James HE
Well-known member
I started playing around with this after reading Fred Forssell's Eq paper.
http://www.forsselltech.com/Evolution%20of%20an%20EQ%20Design2.pdf
I was working on an eq using gyrators and decided to experiment with this one. Forssel writes that this gyrator design "has a distinct advantage over the other" Why? (I hope he or someone will elaborate.) It's straightforward and simple, well, it was until I made it complicated. :razz:
I did a bunch of math figuring out some values. Hell I went to art school I can't do math. :wink: so I went off on a few tangents. I was aiming to have a rotary switch to select frequencys but to split it up in different q's. So let's say with a 2 pole 6 position switch you can have three different frequncy's at 2 different q's. When trying to figure out what Q's and frequncies to use, I realized that the cap values for CS and CI at any frequency with a q of 1 could be swapped around to get a q of .25 and vise versa. These are the only 2 reasonable Q's I've found that swap around. (Well CS = CI with a Q of .5 so swap those if you want) A Q of .25 is about 4 octaves, a q of one is about 1 Octave. Were not going to be doing any surgery with this thing, but we have some flexability.
Now does it really work? Can it be done. I dunno! This is what I got.
so basically flipping the "Q" switch swaps CS and CI around. I figured out how to switch it with a 4PDT switch. I'm sure it's right, but I need someone else to say "yeah that works" to put my brain at ease. One thing I have a question about is what happens to the caps when the switch is flipped, the volages with be going the oppisite way all of a sudden.
The values and frequencies are tentative in the schematic. After figuring out a few frequencies, and a little fudging, I simply plugged in readily available values and went down (er... up?) the sequence. I'm using a 2 - 4 rotary here for the frequency selection. We can parallel caps of course to get what ever value. The Q's generated vary a bit, generally it's around .28 and .9. Good enough I suppose.
I'm very unsure about what will happen with the shelving switches switched in (say that 10 times real fast) (shorting CS or lifting CI) if the CS and CI are flipped. I haven't figured out a way to calculate how the frequency and Q changes. It seems like you'll end up with 8 different shelving frequencies - the Q switch would work more like a "range" switch. I don't know what to assume the Q will be with the gyrator in shelving mode. I have 2 lo and 2 hi shelving sections instead of just one on each end, just cause, I often set up my plugins this way, Good things can happen with one shelf boosting and one shelf cutting.
Another thought is that I would simply set up the frequencies to have a Q of 4 or so and use a pot in series after CS to vary the Q down. This affects the gain. (wouldn't we call this "constant loudness" a feature?) Something that Mr. Forssell dosen't like, but who knows I might like it! Or just make a 10 band graphic EQ with the Gyrator with 2 (or 4) shelving bands?
Or for me, just make the Calrec or the Pultec and stop playing "designer".
So here's my artwork for the day. I worked on a PCB layout for this schem.
(hopefully is clear enough had to do a print screen and use paint)
I just get into the image of the PCB sometimes, I think of it like any other composition. this is just the copper layer. The cap spacing can accommodate 5, 7.5, 10, 15, and 20 mm spacing so I could be flexable to use what I want. Plus it gave me symmetry. Yes it's tiny, only a half rack, too crowded. I usually don't mind small knobs though.
http://www.forsselltech.com/Evolution%20of%20an%20EQ%20Design2.pdf
I was working on an eq using gyrators and decided to experiment with this one. Forssel writes that this gyrator design "has a distinct advantage over the other" Why? (I hope he or someone will elaborate.) It's straightforward and simple, well, it was until I made it complicated. :razz:
I did a bunch of math figuring out some values. Hell I went to art school I can't do math. :wink: so I went off on a few tangents. I was aiming to have a rotary switch to select frequencys but to split it up in different q's. So let's say with a 2 pole 6 position switch you can have three different frequncy's at 2 different q's. When trying to figure out what Q's and frequncies to use, I realized that the cap values for CS and CI at any frequency with a q of 1 could be swapped around to get a q of .25 and vise versa. These are the only 2 reasonable Q's I've found that swap around. (Well CS = CI with a Q of .5 so swap those if you want) A Q of .25 is about 4 octaves, a q of one is about 1 Octave. Were not going to be doing any surgery with this thing, but we have some flexability.
Now does it really work? Can it be done. I dunno! This is what I got.
so basically flipping the "Q" switch swaps CS and CI around. I figured out how to switch it with a 4PDT switch. I'm sure it's right, but I need someone else to say "yeah that works" to put my brain at ease. One thing I have a question about is what happens to the caps when the switch is flipped, the volages with be going the oppisite way all of a sudden.
The values and frequencies are tentative in the schematic. After figuring out a few frequencies, and a little fudging, I simply plugged in readily available values and went down (er... up?) the sequence. I'm using a 2 - 4 rotary here for the frequency selection. We can parallel caps of course to get what ever value. The Q's generated vary a bit, generally it's around .28 and .9. Good enough I suppose.
I'm very unsure about what will happen with the shelving switches switched in (say that 10 times real fast) (shorting CS or lifting CI) if the CS and CI are flipped. I haven't figured out a way to calculate how the frequency and Q changes. It seems like you'll end up with 8 different shelving frequencies - the Q switch would work more like a "range" switch. I don't know what to assume the Q will be with the gyrator in shelving mode. I have 2 lo and 2 hi shelving sections instead of just one on each end, just cause, I often set up my plugins this way, Good things can happen with one shelf boosting and one shelf cutting.
Another thought is that I would simply set up the frequencies to have a Q of 4 or so and use a pot in series after CS to vary the Q down. This affects the gain. (wouldn't we call this "constant loudness" a feature?) Something that Mr. Forssell dosen't like, but who knows I might like it! Or just make a 10 band graphic EQ with the Gyrator with 2 (or 4) shelving bands?
Or for me, just make the Calrec or the Pultec and stop playing "designer".
So here's my artwork for the day. I worked on a PCB layout for this schem.
(hopefully is clear enough had to do a print screen and use paint)
I just get into the image of the PCB sometimes, I think of it like any other composition. this is just the copper layer. The cap spacing can accommodate 5, 7.5, 10, 15, and 20 mm spacing so I could be flexable to use what I want. Plus it gave me symmetry. Yes it's tiny, only a half rack, too crowded. I usually don't mind small knobs though.
