detented pots!!!
|
|
|
|
Post by Blackdawg on Jan 17, 2019 18:36:23 GMT -6
We use the 41 detent pots on some stuff. Then work great. But these days I'm leaning more towards preferring the 21 detents. (I think those are what is on the EQP5 IIRC). I like that they have more deleniated "clunks" and are much easier to recall accurately. That is if you can deal with the increased space between steps, I think they deliver a better "feel" and ease of recall. And even if Eric doesn't like em, I offset his dislike with my like!!! <thumbsup> Bravo detented pots!!! Yeah i was wondering about that. The cool thing is I should be able to limit them. So could do like 30 or something. For something like setting the threshold on a compressor I'm worried that 23 or so might not be enough. But maybe it would be. I certainly enjoy the EQP5's pots. Work perfect for recall and seem to never want something in between a detent. I like them honestly. The 24 step premade 1 pole switches are even only 81 bucks. That seems well worth it. They are also WAY higher quality so even Eric might like them  Hmmm. I'll have to think about which to get..24 might be good. 47 did seem a bit high and would be a pain to count out haha Thanks for bringing me down to earth! Will likely work better. matt@IAAOkay new scenario. This one for Log pots. I bought an SA-3A kit(stereo LA3A kit) and they use 100kLog pots for the gain and reduction knobs. Im not sure if its a feed back setup like you were saying. But if im reading your first post right. i need to know the dB steps i want to use the formula to get the resistances I want yes? |
|
|
|
Post by matt@IAA on Jan 17, 2019 19:47:06 GMT -6
Some of the switches have adjustable stops. Then you could have as many as you like up to the maximum. For the VCA compressor you can set the points anywhere you like -- for example, have nonlinear stuff at the extremes, and finer tune in the middle where you might use it more. Maybe pay attention or look at old recall sheets to see where you're usually setting the threshold. Then you can get more resolution with the same number of stops. For example, if you have that 23 point switch you could do double the steps on either end, then half the steps in the middle. Something like: 910 910 910 453 453 226 226 226 226 226 226 226 226 226 226 226 226 453 453 910 910 910 If the range of the pot is -30 dB to +10 dB for the threshold, your curve would look like this:  Which may be much more useful than 47 linear points. If you know the taper you're trying to copy, it doesn't matter how it's used in the circuit really as long as you hook up 1 to 1 and 3 to 3 and the common to the wiper, ya know? Here's a chart for 100k log taper in a 24 point switch: www.goldpt.com/r_series.htmlThe gain calculation was just specific to if you wanted to swap a mic pre gain knob in a 312 or similar. Just an example. |
|
|
|
Post by Blackdawg on Jan 17, 2019 20:55:21 GMT -6
Some of the switches have adjustable stops. Then you could have as many as you like up to the maximum. For the VCA compressor you can set the points anywhere you like -- for example, have nonlinear stuff at the extremes, and finer tune in the middle where you might use it more. Maybe pay attention or look at old recall sheets to see where you're usually setting the threshold. Then you can get more resolution with the same number of stops. For example, if you have that 23 point switch you could do double the steps on either end, then half the steps in the middle. Something like: 910 910 910 453 453 226 226 226 226 226 226 226 226 226 226 226 226 453 453 910 910 910 If the range of the pot is -30 dB to +10 dB for the threshold, your curve would look like this: Which may be much more useful than 47 linear points. If you know the taper you're trying to copy, it doesn't matter how it's used in the circuit really as long as you hook up 1 to 1 and 3 to 3 and the common to the wiper, ya know? Here's a chart for 100k log taper in a 24 point switch: www.goldpt.com/r_series.htmlThe gain calculation was just specific to if you wanted to swap a mic pre gain knob in a 312 or similar. Just an example. Super helpful! Thanks man! Really appreciate it. Learning a lot. I like the idea of doing the curve with a finer tuned mid point. That could be very helpful. I think I'll attempt doing something like that. Thanks for the info on the log pot. I think i'll try and make my SA-3A stepped as well. That could be SUPER handy for me. Super stoked on this! Thanks for your help. Im sure i'll be asking more questions later as the time comes to build them. Now for more brainstorming. |
|
|
|
Post by Blackdawg on Jan 17, 2019 21:46:19 GMT -6
Oh and just as an FYI.
No way in hell can i fit a stepped pot in the Sb4000. Board put other components in the way and theres like .52'' from the board to the faceplate. Not gonna fit.
The MB2 though will have ample room. So i'll do it on that one and the SA-3A and the Send n Blend.
|
|
