Starting the cabinet all at once. It works! And here's how...

[crashwg]

What happens when you shut down the cab?

I've never taken any electronics classes or anything, but it seem to me that you wouldn't be able to shut down for more than a couple seconds...

What I mean by that is:
When power is applied to the PC, the power button is already being pressed by the cap.  When the cap reaches full capacity it stops pushing the button.

Therefore, when you shut down, the cap would cycle the whole thing over again.  Am I right?  Excuse my ignorance when it comes to electonic components, I can't even correctly solder a modchip to a playstation.  

[danny_galaga]

you

[SirPeale]

When you shut down the cabinet (by turning main power off) the capacitor will discharge.  Since it's such a small cap, it doesn't take very long (less than a minute)

That way, next time you're ready to run the cab again, it all comes on @ once.

[krick]

What happens when you shut down the cab?
<SNIP>
Therefore, when you shut down, the cap would cycle the whole thing over again.  Am I right?

I think what you're asking is (correct me if I'm wrong)...

If you shut down the computer (by clicking shutdown via the start menu), will the capacitor cause the computer start right up again before you get a chance to shut off the main power?

I initially thought this myself at first but I think that what happens is once the capacitor is fully charged, it always acts as an open circuit.  It will not discharge until the mains power is cut off for a significant amount of time.

[SirPeale]

I initially thought this myself at first but I think that what happens is once the capacitor is fully charged, it always acts as an open circuit.  It will not discharge until the mains power is cut off for a significant amount of time.

Less than a minute.  This setup isn't idea for those who are running Windows.

[krick]

Peale,

If you're setting up your cabinet so you can swap in a JAMMA board easily, how are you handling the speaker issues?

I was thinking of something like this...



If you want, you can wire it using stereo plugs and jacks like this...



On the jack connecting to the JAMMA connector, just don't wire the "ring" connection to anything.

I'm pretty sure you can also wire the JAMMA mono signal to both left and right speakers (connect the signal to both tip and ring).  The only caveat is that some JAMMA board amplifiers might not be able to drive two speakers wired in parallel this way.

If you have other boards that have custom stereo output that bypasses the JAMMA connector, you just have to make an adapter cable with a 1/8 inch stereo jack.


Oh, and another speaker related thought.  If you want to avoid having to wire up a power source for the speaker amplifier, you can get a set of USB powered speakers like I'm using.  They tap 5V off of a PC USB port and automatically turn on and off with the computer.  No power brick needed.

http://www.arcadecontrols.org/yabbse/index.php?board=1;action=display;threadid=16381

Just look on ebay for "usb powered speakers".  I bought 10 pairs of Benwin SP691 speakers for $28.97 (shipping included).  I used one pair in my cabinet and sold 6 pairs on eBay.  I have 3 pairs left for future projects.

...
Krick

[SirPeale]

Since the cab uses the original speakers, I didn't have to do anything; I just wired the speaker out jack to the speaker input on the J-PAC.

[krick]

Since the cab uses the original speakers, I didn't have to do anything; I just wired the speaker out jack to the speaker input on the J-PAC.

I'm using the original cabinet speakers too.

There are two problems to handle....

1) sound coming from computer is stereo and un-amplified
2) JAMMA (J-PAC) expects mono and amplified

[SirPeale]

Ah, now I see.  My soundcard has a built-in amp, so I didn't have to do anything.

What you could do is use the amp from a pair of powered speakers, and wire it in that way.  

[krick]

Ah, now I see.  My soundcard has a built-in amp, so I didn't have to do anything.

What you could do is use the amp from a pair of powered speakers, and wire it in that way.  

That's what I'm doing.  My amp is USB powered as well so I don't have a power brick to worry about.

However I'm still curious about how you deal with the mono vs stereo issue.

[SoundDoc]

Just reading over this thread got me thinking, you want to turn the computer on when the power switch is flipped, off when its flipped off, but it would be advisable to also have the computer instead of just powering straight off (for those running windows) to start shutting down instead.

about $10 in components should do it. a 555 timer, a couple caps, resistors, and a relay. Basically, you need a rising & falling edge trigger, with a delayed pulse one shot.

you'd plug the computer into constant power, and the "device" into the switched mains, and a small cable from the "device" to the power on header on the MB.

When you turn on the cabinet, the relay  picks up, triggering the "device", after a second, it would pulse the power on header for a second, starting the computer.
Same thing on power off of the cab, the drop of the relay would trigger the "device", pulsing the power on header, initiating a shutdown, or hibernation, or whatever you've got it programmed to do. No hard crash of the computer.

I agree that the way Peale has it set up must work, and if your not concerned about crashing the machine, its a easy way to go.

