The No.1 Website for Pro Audio
DIY Poor Man's 1073 EQ
Old 23rd October 2018
  #1
DIY Poor Man's 1073 EQ

Hey Folks,

I started a little hardware project and now that most of it is working, I'm confident enough to share it outside of groupdiy.

I've successfully built a 1073 EQ that is very faithful to the original circuit and only consists of resistors, capacitors and standard opamps.

This circuit does not intend to compete with an original 1073 on a performance level. So if you're after the sound of hand wound inductors and the sonic characteristics of discrete class A amplifiers, that's fine and you might still find some useful information in here.

My main goals of this project are to make the 1073 EQ more accessible to people, both in hardware and in the form of documentation. That's why I've also compiled all the bits and pieces of information which helped me gain a better understanding of the original circuit.
I also need to point out that I'm focusing on the EQ part, leaving the choice of an (entirely optional) preamp to you. The schematics are modular, so if you just want to build the HPF, to add it to an existing preamp, go ahead!

You can find all resources on my github repository:
GitHub - ravettel/cheaps73

and here's the groupdiy topic: Poor Man's 1073 EQ

Let me know any questions or further ideas you might have
Attached Thumbnails
DIY Poor Man's 1073 EQ-1073_hpf.jpg   DIY Poor Man's 1073 EQ-1073_presence.jpg   DIY Poor Man's 1073 EQ-1073_shelves.jpg  
Old 24th October 2018
  #2
Lives for gear
Great job KC, thank you for sharing! I'll look at it in more depth, for now I've just played a bit with the simulation and the only issue I see is that the opamp at the top of the middle section sees too heavy a load for a TL082, with the controls as they are it drops to [email protected]~170Hz and stays below ~1.7k all the way up to 20kHz. Note that the datasheet only shows THD for a 10k load, so you can expect it to be rather higher at those loads. A 5532 would work much better here. You should check the one in the input section too for different positions of the pots (especially when R12 becomes small).

Edit: also input impedance is pot-position and frequency dependent and rather low for some settings, as long as you're going with opamps, how about input and output buffers? It's just an additional dual opamp and you can also give the input one some gain to make up for the -2.5dB loss.
Old 24th October 2018 | Show parent
  #3
Quote:
Originally Posted by Cabirio View Post
Great job KC, thank you for sharing! I'll look at it in more depth, for now I've just played a bit with the simulation and the only issue I see is that the opamp at the top of the middle section sees too heavy a load for a TL082, with the controls as they are it drops to [email protected]~170Hz and stays below ~1.7k all the way up to 20kHz. Note that the datasheet only shows THD for a 10k load, so you can expect it to be rather higher at those loads. A 5532 would work much better here. You should check the one in the input section too for different positions of the pots (especially when R12 becomes small).

Edit: also input impedance is pot-position and frequency dependent and rather low for some settings, as long as you're going with opamps, how about input and output buffers? It's just an additional dual opamp and you can also give the input one some gain to make up for the -2.5dB loss.
Thanks! I only chose the TL082 in the simulation, because the NE553x were not included in the standard library. The finished circuit however only uses NE5532 opamps, as they performed much better in my tests.

Now that I have the basic EQ circuit running, I/O stages are next on my list and I've already been trying some things out. I got to admit, that my understanding of impedance/load/driving is still lacking, as I pretty much jumped head first into this project. So your insight is greatly appreciated!

