ADA MP1 EQ portion

[caspercody]

Hello,

I want to build the EQ portion of the ADA MP1, but not sure if if will work? I have attached the schematic showing the EQ portion, and a datasheet for the TC91670 used for the pots. I made a power supply that provides me +16vdc, and -16vdc from a charge pump circuit. I do have these voltages at pin 8, and pin 4 on the IC chips, but zero voltage at the other pins?

How do I connect normal 100K pots instead of using the TC91670 chip?

Do I need a voltage coming in somewhere on the circuit to get voltage to the other pins? I am so use to seeing circuits with just a +vdc and a ground, then a reference (usually half the total voltage) voltage being applied to the op amp input

Thanks
Rob

http://www.datasheet-pdf.com/PDF/TC9170AP-Datasheet-Toshiba-553376

[Rob Strand]

The structure of that EQ circuit isn't too uncommon.

Just wire the pot ccw and cw lines to the "bussed" lines which are already there then
wire the pot wipers to corresponding series R+C networks which hang off the outputs
of the filters.  That's it.

The resistors around opamps handle the biasing.

You can see variations on the structure here,
http://www.dator8.info/pdf/constant/1.pdf


You will need to add a 5k1 in series with C65.  That forms the input line.  The resistor might be on the other sheet.

[Rob Strand]

I went to look at your layout and realized it has 4 opamps and not 6.

The EQ section is this,




You can clean-up the out line with the usual 470 ohm in series with the cap and 100k to ground.

[caspercody]

Thanks Rob for your relpy!

Yes the vero board I made is only the EQ section, it is basically this, with original values:



I am not sure if I do need the other (2) opamps, but based on your reply, I do. Here is a quick hand drawing I made of how I see to add the other (2) opamps:



What do you think??

[caspercody]

Oops, I did not see your last reply before I added mine. Please see corrected hand drawn layout. I am curious about the (2) components I put a question mark by, are these needed? Do the POTS look correctly connected?

[Rob Strand]

I am not sure if I do need the other (2) opamps, but based on your reply, I do. Here is a quick hand drawing I made of how I see to add the other (2) opamps:

Definitely need the other two opamps.   The two opamps are the core of the EQ.  The others are the band filters.

What do you think??

Right idea but needs a few kicks of the tires:
- at the input the 5k1 is in series with the cap; currently drawn in parallel.
- at the output 20k in series with the output is too high, 470 ohm will do.
Your update looks OK.  If you are driving a cable keep a 470R where you had the 20k at the output.

A finer point is the 6k2 on the last stage.   That give a small amount of gain to the EQ.
You can change that to 5k1 for unit gain.   The change will not affect the EQ.
You got that too.

The way it works is this:

For boost mode the filter signals add/mix with the main signal at the last opamp.

For cut mode the filter signal feedback from the output of the first opamp back to the input of the first opamp.
When you do this is creates a cut characteristic which is the mirror image of the boost characteristic.

[Rob Strand]

Sorry I'm a bit out of phase with your replies.

. I am curious about the (2) components I put a question mark by, are these needed? Do the POTS look correctly connected?

Pots look fine.

The 5k1 is where the dry signal feeds through.  In boost mode the filters add to that signal.
(also with this 5k1 cut mode would not pass signal)

The 10uF cap marked (?) is required, that's the feedback path for cut-mode.
This 10uF cap and it's brother 10uF cap at the input of the second opamp are there to help
keep DC offsets down.   Since each filter has a series cap you could remove the 2x10uF,
they are kind of a precaution.  You should keep the 10uF caps at the output of each filter
to prevent scratchy pots.

[caspercody]

Thanks!!!!!!

[Rob Strand]

No worries.  Good luck with it.

FYI, you can very easily put different filters into these types of equalizers.   In the past I've tried different filters by switching the part values or by switching to another filter.

If the pot values are a lot larger than 5k1 (the resistors used around the two main opamps), the control can get cramped near the ends of the rotation.

[caspercody]

I was going to use 100K pots for the controls, is that to high

[Rob Strand]

I was going to use 100K pots for the controls, is that to high

Well it gets a bit messy now.  Here's the dilemma:  Ideally you want 100k so it doesn't change way the EQ works but 100k linear pots are going to cause a cramped control.

If you change the pot values  it has some impact on the way the tone control works.  When you boost and cut more than one band they combine a little differently.

If you look at the datasheet for the digital pot chip you can see the resistance steps aren't even.   So they have allowed for the cramping.

In terms of copying the behaviour of the original there are graphic equalizer taper pots, G taper (for Japanese tapers), which are specifically designed to "hide" the control cramping.  Other companies offer the same taper with a different letter.



So you could draw the line between those two options.

