Triple Power Supply Shortcut Problem

Thread Starter

zaferaltun

Joined Aug 31, 2016
65
Hello

I have a power supply and its model is T50-D, 50W triple power supply, 24V 1A - 12V 1A - 5V 3A, I bought them from AliExpress. When I connect it to RaspberryPi, pi is working normal. But if I connect it to my custom circuit, the current is too high. It is working without problem with powerbanks or 5V adapters like cell phone charging adapters normally.

I checked the power supply pins with a multimeter in diode check mode when the power supply is unplugged and realized there is a shortcut between 12V - 5V and GND pins, 5V is stronger. I have 3 of them and they are all new and giving same result. Should I do something extra to use these PSs? Should I use something like protection diode or something like that? I m not so experienced about power supplies, the answer is simple probably.

I m sharing the picture, any clue can help, thanks.
 

Attachments

MrChips

Joined Oct 2, 2009
30,706
Hello

I have a power supply and its model is T50-D, 50W triple power supply, 24V 1A - 12V 1A - 5V 3A, I bought them from AliExpress. When I connect it to RaspberryPi, pi is working normal. But if I connect it to my custom circuit, the current is too high. It is working without problem with powerbanks or 5V adapters like cell phone charging adapters normally.

I checked the power supply pins with a multimeter in diode check mode when the power supply is unplugged and realized there is a shortcut between 12V - 5V and GND pins, 5V is stronger. I have 3 of them and they are all new and giving same result. Should I do something extra to use these PSs? Should I use something like protection diode or something like that? I m not so experienced about power supplies, the answer is simple probably.

I m sharing the picture, any clue can help, thanks.
You cannot test a power supply reliably with an resistance meter or in diode check mode.
Set the meter to DC VOLTS.
Connect the -VE lead (black) of the meter to COM on the power supply.
With the +VE lead (red) measure the voltage at the three terminals, +24V, +12V, +5V.
 

dendad

Joined Feb 20, 2016
4,451
There usually will be a low resistance reading across the outputs as a load resistor is often included on the board to ensure the minimum loading is met.
What is your custom circuit? Without that info we cannot really help. If your circuit is drawing too much current and is connected correctly and not faulty, then you may just need to buy a power supply that has sufficient current capability. There is probably nothing wrong with the power supplies, it is just they are under rated for the job.
 

Thread Starter

zaferaltun

Joined Aug 31, 2016
65
There usually will be a low resistance reading across the outputs as a load resistor is often included on the board to ensure the minimum loading is met.
What is your custom circuit? Without that info we cannot really help. If your circuit is drawing too much current and is connected correctly and not faulty, then you may just need to buy a power supply that has sufficient current capability. There is probably nothing wrong with the power supplies, it is just they are under rated for the job.
My circuit is a module card, there is AMS1117 5V fixed regulator and 22uF B-Type capacitors 2 side of it. Normally the circuit draws about 50mA in total, because there are only a few ULN2803 driver and a PIC18F87K22 microchip processor, nothing too much. I m using 5V for ULN drivers and 12V into the 5V regulator. So you might be right, but can I use this power supply stable by adding some components to the circuit or should I give up?
 

dendad

Joined Feb 20, 2016
4,451
Looking at your power supply info from the first post, is it -5v and -12V?
If it is, than you cannot use it!
The power supply must be +5V and +12V.
I'm sorry I missed that on the first reading.
 

Thread Starter

zaferaltun

Joined Aug 31, 2016
65
Looking at your power supply info from the first post, is it -5v and -12V?
If it is, than you cannot use it!
The power supply must be +5V and +12V.

sorry but you are on really basic mistakes. do you think rasperry pi can work with - voltage? it is already positive.
 

dendad

Joined Feb 20, 2016
4,451
NO!
If you have -5V and -12V, it indicated the +rails are connected together.
You need a power supply that has +5V and +12V so the -rails are connected together.
Please check it as you will blow things up.

Or is it just the way you have written the info?
Do you have a link to the power supply data?

I actually do know what I'm talking about.
I'll draw a diagram in a bit...
Or maybe not now :)
 
Last edited:

Thread Starter

zaferaltun

Joined Aug 31, 2016
65
NO!
If you have -5V and -12V, it indicated the +rails are connected together.
You need a power supply that has +5V and +12V so the -rails are connected together.
Please check it as you will blow things up.

I actually do know what I'm talking about.
I'll draw a diagram in a bit...
Please check the specification in the attachment. It is T50-D (at right), it seems it can draw 0.2A as minimum while using 12V, my circuit can hold 200 mA since the micro processor peak current is 200mA. Is this can be the problem do you think?
 

Attachments

dendad

Joined Feb 20, 2016
4,451
Ah, it is ok. That is a relief!
The way you wrote the first message looks like you has +24V, -5V and -12V.You may have to add some load to get to the minimum load.
What voltage do you get on the output of the power supply as it is?
Have you measured the volts?
 

Thread Starter

zaferaltun

Joined Aug 31, 2016
65
Ah, it is ok. That is a relief!
The way you wrote the first message looks like you has +24V, -5V and -12V.You may have to add some load to get to the minimum load.
What voltage do you get on the output of the power supply as it is?
Have you measured the volts?
Ofcourse I did. They are normal, they are as they must be. If I right and you too, placing some load might be usefull to hold the current but in series connection right?
 

dendad

Joined Feb 20, 2016
4,451
I'm confused.
When you run it ok is it on separate power supplies?