I like the idea to make up for the gain loss right on the input, as this should not increase the noise floor as much. But I ran into some crazy distortion issues during testing, which I still have not fully figured out:

I would feed the circuit a 1K sine wave and once I started boosting the mid band at either 0,7k or 1,6k, the signal would start to distort at a certain point. The distortion would set in within an instant at a certain level and the threshold between "no distortion" and "Oh look, a saw wave!" was only around the slightest touch of the eq knob. This was regardless of the input level of my audio interface, which I used for testing. The only way to prevent this from happening was to lower the input signal at the circuit. I've attached a screenshot; don't let the label irritate you, I was using the Waves Scheps 73 as a reference point during testing.
Attached Thumbnails
DIY Poor Man's 1073 EQ-1073_gyrator_presence_distortion_04.jpg  
Old 24th October 2018 | Show parent
  #4
Lives for gear
Quote:
Originally Posted by krabbencutter View Post
I ran into some crazy distortion issues during testing, which I still have not fully figured out
Bear in mind that with +/-12V supplies a normal (not rail-to-rail) opamp may swing to around +/-10V max peak, so it doesn't take a lot of input signal to reach that point with the mids boosted. If you run the tran simulation with the mids at max, input signal at an amplitude of 2V and then again at 2.2V (which is around 1.4-1.5VRMS, not that much), you will see clearly the output of the top opamp in the mids section go from reasonably clean to clearly clipping (*).

If you want to stick with normal opamps (I would), increasing the supplies to +/-15, or even 17 (still within the specs of most opamps) will help, and if you include the buffers, you can also trade off headroom for noise by attenuating the signal at the input and making up for it in the output buffer.

(*) that's with the 5532 model I'm using, attached in case you want to give it a try
Attached Files
File Type: txt NE5532.txt (1.3 KB, 127 views)
Old 25th October 2018
  #5
Lives for gear
A quick one: the bass control topology looked a bit odd to me so I double cheked with the PDF schematic and found a small mistake in the LTSpice file: the top terminal of C4 (10n) should go to the pot wiper rather than the left side of the pot.

Since you've already built it I assume you're happy with the noise performance, but there's quite a lot of room for improvement in the mid control gyrator, R6 being the main culprit. It's counterintuitive but due to the way a gyrator works, the smaller that resistor, the higher its noise contribution. If you simply make it 360R and make C2=10n, noise drops by almost 20dB and the filter response doesn't change all that much, you could probably fine-tune it easily by adjusting a bit the other components, R2 in particular.
Old 25th October 2018
  #6
wow, thanks a lot for your contribution! The version I built was just a proof of concept and I also built it, because I'm equally excited, as I am impatient :D
I still consider this an ongoing project and I'm all in, in regards of improving the circuit. I just picked up some missing capacitors and will have a go on my breadboard, as soon, as I'm home.

Update: just tested the circuit on my breadboard. That's IMPRESSIVE, to say the least! Maximum gain got reduced by ~1dB and the curve got shifted a tiny bit, but looking at the dramatic reduction in noise, I'd say that's absolutely worth it!
Attached Thumbnails
DIY Poor Man's 1073 EQ-gyrator_update_02_reference.jpg   DIY Poor Man's 1073 EQ-gyrator_update_03_36r.jpg   DIY Poor Man's 1073 EQ-gyrator_update_04_356r.jpg   DIY Poor Man's 1073 EQ-gyrator_update_01.jpg  
Old 25th October 2018
  #7
I've been playing around with the simulation a bit more. And while we're at it, why not compare it to the real deal right away? As it turns out, the center frequency of the sub <100R gyrator configuration was already shifted and increasing the series resistance, as you suggested, brings it mostly back in line with the inductor configuration.

The available gain will be a bit lower, but this will only become apparent around the last +-/5% of the gain pot. Lowering the 2k7 resistor to 2k4 or 2k2 would increase the maximum q again, but keeping the already sensitive headroom in mind, I don't think it would be that big of a deal to have 'only' 15dB of gain available
Old 25th October 2018
  #8
Lives for gear
That's excellent! You could also play a bit with the resistor below the gyrator (R7) which should be easy to fine-tune since you've done it with two resistors in series, but you've obviously done a great job nailing the filter shapes and, given the tolerances of caps, pots and inductors, I wouldn't be surprised if your circuit was closer to average than some actual 1073 units out there.

