Lots of good questions and I actually have some of the answers already. Please excuse my novice approach and language regarding components, I don't know all the lingo .

I will have (4) parts to this setup
1. Power supply
2. Relay board (next board will use the ULN2803 chips)
3. Sound card
4. PCB board where the other components and relays will be soldered in. I have also dabbled in etching a circuit board in the past but did not go that route with this project because I did not think it would grow this large . I may try that next time around.

I will be using THESE for inputs and outputs on the PCB and relay boards to ensure flexibility when it comes to taking things apart. Essentially making my own ribbon cables I suppose. I will be cutting the connectors off of a few ribbon cables to make the jumps between boards and the wires will be connected to those terminal blocks. You can kind of see my idea for the PCB board in the attached drawing.

Obviously I have made some revisions to the design since that drawing but thats basically it.

 

Your diodes across the relays are backwards.

You could put red/green/yel LED PCB indicators across the relays in the direction you have the diodes installed now.

Red for the red, grn for green and yellow for the other one.

How about a real power inlet connector for 120 V in. They can be bought with switches and fuses. With the LEDS in place, the 120 VAC cord could be removed and the board could be more safely worked on. You could investigate the smaller versions used on laptop bricks. A 120 VAC indicator might be nice too.

I would not be crazy with all of the connectors being 2 pin nor would I like the ability to apply 120 VAC to the 12 V DC port. 12 and 120 look alike after a few beers.

You, at least have a plan.

So. your lights are 120 VAC? Could some sort of low voltage LED system work?

 

Good call on the 120v / 12v connectors being the same. I will definately change those inputs so they can not be switched. I will also fix the diodes, didnt even notice that. I am going with 120v bulbs because I want the large incandescent look for the right/wrong answer indicators.

I have attempted to put together the last schematic and layout, and I will attach both here. I am VERY tempted to just wait for a few ULN2803 chips to come in considering how much time and space it will save me. I am assuming that if I cut the ground to the ULN chips they will not work correct? I have them layed out so they only have a ground connected if the answer is correct. Have a look and see if I messed anything up, Im sure I got something wrong here..

**EDIT***
I know I can substitute a DPDT relay for two of those singles as well.

 

Write up a circuit description as best you can.

What I don't like is the removal of ground to the driver chips. Not sure if it's the right thing to do.
The chips output will be an open circuit when the input is an open circuit or the input is grounded.
The chips output will be grounded when the input is greater than 2.4 Volts.

When used as a "relay driver" the (+) side of the relay is always connected to +12
(COM) of the Driver IC goes to +12
A voltage greater than 2.4 V on the input grounds the (- )relay coil turning it ON

The first time I saw this chip (the ULN2003 anyway) was back in the 80;s. That's like 26 years ago. It confused me for a while.

For a DC relay, there really isn't a (+) and (-) side, HOWEVER many sockets force the convention. The sockets sometimes contain the free wheeling diode hence the importance. It's also important for magnetic latching relays

A diode drops about 0.6-0.7V when it conducts.

You know about free-wheeling diodes.

I'll leave you with that.

Now, come up with a circuit description:
Your best shot anyway.
If your lucky, you'll discover mistakes.

For me:
a) Unclear how it's SUPPOSED to operate.
b) Isolating with diodes generally means two per input.
c) Don't know if it would be worthwhile to drive relays and even LEDs with another driver?
d) An uncommited driver attached to a LED MIGHT provide a troubleshooting tool.
e) I'd still like to see a LED+Resistor across the coils so you can debug without 120 V power.