Does the power supply run the Pi ok by itself, without your board connected?
And does it run your board ok by itself without the Pi connected?

Then does it overload when you connect them together?
If so, you have a wiring problem
It would help if you could show me a picture and circuit of it all.
 

dendad

Joined Feb 20, 2016
4,451
As a test,
With the supply disconnected and turned off.
Check there is no voltage left on the supply.
Then measure the ohms across the 0V to +24V
Then measure the ohms across the 0V to +12V
Then measure the ohms across the 0V to +5V

Then still NOT powered.
connect the boards together.
Repeat the...
Then measure the ohms across the 0V to +24V
Then measure the ohms across the 0V to +12V
Then measure the ohms across the 0V to +5V

If any of these readings is now much lower than before, you have a wiring error.
I'll be interested to see what your readings are.
 

Thread Starter

zaferaltun

Joined Aug 31, 2016
65
I checked and they are stable. I was not know this technique, now I learnt something new, thank you.

BUT, the 5V regulator on the module card is not stable, it has gone while testing I think. It is taking 12V input and its output pin is giving almost 12V instead of 5V since the regulator is not working and so it is causing the current is much more than it must be.

As summary, the power supplies are ok, my mistake, I 'm so sorry.
Best
 

dendad

Joined Feb 20, 2016
4,451
Have you checked the pinout of the AMS1117 as it is different from the 78xx series.
Also, did you insulate the tab of the regulator? It is not 0V, but is the output volts.
This is a bit of a trap as there are a few different pinouts for regulators.
 

Thread Starter

zaferaltun

Joined Aug 31, 2016
65
You are right, I 'm not using this regulator for the first time, I 'm sure from the outputs. I know why it is burning: It is about the "heat" because of the voltage difference between input and output, there is 7 volts difference and it is so hot when I touch, event my circuit drawing 80mA.

What can I do except placing a resistor between input and the regulator input pin you think? Because the resistor value must be different according to the circuit load.
 

dendad

Joined Feb 20, 2016
4,451
Have you measured the voltages?
Also, a circuit diagram would be a good idea to help. I'm still a little confused as to what the problem is. Originally I thought you said you had a "shortcut" problem, and I took that to mean a short circuit problem.
And 80mA with 7V drop is only a bit over half a watt. Does the regulator have a heat sink? That should be well within it's rating.
I've taken to using switch mode regulators on most of my boards now. No heat sink required and a smaller power consumption. They do cost a bit more.
Here are a couple of links..
http://uk.farnell.com/xp-power/tr05s05/dc-dc-converter-0-5a-5v-sip/dp/2319830
http://au.mouser.com/ProductDetail/...=sGAEpiMZZMvfpQN6QVmrfKHR1X8i8dWBvz/2JDG7tt0=

To drop the voltage to the regulator a bit, a series diode or a few will lower it about 0.7V a diode, but at the current you are saying of 80mA, it should not need anything other that adding a heat sink.
Just remember to insulate the heat sink otherwise you will have a potential trap with the 5V on the heat sink.
So you need to either isolate the regulator tab from the heat sink, or the heat sink from touching anything else.
 

AnalogKid

Joined Aug 1, 2013
10,986
Please post a photo of the power supply connector block showing the labels.
Or a datasheet with the connector pinout.

With the power supply off and disconnected, measure the DC resistance between terminal block pins 5 and 7.

ak
 
Last edited:

Thread Starter

zaferaltun

Joined Aug 31, 2016
65
Have you measured the voltages?
Also, a circuit diagram would be a good idea to help. I'm still a little confused as to what the problem is. Originally I thought you said you had a "shortcut" problem, and I took that to mean a short circuit problem.
And 80mA with 7V drop is only a bit over half a watt. Does the regulator have a heat sink? That should be well within it's rating.
I've taken to using switch mode regulators on most of my boards now. No heat sink required and a smaller power consumption. They do cost a bit more.
Here are a couple of links..
http://uk.farnell.com/xp-power/tr05s05/dc-dc-converter-0-5a-5v-sip/dp/2319830
http://au.mouser.com/ProductDetail/RECOM-Power/R-78E50-10/?qs=sGAEpiMZZMvfpQN6QVmrfKHR1X8i8dWBvz/2JDG7tt0=

To drop the voltage to the regulator a bit, a series diode or a few will lower it about 0.7V a diode, but at the current you are saying of 80mA, it should not need anything other that adding a heat sink.
Just remember to insulate the heat sink otherwise you will have a potential trap with the 5V on the heat sink.
So you need to either isolate the regulator tab from the heat sink, or the heat sink from touching anything else.
Yes I measured them, I replaced the regulator with new one and its output is 5V, it is ok. But the heat. I havent a heat sink and maybe helpful as you said if I ensulate it as well also. I just dont understand why do I need a heat sink becuase of just 80mA, because it says it is 1A 5V regulator. You wanted the diagram, unfortunately I cant share it since it is a secret project but let me say something: I tried this regulator with a very simple circuit before and its heat problem was there also, the simple circuit like this ordered from input to output: 12 dc input, then 22uf b type capacitor, then the regulator, then 22uf b type cap again, and the output. I think it can be better using regulators like you give the link, these AMS regulator are just cheap and undurable I think.

The power supply seems still shorting between the positive pins and gnd but working good with all devices, also there is no resistance difference as you suggested the test. I really didnot understand, I m still in doubt its minimum current draw, 0.2 A.

I think I will use some diodes, regular or zener. It is better the heat sink I think.
 
Top