Nyan 1073 EQ =^x^= (formerly Cheap's73 / Poor Man's 1073)

[MickMad]

Hey there, great project! I was myself looking at gyrators and eq designs to make my own dream eq and I stumbled upon this, you did a nice job, especially so for being a "noob" in DIY electronics! Keep it up!

If I may, I'd suggest using a Teensy 3.2, it's much more powerful and faster than an Arduino Uno; it also has the possibility to change USB functionality (like USB MIDI) by just choosing an option in the IDE It may make your life easier!

 

[Rocinante]

I agree with using the teensy. You can still use arduino coding with it too.

 

[krabbencutter]

Hey there, great project! I was myself looking at gyrators and eq designs to make my own dream eq and I stumbled upon this, you did a nice job, especially so for being a "noob" in DIY electronics! Keep it up!

Hey, thanks a lot

About the teensy: I know everybody loves them, but take a look at my current main:

void loop() {

  val1 = analogRead(analogPin1);
  val2 = analogRead(analogPin2);
  val3 = analogRead(analogPin3);

  Serial.println(val1);
  Serial.print(val2);
  Serial.print(val3);
  
  display.clearDisplay();
  display.setTextSize(2);
  display.setTextColor(WHITE);
  display.setCursor(0,0);
  
  display.println(val1);
  display.println(val2);
  display.println(val3);
  
  display.display();
  
  delay(1);
}

The arduino is mostly busy doing nothing and a teensy would just do 'more nothing' but for double the cost.
Although I'm having a couple small additions to the code in mind, it would still be just very basic stuff.
So from my current perspective, a cheap arduino nano clone is all that is needed.

However ... higher resolution ADCs could of course be a benefit but there's a catch:
although I'm using ALPS pots, there's still quite a bit of variance between the channels. So unless you know how to carefully measure & calibrate each potentiometer and input, a better ADC will just give you incorrect values at a higher resolution ;D
And remember, whatever you do, somebody will eventually come along and shout "use your ears!" anyway 

 

[krabbencutter]

Quick update:
- I've finally re-built my PSU and the voltage regulators are now inside the EQ case
- while reworking the power supply I've also put some more thought into the grounding layout. The output of the voltage regulator PCB is now the common point where Chassis/Signals/Power/Earth come together

'Unfortunately' this EQ is so quiet now, that I can't accurately measure THD+N with my audio interface anymore (Presonus Studio 26). THD+N is at minimum -75dB but if I increase the input signal even further, I'm ending up at the same -90dB as my audio interface.

 

[krabbencutter]

Thing's are looking good  8)

The case is complete, although a bit sketchy. I might build a new one someday, but for now it's sturdy enough and sufficiently protects against any outside interference. The OLED display has been moved to the front, which taught me an important lesson about shielded cables ;D

Yesterday, I've updated the arduino sketch. Potentiometer readings are now smoothed and calibrated to compensate any 0dB offsets in my unit. Display readout is now in dB, the values however are just basic approximations.
Given that the advertised gain values of a real 1073 are all over the place anyway, I'm fine with that for the time being

Now that everything's in place I've also uploaded new measurement data. If you look at the mid band, you'll notice that all frequency settings are actually reaching full +/-18dB.
This is because each frequency setting has a trim-pot, which makes it possible to calibrate the maximum available gain for each band (and of course it would be just as easy to recreate the default behaviour).

Next up:
- I'm going to order some new knob caps
- If I ever re-build the case, I'll add the frequency selectors & the HPF switch to the arduino as well
- I'm thinking about doing a eurorack compatible 'light' version called tiny73 (single PCB, no HPF, 2 selectable frequencies for mid & low)

 

[TheseGoToEleven]

Amazing project! Love the digital recall.

Could I ask about a couple of resistors in the hi/lo shelving section? The 1073 schematic that I've seen has 39k from the B205's input pin F to ground. On your KiCad schematic at GitHub, there is a 39k but it is going from pin K to ground instead. Similarly, KiCad has a 12k (R6) from pin V to ground, which is K to ground on the 1073. Just wondering about the reason for this difference.

 

[krabbencutter]

Amazing project! Love the digital recall.

Could I ask about a couple of resistors in the hi/lo shelving section? The 1073 schematic that I've seen has 39k from the B205's input pin F to ground. On your KiCad schematic at GitHub, there is a 39k but it is going from pin K to ground instead. Similarly, KiCad has a 12k (R6) from pin V to ground, which is K to ground on the 1073. Just wondering about the reason for this difference.