I've seen too many times though a machine just getting its plug pulled killing the HD, or the CPU, as it doesn't get a chance to shut down gracefully. When you just pull the plug on the PSU, the 12v and 5v lines surge momentarily, and can kill components (HD's usually die first).

If anyones interested, I'll draw it up quick. Everything for it would be availiable at radio shack, and for probably less than $10. You'd need to be able to solder though.

SD

[krick]

I agree that the way Peale has it set up must work, and if your not concerned about crashing the machine, its a easy way to go.

Peale is using DOS so there's not likely to be a problem with just "pulling the plug".  If I was building a cabinet for a family member, I've probably use DOS and do it the way he does.  Less stuff to go wrong.  


On the other hand for those of us who need to use Windows....

If anyones interested, I'll draw it up quick. Everything for it would be availiable at radio shack, and for probably less than $10. You'd need to be able to solder though.

YES PLEASE!  If this works, you'll be famous....  Well, with me anyway.

[SoundDoc]

Ok, give me a day here to draw it up.
I still hate just pulling the plug on a machine. When the PSU just looses power, the voltages go all out of whack, and components can take damage. when the control signal from the computer shuts down the psu, it "turns off" the power supply. Lets call it a soft off, vs a hard off.

Anyway, I'll draw it out tomorrow, and calculate out the values.  Should have something by the evening.

SD

[krick]

I found this.  You might find it helpful...

http://www.schematica.com/555_Timer_design/555.htm

[danny_galaga]

going slightly back, i just happened to notice that very occasionally maybe one time in fifty that my cab doesnt power up when i turn it on. you might recall that i am using a relay on its lonesome. i suspect this is because it must be just on the edge of not working (or working i guess). in time it will either get better or worse i imagine as the relay physically wears out. ill keep you posted.
still, when it doesnt power on, i just turn it off and on again and its away!

[krick]

More interesting stuff that someone might find useful:

I found a thread where a guy mentions using a 4.7uf capacitor across the power switch.  Then he mentions that some motherboards don't work with a capacitor and then goes on to  describe a circuit that will work by tapping 5V off of the "wake on lan" header that is present on most modern motherboards.  Later in the thread, he discusses other solutions involving relays, zener diodes, and/or 555 timers:
http://lists.exploits.org/ups/Feb2003/00018.html
http://french.moren.ca/~kenm/cct/index.html
http://lists.exploits.org/ups/Feb2003/00019.html

I've quoted the complete text below for your convenience...

Hi All...

I'm new to the list.  I've been a Linux/Unix sysadmin for several years,
but my main background has been electronics/hardware.

recently I installed NUT for a client and ran into a problem with his
hardware, namely, his ATX motherboard had no provision to auto-startup
after a power failure: no jumper; no BIOS setting.  Experimenting on a
few of  the machines I had sitting around, I found that connecting a
4.7uf. capacitor across the power switch connector on the motherboard
did a fine job of simulating a momentary switch closure.  It would be
best to use a non-polar capacitor (hi-K ceramic or mylar) to save having
to discover what the polarity of the switch connector is and avoid a
polarity mismatch.

However, in this case it did not work on my client's system, whose
motherboard was one of the small percentage which require a short delay
before applying the switch closure :-(

So I developed this circuit:

http://french.moren.ca/~kenm/cct/index.html

Hope it works for you if you need it...

Regards,
Ken


Charles Lepple wrote:

>Out of curiosity, did you look at feeding the wakeup signal back through
>the WOL port? At first glance, I'd say that you could get by with just a
>555 timer and 3 or 4 passive components, but I'll readily admit that I
>haven't taken a look at this in any great detail.
>
>  


I had not thought of trying to use that at the time, however I expect I
would run into the same issue, that of needing a delay after powerup.  
Not knowing the specifics of the WOL interface standard, I didn't go
there...

 For most motherboards, the simple capacitor suffices.

My circuit was based on what I had at hand. Certainly with a little
ingenuity, one could probably build a working device around a single
555... I didn't have one handy.  With the 4528, which I did have, it was
relatively easy: the first one-shot to set the delay; the second to set
the output pulse length. I used the little relay because I had it, and
to eliminate any polarity issues at the motherboard power-on connector.

If one used the relay with a 555 to provide the delay, they could use it
to switch in a capacitor to the motherboard connector as in the simple
solution... in that case you could eliminate many of the components of
my circuit as the 555 could drive the relay directly without the
buffering transistor that I used. A 555 used as a one-shot uses one more
capacitor than its 4528 equivalent. It may not have the problem that the
4528 has of delivering an immediate power-on output pulse in addition to
the delayed pulse, which I devoted two capacitors and a resistor to solving.