For these EQs, in the past I've used 10k feedback resistors and 10k pots and it works fine.  It helps to have low noise opamps for the main two opamps as low value pots can boost the noise floor a little bit.    In other cases to avoid cramping I've tried to make the pots no more than 4 times the feedback resistors:  in your case 5k1 feedback resistors, 4x5.1k = 20k to 25k pots.

So we have a few choices:
- 100k G tapers.   Best way to mimic original EQ.   Possibly not so easy to get from your normal supplier.
- 100k linear.  Mimic original but cramping will be annoying.
- 50k linear.   Lose a  bit of originality, but a little less cramped
- 20k/25k linear.  Usability OK, but some impact on response.
- 10k linear.  Good useability,  small increase in noise, possible different behaviour to the original.

Not a clear choice because it's about making a compromise.
You would need to do a spice simulation to judge how much a 20k linear differs from the original response.

[Rob Strand]

Here's a simplified example of the pot camping issue.

EQ +/- 15dB.

When we set the pot to half boost or cut we expect about +/- 7.5dB.

In reality the amount of boost or cut is somewhat less.   

Plots are for pot values: 10k, 20k, 50k, 100k.



For the 100k case you only get +1.6dB boost instead of +7.5dB boost and the other (15dB-1.6dB) = 13.4dB is cramped  into the last 1/2 turn of the control.

[caspercody]

Thanks, what value would you recommend to use

[Rob Strand]

If you just want to use easy to get linear taper pots and you aren't ultra-fixated on matching the original to below the 1dB level then the 20k/25k option should have acceptable cramping and the response won't be too far off the original.   I'd call that the best approximation option.

[m4268588]

When couldn't get "S taper", I thought about VR=R.
VR may be too small.

[Vivek]

Anyone have the frequency response curves of the mid control ?

[Vivek]

Great project CasperCody !!!

I wish you success !

I have been looking at the Studer 169 EQ

https://groupdiy.com/threads/studer-169-961-eq-again-stepped-midrange-4x-in-1ru.63563/
https://groupdiy.com/threads/studer-169-eq-in-api-500-format.36522/
https://groupdiy.com/threads/diy-equalizer-based-on-studer-169-standard-and-mastering.33996/

[Rob Strand]

I have been looking at the Studer 169 EQ

It's like a modded Baxandall EQ.   I know I've played with stuff like that in the past, mainly when experimenting with 3-band 1-opamp EQ's.   Not sure if it offers anything over the Baxandall - the back of my mind says not.

[caspercody]

I go it to work, thanks!! Here is a video of turning each EQ pot showing the range:

https://youtu.be/IE0w_BHDrbg

I have the MP1 input buffer, going into the U4 opamp stage (page 5), then to the U18 opamp right before the EQ opamps, then to U13 output opamp after the EQ (page 7).

I removed the 20K resistor in the U4 opamp stage and wired the 20K pot to pin (2) of U4 (page 5). This allowed me to use it as a volume control, but in the video even with it set to minimum it still has sound coming through. It is very loud as I turn up the 20K pot.

I want to try to recreate the distortion channel of this thing, but have questions:

1 - What does U5, and U3 (ZETA C4) do, or how do I replace these? Page 5
2 - Is the circuit starting at U6 needed? Page 5
3 - Bottom left of Page 5, who would these pots be wired instead of using U14 (TC9176 - link to data sheet below)
4 - HDRM is mentioned at both U4's, what is this?

http://www.datasheet-pdf.com/PDF/TC9176P-Datasheet-Toshiba-506423

[ElectricDruid]

1 - What does U5, and U3 (ZETA C4) do, or how do I replace these? Page 5

It looks like dynamics control to me. The resistor (LDR presumably) inside the ZETA acts as the input resistor to the inverting-op-amp-based U4 distortion stage. *I think* more signal coming in would increase the brightness of the LED, decrease the resistance of the LDRs, and therefore increase the gain, but that's the inverse of a typical compressor. So it might be a log amp? Or I might have it back to front? U9 is acting as a basic half-wave rectifier to drive the ZETA opto's LED though, that much is clear.

2 - Is the circuit starting at U6 needed? Page 5

The whole U6/U7 stage seems to be a more-complicated-than-usual compander built around the '572. As such, it's for noise reduction. If you've eliminated some of the noisy parts of the circuit, perhaps you can live without it. Since it will mess with the dynamics and the levels going to other parts of the circuit, it will affect the sound in several complex ways, so it depends how authentic you want to be.

3 - Bottom left of Page 5, who would these pots be wired instead of using U14 (TC9176 - link to data sheet below)

That weird arrangement of dual pots with one hanging off the wiper of the other, you mean? It could just be a way to get a "log/audio" response from a chip that only has linear pots. You'd have to do some calculations or a sim to know for sure - plot a graph of the response when both sections move in tandem. But it's possible that it would simplify to just one ordinary pot.

4 - HDRM is mentioned at both U4's, what is this?

I'd guess it's level monitoring, perhaps a bargraph meter or perhaps just a simple clip LED.