midwayfair
Well-known member
I know, I'm posting a bunch of questions about this. I've done a ton of searching though and I don't think these have been adequately covered.
Using MNAT's schematic: http://mnats.net/files/1176REVD_VOLTS.pdf
But there's content from multiple projects here (including Gyraf's).
----
Where's the hold cap for the release and attack?
I tried figuring out the attack time. Time constants are usually just
Time in milliseconds = capacitance * resistance
There's a 220nF cap after the hold cap, and in parallel (sorta) with the release pot. Except for a 220pF, there aren't any other caps in the circuit right there when the release pot is at max.
If I calculate the release pot range (270K-5.27M), I get
59mS fastest release
1159.4 slowest release
So far so good -- almost exactly what it says in the manual.
But if I calculate (note Q1 below) the attack pot range, I get
5.5mS longest range
~53uS assuming that the output impedance of the envelope is only a couple dozen ohms.
That's almost 7 times longer for the longest attack time, and while the lowest attack time is still ridiculously fast, it's not 20mS.
To get close enough to the 800mS quoted in the manual, I need to plug a 33nF into the equation. If I then use that same value to figure out the minimum attack's lowest impedance, I get 600R, which seems maybe a little high for a pair of emitter followers in parallel, but not totally out of bounds.
So my questions:
1) Did I just completely screw up the formula for the attack? My brain is telling me that I've used the attack and decay resistances in parallel in the past when calculating the attack, but that doesn't give me the right number either ... I end up with 3.6K for the highest attack setting, and I don't see any way that the 25,000 Ohm pot gets divided that low.
2) If I DIDN'T screw up the formula, where on earth do the manual's stated attack times come from?
3) As an incurable "can't build it the way it was designed" DIYer, I'd like to use the switch that's commonly on the attack control to switch in a cap to lengthen the attack times, as well as use a larger (but not gigantic) attack pot. I know I can easily get away with a 250K attack pot (because people have done it) and that a loss of half the control voltage can be easily dealt with in this design, but even then I'm only getting to 8mS. I'd like to get above 20mS if possible, which means switching a cap in parallel with whatever the attack pot is interacting with. (I figure if I switch in a 470nF, the lowest decay setting of ~150mS is perfectly acceptable).
------
Ratio stuff:
1) It seems silly to me that there are multiple ways to shut off the compression: all buttons out has no connection to the control amp, and the attack pot has a switch to ... disconnect from the control amp.
I don't like duplicate functions. So if I were using the switch bank it seems obvious to me that ditching the attack pot switch, or using it for something else (like a time constant cap change) is a good call. I'm actually a little surprised that some projects leave them both in, but I guess it's so people can build functional clones.
2) I'm going to do a rotary for a few reasons. I didn't really want to have to drill square holes for the switch bank, but 2P6T metal switches are a lot cheaper than 4 toggles (as cool as it would be to use on-off-on switches and double the number of ratios I can get; even just a pair of DPDT toggles covers all of the original's meaningful ratio settings), so I'm doing 6 ratios instead of 4. I do lose the ability to press 4 and 12 at the same time, which some people report sounds different. Oh well. It's cleaner on the panel to use a rotary and also easier to drill than a switch bank. The nice thing is that I can add 2:1 (I use this ratio a lot in plugins) at least and I'll probably add 6:1 (usually my pre-limiter limiter on bus plugins) and there aren't any ratio settings where nothing's happening.
I lose "all buttons in" without a separate switch.
3) But ... uh. Why does it have to be ALL buttons?
Gyraf discussed how the "Nuke" has to be implemented with a separate switch:
http://www.gyraf.dk/gy_pd/1176/slam.htm
I know sometimes I miss something, but this looks an awful lot to me that "all buttons in" is electronically identical to simply pressing 4 and 20. In both cases the first 56K is shorted to the 47K (lowering the threshold), and the 1.5K is shorted to the 150R.
Shorting stuff in between sure doesn't look like it removes any additional components from the circuit.
That's what it LOOKS like. But I would think that if that were the case, then people would have reported that simply pressing 4 and 20 at the same time is the same sound as pressing all the buttons.
One way would require a 4-pole switch and the other only needs a DPDT.
So am I missing something? Is the Gyraf plan simply an approximation?
One thing I was wondering about was that it seems silly that there's a position where no ratio buttons in isn't doing anything. But if a milder compression ratio, say, 2:1, were hard-wired in, then there WOULD be a problem where the highest ratio setting turns into nuke even with the intervening buttons out.
If it's convenient for anyone, once I figure out the ratio resistors for the additional modes, I can post them here, but it'll be a while.
--
Here's a preliminary (untested) adaption to a 12-position 2-pole rotary (Hairball sells a Grayhill switch that I think can do this ... but any six could be picked for those $2 open frame metal rotaries as well), including 2:1 ratio at the lowest setting and most ratios up to the limiter settings. Not sure what I want to do with the last pole.
https://dl.dropboxusercontent.com/u/9878279/Jon%20Patton%27s%20layouts/Circuit%20ideas/1176%2012-pole%20ratio%20switch%20wiring.png
Using MNAT's schematic: http://mnats.net/files/1176REVD_VOLTS.pdf
But there's content from multiple projects here (including Gyraf's).