Thinking about it now, I can visualize a circuit using a resistor, two
capacitors, a zener diode and the relay that would do the job, if we
wanna talk "minimum component"  ;-)

[SirPeale]

On the other hand for those of us who need to use Windows....

The only reason to use Windows that I can think of (unless you're using another emulator that is Windows only) would be sound driver issues.  There aren't many DOS sound drivers for current stuff.  Unless they hide really well.

[SoundDoc]

Haven't built it up to test yet.  Just did it on a coffee break, so its pretty rough, but this should do the job. Decided to steal power from the WOL header instead of a wire from the PS connector, just makes it a bit neater. And I was just going to use a transistor to switch the power connector, but this way its fully isolated from the pc and can't hurt anything if you build it wrong. (well, it could hurt you   )
Some MB's also have the 5v standby power availiable on a pin on the MB header, so you could get it from there as well.
If I get a chance tonight, I'll throw one together and test it out. I've got to do something while I wait for the paint on my cab to dry.....  

SD
*edit* deleted the pic as it REALLY messed up the thread.....  
look futher down for the updated circuits......

[krick]

Haven't built it up to test yet.  Just did it on a coffee break, so its pretty rough, but this should do the job. Decided to steal power from the WOL header instead of a wire from the PS connector, just makes it a bit neater. And I was just going to use a transistor to switch the power connector, but this way its fully isolated from the pc and can't hurt anything if you build it wrong. (well, it could hurt you   )
Some MB's also have the 5v standby power availiable on a pin on the MB header, so you could get it from there as well.

You are the man.  I can't wait to get some parts together and give it a try.

I hadn't thought about the need for 5V of standby power.  Do all current motherboards have WOL connectors?    Do PS/2 keyboard connectors supply 5v standby?   I think some USB ports do.

If you can get this design polished and it works well, you might want to talk to someone about getting some boards made.  I'm sure you could sell them.

[SoundDoc]

Lots of new motherboards do have the WOL connection, as do most atx style older ones. If the MB has onboard Lan though, they usually leave it off. its hit or miss otherwise.
Some have it on the main button header, others if they have power on KB always power the ps/2 keyboard port, and some do provide it on the USB connectors, usually with a jumper But you can always find it on the MB power conector.
Some people though I don't think are confident/savy enough to solder/splice into their motherboard, so I threw the WOL connector possibility out there so you wouldn't have to modify the MB or PSU.
What MB are you using? I'll take a look and find the best place to get it from.

SD

[krick]

What MB are you using? I'll take a look and find the best place to get it from.

I'm not really worried about me.  I'll find standby power somewhere.   Even if I have to dive into the power supply to find it.  

I'm just concerned about other people who might not be a technically savvy.

I have 3 Pentium4 motherboards in my house.  All have integrated LAN.  Most new boards do as well.   So the WOL header might be a problem.

My MAME cab is currently running on an ASUS P4S533 (the original version) but I may be swapping it with an ABIT IS7-E (original version) in the future.

UPDATE:

I found this diagram...

...on this page...
http://www.duxcw.com/faq/ps/ps1.htm
...that shows which wire (#9) on the ATX connector is the 5V standby.  If all else fails and I can't find a convenient spot on the motherboard, I'll tap off of that wire.

I had a thought.  It would be pretty cool to implement this circuit on a card that can be plugged into a PCI slot (without making any electrical connections).  

[krick]

Haven't built it up to test yet.  Just did it on a coffee break, so its pretty rough, but this should do the job.

What's the deal with the 3 diodes in the circuit?
They're not listed in the parts list.

[SoundDoc]

I'll have to fix that...  
The three diodes make up part of the trigger circuit. 2 isolate the two "triggers" and the third clamps the trigger signal from flying too high above vcc.

I'm changing it a bit as I go as well, adding here and there to make it more user friendly.

sd

[SoundDoc]

I just wired it up, and it works just fine.  
Actually, heres a question on how you want it to operate.

1) It can trigger on every flick of the switch, no matter how fast you flick the switch.
2) it can trigger on every flick, but takes about 15 seconds for each "direction" of the switch to recycle before triggering again. This may come in handy if the computer on/off got out of sync with the cab on/off. (for in a case like Peales)

Actually, out of curiosity, because I've seen a couple other posts of yours, do you want the cab to turn on the pc, or the pc to turn on the cab?

This circuit will work great in a case where you don't want to mod the cabinet, and when it powers up the computer will, and when it turns off it will cycle the power button again, shutting down the computer. (I can also adjust it so a power on "presses" the power button for .5 seconds, and a power off "presses" the button for 4 or 5 seconds (hard power off).