I've played some more with the simulation too and, other than the input and output buffers, I don't really see much more room for improvement. LTSpice always underestimates THD by about one or two orders of magnitude, but it's good to run it just to see that the circuit is behaving ok and I'm getting 0.000024% with flat controls and 1V input. As for noise (which LTSpice pretty much nails) I'm getting a more than respectable -86dB. For scaling resistors down / caps up to make a significant difference, you would have to go to values that would load even the best opamps too much, and using e.g. LM4562 for the mid section gyrator will only buy you another couple of dBs, so it may not be worth it.

As for the buffers, although generally it makes sense to put the make-up gain at the input one for noise reasons, this will eat up some headroom and in fact it looks like putting it at the output only increases noise by 1dB or so (the dominant contribution is still the mid gyrator), so I would put it there.
Old 25th October 2018
  #9
Thanks man I was just about to update my reply above. I had enough space left on my perfboard, to swap out the series resistors and capacitors of the mid band gyrators. But as I didn't have all desired resistors around, I had to improvise a bit. Unfortunately the margin of error of my inductors is now 10x as high as before and has jumped from 0,1% to a 'whopping' 1%

How did you actually figure out, that the mid band gyrators were the source of the noise? I noticed before, that the SNR was quite bad actually, but I thought, that this was due to my really messy perfboard layout and the unbalanced interfacing
Old 26th October 2018 | Show parent
  #10
Lives for gear
Quote:
Originally Posted by krabbencutter View Post
How did you actually figure out, that the mid band gyrators were the source of the noise?
When you run a noise simulation (.noise V(output) V3 oct 24 20 20k), if you probe the output you get the overall noise (and if you ctrl-click on the name of the trace it integrates for you the total RMS noise over the simulation bandwidth), but then you can also click on each resistor and see a graph of its contribution. I just poked around, saw that R6 was the main culprit, scaled its value and that of C2 to keep the inductance fixed, checked that the response was more or less the same, run the simulation again and saw that whopping 20dB improvement.

Messy perfboard? I love those! Pictures!
Old 26th October 2018
  #11
Here you go :D
Attached Thumbnails
DIY Poor Man's 1073 EQ-img_20181026_082726848.jpg  
Old 28th October 2018
  #12
Played around with the I/O stages a bit, images attached. What do you think?

I've also updated all files on my github to reflect the recent changes and included a full set of measurement data. The 10nf cap in the simulation has also been fixed, thanks for the find
Attached Thumbnails
DIY Poor Man's 1073 EQ-cheaps73_inputbuffer.jpg   DIY Poor Man's 1073 EQ-cheaps73_outputstage.png  
Old 29th October 2018
  #13
Lives for gear
I've realized there's something I overlooked before, DC across the pots that may cause scratching noises when moving them. This may be ok now but generally it gets worse over time.

The problem is the high input bias current of the NE5532 and bipolar opamps in general (with the notable exception of the LM4562), which is why you rarely see them used in Baxandall type tone controls. With the circuit as it is, both the bass and mid pots will have very significant voltages across them, ~30 and ~50mV respectively. You really want to keep those voltages below 1mV or so.

An easy solution would be to use something like an OPA2134 for the high/low shelf stage and for the top opamp in the mid stage. It doesn't have the stellar drive capability of the 5532 but it can still drive 600R loads with very low distortion, and interestingly the impact on noise is negligible (the mid gyrator is still dominant noise-wise). Or, if you don't mind the expense, the LM4562 is as good as it gets and its input bias current is unusually low for a bipolar, so it would work too.