There are versions of this http://www.alliedelec.com/schneider-electric-magnecraft-725bxxsc3ml-12d/70185418/ relay I like, but I'm mostly OK with what your doing. I'd prefer you use a technique appropriate for 120 VAC (conduit/romex) and another for low voltage (exposed). The latter is called NEC Class II wiring (It covers thermostats and doorbells). The "interface" between the two worlds have to be carefully thought out. Meaning you can put two relays in a J-box, wire the 120 side with Romex and/or conduit. You can likely get away with a connector for the 12 V and wire inside the box rated for 120 V service. Outside the box, you can wire how you would like because your reasonably assured that 120 V won't contact the 12 V DC wiring (inside the box) AND the 12 V DC circuit is power limited to 100 W or less.
Inside that J-box you need physical separation and/or wire rated for the highest expected voltage.

 

................
a) Unclear how it's SUPPOSED to operate.
b) Isolating with diodes generally means two per input.
.................

a) Unclear about how what is supposed to operate?
The whole circuit or some specific part?
b) Isolation of what?

 

a) A blurb by the TS will help HIM understand. I've had a cold after a cold for seeming 8 weeks. Digesting all of the info from the beginning is hard. So things like a connection between x an y (may be use a dashed or green line) is an example of a right answer and the green dotted line shows a typical connection for a right answer. RY? and RY? drive are 120 VAC contact closures for the lamps for the right and wrong answers.
When a right answer occurs: Blurb
When a wrong answer occurs: Blurb

b) I thought there was a discussion of isolation or "Diode -OR gates earlier. A 2 input diode OR gate usually requires 2 diodes. That was a comment, and not necessarily the wrong answer.

You get novice programmers to find their mistakes by asking them what their code does. Generally, you don't even have to pay attention, They find their own problems. Not always true, but it worked more times than I could count.

 

I think a 120V bulb is a bad idea unless it will be in a protective housing. Kids try taking stuff apart and, if they can access the bare bulb, they have access to 120V (I know outlets are 120 V but those are hard to put your finger into.

Also. Where is your fuse on the 120V? Should you have one?

There are some cool looking bayonet style incandescent 12V bulbs. Might want to look into those.

 

a) A blurb by the TS will help HIM understand. I've had a cold after a cold for seeming 8 weeks. Digesting all of the info from the beginning is hard. So things like a connection between x an y (may be use a dashed or green line) is an example of a right answer and the green dotted line shows a typical connection for a right answer. RY? and RY? drive are 120 VAC contact closures for the lamps for the right and wrong answers.
When a right answer occurs: Blurb
When a wrong answer occurs: Blurb

b) I thought there was a discussion of isolation or "Diode -OR gates earlier. A 2 input diode OR gate usually requires 2 diodes. That was a comment, and not necessarily the wrong answer.

You get novice programmers to find their mistakes by asking them what their code does. Generally, you don't even have to pay attention, They find their own problems. Not always true, but it worked more times than I could count.

Okay, so basically the board works like this....

1. There is a row of names on the left side of the board and there are pictures on the right side of the board.
2. The junior ranger has to match the correct name with the correct picture. For this particular board we are matching bird names to bird pictures.
3. Each name and each picture have a metal disc the size of a quarter next to them. Behind the board the metal discs that are matching pairs are tethered by a conductor.
4. Out front there are two metal pens that when they touch the correct name with the correct picture, complete the circuit behind the board. Those pens terminate on the PCB board and and can be identified on the PCB board by the names "Quesion" and "Answer." To assist the junior ranger with the selection, the pens are magnetized and stick to the metal discs so they dont have to stretch to reach the answer .

The rub here was that they wanted much more logic behind what happened when the circuit was completed or not completed. They want a greeen light and a sound thats associated with the correct selection. So when they select the name Blue Jay with the pen on the left and then correctly selecct the picture of the Blue Jay on the right, they hear the sound of a Blue Jay singing and a green light lights up. If they select incorrectly, they get a red light and possibly a buzzer sound.

Some challenges to the project (for me), were to figure out how to make the board know when a wrong answer was selected. On the first box of this type we had nothing identifying when a wrong answer was selected. If you made the correct selection, a green light would come on. If you made a wrong one, nothing happened. Knowing when the correct answer was selected was easy, but identifying when a wrong selection was made and alerting the user to it was much more difficult. @crutschow came up with the original design to make it work, I just elaborated (way to much) on it.