Nice find. I accidentally swapped the 12K & 39K resistor on the latest revision. And pin F ... well I gotta admit that I might have had it all wrong right from the start and never noticed it. I will look into it and update the repository. Thanks a lot

 

[TheseGoToEleven]

Nice find. I accidentally swapped the 12K & 39K resistor on the latest revision. And pin F ... well I gotta admit that I might have had it all wrong right from the start and never noticed it. I will look into it and update the repository. Thanks a lot

Glad to help. You seem to have had great results with your build. Which way did you build it?

In my own simulations of the 1073 high/low shelving section, the 39k from B205 pin F to ground doesn't seem do much of anything to the frequency response, at least not above 5 Hz. So I am curious what its purpose is.

The value of the 12k from B205 pin K to ground definitely affected the low frequency shelf in my simulation, shifting it left or right. It was especially noticeable for moving the peak that I've circled in yellow (this is with the low frequency knob set to 200nF / 35Hz.)

 

[krabbencutter]

I finally found the time to revisit the circuit today and have updated the schematics. Th 39k resistors are gone now. The arduino sketch has also been updated with a "dead zone" variable, which makes it easier to keep the display at 0. Last but not least I've ordered a T1280 Inductor from don-audio.com, so that I can do some more research and maybe find some ways to optimize the gyrator circuits.

 

[Ricardus]

I somehow missed this project this whole time. I definitely want a bunch of these in my rack.

 

[MidnightArrakis]

Hey thanks I have no interest in doing a PCB myself (hence why I put additional effort in making my perfboard as PCB-like as possible) but if anybody wants to pick up this task, I'll be happy to assist.

Aout the Arduino:

HEY!!! Just a short while ago from the time I am writing this message to you, I received a message from someone on this forum who told me to check out this thread of yours, as you are possibly needing some assistance with designing the PCBs for this project. Since I work as a "Senior Electronics Mechanical Packaging & PCB Designer" as a "Consultant" for a number of aerospace/avionics companies, defense contractors, medical electronics firms, NASA, R&D laboratories >> AND!!! << I have also once worked with "one of the largest concert sound-reinforcement companies in the world".....I believe that I may be able to help both you and the followers of your thread here!!! Essentially, my job is to take -- completed schematics -- and turn them into a "working physical reality". I will attach a couple of PDF files that will show you some of the electronic equipment and PCBs that I have personally designed.

My PCB design software is a "professional" industry-standard program (i.e., not FREE!!!) but it can also import both the KiCAD schematics and PCB design files. So, if you could share any of your KiCAD design files, along with a detailed BOM with me, I could review them and see what it would take to get started with designing your NYAN 1073 EQ PCBs. If any of the followers of this thread only wish to purchase a bare PCB and work out the rest of the project for themselves, then that could be possible as well.

Since I am -- also -- a "Senior Electronics Mechanical Designer", if you could either provide me with or direct me to a place where I can obtain the necessary mechanical details about the 1073 and/or about the Euro-Rack enclosures, then I could also start designing the required mechanical details for this project. Due to the nature of my work, I design electronic equipment > as a piece of "real" electronic equipment < and -- NOT -- as a thrown-together basement job!!! So, the more data and information that you can throw my way, the better I can get a handle on what it will take to make your project a "reality"!!!

/

 

[oran.outan]

Hi

I understand it's made and work for +12/-12v but could it work fine on +16/-16, like in a 500 series ? I see on previous post that voltage affect headroom and some question about that...
thanks

 

[Ricardus]

I just read the GearSpace thread. Holy cow that user Cabirio helped a lot. Awesome work from the both of you.

NOW TAKE MY MONEY!!!!!

 

[krabbencutter]

Hi

I understand it's made and work for +12/-12v but could it work fine on +16/-16, like in a 500 series ? I see on previous post that voltage affect headroom and some question about that...
thanks

Sure, +/-16V is fine for the NE5532 as well as the LM4562 and it should give you more headroom. But to put things into perspective: the original 1073 operated on a single +24V power supply, so don't worry too much about headroom. Use whatever you have around, or if you're buying/building a new power supply, use the voltage that makes the most sense for your projects.

 

[krabbencutter]

HEY!!! Just a short while ago from the time I am writing this message to you, I received a message from someone on this forum who told me to check out this thread of yours, as you are possibly needing some assistance with designing the PCBs for this project. Since I work as a "Senior Electronics Mechanical Packaging & PCB Designer" as a "Consultant" for a number of aerospace/avionics companies, defense contractors, medical electronics firms, NASA, R&D laboratories >> AND!!! << I have also once worked with "one of the largest concert sound-reinforcement companies in the world".....I believe that I may be able to help both you and the followers of your thread here!!! Essentially, my job is to take -- completed schematics -- and turn them into a "working physical reality". I will attach a couple of PDF files that will show you some of the electronic equipment and PCBs that I have personally designed.