The buffers look fine but you should use a DC blocking cap at the input (10u will work) and, if you use NE5532s, also between the output of the input buffer and the next stage. Again 10u will work here. As for the output buffer, it's already DC-blocked at the input with the HPF engaged but, according to the schematic, not if the HPF is bypassed, so you should put another 10u cap in the bypass link, otherwise the output DC offset will jump when engaging/disengaging the filter and you may get switching noises. In my simulation the gain is a bit too high, I get it to make up for the insertion loss exactly with R62=20k and R63=6.8k and, since the parallel of those is ~5.1k, output DC offset is negligible even with the 5532, so no output cap is needed. Also a 10p film or C0G ceramic cap across R63 is always a good idea for stability.
Old 29th October 2018
  #14
Someone over at groupdiy asked, if I thought about making pcbs. I guess "not as long as this one guy over at gearslutz keeps finding issues I didn't even know existed in the first place" would make for an appropriate answer
Srsly tho, thanks for contributing so much technical knowledge to this project!

- 10uf decoupling caps were already in place, I just didn't draw them in
- I'll give your gain setting a try. My current values were a rough estimate and kind of a compromise, because the EQ is never truly flat.
- shouldn't those 4M7 resistors already prevent any switching noise in the HPF? Might be I'm wron on this, I just implemented them, because they're also used with the frequency switches
- speaking of the HPF: when I updated the BoM yesterday, I realized I overlooked a 10k resistor in the original B182C circuit I've added it to the schematic, but didn't have time yet to check, what the effect will be
- the amplifier for the High/Low & Presence is a single dual opamp, so replacing that one with an lm4562, which would only cost 3 bucks more, would be absolutely fine. I'll make sure to order some, to give it a try!
By the way: that made me realize that not a single component (pots, switches & psu aside) of this circuit did cost me more than 1€ so far
Old 30th October 2018 | Show parent
  #15
Lives for gear
Quote:
Originally Posted by krabbencutter View Post
Someone over at groupdiy asked, if I thought about making pcbs. I guess "not as long as this one guy over at gearslutz keeps finding issues I didn't even know existed in the first place" would make for an appropriate answer
I've run some stability simulations and all looks good on that front. Don't forget that 10p cap across R63 and of course proper supply bypassing as close as possible to each opamp, as well as 2x100u caps to ground at the supply entry point on each separate PCB. The NE5532 is pretty tame and a single 100n (use ceramic X7R caps for this) directly between the V+ and V- rails seems to work best, although I've also used it with the more conventional 2x100n from each rail to ground with no problems. The LM4562 is touchy and it seems to prefer one 100n between rails and another one from V- to ground. Usually when making a PCB I provide for the two caps from each rail to ground just in case, and if I need one between rails, I solder it directly to the pins on the copper side so the connections are as short as possible.

Quote:
- shouldn't those 4M7 resistors already prevent any switching noise in the HPF? Might be I'm wron on this, I just implemented them, because they're also used with the frequency switches
They're good to make sure the caps hold no charge when switching between them, but in the HPF bypass position they are shorted out and, since the output buffer is now DC coupled to the previous stage, the output DC offset will jump from pretty much 0V to about 2.4mV. It's not a lot but depending on the gain of what comes after the circuit you may get a subtle click or a loud pop, so better to put another 10u cap there too.

Quote:
- the amplifier for the High/Low & Presence is a single dual opamp, so replacing that one with an lm4562, which would only cost 3 bucks more, would be absolutely fine. I'll make sure to order some, to give it a try!
Great! I attach the model I use in case you want to play with it in your simulations. It's the old one from National, TI has a newer one but it's buggy and this one works just fine.
Attached Files
File Type: txt LM4562.txt (3.0 KB, 149 views)
Old 31st October 2018
  #16
I was trying to get a better understanding of the baxandall-problem and stumbled upon this
Quote:
[...]Substituting a bipolar opamp such as the 5532 will improve the noise ad distortion performance markedly, at the expense of the need to make provision for the much greater bias currents by adding DC-blocking capacitors.
Source: Small Signal Audio Design - Douglas Self - Google Books
Does that mean I could simply add some caps on both outputs of the U1A/B opamp instead?

