EQ Question with Baxandall and... other?

Started by bushidov, April 05, 2020, 08:02:04 PM

Previous topic - Next topic

bushidov

Hi All,

I got another "how does this thing work" style question.

Below is a schematic of a known working EQ circuit.


I am familiar with the stereotypical "look" of a Baxandall circuit, and I think this one appears to be active, as feeds into an inverting op-amp. However, the first part of this EQ (the Bass and Treble part) looks like what I would see in a text book on how the schematic is laid out. The second part (the one with the mids potentiometer), I am told, is also a Baxandall. I tried moving parts around, but I cannot get it to look like first half. I was told that some of the values are "hard wired" on the mids Baxandall, but I am just not seeing it.

Anyone willing to show me how the mids section could be laid out so that it at least looks like the first section's baxandall, schematically?
"A designer knows he has achieved perfection not when there is nothing left to add, but when there is nothing left to take away."

- Antoine de Saint-Exupéry

swamphorn

#1
This type of treble-bass equalizer comes in two varieties. First is the James tone control which your circuit seems to be based on. I've only seen this as a passive circuit. Your equalizer is still mostly passive as there is no feedback from the op amp; it is only maintaining a virtual ground, isolating the treble and bass sections from each other.

Then there's the Passive Baxandall which has fewer parts at the cost of a smaller adjustment range.

And finally, there's the Active Baxandall, here with a mids control as well.


You can't rearrange your mids control to look like your treble-bass controls because your treble-bass controls are of the James topology but your mids control is Baxandall topology.

1 Images from https://sound-au.com/articles/eq.htm

PRR

#2
> look like first half.

That first half is NOT a Bax. More like a Picasso of a Bax.

> your treble-bass controls are of the James

To even start to be a James it would have to be grounded at the bottom, and feed a hi-Z not a low-Z input. (Yes, a low-Z loaded "James" is possible as a dual, but this is not that.)

"Passive Bax" is nonsense.
  • SUPPORTER

swamphorn

You're right; I'm more referring to the arrangements of the capacitors for taxonomy's sake. If there's a proper term for a treble-bass equalizer with the same resistor-capacitor arrangement as a Baxandall equalizer, but without negative feedback, I'd be happy to use it.

Rob Strand

#4
At first glance it looks like the junction of R9 and C9 should connect to the output of U1B.
I'm fairly sure that's the case.

The resistor R10 then looks odd.  Is it really there?
In some cases you might find this resistor set to a large value  like 1M to 10M,
but in these cases there is usually an extra cap around pin 6; there's a few arrangements.

Example of correctly drawn first stage.  [Re you ckt: without the 10k (R10). The 47pF (C9) is OK to leave in.]
https://www.petervis.com/record_players_and_turntables/baxandall/baxandall-tone-circuit.html
Send:     . .- .-. - .... / - --- / --. --- .-. -
According to the water analogy of electricity, transistor leakage is caused by holes.

bushidov

#5
Alright, I redrew the schematic


QuoteExample of correctly drawn first stage.
So, yeah, I made it look closer to how this one was drawn

QuoteFirst is the James tone control which your circuit seems to be based on.
It appears you are correct. I've been using James tone stack and Baxandall tone stack interchangeably, which I am now discovering was a mistake.

QuoteTo even start to be a James it would have to be grounded at the bottom, and feed a hi-Z not a low-Z input. (Yes, a low-Z loaded "James" is possible as a dual, but this is not that.)
Paul, that seems fair. However, with the way its laid out, what would you call it? "James Tonestack minus a ground and fed into an inverting op-amp for amplification?" I know it physically works, but if I am understanding your posts, I probably should be asking, "why is it working?"

QuoteThe resistor R10 then looks odd.  Is it really there?
In some cases you might find this resistor set to a large value  like 1M to 10M,
Yeah, it is for sure a 10K resistor. The EQ section originally belonged to a Darkglass B7K, but I applied it to an MXR Micro Amp, by adding a DPDT switch. On the Darkglass B7K, it was a 1M resistor, but when I made this add on PCB, I found that it made the volume jump too high when going from MXR Micro Amp only to MXR Micro Amp, plus an EQ section. After reducing the resistor to 10K, I found the volumes "matched" close enough. See below for the full schematic.


So, I guess my question is, "ok, so I know it works...", and I will admit in this case I do not know "exactly" what I am doing, so how is it working and how would one discover what frequencies are being effected and by how much?

I know I based the EQ section off of my Darkglass B7K's Low, Low Mids, and High circuits, which would mean, orginally the low is +-12dB @ 100Hz, the high is +-12dB @ 5kHz, and the low mids are +-12dB @ 1kHz. However, as mentioned above, I changed R10 from 1M (like on the Darkglass B7K) to a 10K to get some sort of "unity" going on between the two switch settings. Also, the Darkglass B7K included a final "high mids" EQ that came after the low mids output pin on the op-amp that focused around +-12dB @ 2.8kHz that I didn't include.

