Mesa Boggie Mark series EQ question

[Phil smith]

Hey guys!

Assembling my Mesa Mark IIc+ clone I found some problems... with EQ sliders!
Here's my circuit diagram, which is identical to original mark II / IV:



But my eq doesn't  behave as it supposes. Fortunately I got real Mesa Studio  and Quad pre's with the same EQ. So I opened up Studio preamp and started to measure some resistance values (in circuit, power off).

Here's what I measured:

Test point 1:



At any slider resistance values are:
~80-100R (slider's all way up),
~13-16k (slider's at the middle),
~6.1k (slider's all way down)

Test point 2 (2nd lead attached to the chassis):



At any slider resistance values are:
~4.8k (slider's all way up) - I think it is a value of "EQ in/out" LDR shunt resistor (in my case it should read 47k),
~15-18k (slider's at the middle),
~4.8k (slider's all way down).

So is this 50k slider not just usual linear potentiometer?

I did some research and found link to the mesa boogie official store, which says "50k pot, special taper"
https://store.mesaboogie.com/products/pot-588535-slider-50k-special-taper.html

What kind a special taper it is? Like a bell shape resistance curve - 0-50k-0?
Does anybody know any information related to this slider (datasheet, etc)?

Will appreciate any info much!

Regards,
Phil

 

[ruffrecords]

This is a fairly common EQ topology. It normally uses linear pots.

Cheers

Ian

 

[Phil smith]

This is a fairly common EQ topology. It normally uses linear pots.

Cheers

Ian

Yep, I always thought like that, but measuring eq on the real mesa got me confused a bit, and now I'm curious what is it a special taper pot?

 

[CJ]

if you find that after completion, you use the amp with the eq bypassed, then you probably have a good set of ears, the word "harsh" comes to mind, maybe check out some other amp schematics, the filter circuit might be the same but the transistor circuit might be better in something like a Peavey Muscian or maybe a Carvin or possibly an old Acoustic  or maybe even an old Sunn Beta Lead

or maybe the DIY version we built over here back in the 70's was not up to snuff, you might have better luck,

 

[abbey road d enfer]

Here's what I measured:

These measurements look normal to me. When the faders are down, they are connected to other parts of the circuit.

What kind a special taper it is? Like a bell shape resistance curve - 0-50k-0?

Very likely "S" taper, i.e. linear with slower taper near the ends of travel to avoid the gain jump there. Linear tapers are often good enough if you are moderate with the amount of EQ.

 

[Phil smith]

Hey!
Thanks every one for help!
In my case issue was an incorrect pcb footprint of a slider  I made in Altium library, now I've re-etched the board. But got another issue - the 0.033 H coil.
They are all diy, winded by some local guy for me several years ago. When I's installing em into the new pcb, I checked them all, but the 0.033 one had no resistance - so coil wire is broken somewhere.. I opened it and, sure enough, it's snapped inside and lost.  jeez! Now this guy is outta businesses and local electronics store has max 0.015 H coils. So can I connect 2 such coils in series and install them into the eq? L_total = 0.015+0.015 = 0.03 H. Will be there sonically any downsides of this "hack"?

 

[abbey road d enfer]

Hey!
Thanks every one for help!
In my case issue was an incorrect pcb footprint of a slider  I made in Altium library, now I've re-etched the board. But got another issue - the 0.033 H coil.
They are all diy, winded by some local guy for me several years ago. When I's installing em into the new pcb, I checked them all, but the 0.033 one had no resistance - so coil wire is broken somewhere.. I opened it and, sure enough, it's snapped inside and lost.  jeez! Now this guy is outta businesses and local electronics store has max 0.015 H coils. So can I connect 2 such coils in series and install them into the eq? L_total = 0.015+0.015 = 0.03 H. Will be there sonically any downsides of this "hack"?

You copuld do that; provided the total DC resistance is close enough to that of the original design, it should work well. Another possibility is that you rewind your inductor; it's not so hard to do.

 

[Phil smith]

You copuld do that; provided the total DC resistance is close enough to that of the original design, it should work well. Another possibility is that you rewind your inductor; it's not so hard to do.

Thanks! Yep, I think of winding it myself too, cause I want to clone mesa bass 400+ preamp & eq in future and this beast has 7 bands, but fortunately same values - 0.033 / 0.068 / 0.22 / 0.39 / 1 H - so DIY is a way to go I guess
Also that guy made axial style instructors (25 mm long and 13 mm in dia), I much prefer more compact radial design Mesa uses

 

[Phil smith]

Hey!
I've finally bought these ferrite cores for coils (outside diameter is 14 mm, so it's very close to original ones installed in Mesa EQs):

Permeability of the core = 2000.

Just for info - original EQ inductor enclosure dimensions:
http://www.tubeampdoctor.com/images/gallery_images/Z-MB-EQI1000.jpg

So I started to design a coil.

I found a piece of software called Coil32 (http://coil32.net/) and there is my case - so I entered all my parameters, required inductance (in my case it's 0.033 H) and it gave my the answer of 30 turns. So I just made this coil and measured the inductance with me LCR T4 meter - it read  2.5 mH! I thought" what?!")))
I verified my measurements with this technique (using known cap, sig.gen. and scope):
https://www.youtube.com/watch?v=GF4AbbBGa5M

And sure enough it was in the same ballpark - around 2-2.5 mH - so, clearly, the software got some error..

Is there any proven method of calculating number of turns of a potted core coil??

Will appreciate any help, thanks!

 

[abbey road d enfer]

Hey!
I've finally bought these ferrite cores for coils (outside diameter is 14 mm, so it's very close to original ones installed in Mesa EQs):
Permeability of the core = 2000.
So I started to design a coil.
I found a piece of software called Coil32 (http://coil32.net/) and there is my case - so I entered all my parameters, required inductance (in my case it's 0.033 H) and it gave my the answer of 30 turns. So I just made this coil and measured the inductance with me LCR T4 meter - it read  2.5 mH! I thought" what?!"))) 

