could a BAXANDALL EQ be scaled down like Marshall EQ?

[richon]

could a BAXANDALL EQ be scaled down like Marshall EQ?

i've seen it with the marshall EQ when multipliing the caps by 10 and resistor divided by 10


could this be done in the baxandall also?

[R.G.]

It can be done with all RC (and L!) networks as long as the driving and driven impedances don't matter to the circuit operation, or if the driving and driven impedances are scaled with the network being scaled.

[richon]

the output of a LM386 is 100ohms....

the baxandall likes to see 15K input and 1M or more on the output...   so i was planing to scale it down 10 times to make the 386 a little "happier" with the EQ....   Am I right?

[R.G.]

the output of a LM386 is 100ohms....

Well, maybe open loop. Closed loop, it's significantly less than 8 ohms because of the feedback. It's quite low.

the baxandall likes to see 15K input and 1M or more on the output...   so i was planing to scale it down 10 times to make the 386 a little "happier" with the EQ....   Am I right?

I'm not sure. How were you planning to hook them up? With the LM386 ahead of or behind the baxandall, or with an active baxandall around the LM386?

There are active and passive versions of the Baxandall EQ. All of them want a source driving them that is much, much (that is "10x for each "much") lower than the impedance of the Baxandall input, and much, much (that is, more than 100X) higher than the output of the Baxandall network. Scaling the whole Baxandall network down lowers both its input impedance and output impedance by the amount you scale it. The new, scaled network needs corresponding limits on the driving impedance and input that the network drives.

Example. If your Baxandall network wants no higher than 15K driving its input and no lower than 1M loading its output, and you scale it down by 10X, then the new scaled network will need something driving its input to be no higher than 1.5K (that is, lowered by the same ratio) and loaded by no less than 100K (also lowered by the same factor). If you drive the input with an LM386 which is much smaller than 8 ohms on its output, then the new, scaled Baxandall will not care, since much less than 8 ohms is also much less than 1.5K.

Ask questions where this is confusing. I don't understand why you're trying to scale a Baxandall to match an LM386.

[PRR]

I'm sure you mean "James". Baxandall's circuit is a radical re-wiring of the James, using an amplifier and different ratios/tapers.

> the output of a LM386 is 100ohms....

The LM386 output impedance is far-far under 1 ohm to over 100KHz. (Look at LM380 datasheet bottom of page 5, infinite or 8 ohms makes no difference; LM386 is a small LM380.)

LM386 strains to drive less than 8 or 4 ohms, but is perfectly happy (and maybe cleaner) driving higher loads (to infinity).

It won't be any happier with smaller network impedance.

If the LOAD is heavy, lower impedances will be better.

> multipliing the caps by 10 and resistor divided by 10

Yes, that will be workable for typical guitar-cord loads. It won't be "flat" because James (and all passive tone networks) is very sensitive to capacitive loading such as cables. A buffer may be nice.





For very low-Z loads.....

You "could" design a James with 10 ohms, 100 ohm 10% audio pot, 1 ohm. The load could now be as low as 500 ohms without great error.

100 ohm 10% audio taper is a VERY rare pot.

Cap values will be 1uFd to 50uFd and must be bi-polar... more expensive than standard-scale values.

Sub-ohm ground resistances will matter.

LM386 supply current (battery life) driving high levels into 10 ohms will be much higher than when driving 100K-1Meg network.

Voltage noise (hiss) in the network will be lower.... possibly 0.1 microvolts. Yay? No, the LM386's input hiss is over 1 microvolt, amplified by 20 then lossed by 10, 2 microvolt out of the network. The ultra-low network values are no noise advantage. 10K impedance would be just as quiet. (As we might guess from the LM386/LM380's internal 15K/25K resistors.)

[richon]

this is what i was planning to use (and I'm ussing right now) so it seems like i'm super LOST on this... and have to re-think everything again to have it to work right 

[R.G.]

this is what i was planning to use (and I'm ussing right now) so it seems like i'm super LOST on this... and have to re-think everything again to have it to work right 

Now that I see what you're doing, no, you don't have to re-think everything. However, if you were going to drive a speaker, you have a problem.

The LM386 will drive the (passive) Baxandall setup you have just fine. However, you need a high impedance load after it, and some gain to make up the approximately 20db (10:1) loss of signal voltage in the Baxandall. So you can't drive a speaker from the output on the volume control. You will need a follow-on opamp input with an impedance of 1M or more to use the network with those impedances.

[richon]

ok....  but i'm planning to use it as a PEDAL/STOMPBOX...(not driving a speaker)

should I put a buffer at last (after the VOLUME pot) to have a high Z load (like a JFET buffer) to make the EQ work right with the right impedancies 

(and by "right" I mean: avoid the problem that the EQ changes as I rotate the VOLUME pot)

[RedHouse]

I'd also suggest you start with a Baxandall that is optimized for guitar tone before re-scaling, this one is a modified  Baxandall type from the old Silvertone 1484 and is a really good sounding circuit for guitar.

[richon]

@Redhouse:

It seems that the picture you loaded puts the EQ (baxandall) with hi load on input and on output...

besides the Vol Pot seems to interact with the Treble pot, isn't it?

[RedHouse]

Yeah, Nate was a crazy guy, but the circuit sounds really good.

[PRR]

> the Vol Pot seems to interact with the Treble pot, isn't it?

SAME as the plan you present.

Either of the plans I presented will give output impedance low enough for guitar-cord work.

You NEED a coupling cap between the DC level at LM386 output and the tone network. Otherwise the Bass pot will scratch badly when you turn it.

[richon]

as test go....  i found the exact problem

The output of the 386 was going direct to the EQ (no caps in between) and after the EQ direct to the volume pot (also, no cap in between)

so when turning the pot affects more than just the EQ, it affects the 386 output...

so, after testing some mods on the EQ and puting a Cap between the 386 output and the EQ, everythings sounds awesome..

it works perfect now!!!

son will record a video"demo" to show how much the EQ could do.