I hope this helps clear up the idea(s) behind the project. I started building the wooden portion of the cabinet today and should be done most of it by tomorrow. After thats done I will begin on the electronics part of it.

I am attaching a picture of an old board we have that lets a junior ranger try to match tree names to samples of trees. You can see the pens at the base of the box in their holders and you can see the little metal discs beside the names on the left and the samples on the right.

 

I think a 120V bulb is a bad idea unless it will be in a protective housing. Kids try taking stuff apart and, if they can access the bare bulb, they have access to 120V (I know outlets are 120 V but those are hard to put your finger into.

Also. Where is your fuse on the 120V? Should you have one?

There are some cool looking bayonet style incandescent 12V bulbs. Might want to look into those.

The bulb(s) will be all the way on the top of the board, about 7' off the ground and shoud be inaccessible to little fingers . I may eliminate the 120v portion of this and just go with 12v bulbs or large LEDs the more I think about it.

 

Also, if interupting the ground to the ULN2803 chips is not a good idea I may just switch the + lead of the speaker with a relay instead as @crutschow suggested awhile back.

 

Also, if interupting the ground to the ULN2803 chips is not a good idea I may just switch the + lead of the speaker with a relay instead as @crutschow suggested awhile back.

OK, Why does the speaker have to be turned off?

How do you use the signal ANSWER?

How do you use the signal SELECT?

So, pretty much you created a design spec. For the most part, I got that. Although the actual sounds for a right answer and a sound for the wrong answer I must have missed.

So, sound termination (Play for x time) is in the sound card?

You sort of end up with a "contact" closure input for a right answer and a "contact closure" input for a wrong answer. That's what I don't see. Those are the dots I can't seem to connect.

So, where is your "wrong answer" sound?

So, anyway I see if you have a signal at one of the drivers, you get an independent input to the sound card, and effectively a "correct answer" signal. I could see where this could light the proper bulb.

What's the deal with "not correct"? Sound card "not correct" input is where? So, wouldn't I see somewhat of a similar deal except, only one input to the sound card. Wouldn't all of the "wrong answers" be "in effect" "parallel switches"

 

Im sorry but I only have a quick moment before I have to run out but I will be back later to try and explain more thoroughly.

1. The sound card is a stand alone over the counter PCB, its a 20 channel sound card that is triggered by a ground, hence the ULN2803 chips. The run time, volume, sound selection, etc, us all handled on board that card. HERE

2. The "select" and "answer" terminals on the circuit board in my schematic represent simple contact switches in the form of magnetic pens. The "answer" and "question" terminals each have a wire that comes off of them that goes to a magnetic pen.

When I get back I will try to comment out the schematic so it makes more sense.

 

Thanks. That was "just enough" to help me fill in the pieces.

Comments are now:

1. Probably you can get away with interrupting the ground on the chips. See 1a.

1a. Being the Devil's Advocate, you MAY need a leakage resistor for the off state.
e,g, The off state input current is 65uA per port. You need a place for 8 times that current to go. So R<=(12/ (65uA *8) since there is no ground/

2. The sound board doesn't tell you what LOW is. Sometimes connecting to ground requires a FET rather than a transistor. There's a good possibility that it's a TTL compatible input which means a transistor will work. The drop will likely be around 0.6 V, not ground.

3. I don't see a "wrong answer" input to the sound card. I would guess that the RED LED, THE DRIVER POWER and a sound card input have to be created.

4. FWIW: A Driver chip can act as an inverter with a pull up resistor. (A resistor to +!2) e.g. 10K

5. I LIKE the fact that the magnetic pen activates a relay directly. It's a good thing.

6.I don't like 120 V lamps. You can use them if the 120 stuff is moved off board and enclosed. Look at automotive LED bulbs as a possibility.