I get typical passive low pass and high pass filters (just basic RC filters). I am also familiar with the filter calculator from Okawa Electric (http://sim.okawa-denshi.jp/en/Fkeisan.htm), but this is the next step of my self-taught education, and I'm going to need a little "schooling" here.

Thanks for being patient with me.
"A designer knows he has achieved perfection not when there is nothing left to add, but when there is nothing left to take away."

- Antoine de Saint-Exupéry

iainpunk

isn't that drawing missing a ground at the tone control?
friendly reminder: all holes are positive and have negative weight, despite not being there.

cheers

PRR

> "ok, so I know it works...", and I will admit in this case I do not know "exactly" what I am doing, so how is it working....based the EQ section off of my Darkglass B7K...... However,..., I changed R10 from 1M (like on the Darkglass B7K) to a 10K to get some sort of "unity" going on

OK, I think I see. It is mis-drawn and apparently mis-wired. The 1Meg should NOT set audio gain. Audio gain is set by feeding-back from amp output to the other end of the pot/cap network. Which means this wire is missing:


Some folks here have {data} network experience. You mess with a lot of them, some trends stick in the mind. If the data-wires don't go to the internet, or don't go to the unhappy-user's desk, the network won't work, or not "right". Maybe it passes signal indirectly through a neighbor's wi-fi. Networks mostly are smart enough to block that. Audio networks are also susceptible to experience. Audio networks not as smart as data-boxes, often more diverse, and mis-wires are prone to "work" through stray paths.

> "ok, so I know it works..."

AFAICT, "your version" can't properly boost/cut. However you are satisfied "it works" and I have to accept that it serves *your* needs.

There are "half Baxes" which only boost. My bad-morning intuition/fog says this one may only cut, but I need a meal and a nap before thinking on that.
  • SUPPORTER

Rob Strand

QuoteYeah, it is for sure a 10K resistor. The EQ section originally belonged to a Darkglass B7K, but I applied it to an MXR Micro Amp, by adding a DPDT switch. On the Darkglass B7K, it was a 1M resistor, but when I made this add on PCB,
I found that it made the volume jump too high when going from MXR Micro Amp only to MXR Micro Amp, plus an EQ section.
The 1M on the  Darkglass B7K makes sense.
Quote
After reducing the resistor to 10K, I found the volumes "matched" close enough. See below for the full schematic.
Unfortunately using that resistor to trim the gain isn't the right thing to do.  It messes with the way the Bass/Treble controls work (in a noticeable way).    Perhaps a better way to gain-match would be a divider before IC1D.

BTW your schematic is missing a bias connection on IC1D.
Send:     . .- .-. - .... / - --- / --. --- .-. -
According to the water analogy of electricity, transistor leakage is caused by holes.

bushidov

QuoteBTW your schematic is missing a bias connection on IC1D.
Odd, it seems to be working that way, though. But I do see what you are talking about. Odd.

QuoteUnfortunately using that resistor to trim the gain isn't the right thing to do.  It messes with the way the Bass/Treble controls work (in a noticeable way). Perhaps a better way to gain-match would be a divider before IC1D.
I'll try drawing that out tomorrow and testing it on my PCB

QuoteOK, I think I see. It is mis-drawn and apparently mis-wired. The 1Meg should NOT set audio gain. Audio gain is set by feeding-back from amp output to the other end of the pot/cap network. Which means this wire is missing:
I want to try that out as well. I think Paul might have stumbled onto something else I was trying to figure out.

Thanks again for your patience. This COVID-19 thing has gave me some extra time to work on side projects outside of working from home.
"A designer knows he has achieved perfection not when there is nothing left to add, but when there is nothing left to take away."

- Antoine de Saint-Exupéry

bushidov

Well, Paul and Rob, you guys nailed it:


QuoteOK, I think I see. It is mis-drawn and apparently mis-wired. The 1Meg should NOT set audio gain. Audio gain is set by feeding-back from amp output to the other end of the pot/cap network. Which means this wire is missing:
First, I made the wire connection, per Paul's instructions and changed R10 from 10K to 1M. Now the volume is matched and I got a lot more clarity out of it.

QuoteBTW your schematic is missing a bias connection on IC1D.
I added R19, a 1M resistor to pull up to bias. And I got a bit more clarity on my pick attack.

Thanks again for finding my mistakes and giving me the solutions to fix it.

So, I believe the circuits are either identical, or at least very very close (these fixes were not present on the Darkglass B7K, except for the 1M resistor was in place), the assumed frequency bands are:
The low is +-12dB @ 100Hz
The high is +-12dB @ 5kHz
The mids are +-12dB @ 1kHz

Doing the math, 12dB would be a gain of roughly 3.981072, so a +- 3.981072 gain.