I also tried your gain values, but the ouput signal ended up being a bit too low
Attached Thumbnails
DIY Poor Man's 1073 EQ-cheaps73_makeupgain.jpg  
Old 1st November 2018 | Show parent
  #17
Lives for gear
Quote:
Originally Posted by krabbencutter View Post
I was trying to get a better understanding of the baxandall-problem and stumbled upon this
Source: Small Signal Audio Design - Douglas Self - Google Books
That's an excellent book, highly recommended. You may already know it but there's also this free one by Walt Jung, opamp specific and a great both as a reference and for its practical content.

Quote:
Does that mean I could simply add some caps on both outputs of the U1A/B opamp instead?
It's not only the DC offsets at the outputs of the opamps but also the fact that they are taking their input bias currents through a path that includes the pots (except the high shelf, which is DC-isolated) and, in the case of the shelves opamp, also the required negative feedback at DC that keeps the output at ~0V. This means that you would have to:

- DC-isolate the bass and mid pots with caps large enough that they won't have an effect on your carefully tailored filter responses for any position of the pots

- For the shelves opamp, provide DC feedback by putting a resistor in parallel with the 470p cap, again large enough not to have a significant effect on your FR

- For the mid gyrator, put a resistor to ground at the + input to provide the input bias current, again large enough etc.

You wouldn't really need one cap at each opamp output and one at each pot terminal, but I don't think you'd be able to get away with less than a couple of caps in each section placed at the right spots. I'd probably just use the LM4562 and be done with it rather than complicate the circuit more than it is, but you can simulate and see, it's always good to experiment.

Quote:
I also tried your gain values, but the ouput signal ended up being a bit too low
That's strange, there must be some extra signal attenuation somewhere... Anyway, if your original 10k/6.2k worked well, 13k/8.2k will give you the same gain and the parallel is close enough to 5.1k to keep the output DC offset in the uV range.
Old 1st November 2018
  #18
Thanks for your explanations, that definitely sounds like way too much trouble. My LM4562 opamps should arrive next week

That Opamp handbook from AD is great! I stumbled upon it a while ago and should now be in a much better position to make sense of it.

I also think, I found the issue with the gain settings. Adding an Opamp on the output messes with the loading of the HPF. Keeping the 5k1 load resistor on the HP output in place and adding a 10k load to the opamp input seemed to get the gain & filter shape back to where they were. I guess your values should work now?

While I was looking through the og schematics, I noticed, that there was a 22uF bypass cap in the off position of the HPF all along and I was simply too ignorant, to pay it any attention :D

I've added all this to Rev:06c on my GitHub.
Old 1st November 2018 | Show parent
  #19
Lives for gear
Quote:
Originally Posted by krabbencutter View Post
I also think, I found the issue with the gain settings. Adding an Opamp on the output messes with the loading of the HPF. Keeping the 5k1 load resistor on the HP output in place and adding a 10k load to the opamp input seemed to get the gain & filter shape back to where they were. I guess your values should work now?
I'm not sure, I attach my sim file (basically yours but with the input and output buffers) for you to check. Using the 5.1k as the input bias resistor for the output buffer and the 20k/6.8k values gives the required 0dB gain. If I add another 10k there the gain drops, as you would expect, and the curve is not as flat, though I don't know which is the right shape... Maybe there's some component (probably a resistor) that's different in the sim/schematic vs. the real thing?
Attached Files
File Type: zip gs1073.zip (2.0 KB, 30 views)
Old 1st November 2018 | Show parent
  #20
Quote:
Originally Posted by Cabirio View Post
If I add another 10k there the gain drops, as you would expect, and the curve is not as flat, though I don't know which is the right shape...
If it's not flat, it's the right one
see UA WebZine "Plug-In Power" December 06 | Why Do I Need Both the 1073 and the 1081?
The 1073 HPF is kinda chebyshev-ish (?) in it's response, which can be best observed at the 300Hz setting.