Apparently, this software does not provide calculation for ferrite pots.

Is there any proven method of calculating number of turns of a potted core coil??

Very simple, when you know An; in your case An=2000, but 2000 what?
An is in nH per square root of turn number.
That means if you wind just one turn, the inductance will be 2000nH, or 2uH; well, in practice, that would probably be out by several hundred percent, but for a large number of turns (about 100+), the formula works quite well.
L= An. (N)²
If you reverse the formula, you end up with N=sqrt(L/An)
Entering your values, you end up with N=sqrt(33. 10E-3/2.10E-6)=sqrt16500=128

 

[Phil smith]

Thank you very much!
I think guy who wrote the software messed up some coefficients, I found his original article telling about formulas used in the code,  check it out:
http://coil32.net/ferrite-pot-core.html
There is some coef. k - 19.74 for ferrite cores and 60 for iron powder cores - I guess, here's a problem.

So number of turns for all coils, based on your calculation:

1H - sqrt(1.0 / 2e-6) = 707 turns;
0.39H - sqrt(0.39 / 2e-6) = 442;
0.22H - sqrt(0.22 / 2e-6) = 332;
0.068H - sqrt(0.068 / 2e-6) = 184;
0.033H - sqrt(0.033 / 2e-6) = 128.

I guess I should buy quite fine wire, cause now smallest diameter I got is Ø0.3 mm and I can barley wind 100 turns with it on a such a small bobbin..

PS. Found datasheet for my cores:
http://www.mmgca.com/catalogue/MMG-Ferrite-PT2.pdf
see page 4 - Ø14x8 core, material - F44

 

[abbey road d enfer]

Thank you very much!
I think guy who wrote the software messed up some coefficients, I found his original article telling about formulas used in the code,  check it out:
http://coil32.net/ferrite-pot-core.html
There is some coef. k - 19.74 for ferrite cores and 60 for iron powder cores - I guess, here's a problem.

The problem is that this coefficient is valid only for one type of ferrite, and there are many different, AND the formula is for an ungapped core. Many cores are gapped, because it gives a better tolerance (ungapped cores are often ca. +/-30% tolerance) and better saturation performance, at teh expense of increased quantity of copper.

So number of turns for all coils, based on your calculation:

1H - sqrt(1.0 / 2e-6) = 707 turns;
0.39H - sqrt(0.39 / 2e-6) = 442;
0.22H - sqrt(0.22 / 2e-6) = 332;
0.068H - sqrt(0.068 / 2e-6) = 184;
0.033H - sqrt(0.033 / 2e-6) = 128.

Seems right.

I guess I should buy quite fine wire, cause now smallest diameter I got is Ø0.3 mm and I can barley wind 100 turns with it on a such a small bobbin..

This type of inductors commonly uses wire between 0.05 and 0.1 mm diameter (AWG 44-38)

 

[Phil smith]

The problem is that this coefficient is valid only for one type of ferrite, and there are many different, AND the formula is for an ungapped core. Many cores are gapped, because it gives a better tolerance (ungapped cores are often ca. +/-30% tolerance) and better saturation performance, at teh expense of increased quantity of copper.

Seems right. Anyway,  a "trial & error" method is the way to go 

This type of inductors commonly uses wire between 0.05 and 0.1 mm diameter (AWG 44-38)

Thanks!
I've made some diagrams showing coil former dimensions and max number of turns with 0.3 and 0.1 mm wire


Note: All dimensions are in mm.

So I'll probably stick with a 38 gauge which gives me 903-707 = 196 turns margin

 

[CJ]

wind 500 turns and see where it gets you,

remember that inductance goes up as turns squared,

best way to check inductor is combine it with the cap used in the actual circuit and see where the peak is at by using a signal generator and using a sense resistor to measure voltage drop across, maybe 10 ohms,

lowest voltage will be at the center freq, use about 1 volt rms

 

[abbey road d enfer]

Seems right. Anyway,  a "trial & error" method is the way to go 

Indeed! See how the tolerance goes +30/-20%.
Also, I believe you'll soon figure out how the final inductance can be changed and somewhat adjusted by rubbing the half-cores against each other.

 

[abbey road d enfer]

best way to check inductor is combine it with the cap used in the actual circuit and see where the peak is at by using a signal generator and using a sense resistor to measure voltage drop across, maybe 10 ohms,

lowest voltage will be at the center freq, use about 1 volt rms

That's an excellent recommandation. The only accurate way of testing inductors is in-circuit in conditions as close as possible to intended.

 

[Phil smith]

wind 500 turns and see where it gets you,

remember that inductance goes up as turns squared,

best way to check inductor is combine it with the cap used in the actual circuit and see where the peak is at by using a signal generator and using a sense resistor to measure voltage drop across, maybe 10 ohms,

lowest voltage will be at the center freq, use about 1 volt rms

Thank you, I'll check it out!
I did something similar to verify my LCR-T4 measurements.
I found this method some time ago on Youtube and it's really quite simple and:
https://www.youtube.com/watch?v=GF4AbbBGa5M

Indeed! See how the tolerance goes +30/-20%.
Also, I believe you'll soon figure out how the final inductance can be changed and somewhat adjusted by rubbing the half-cores against each other.

Oh, shall see, thanks!

 

[Phil smith]

Hey everyone!

Sorry for the delay - got a quite busy year...
Anyway! Back to the inductor
I've ended up winding 230-240 turns to get 33mH)) And now it's on its place and EQ works hunky dory 8)
So thanks you guys for your input!

Cheers!