IPAC and flashing LEDs

[ahofle]

Hello all, I have a question regarding the IPAC's support for the flashing numlock, capslock, and scroll lock LEDs.  I recently hooked up an LED driver board (thanks whammoed) to the IPAC to drive some superbrights in MAME to flash the start buttons when credits are inserted for games like Asteroids, Centipede, etc.  Well the lights do work and flash, but they are the inverse of what they should be (ie. they are on when they should be off and vice versa).  Using a multimeter on the IPAC's LED harness connector, I can see that it's putting out 5V when the LED should be off and 0V when on.  Is this normal?  Is there something in MAME I should be setting so this is reversed?  Any help would be appreciated!

Thx,
Andy

[BobA]

Looking at the diagram on ultimarcs site the LEDs are fed +5V all the time and the go to ground to turn on. 
So it sounds like you have the LEDs backwards. Diagram was from IPAC info not sure why it says OPTIpac

[ahofle]

I'm confused what you mean by backwards.  You mean the anode and cathode are backwards?  I thought they would not even light if backwards.  I'm actually using a driver circuit like Oscar's below, so the +5V to the LED is actually going to a transistor on the driver circuit.  I'm not very knowledgable on this stuff though -- could the transistors be backwards?

EDIT: here is Oscar's page on the driver: http://www.oscarcontrols.com/led/index.shtml

[BobA]

What I meant by backwards was wired with gnd as the common not +5v.  Since you are using an interface instead
that complicates matters.   All I was pointing out is that the terminal will go low when the LED is on.  This is the proper action of the IPAC and seems to be what you saw when you measured the volages.  If you built the interface then a low signal on the input should turn the transistor on since the PNP will drive 5V thru the resistor and LED to gnd where the cathode of the LED is terminated.   

Why not check the IPAC with a single LED and resistor to check when it is on and off and trouble shoot your interface from there.

[ahofle]

It's actually sounding like I need to use the LED 'shared' inputs instead of the LED header pins on the IPAC.  In other words, wire from Coin1 and Coin2 on the IPAC to the driver board.  That's what Oscar did it looks like.  I suppose that will flash the LEDs when a credit goes in, but now that I think of it that doesn't bother me at all.    I'm gonna try it when I get home.

Thx!

[The 80s Man]

Well the lights do work and flash, but they are the inverse of what they should be.
Thx,
Andy

The same thing happend to me the first time I made a driver board. My problem was that I use NPN transistors in stead of PNP ones. Dumb me..

[ahofle]

LOL OK that's my problem too then (this board has NPN transistors). 
Thanks for that clarification!  Off to radio sack I go.

[whammoed]

Well the lights do work and flash, but they are the inverse of what they should be.
Thx,
Andy

The same thing happend to me the first time I made a driver board. My problem was that I use NPN transistors in stead of PNP ones. Dumb me..

Yep, thats probably it.  The first sites I read simply said that PNP and NPN transistors "did the same thing".  Didn't actually matter for me since I am using advmame and have greater control over whether the leds are off or on.  Switch out the transistors and all should be well I would think.

[whammoed]

Well the lights do work and flash, but they are the inverse of what they should be.
Thx,
Andy

The same thing happend to me the first time I made a driver board. My problem was that I use NPN transistors in stead of PNP ones. Dumb me..

Also, I don't think it was too dumb to think you could substitute one for the other.  A few popular hits from googling don't quite reveal the difference between the two.  Here is a site that does:
http://encyclobeamia.solarbotics.net/articles/bip_junct_trans.html
There's at least two of us who thought it would be just fine. 
Yes, I am still dumb, but I'm just not sure this is a defining moment...