I've been working on following the design found here: 5 Band Audio Equalizer Circuit using LM833 - Engineering Projects
The schematic has a couple of problems I noticed in that it leaves out the feedback for IC4:A and there should be a cap (I think) to ground for pin 2 of IC2:A of .1uF. And when I checked the frequencies, some of the bandpass filter values were incorrect.
A fuzz bit of background, I laid out and ordered PCBs from ExpressPCB. And I've got some experience with board layout and circuit assembly. I've got a Fluke but no o-scope. I'm having some challenges with this circuit and am looking for some advice to make sure I haven't missed something really obvious. I'd hoped this was a cookbook design but it really isn't working out to be that way.
I'm using the headphone jack from a laptop (a Surface sometimes and a Mac at other times). I usually feed amps/receivers with the laptop output at 100%. If I hook this EQ inline then I get a lot of distortion and an unexpected gain. If I turn the output of the computer down to 50% then some of the distortion goes away. I think the gain is causing quite a bit of the problem here.
I tried to use decent components. The only aluminum electrolytic cap I used was for the 100uF C2. All of my resistors are 1/4w 1%. I used tantalum capacitors for the 10uF caps. And cermet pots.
I haven't had a bunch of time to try to measure anything yet (again, wondering if there's something obvious I missed) but am likely to start heading down that route next.
Any help would be very appreciated.
The schematic has a couple of problems I noticed in that it leaves out the feedback for IC4:A and there should be a cap (I think) to ground for pin 2 of IC2:A of .1uF. And when I checked the frequencies, some of the bandpass filter values were incorrect.
A fuzz bit of background, I laid out and ordered PCBs from ExpressPCB. And I've got some experience with board layout and circuit assembly. I've got a Fluke but no o-scope. I'm having some challenges with this circuit and am looking for some advice to make sure I haven't missed something really obvious. I'd hoped this was a cookbook design but it really isn't working out to be that way.
I'm using the headphone jack from a laptop (a Surface sometimes and a Mac at other times). I usually feed amps/receivers with the laptop output at 100%. If I hook this EQ inline then I get a lot of distortion and an unexpected gain. If I turn the output of the computer down to 50% then some of the distortion goes away. I think the gain is causing quite a bit of the problem here.
I tried to use decent components. The only aluminum electrolytic cap I used was for the 100uF C2. All of my resistors are 1/4w 1%. I used tantalum capacitors for the 10uF caps. And cermet pots.
I haven't had a bunch of time to try to measure anything yet (again, wondering if there's something obvious I missed) but am likely to start heading down that route next.
Any help would be very appreciated.
Last edited:
IC2A is the summing amp, why add a capacitor at pin 2, that's the summing point?
Measure the voltage at the top of C2 - I reckon it will be about +4V with a 12V supply, and you can usefully reduce R1 to 47k to bring that point closer to mid-rail which is where it ought to be I think.
R3's only function seems to be to attenuate the input, but it will add a lot of noise. Try reducing it to 1k and feeding a lower signal level into the circuit.
Another way to reduce gain is remove R4 and R5 (keep R1 at 100k if you do this).
The whole input section is a little odd.
Measure the voltage at the top of C2 - I reckon it will be about +4V with a 12V supply, and you can usefully reduce R1 to 47k to bring that point closer to mid-rail which is where it ought to be I think.
R3's only function seems to be to attenuate the input, but it will add a lot of noise. Try reducing it to 1k and feeding a lower signal level into the circuit.
Another way to reduce gain is remove R4 and R5 (keep R1 at 100k if you do this).
The whole input section is a little odd.
The whole circuit is odd. It is boost-only in each band. Like the original Audio Baton. Equal boosts in each band will not give a smooth flat response.
Such all-bumps EQs are sometimes favored by musicians. But there's enough odd stuff there that it will be much work to suit it for guitar work.
I'd chalk it up to "learning" and go steal a REAL audio EQ. RANE was a good company and all their plans used to be on the Web. Dennis sold-out and you may have to hunt.
Such all-bumps EQs are sometimes favored by musicians. But there's enough odd stuff there that it will be much work to suit it for guitar work.
I'd chalk it up to "learning" and go steal a REAL audio EQ. RANE was a good company and all their plans used to be on the Web. Dennis sold-out and you may have to hunt.
I'll measure and try some of these things out. I'm working on some projects with my daughter to show her how things work and that you can build stuff on your own. Guess she gets to learn more about modifications.
The reason I added the .1uF cap is because I saw someone else add one there in a very similar circuit setup. Easy enough to remove.
I'm hoping I can tweak enough to get this to work. As I noted above, I'm working on some projects with my daughter while we're all stuck at home. We built Hercules MTs from DIYSG (which perform way better than my expectations). And we've also built a two channel amp using an LMT1875. It sounds alright but not as good as the HK 330B that I have setting around.. So the 5 channel EQ was meant to help that out a bit. But, if after some tweak we may chuck this 5 channel EQ up to "learning" for sure.
The input stage is an unnecessary puzzle but 'should' work. It is a unity gain buffer at half supply, but tarted-up with 0.5 and 2.0 factors. If you are deathly afraid of common-mode distortion, this cuts it in half; but why?? If I have missed a mistake, then just tack-solder resistors and jumpers for a no-frills half-supply unity gain buffer.
If the gain is too high, clipping in the bandpass filters, put a passive divider (or pot) in front of it all.
You have identified the mis-draw at the bottom. There's another in the middle.
The article has filter math. Did you run the numbers for gain and frequency? Clearly the graphic has errors of connection and could have errors of values.
If the gain is too high, clipping in the bandpass filters, put a passive divider (or pot) in front of it all.
You have identified the mis-draw at the bottom. There's another in the middle.
The article has filter math. Did you run the numbers for gain and frequency? Clearly the graphic has errors of connection and could have errors of values.
