I have a few 19.5V power supplies with damaged connectors that I'd like to repurpose for trying out class D amp boards that require +/- voltages. As they are, the - side is connected to earth ground (3rd wire on the wall outlet). I wanted to see if it was possible to make them floating/isolated for the - supply. So I opened one up. It looks to me like this will be very simple.
On the back of the board, there's a D7575 chip.
I looked up the datasheet, and found their example circuit, with separate grounds for input and output, separated by a transformer and an optocoupler. The transformers on the board are obvious, and I also found a BL817 optocoupler (circled in red) and a green/yellow wire from the earth ground input to the negative side of the output.
If I cut the green/yellow wire, or put a ~megohm resistor in the middle of it, will this be a workable negative supply?
On the back of the board, there's a D7575 chip.
I looked up the datasheet, and found their example circuit, with separate grounds for input and output, separated by a transformer and an optocoupler. The transformers on the board are obvious, and I also found a BL817 optocoupler (circled in red) and a green/yellow wire from the earth ground input to the negative side of the output.
If I cut the green/yellow wire, or put a ~megohm resistor in the middle of it, will this be a workable negative supply?
It sounds to me like the green/yellow wire is for safety in the event of a failure of the insulation of the switching transformer or other circuitry we are not aware of that bridges the isolation barrier such as feedback controls for voltage regulation. It may also prevent the output (if it was floating) from assuming a very high potential with respect to ground which is something that SMPS can commonly do if floating.
So the simple answer is that it would be unwise to do what you propose.
So the simple answer is that it would be unwise to do what you propose.
+ cap to - of other cap to get false earth.
Means + rail to first cap +, that cap - joined to + of second cap, that junction is Earth, and - of that second cap to - rail of SMPS / brick.
To split the supply.
Do not damage the safety wire, you will be deemed to have caused grievous harm if something goes wrong.
This will give you + / - 9 or so, not very high, adequate for most rooms.
2200 or 3300 / 35 V will be more than sufficient.
Means + rail to first cap +, that cap - joined to + of second cap, that junction is Earth, and - of that second cap to - rail of SMPS / brick.
To split the supply.
Do not damage the safety wire, you will be deemed to have caused grievous harm if something goes wrong.
This will give you + / - 9 or so, not very high, adequate for most rooms.
2200 or 3300 / 35 V will be more than sufficient.
Last edited:
A datasheet, schematic P13.
http://file.remont-aud.net/baza/ic_power/sm_data/LD7575_data.pdf
I guess what the OP was asking is, will it work (and safely) if the ground wire is connected to + output instead of - output.
Something I would like to know myself.
http://file.remont-aud.net/baza/ic_power/sm_data/LD7575_data.pdf
I guess what the OP was asking is, will it work (and safely) if the ground wire is connected to + output instead of - output.
Something I would like to know myself.
I was thinking of the megohm resistor as a way of keeping the output from floating to too high a potential, but I hadn't thought of the failure mode where the isolation barrier fails. Doing a little math, if I instead used a 6.8 kOhm it would trip a GFCI while drawing only 56 mW.It sounds to me like the green/yellow wire is for safety in the event of a failure of the insulation of the switching transformer or other circuitry we are not aware of that bridges the isolation barrier such as feedback controls for voltage regulation. It may also prevent the output (if it was floating) from assuming a very high potential with respect to ground which is something that SMPS can commonly do if floating.
So the simple answer is that it would be unwise to do what you propose.
But in the configuration I'm thinking of, there would also be a second power brick for the positive supply. I would connect the + side of the modified one to the grounded - side of the unmodified positive supply. I think that would take care of the problems you brought up.
i see no reason why you can not series connect outputs of laptop psu if need to....the secondaries are isolated from the primary hot side...
Yeah. But if the negative side of the output is connected to mains ground you end up shorting out the V- supply.i see no reason why you can not series connect outputs of laptop psu if need to....the secondaries are isolated from the primary hot side...
There could be multiple reasons why the output of the laptop brick is grounded. The more likely reason in my view is that it was too expensive to build the supply to meet electrical safety Class II (double insulation), so grounding any metal the end user could touch became required. I would not simply cut the ground wire.
See if you can find a pair of Class II switching bricks instead. Mean Well makes some good ones.
Tom
i will lift the ground connection on the upper brick...you only need one brick to connect to mains ground...
look at the Dynaco ST70 and others, the power cord is simply two wires and not safety ground third wire, are they no good?
look at the Dynaco ST70 and others, the power cord is simply two wires and not safety ground third wire, are they no good?
All the Meanwell SMPS can be treated as fully isolated, and can be stacked, or reversed, at will.
i will lift the ground connection on the upper brick...you only need one brick to connect to mains ground...
look at the Dynaco ST70 and others, the power cord is simply two wires and not safety ground third wire, are they no good?
https://www.xppower.com/resources/blog/iec-protection-classes-for-power-supplies
Your comment got me looking more closely at the labels. I have 4 of these power supplies, and was assuming they were all the same. It turns out that two of them, but not the one I opened up, have a Class II with ground symbol, upper right:The more likely reason in my view is that it was too expensive to build the supply to meet electrical safety Class II (double insulation), so grounding any metal the end user could touch became required. I would not simply cut the ground wire.
See if you can find a pair of Class II switching bricks instead. Mean Well makes some good ones.
Tom
And I measured no continuity from earth ground to negative! Looks like a viable negative supply.
The utility ground is connected to the secondary return, so that any metallic parts like enclosure, cable returns etc. maybe safely touched by the user. If the power supply indeed satisfies IEC 60950-1 (attached below) as claimed and you know your thing, then you may try separating the grounds without much problem.
Please read Section 0.2.1 (page 23/250) of the following document to get an idea of things.
Please read Section 0.2.1 (page 23/250) of the following document to get an idea of things.
Attachments
The one that I was thinking of altering is different. It doesn't claim to satisfy IEC 60950-1.If the power supply indeed satisfies IEC 60950-1 (attached below) as claimed...
Though without opening it up, I don't see how you could shock yourself with it. 19.5 V is touch safe. A pair of these connected in an enclosure, one grounded only through the other, in combination with a GFCI, looks pretty safe to me. That said, an unaltered Class II supply is both easier and safer.
Yes, in that case, it may be better to use the IEC ones for the negative rail. Sorry for not following the story properly.The one that I was thinking of altering is different......an unaltered Class II supply is both easier and safer.
- Home
- Amplifiers
- Power Supplies
- Laptop power brick modification for +/- power supply