Cheapo chinese pedal supply - traced purely for fun

[JustinFun]

Here's a dirt cheap 10 output 9v DC supply aimed at the guitar pedal market. Regulated from a hardwired-in 1a 15v smps (which feels light as a feather...)
I was expecting the filtering to be very limited, which it is, but what I wasn't expecting was for such a simple board to be so modified at the factory (this was bought new).
The 9v plane has been cut between the two sides of 5x9v sockets (isolating the output of the regulator), and then bridged again with 2 wires.
The ground connection from the regulator has been cut, and then bridged with a 1.2k resistor. I've not seen this done before - not sure what it does?
The 220uf coupling capacitor on the output has had a 1uf electro added in parallel. I've seen poly or ceramic caps of lower value used in this way to help filtering, but I'm not sure that a 1uf electro will do anything useful?
Anyway, pics...

(edited to add - LED indicator wiring omitted from the schematic - it runs straight from the 15v via a current limiting resistor)

[Rob Strand]

It's all screwed up.

To me it looks different to your schematic:
- the ground lead of the regulator has been cut completely
- the resistor connects from regulator output

I'm reading it as:
- the regulator has been knee-capped by lifting the ground lead, so it just passes the supply
- the resistor is placed in series with with regulator output

pins on L7809CV: in, ground, out  when looking at the face and from left to right

All in all the regulator is stuffed and they put a 1k2 resistor in series with the 15V supply
so the open circuit voltage is 9V but when you put load on it the output drops to 0V.

All looks very suspicious, like badly hacking a failed unit.   I could be wrong.  If anyone sees different
please set me straight!

[antonis]

The ground connection from the regulator has been cut, and then bridged with a 1.2k resistor. I've not seen this done before - not sure what it does?

It offsets (upwards) output voltage by I_GND X R1..
(but it's unpredictable due to GND pin current spread..)

You see, those fixed regulators develop a potential difference between pin3 (OUT) and pin 2 (GND)..
In case of pin 2 is grounded, that voltage is the same with the output one..
In case of a voltage added to pin 2, output voltage is the sum of fixed voltage and added one.,.

edit: The above stand only in case of your 2^nd schematic is correct..
(for which, Rob has his doubts..)

P.S.
You can turn a fixed voltage regulator into a variable one by wiring it like an adjustable regulator (e.g. LM317) and replacing 1.25V with particular fixed voltage in the V = 1.25 (1 + R2/R1) + I_GND*R2

[ElectricDruid]

I think I'm with you, Rob. On this image, you can see the track cut *both sides* of the ground leg:

https://postimg.cc/ykFdKdqR

This leaves no ground connection. Which might explain why it looks like something was soldered to the ground pin at some point:

https://postimg.cc/4mCdjc0Z

It looks to me like it had futher modifications which are no longer there. Otherwise why solder to the boardside pins of the regulator? As Rob so eloquently put it, "It's all screwed up". Yes, it is.

[Rob Strand]

I think I'm with you, Rob. On this image, you can see the track cut *both sides* of the ground leg:

https://postimg.cc/ykFdKdqR

I'm a bit confused because that pic shows no ground connection but this one

https://postimg.cc/bZpvwh8B

has wire connecting to the regulator ground!

[JustinFun]

Yes, the wire fell off while I was taking the pictures! 

The schematic is wrong - I got my pins mixed up. Should look like this:

[JustinFun]

So it's just series resistance for filtering? 1k2 seems a lot given that it's only a 1A power supply into 10 sockets (supposedly 100mA each).

[JustinFun]

(also very ashamed to have screwed up such a simple trace...)

[antonis]

1k2 seems a lot given that it's only a 1A power supply into 10 sockets (supposedly 100mA each).

Nahhh...

A 1.2kV drop should be considered reasonable..
(although it doesn't look like a 1.5kW resistor to me..)

[ElectricDruid]

Yes, the wire fell off while I was taking the pictures! 

The schematic is wrong - I got my pins mixed up. Should look like this:

So why did they cut the trace both sides of the ground pin, only to then replace that trace with a bit of wire that does exactly the same job? As shown here:

https://postimg.cc/bZpvwh8B

Makes no sense to me.

[JustinFun]

Yes, the wire fell off while I was taking the pictures! 

The schematic is wrong - I got my pins mixed up. Should look like this:

So why did they cut the trace both sides of the ground pin, only to then replace that trace with a bit of wire that does exactly the same job? As shown here:

https://postimg.cc/bZpvwh8B

Makes no sense to me.

Worried about a ground loop maybe? In the revised version the ground follows a different route.

[JustinFun]

1k2 seems a lot given that it's only a 1A power supply into 10 sockets (supposedly 100mA each).

Nahhh...

A 1.2kV drop should be considered reasonable..
(although it doesn't look like a 1.5kW resistor to me..)

Ha! Your sarcasm is justified. Back to Ohm's law remedial class for me.

[duck_arse]

looks as tho they haven't read the datasheet for the regulator, otherwise the 220uF cap would be on the in pin, not the out pin.

[Phend]

Plug it in. Does it work?

[Rob Strand]

looks as tho they haven't read the datasheet for the regulator, otherwise the 220uF cap would be on the in pin, not the out pin.

Early on I couldn't make sense of the unit so I looked at the datasheet only to be confused further.
Check out the odd numbering on the example circuits. There's no other info showing which physical
pins those numbers correspond to - actually *some* of the mechanical drawings have the common
123 pin numbering.

https://pdf1.alldatasheet.com/datasheet-pdf/view/22639/STMICROELECTRONICS/L7809CV.html

For such a simple circuit it does my head in a bit.

Plug it in. Does it work?

Good idea:
- What's the output voltage?
- What's the output voltage under load, like 1k load?

If the circuit is wacky like the updated schematic it the output should drop to from 9V to 4.1V.

A later datasheet(?), showing the normal/expected pin outs,
https://www.st.com/resource/en/datasheet/l78.pdf
https://www.farnell.com/datasheets/2307057.pdf

[Rob Strand]

Yes, the wire fell off while I was taking the pictures! 

The schematic is wrong - I got my pins mixed up. Should look like this:

Maybe like this:

[PRR]

If it were a 7805 (far more available), a 1.2k in the "GND" leg would make sense as a stupid way to jack-up a few Volts to make more like "9V". And that would be steady-ish with load, maybe Vin (but not temperature or lot).

If instead it is ~~1k in *series* with the output a 1k-2k load will instantly reveal this by dropping from 9V to ~~4V.

[Rob Strand]

If it were a 7805 (far more available), a 1.2k in the "GND" leg would make sense as a stupid way to jack-up a few Volts to make more like "9V". And that would be steady-ish with load, maybe Vin (but not temperature or lot).

If instead it is ~~1k in *series* with the output a 1k-2k load will instantly reveal this by dropping from 9V to ~~4V.

Pretty sure at this point the resistor in the ground lead was a tracing error.
(It is odd the resistor works out that way though. Earlier on I got 5.0V across a 1.2k resistor.)