Connecting external power

The Dot Matrix Display (DMD) is a 32x16 array of high-brightness LEDs for visually striking effects. [Product Page]
Post Reply
CodeStorm
Posts:8
Joined:Sun Aug 05, 2012 12:17 pm
Connecting external power

Post by CodeStorm » Mon Jul 07, 2014 3:10 pm

Hi all, I have just finished testing my setup with a single panel and have mostly finished making a housing for 3 panels. When I get around to connecting the 5v terminals to my ATX supply, I'm not sure whether it would be isolated from the Arduino's own power. I'm using an EtherMega with power supplied over USB through a powered hub to protect my PC. I am unsure whether I can simply hook up the ATX supply and leave everything else as-is. Since I will be sharing the Arduino between other projects, I have decided to eventually feed 12v from the ATX PSU into the external power jack and jumper it to supply its own power through its regulator. This is so that I don't risk damaging the regulator by supplying 5v directly through a shield and forgetting to re-jumper the power selector, the worst case scenario should be it getting no power if I jumper it incorrectly this way. I'm just a little concerned that the potential may exist for the power from the ATX supply to feed back through the DMD cable to the regulator. Also, this may be a stupid question but where does it actually get its power from? is it AREF?

angusgr
Freetronics Staff
Freetronics Staff
Posts:853
Joined:Tue Apr 09, 2013 11:19 pm
Location:Melbourne, Australia
Contact:

Re: Connecting external power

Post by angusgr » Tue Jul 08, 2014 6:14 am

Hi Codestorm,

The DMDCon adapter is not fully isolated. However powering it via the 5V rail of an ATX supply should be perfectly fine, just make sure ground from the ATX supply is also connected to the DMD display ground lug.

You may find that if you power off the Arduino but leave the DMD powered then it may backpower the Arduino parasitically through the DMDCon board. I would recommend actually connecting the 5V from the ATX to the +5V pin on the EtherMega as well so they're powered in tandem, but this is up to you. If you remove the "USB Power" jumper from the EtherMega then this will prevent the ATX from possibly backpowering the PC (if you're injecting power directly into the 5V pin then you can leave the power jumper off entirely.)
CodeStorm wrote: this may be a stupid question but where does it actually get its power from? is it AREF?
It's not a stupid question at all! There are no power connections on the DMDCon adapter board. We don't recommend connecting the DMDCon without separately powering the DMD board via the 5V/GND terminal lugs. It turns out that the DMDCon will actually parastically power the board via the data connections and it updates correctly, but this is not recommended (that said it doesn't seem to usually cause any problems, but it might cause damage.)

User avatar
stryker
Posts:282
Joined:Sat Jan 14, 2012 2:44 pm

Re: Connecting external power

Post by stryker » Tue Jul 08, 2014 6:40 am

angusgr wrote:It turns out that the DMDCon will actually parastically power the board via the data connections and it updates correctly, but this is not recommended (that said it doesn't seem to usually cause any problems, but it might cause damage.)
I don't believe I've seen damage from this, but the brightness of the display certainly suffers when it's powered that way.

An ATX power supply does a great job, but don't follow my lead and use the 12V output. 4 of my 6-DMD array aren't quite right after getting that extra boost. From talking to others that seems to be the most common mistake - and a trap for late night tinkerers. I can confirm 12V backwards lets the magic smoke out too.

Geoff

CodeStorm
Posts:8
Joined:Sun Aug 05, 2012 12:17 pm

Re: Connecting external power

Post by CodeStorm » Wed Jul 09, 2014 5:13 am

Thanks for all the replies.
Hi Angus, it's funny you mention the parasitic supply as I had suspected this in a previous project involving several 74HC595s where a rather major modification I made to a breadboard resulted in the ground connections being severed from one of the chips.

For some reason I blew the chip but I didn't notice the mistake and suspected a fault elsewhere. I rectified the suspected issue and found it worked perfectly for ages with slight 'leakage' on the outputs which I could never quite explain until I stumbled across the error and jumpered the ground back.

The board was driving several bridged ULN2803s to power some hefty LED arrays and I replaced these boards with TIP122 (off memory) based drivers, suspecting a failed 2803. The outputs drove the TIP122s through resisters perfectly but I also had them connect to ground through LEDs/resisters as a display on the chassis so the operation of the controller could be monitored.

These 'monitor' LEDs never turned fully off until the ground connection was restored but, in spite of this leakage, the TIP122 driven LED arrays never came on incorrectly at all.

Anyway, thanks again for the advice.

Post Reply