DIYstompboxes.com

DIY Stompboxes => Building your own stompbox => Topic started by: bluelagoon on November 24, 2020, 12:53:30 AM

Title: Does Ground reference always need to be 0 Volts?
Post by: bluelagoon on November 24, 2020, 12:53:30 AM
just curious as were contemplating using the Klon power supply with a +18 volt and a -9 volt from the MAX1044 and using the the 2 rails as positive and (negative ground) through a resistor divider for Vref (Vdd/2) Not sure if it is possible, am still learning, but some things with power supplies aren't so readily identified through pages of reading and hours Googling on the topic. So please educate me, I implore you and thankyou.
Cheers, TTay :)
Title: Re: Does Ground reference always need to be 0 Volts?
Post by: merlinb on November 24, 2020, 03:17:56 AM
Quote from: bluelagoon on November 24, 2020, 12:53:30 AM
+18 volt and a -9 volt ...
...using the the 2 rails as positive
Do you mean use the -9V rail as 'ground', to create +9V and + 27V? It might be possible, but sounds very inconvenient...
Title: Re: Does Ground reference always need to be 0 Volts?
Post by: 11-90-an on November 24, 2020, 03:51:08 AM
If I can ask, why do you want to do this? There might be some other things tried and tested by other people for your application/curiousity.

Quote from: merlinb on November 24, 2020, 03:17:56 AM
Quote from: bluelagoon on November 24, 2020, 12:53:30 AM
+18 volt and a -9 volt ...
...using the the 2 rails as positive
Do you mean use the -9V rail as 'ground', to create +9V and + 27V? It might be possible, but sounds very inconvenient...

op said he wanted a vdd/2 so I would estimate a 13.5v VREF?
Title: Re: Does Ground reference always need to be 0 Volts?
Post by: teemuk on November 24, 2020, 04:26:22 AM
Yes, it is possible. Any arbitrary voltage potential can be chosen as "common" reference. It's really all about voltage potentials and their relations to each other in the first place.

But you of course need to acknowledge that your "common" is really that only for your particular circuit section and it will not be the same as true 0V ground. So, for example, inputs and outputs referencing to that 0V ground must be AC coupled to interface to your circuit with different common reference voltage.

Just like in using a +4.5V common derived from 1/2 Vcc of 9V battery. +4.5V is "ground" in sense that battery's terminals appear as +4.5V and -4.5V voltage sources in reference to it (and there is 0V potential between two +4.5V "common" points), but it's not same voltage potential as true zero volt ground. So one must AC couple stages with +4,5V common and 0V ground from each other in order to to isolate the DC offset.
Title: Re: Does Ground reference always need to be 0 Volts?
Post by: antonis on November 24, 2020, 04:54:07 AM
Quote from: 11-90-an on November 24, 2020, 03:51:08 AM
op said he wanted a vdd/2 so I would estimate a 13.5v VREF?

OP is confused a bit but you're confused a lot.. :icon_lol:

+4.5V is the midpoint (VREF) for a +18v/-9V bipolar supply.. :icon_wink:
(13.5Vp is the positive/negative headroom for a 27Vp-p signal amplitude..)

P.S.
Arithmetic "mean" is the algebraic (including +/- sign) sum divided with the participant number..
i.e. [+18 + (-9)] / 2 = 4.5
Title: Re: Does Ground reference always need to be 0 Volts?
Post by: 11-90-an on November 24, 2020, 05:12:52 AM
Quote from: antonis on November 24, 2020, 04:54:07 AM
Quote from: 11-90-an on November 24, 2020, 03:51:08 AM
op said he wanted a vdd/2 so I would estimate a 13.5v VREF?

OP is confused a bit but you're confused a lot.. :icon_lol:

+4.5V is the midpoint (VREF) for a +18v/-9V bipolar supply.. :icon_wink:
(13.5Vp is the positive/negative headroom for a 27Vp-p signal amplitude..)

ooh right... I was referencing the voltages to the -9v as 0v ground (and the 18v as 27v)... not confused, but should've specified...  :icon_lol:

let my error be a lesson, OP... depending on what you call "ground", the voltages "vary" somewhat...  :icon_mrgreen:
Title: Re: Does Ground reference always need to be 0 Volts?
Post by: merlinb on November 24, 2020, 10:41:13 AM
I think the answer we're all spiralling towards is:
Yes it is technically possible but you'd better have a realy really good reason to do it, because it's going to be a headache for no obvious reward
Title: Re: Does Ground reference always need to be 0 Volts?
Post by: R.G. on November 24, 2020, 11:45:13 AM
First some basics. "Ground" (earth as our English cousins call it) used to mean literally "connected to ground/earth by a wire to a metal stake in the ground". It still means that for many AC power wiring situations. Later when humans did more adventurous electronics, we figured out that "ground" could just be any circuit point we decided to define as "the source of truly zero volts as a reference point, in this one circuit". That is, ground inside any one circuit is any circuit place you say it is, as long as you're consistent about it and measure all other voltages from that point.

OK so far - each separate box can have it's own unique definition of ground. What happens when we connect two or more boxes together with signal+ground cables? The signal flows from box to box and back, and both "grounds" are forced together by the ground wire of the cable.

It's that third conductor between two boxes that gets you into trouble. If it's a common connection to an external power supply, then the external power supply had better be right for both boxes at the same time. The classic problem in pedals is one pedal using the negative power supply voltage defined as "ground", and a second one which wants to use ground and a minus voltage, and defining the positive power supply voltage as "ground". If you connect these two together to one single power supply, you're effectively connecting the external power supply's positive and negative terminals together through the ground wires, and neither pedal works, while the external power supply suffers.

This idea of "ground" being whatever you say it is is the idea behind DC biasing of circuits. An opamp might need a DC bias in the middle of a single DC power supply. So we construct a resistor divider to create a source of 1/2 the DC power supply, and tie that point to DC ground with a BFC (..er... Big Freaking Capacitor) so that the signal at the 1/2 voltage point is tied to circuit power ground with a very low capacitive impedance. It's effectively "ground" for AC signals, but with a DC offset.

The Klon power supply setup is a tricky one. The design uses a single charge pump to full advantage to create 3 x 9V. You are free to label any point in the power supply string of outputs with "ground" for use in your circuits. IN the Klon setup, you have two incoming power wires, + and - , and they must connect to the points that power the charge pump. The charge pump makes 9V more positive than its positive supply wire, and 9V more negative than its negative supply wire. Everything works inside this specific box no matter which of the power supply terminals (or any other terminal) is labeled "ground".

But if you connect it to external boxes with three wires - signal ground, signal, and an external power supply with + or - voltage, your definition of ground must not be in conflict with the external box AND the external power supply.

The Klon pedal uses +18 and -9 to get a 27V  total power supply swing, but uses +9 and 0V to be the power supply to the IC charge pump chip. This lets it connect external boxes to the same 0V point for DC power and signal ground and work in harmony with other pedals. A Klon style power supply will work fine re-labeling any one of the charge pump outputs "ground", but this has to match what any external power supply the external boxes label "ground" too.

I hope this is making sense. I've tried to squeeze it down a lot. Ask questions where it's mystifying.