I'd send you a simulation, but LTspice is freaking me out right now xD
Old 1st November 2018
  #21
Lives for gear
I see! In that case your values definitely work better. Try this: merge the two bias resistors (I mean the existing 5.1k and the additional 10k) into a single 3.3k one (why not) and use 10k/5.1k in the buffer. I think that's as close as you will get to the right FR shape, 0dB gain and no DC offset at the output.
Old 1st November 2018
  #22
I think we got it
Update: frickin image compression! https://1drv.ms/u/s!Aob75XFkRCIip4F5jXfQDPHiZi5f7w
Curves 1-3 are the HPF and 4-5 are the unity gain measurement
Attached Thumbnails
DIY Poor Man's 1073 EQ-cheaps73_output_debug.jpg  
Old 4th November 2018
  #23
As I'm still waiting for my LM4562 delivery, I was giving the TL072 a try, just to see what would happen. Because, If I understand that segment from Douglas Self's book correctly, it should minimise the bias currents at the expense of performance.

And I hate to admit it, but from what I can tell, it's kicking the NE5532's butt, when used in the U1A/B position
- SNR was on par with the NE5532 and I was not able to measure any increase in the noise floor with all bands at unity / full boost / full cut
- the measured DC across the mid pot (33mv) and low pot (14mv) however was reduced to <1mv

Is this too good, to be true, or am I just unaware of some potential pitfalls with the TL072 in that position? I mean, sure, an LM4562 should still result in a better/safer performance overall but could the TL072 be considered a 'gets the job done' alternative?
Old 4th November 2018
  #24
Lives for gear
Quote:
Originally Posted by krabbencutter View Post
- SNR was on par with the NE5532 and I was not able to measure any increase in the noise floor with all bands at unity / full boost / full cut
- the measured DC across the mid pot (33mv) and low pot (14mv) however was reduced to <1mv

Is this too good, to be true, or am I just unaware of some potential pitfalls with the TL072 in that position?
It's what you would expect given that:

- as we've found out, noise is dominated by the mid gyrator, so even though the TL072 is much noisier, it has negligible effect

- the NE5532 is a bipolar and the TL072 a JFET (the LM4562 is also a bipolar, but as I said it has an extremely low input bias current, so it works fine here)

The main pitfall of the TL072 is what I said in post #2 above about the TL082, its poor drive capability, and that's very relevant in this particular circuit due to the loads the opamps see.

You can only trust datasheets so much, so how about some real measurements: check out this very thorough analysis of opamp distortion and compare the comments and figures for the LME49860 (identical to the LM4562) and TI TL071 (the single version of the TL072), especially the figure at the top right for each opamp. Even at a light 2.2k load, the TL averages 0.25%, while the LM starts at 0.01% and rises to 0.06% at 20kHz, and with a 600R load it isn't much worse, while the TL goes above 5% (!). While you're at it, you can check out the NE5532 too: not as spectacular as the LM4562 for sure, but still in a different league.
Old 7th November 2018
  #25
Thanks, completely forgot that you mentioned the TL opamps already at the beginning of this thread. Anyway, LM4562 is up and running and DC is a bit higher but still below 2mv. The only issue I'm still having is popping of the mid band selector. This might be, because I have to switch the ins & outs, due to the layout of the gyrators, whereas the original circuit needs only switching on the inputs.
But I've been thinking about replacing the mechanical switches with encoders + ICs anyway. This would enable recall & automation and the amount of center frequencies would not be limited by the switch anymore. But I'm still unsure about what ICs would be the best option for this. Guess I need to familiarize myself with the AD product lineup & LTSpice a bit more
Old 8th November 2018 | Show parent
  #26
Lives for gear
Quote:
Originally Posted by krabbencutter View Post
The only issue I'm still having is popping of the mid band selector. This might be, because I have to switch the ins & outs, due to the layout of the gyrators, whereas the original circuit needs only switching on the inputs.
Yup. That's one of the problems with simulation, you don't see the whole circuit, switching and all... The gyrator opamp is taking its input bias current through the pot wiper, so when you switch it out and effectively add 4.7M to the bias path, DC offset will jump quite a lot. I'll simulate later at home to see but I suspect that even the LM4562 will draw too much current to avoid this. Off the top of my head, the LT1113 would be ideal here, it's the lowest noise JFET opamp I can think of (in fact even better than the NE5532) in a dual PDIP package and Mouser has stock at ~€5. The model should be included in the ltc.lib standard library, if you want to give it a try.