However, I don't like to assume anything. How would I be able to know? There should be a calculation I can use, like in other filters?
"A designer knows he has achieved perfection not when there is nothing left to add, but when there is nothing left to take away."

- Antoine de Saint-Exupéry

Rob Strand

#11
QuoteHowever, I don't like to assume anything. How would I be able to know? There should be a calculation I can use, like in other filters?
Just chuck it into a circuit simulator.

There's been a few threads giving equations for EQs but they don't cover all the different variants.   The equations themselves are approximations anyway.   Also the addition of the 1M means you have to re-jig all the equations.     Use the sim!!!

The frequency specs are very nominal.  You can assign many meaning to EQ's which don't have peaks.    Even the specs are inconsistent in this - simply take a circuit with specs put it in a sim, then you will will see the frequency points have no definite meaning!
Send:     . .- .-. - .... / - --- / --. --- .-. -
According to the water analogy of electricity, transistor leakage is caused by holes.

bushidov

Hi Rob,

Which simulators would you recommend that would do something like this attached schematic?

Thanks!
"A designer knows he has achieved perfection not when there is nothing left to add, but when there is nothing left to take away."

- Antoine de Saint-Exupéry

Rob Strand

QuoteWhich simulators would you recommend that would do something like this attached schematic?

LT Spice is pretty common and free.

PRR, mentioned Micro-cap is now free,
https://www.diystompboxes.com/smfforum/index.php?topic=123073.0

These tools can be frustrating at first and there's always a learning curve.   It takes a good deal of perseverance and time to familiarize yourself.    You only know which one suites you by trying out a few and dumping the ones that frustrate you! 

It's a long time since I've looked at any of these.  I use a very old spice program,  I use it because it has particular features I use and I'm familiar with it.
Send:     . .- .-. - .... / - --- / --. --- .-. -
According to the water analogy of electricity, transistor leakage is caused by holes.

bushidov

I've heard a lot of good things about LT Spice. It's been on my to do list of things to learn. I guess now I have a really good excuse to do so.
"A designer knows he has achieved perfection not when there is nothing left to add, but when there is nothing left to take away."

- Antoine de Saint-Exupéry

edvard

Quote from: bushidov on April 07, 2020, 08:46:33 PM
I've heard a lot of good things about LT Spice. It's been on my to do list of things to learn. I guess now I have a really good excuse to do so.

LTSpice is very popular, so there are a lot of contributed models and a pretty big discussion group use to be on Yahoo, but has now moved to https://groups.io/g/LTspice
Texas Instruments has Tina-Ti, which works well and some things are much simpler (capturing output impedance is as easy as placing a multimeter symbol, IIRC), but it's not as documented as LTSpice.
Download and try them, see what works for you.
All children left unattended will be given a mocha and a puppy

PRR

>>     your schematic is missing a bias connection on IC1D.
> I added R19, a 1M resistor to pull up to bias.


What is IC1D doing that IC1C couldn't do already?
  • SUPPORTER

bushidov

QuoteWhat is IC1D doing that IC1C couldn't do already?
Technically, nothing. On the PCB, I didn't have the room to run it off and help with the bias, but I probably could have just done the old "tie the inverting input to output and run straight bias into the non-inverting pin" hat trick.
"A designer knows he has achieved perfection not when there is nothing left to add, but when there is nothing left to take away."

- Antoine de Saint-Exupéry

Smegma

Quote from: bushidov on April 06, 2020, 09:54:07 AM
I know I based the EQ section off of my Darkglass B7K's Low, Low Mids, and High circuits, which would mean, orginally the low is +-12dB @ 100Hz, the high is +-12dB @ 5kHz, and the low mids are +-12dB @ 1kHz. However, as mentioned above, I changed R10 from 1M (like on the Darkglass B7K) to a 10K to get some sort of "unity" going on between the two switch settings. Also, the Darkglass B7K included a final "high mids" EQ that came after the low mids output pin on the op-amp that focused around +-12dB @ 2.8kHz that I didn't include.
Would you mind sharing the components values for the high mid band of the Darkglass B7K? The complete schematic would be nice either, if you have it. Thanks.

Vivek

Quote from: Smegma on November 29, 2021, 05:09:59 AM
Would you mind sharing the components values for the high mid band of the Darkglass B7K? The complete schematic would be nice either, if you have it. Thanks.

I recently assisted an effect PCB manufacturer to add switches for mods to the Darkglass (to shift the freq of the mid controls)

He had begun with a schematic that was posted in this forum earlier

https://www.diystompboxes.com/smfforum/index.php?topic=126970.msg1215887