Options for dividing voltage for higher amp circuits

Tutorials, tools, techniques, methodologies......

Moderator: Modulators

Post Reply
User avatar
FAP
Merzwow
Posts: 539
Joined: Wed Jan 08, 2020 11:50 am
Has thanked: 106 times
Been thanked: 228 times

Options for dividing voltage for higher amp circuits

Post by FAP »

I'm working on a project wherein I want to send 9v to one part of the board ("Part A"), and 4.5v to another ("Part B").

The short version is: I've been trying to do this with a voltage divider, but it's just not working.

Part A is an EHX LPB-1 Booster clone circuit: I measure ~0.365mA max going through it.*
*Interestingly, the official specs list 1mA of current draw.

Part B is a karaoke microphone w/ echo: I measure 108.5mA max going through it.†
†varies based on volume of input source, delay time, feedback amount, etc. The minimum I read was ~99.6mA.

Skipping ahead a bit, it's now come to my understanding that, according to this video, a voltage divider generally shouldn't be used if ≥10mA needs to be supplied.

So what are my options?



For context (though it may just confuse readers more than it helps) here's my math & notes:

- - -

As I understand it, 4.5v could be attained simply by putting two equal resistors in series between 9v and GND, then attaching a lead to where the two resistors meet: makes sense.

One of the earliest problems I came across was, the higher the resistor values, the lower the supplied current: okay, so that just means lower resistor values, right?

Well, I initially came up with a value of 42Ω for the resistors (R = V/I = 4.5v/0.1085A = 42Ω): this seemed suspiciously low to me and, predictably, it didn't work. There was obviously some current being drawn for Part B, but not enough. At the same time, I didn't want to risk lowering it any more and frying the circuit (the max voltage for the primary IC in Part B is ~5.5-6.0v).

I tried simplifying my calculations with the Rthevenin equation from the first video and came up with 21Ω:

Rth = (Rtop * Rbottom)/(Rtop+Rbottom) = (42 * 42)/(42+42) = 1764/84 = 21Ω

Meanwhile, I couldn't get a solid read on the cumulative resistance (Rload) of Part B's circuitry. Again, it's to my understanding that DMM measurements shouldn't be trusted when calculating Rload, so I went with the last method highlighted in this article and came up with 45Ω, like so:

Runknown = Rload = the cumulative resistance of Part B's circuitry
Rknown = 38.7Ω
Vsource = 4.65v (in this case, I'm omitting Part A entirely and only powering Part B)
Measured voltage across Rknown w/ Rload in series = 3.97v

Current of Rknown:
I = V/R = 3.97v/38.7Ω = 0.10258A = 102.58mA

Resistance of Rload:
R = V/I = 4.65v/0.10258A = 45Ω

So then, plugging this back into the Rthevenin calculations I got:

Rth = 21Ω
Rload = 45Ω
Rth + Rload = 66Ω

Vload = Vin * ((Rload)/(Rload + Rth) = 4.5 * (45/66) = 3.068v

Plug that into Ohm's law:

I = V/R = 3.068v/45Ω = 0.06818A = 68.18mA

WTF this doesn't make sense: the only way I'd even get close to 100mA is if Rth = 0, and if Rth = 0, then that would effectively mean 9v straight to Part B, which would fry the thing.

- - -


As you can imagine, I've spent enough time on this already: is there a simpler solution I could try?
I've got a few standard op-amp chips on hand (e.g. TL072): would they help?
I mean, I guess I could just do 6xAA's with a center-tap, but it'd be nice to be able to power this thing through a standard 9v wall wart [like a guitar pedal].

One more bit of context: the only reason I'm adding Part A at all is because the output signal on Part B is so low that I need a way to boost it. Part A actually comes after Part B in the chain.
User avatar
crochambeau
Merzwow
Posts: 512
Joined: Fri Nov 22, 2019 11:21 am
Location: Cascadia
Has thanked: 167 times
Been thanked: 147 times
Contact:

Re: Options for dividing voltage for higher amp circuits

Post by crochambeau »

Please forgive me, I did not thoroughly read the in depth portion of your post.

Is the 4.5 volts you need a "virtual ground" for use in a bipolar circuit? Or is the Vcc power rail not to exceed 4.5 volts?

If we're talking a virtual ground, all of the heavy lifting is done on the power rails, and you just need to hold "virtual ground" steady for small signals. This can be done with a voltage divider and cap drain to (actual) ground 95% of the time - but it's not a high current solution.

You're registering current pull in the 100 mA range?

That sounds like 4.5 volts is Vcc to me.

Assuming your 9 volt rail is stout enough to deal with all this shit to begin with, you could employ a 7805 voltage regulator to spit out +5. If that's too much, simply slap a 1N4001 or similar diode in series to magically drop a half volt or so.

If you want to get PRECISE, you can use an opamp and simply reference your desired voltage in a non-inverting DC pass circuit, and select your opamp to be able to deliver the goods. 100 mA is veering into pass transistor territory, and it would be simpler to just exploit a 7805 (my opinion, it really depends on what's in your parts bin, doesn't it?).

Plenty of ways to skin this cat though. You're 100% correct that a resistive voltage divider blows at these currents.
When in doubt, add resistance.

http://www.rochambeau.net/
User avatar
FAP
Merzwow
Posts: 539
Joined: Wed Jan 08, 2020 11:50 am
Has thanked: 106 times
Been thanked: 228 times

Re: Options for dividing voltage for higher amp circuits

Post by FAP »

crochambeau wrote: Sun Sep 25, 2022 7:08 pmI did not thoroughly read the in depth portion of your post
Totally understandable, that's just here for reference. Your help is always welcome :mrgreen:

Speaking of, it seems the 7805 solution works perfectly so far (though I'm technically using a 78L05) so thanks for that!
User avatar
crochambeau
Merzwow
Posts: 512
Joined: Fri Nov 22, 2019 11:21 am
Location: Cascadia
Has thanked: 167 times
Been thanked: 147 times
Contact:

Re: Options for dividing voltage for higher amp circuits

Post by crochambeau »

Nice! Yeah, 78L05 is the same thing in a different package. Glad it's working out.
When in doubt, add resistance.

http://www.rochambeau.net/
Post Reply