7. Your full schematic should show dotted lines or something labeled say Pen A (Right answer) Pen B (Right answer) and the same fro wrong answer. Just to show what one right and 1 wrong input looks like.

8. I'd sill like to see relay indicators. Where RED and GREEN are used for the right and wrong relay coils.

 

Hi

You might want to check the current load on each of the 2803 outputs driving a lamp. I didn't check the 2803 but the 2003 can support about 80mA on each port with all 7 ports used.

 

eetech:
The 2803's are driving what we believe is a TTL compatible input . It's described as a contact closure to ground.

 

3. I don't see a "wrong answer" input to the sound card. I would guess that the RED LED, THE DRIVER POWER and a sound card input have to be created.
.

I believe that there is not a requirement for a "wrong answer" input to the sound card. The OP stated that this was optional. "Possibly a buzzer sound" is the closest to a requirement per his design criteria.

 

Wow, there is a lot going on here...LOL. Let me try to take it one at a time.

1. If I do chose to use a wrong answer buzzer I will simply tie it into the "wrong answer" relay. That should be the easiest way.
2. So I can cut the ground off to the ULN chips in order to disable the sound card? What are you recommending I use if I do that? A bleed out resistor? Can you explain? Are you now saying that the ULN chips wil NOT output a gound? I dont quite follow your statement about a FET or 0.6v.
3. Are you saying that a ground does not trigger the sound card? I though thats how I read it on the data sheet?

You will need to dumb it down a little when replying..LOL. I dont have the electronic background some of you have. I am just a backyard hobbyist..

I got the contruction part just about done today. I will post a pic shortly.

 

I looked at the datasheet for the sound card.

This is the answer to 2 and part of 3.

Are you now saying that the ULN chips will NOT output a ground?

No, it's not the same as "ground" to a purest. Meaning, it's not zero Volts. There will be a SMALL voltage drop across that transistor that will increase with the amount of current drawn and will decrease with temperature and it varies a LITTLE per device.

I had to ask the manufacturer of a demand side management system what the input circuit looked like. I would not be able to use a simple transistor based on their design. I think it's good practice for a manufacturer to show you what the equivalent circuit is for any input or output that you connect to.

In one case, I was buying potentially 6 power supplies and I needed access to the current sense resistor. It turned out it was available on the back panel. I also special ordered them to operate at 240 V rather than 120 V. A number of years later supplies from another manufacturer were available with a wide range of AC input, had digital meters and had 0-5 V outputs. Isolated outputs cost more. They were also 1 Rack Unit high.
We would have liked a power meter. Technology changed.

To "some" electronics, ground really has to be almost zero volts. A silicon transistor can't pull it's output to 0 volts. It will pull it to around 0.6 volts or what's called a "diode drop". The 0.6 V is a number at room temperature and it's approximate.

So, if the sound card works when you take a diode to ground rather than a wire, it will likely be fine. If it works with 2 diodes in series it will definitely be fine.

A FET can. When we talk about FETS, we talk about Rds(on) or the effective resistance from the drain to the source which are terminals of the FET. That number can be in the milli-ohm range. No diode drop.

I'm nearly certain that it's not an issue, like 99%.

 

2. So I can cut the ground off to the ULN chips in order to disable the sound card?

Their might be issues. There might not. Look at the voltage on an input relative to your power supply ground after it's been on for some time. It might creep up with the ground not connected. Look at the output as well.

If it's a problem, I would try connecting the ground of the chips to a weak ground when they are supposed to be off and real ground when they are supposed to be on. e.g. Short out the resistor, I gave you how to calculate a value for one chip. For two chips, half it.

I have a ULN2003 somewhere, but I don't have time to play.

 

eetech:
The 2803's are driving what we believe is a TTL compatible input . It's described as a contact closure to ground.

oh..ok..i thought in an earlier post it was mentioned replacing the lamps with leds and driving them directly with the 2803's.