----
Where's the hold cap for the release and attack?
I tried figuring out the attack time. Time constants are usually just
Time in milliseconds = capacitance * resistance
There's a 220nF cap after the hold cap, and in parallel (sorta) with the release pot. Except for a 220pF, there aren't any other caps in the circuit right there when the release pot is at max.
If I calculate the release pot range (270K-5.27M), I get
59mS fastest release
1159.4 slowest release
So far so good -- almost exactly what it says in the manual.
But if I calculate (note Q1 below) the attack pot range, I get
5.5mS longest range
~53uS assuming that the output impedance of the envelope is only a couple dozen ohms.
That's almost 7 times longer for the longest attack time, and while the lowest attack time is still ridiculously fast, it's not 20mS.
To get close enough to the 800mS quoted in the manual, I need to plug a 33nF into the equation. If I then use that same value to figure out the minimum attack's lowest impedance, I get 600R, which seems maybe a little high for a pair of emitter followers in parallel, but not totally out of bounds.
So my questions:
1) Did I just completely screw up the formula for the attack? My brain is telling me that I've used the attack and decay resistances in parallel in the past when calculating the attack, but that doesn't give me the right number either ... I end up with 3.6K for the highest attack setting, and I don't see any way that the 25,000 Ohm pot gets divided that low.
2) If I DIDN'T screw up the formula, where on earth do the manual's stated attack times come from?
3) As an incurable "can't build it the way it was designed" DIYer, I'd like to use the switch that's commonly on the attack control to switch in a cap to lengthen the attack times, as well as use a larger (but not gigantic) attack pot. I know I can easily get away with a 250K attack pot (because people have done it) and that a loss of half the control voltage can be easily dealt with in this design, but even then I'm only getting to 8mS. I'd like to get above 20mS if possible, which means switching a cap in parallel with whatever the attack pot is interacting with. (I figure if I switch in a 470nF, the lowest decay setting of ~150mS is perfectly acceptable).
------
Ratio stuff:
1) It seems silly to me that there are multiple ways to shut off the compression: all buttons out has no connection to the control amp, and the attack pot has a switch to ... disconnect from the control amp.
I don't like duplicate functions. So if I were using the switch bank it seems obvious to me that ditching the attack pot switch, or using it for something else (like a time constant cap change) is a good call. I'm actually a little surprised that some projects leave them both in, but I guess it's so people can build functional clones.
2) I'm going to do a rotary for a few reasons. I didn't really want to have to drill square holes for the switch bank, but 2P6T metal switches are a lot cheaper than 4 toggles (as cool as it would be to use on-off-on switches and double the number of ratios I can get; even just a pair of DPDT toggles covers all of the original's meaningful ratio settings), so I'm doing 6 ratios instead of 4. I do lose the ability to press 4 and 12 at the same time, which some people report sounds different. Oh well. It's cleaner on the panel to use a rotary and also easier to drill than a switch bank. The nice thing is that I can add 2:1 (I use this ratio a lot in plugins) at least and I'll probably add 6:1 (usually my pre-limiter limiter on bus plugins) and there aren't any ratio settings where nothing's happening.
I lose "all buttons in" without a separate switch.
3) But ... uh. Why does it have to be ALL buttons?
Gyraf discussed how the "Nuke" has to be implemented with a separate switch:
http://www.gyraf.dk/gy_pd/1176/slam.htm
I know sometimes I miss something, but this looks an awful lot to me that "all buttons in" is electronically identical to simply pressing 4 and 20. In both cases the first 56K is shorted to the 47K (lowering the threshold), and the 1.5K is shorted to the 150R.
Shorting stuff in between sure doesn't look like it removes any additional components from the circuit.
That's what it LOOKS like. But I would think that if that were the case, then people would have reported that simply pressing 4 and 20 at the same time is the same sound as pressing all the buttons.
One way would require a 4-pole switch and the other only needs a DPDT.
So am I missing something? Is the Gyraf plan simply an approximation?
One thing I was wondering about was that it seems silly that there's a position where no ratio buttons in isn't doing anything. But if a milder compression ratio, say, 2:1, were hard-wired in, then there WOULD be a problem where the highest ratio setting turns into nuke even with the intervening buttons out.
If it's convenient for anyone, once I figure out the ratio resistors for the additional modes, I can post them here, but it'll be a while.
--
Here's a preliminary (untested) adaption to a 12-position 2-pole rotary (Hairball sells a Grayhill switch that I think can do this ... but any six could be picked for those $2 open frame metal rotaries as well), including 2:1 ratio at the lowest setting and most ratios up to the limiter settings. Not sure what I want to do with the last pole.
https://dl.dropboxusercontent.com/u/9878279/Jon%20Patton%27s%20layouts/Circuit%20ideas/1176%2012-pole%20ratio%20switch%20wiring.png