My PCB design software is a "professional" industry-standard program (i.e., not FREE!!!) but it can also import both the KiCAD schematics and PCB design files. So, if you could share any of your KiCAD design files, along with a detailed BOM with me, I could review them and see what it would take to get started with designing your NYAN 1073 EQ PCBs. If any of the followers of this thread only wish to purchase a bare PCB and work out the rest of the project for themselves, then that could be possible as well.

Since I am -- also -- a "Senior Electronics Mechanical Designer", if you could either provide me with or direct me to a place where I can obtain the necessary mechanical details about the 1073 and/or about the Euro-Rack enclosures, then I could also start designing the required mechanical details for this project. Due to the nature of my work, I design electronic equipment > as a piece of "real" electronic equipment < and -- NOT -- as a thrown-together basement job!!! So, the more data and information that you can throw my way, the better I can get a handle on what it will take to make your project a "reality"!!!

/

Sounds great! You can find the kicad files on my github: GitHub - ravettel/Nyan-1073-EQ: Not yet another 1073 EQ
But I'm planning to rework some parts of the circuit in the coming weeks. The implementation of the mid band gyrator has several flaws and is also technically incorrect. I've ordered a third party 200mH Neve inductor so that I can do a bit more research on a better implementation.

 

[oran.outan]

Hi
Thanks for the reply...
I don’t care at all about the headroom as it’s really not a matter...
Was more to know if tr circuit is fine working on a +/-16v as it’s usually what I use and what I have on my ‘test supply’... and the voltage used on 500 series, as I be interesting to build for...

Good work anyway and thanks
F

 

[Monte McGuire]

You can always check the datasheet for the ICs used in the device. With an op amp circuit, the exact supply voltage is not nearly as critical as it is with a discrete circuit, where the biasing will change sometimes greatly when the supply voltage is changed. An op amp uses current sources to set its internal bias, so changing the supply voltage, within an acceptable range, will change very little. The values to check are the absolute maximum supply voltage (usually ±18-20V), which you need to stay at least a volt or two away from, and the minimum supply voltage (usually ±4-5V), which you also want to stay away from to get high performance. Between those extremes, the only differences when the supply voltage is changed will be output headroom and internal dissipation, usually only important for low impedance loads.

 

[krabbencutter]

My inductor has arrived and I've finally found a solution for the mid band I'm changing the filter topology of the mid band to a more "classical" design* and the single opamp gyrators will be replaced with FDNRs. The downside is, that the circuit is going to need 5 additional opamps. But this should improve the frequency response to the point where it's 99,9% identical to an "ideal" 1073. Noise should see an improvement as well and the gyrators/fdnrs will probably work with TL074s instead of NE5532.

The screenshot shows my current simulation for the mid band EQ, with the original Neve circuit on the left and the improved gyrator circuit on the right. (edit: gdiy's image compression is abysmal. I've attached the simulation as well)

* see Fig.16 RANE Commercial - Knowledge Base

 

[krabbencutter]

My inductor has arrived and I've finally found a solution for the mid band I'm changing the filter topology of the mid band to a more "classical" design* and the single opamp gyrators will be replaced with FDNRs. The downside is, that the circuit is going to need 5 additional opamps. But this should improve the frequency response to the point where it's 99,9% identical to an "ideal" 1073. Noise should see an improvement as well and the gyrators/fdnrs will probably work with TL074s instead of NE5532.

The screenshot shows my current simulation for the mid band EQ, with the original Neve circuit on the left and the improved gyrator circuit on the right. (edit: gdiy's image compression is abysmal. I've attached the simulation as well)

* see Fig.16 RANE Commercial - Knowledge Base

Everything was fine ... until I increased the simulation to 3MHz and discovered some really bad resonance. The breadboarded circuit shows the same behaviour at high-ish gain settings. Adding two 100pf capacitors like here (fig. 5) seems to keep everything in check.

 

[krabbencutter]

After experimenting around for the past days, I really do like the improvement with the GICs. First of all they need a lot less fiddling to match the frequency response of a 1073. You can read a bit more about it here: GIC Transform Fundamentals . The GIC in the screenshot will give you a 200mH inductance while keeping resonances in the MHz range comparatively low. It's also easy to tune because you simply need to change R45.
The gyrators took a lot more trial & error to get the Q & maximum gain right. On top of that the GICs produce a lot less noise. The current gyrator implementation at full boost produces an RMS noise of 174uV, whereas the GIC circuit only produces 24uV of noise. That's still 16uV more than with a real inductor but a 150uV improvement nonetheless.

And now that I've found out about the correct placement for the bypass capacitors, I might be ready to build a prototype soon.