On second thought: have you tried not switching the pot wiper, just the top of the gyrators? Simulation and measurements will tell, but I don't think some extra gyrators connected there but open at the top should affect the filter shape and that would avoid the DC offset jumps...
Old 8th November 2018 | Show parent
  #27
Quote:
Originally Posted by Cabirio View Post
Yup. That's one of the problems with simulation, you don't see the whole circuit, switching and all... The gyrator opamp is taking its input bias current through the pot wiper, so when you switch it out and effectively add 4.7M to the bias path, DC offset will jump quite a lot. I'll simulate later at home to see but I suspect that even the LM4562 will draw too much current to avoid this. Off the top of my head, the LT1113 would be ideal here, it's the lowest noise JFET opamp I can think of (in fact even better than the NE5532) in a dual PDIP package and Mouser has stock at ~€5. The model should be included in the ltc.lib standard library, if you want to give it a try.

On second thought: have you tried not switching the pot wiper, just the top of the gyrators? Simulation and measurements will tell, but I don't think some extra gyrators connected there but open at the top should affect the filter shape and that would avoid the DC offset jumps...
Oh well, that was too easy I was so focused on the og schematics that it didn't even occur to me, to keep the pot connected and just switch before R18. I just ran the simulation with all gyrators connected to the pot and it doesn't have any effect on the filter response.
Old 8th November 2018
  #28
This is it! Rev: 07 is now on my Github. I've modified & tested my perfboards and everything's working like a charm! No scratching pots, no popping switches. Build documentation will be next and maybe (but just maybe) I'll try a PCB layout. Anyway, thank you so much for your help & patience!
Attached Thumbnails
DIY Poor Man's 1073 EQ-wp_20181108_19_31_23_pro-2-.jpg  
Old 10th November 2018
  #29
Lives for gear
My pleasure, it's been a lot of fun! I'm glad all is working well now. A PCB would be great and shouldn't be too difficult, it looks like it could be done single sided (if you don't mind the odd jumper here and there) and small enough to fit neatly in a 500 module or small Eurocard box.

If you go for it, here's a tip: place the opamps first (in this case arranged in two rows of 3 would make sense), then layout the supply traces (V+ / GND for the bypass caps / V-) as short and straight as possible (all the way through the opamps, GND passing uninterrupted between the two rows of pins of each opamp, as in the attached example), place your local supply bypassing as close as possible to each opamp, then layout everything else around that. Pay attention to grounding too: join that ground with the signal one only at the point between the 100u bypass caps at the point of entry of the supplies or, if you include a power supply in the box, at the star ground there. Interestingly, opamps that work rock solid stable on a messy perfboard can start oscillating on a PCB due to poor supply layout, I learnt that the hard way a long time ago...
Attached Thumbnails
DIY Poor Man's 1073 EQ-pre-pcb.jpg  
Old 12th November 2018
  #30
Gear Nut
I'd just like to thank krabbencutter for sticking with this! I had the same idea around a year or two ago but got bogged down somewhere along the way. I just tried to find my LTSpice files from back then but they seem to have vanished into the aether :(

I may have to buy in some parts and build one of these up! Don't hold your breath though; too many hobbies, too little time!
πŸ“ Reply
Topic:
Post Reply

Welcome to the Gearslutz Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…
πŸ–¨οΈ Show Printable Version
βœ‰οΈ Email this Page
πŸ” Search thread
πŸŽ™οΈ View mentioned gear
Forum Jump
Forum Jump