But if your going to Mod the cabinet, to have the pc power also turn on the monitor, and marquee, wouldn't it be easier to just put in a momentary button, or if you wanted a toggle switch, use a STDP switch and wire the two ons to this circuit? You'd save 120vac wiring, and a 120vac relay.

Just curious what your frankensteining.....

sd

[krick]

Just curious what your frankensteining.....

I, like Peale, want to keep the cabinet as close to stock as possible so that I can easily change it back to a dedicated single JAMMA game in the future.  So I'd really like to have the cabinet control the PC via the stock toggle switch.

However, if that is too difficult to implement, I know that having the PC turn on the cabinet is far easier.  Essentially, it requires a momentary pushbutton switch and a relay with a 12V coil that can be triggered off an ATX power connector and is capable of switching 120V AC.

My latest brainstorm is using an ATX power supply that has an auxillary AC outlet on the back.  Vantec makes one that I *believe* is switched on and off with the PC.  I think they just integrated a relay into the PS.

So, yes, as you've seen in some of my other posts, I'm exploring all avenues and will probably try them all see which one I like best.

As for your circuit and how it triggers...  I don't think that triggering on every flick is a good idea for the reasons you mention.  Basically, what happens when the PC on/off gets out of sync with the cab on/off?  Like you say, there needs to be some sort of cycle time.

The idea about using the STDP toggle to trigger your circuit a good one too.  In that case, since it's not connected to the mains power, you could use a transistor instead of a relay to make it cheaper.


Maybe I'll start a web page to document all these different methods so other people don't have to keep re-inventing the wheel.

[SoundDoc]

I'm actually working on a website for it right now...

Its working great, sitting on the table in front of me. It isn't turning the computer on and off, but the relays closing and opening like its supposed to, I also added a supervisory led to show when its active, and a small pot to adjust the pulse length. Gimmie a min and I'll have the pic posted.
I just looked through my "junk" and I've got maybe 8 or 10 power supplies with the switched output plug. Their all bad though, and for parts.  

sounds like your looking at plugging the rest of the cab into the power supply to switch it. In that case it couldn't get out of sync.

You could also use a scr, or small 12V relay powered off a HD connector to switch the cab as well.

SD

[SoundDoc]

heres a quick pic.
lets see, top left - simulated poles of the 120vac switched relay, bottom left 5A reed relay for power switch, and lit supervisiory led, right - adjustable pulse, .01 sec to 1 second. forget the top right, just other parts.

Have to go put another coat of paint on the cab, I'll throw together a version after that has seperately adjustable on and off pulse lengths.

sd

[SoundDoc]

Still havent gotten the cab painted....

Here we go:
- individually adjustable on and off pulse lengths, 0-10 seconds.
- selectable recycle incase the computer gets out of sync.
- supervisory led to see if the things actually working.

I should have a webpage for it later tonight, or tomorrow morning with schematics and parts info. I'll post it when its up. If anyone really wants one pre-built, let me know and I'll see what it would cost to do.

sd

[SoundDoc]

Still haven't gotten that coat of paint on the cab.....
Maybe tomorrow night.....

*edit*
as for the pci card..... ummm... no, I'm not going to go there. It would be nice to build it that way, but this is just a quick circuit I threw together, and don't want to get into trying to extend it that way. I think its best to keep it as a completly isolated device to the system. You could always build it onto a board that takes up a rear slot location, but the signals can't easily be fed into the MB that way. (plus the support nightmare that would result would kill me... )
/*edit*

I've thrown the designs I did up on my cabs page:
http://daCab.DatPlace.com

the link straight to the circuits is:
http://dacab.datplace.com/Design.htm
(about half way down)

*edit*
play with it at your own risk, I take no resposibility if you kill yourself... )
/*edit*

I've tested them both now, both on pc's, and they work great.   no problems after around 20 cycles with it getting out of sync. (asus p4c800-e-dx MB). System hibernated on power off, resumed on power on each time.

If you use them on your cab, please throw up a credit on your website for me, and enjoy!

If your not that tech. savy, I can supply them on a limited basis.  Contact me if your interested: DaCab@datplaceREMOVEME.com

sd

[krick]

You ARE the man.  When I get a chance, I'll swing by Radio $hack and pick up some parts and give it a go.

as for the pci card..... ummm... no, I'm not going to go there. It would be nice to build it that way, but this is just a quick circuit I threw together, and don't want to get into trying to extend it that way. I think its best to keep it as a completly isolated device to the system. You could always build it onto a board that takes up a rear slot location, but the signals can't easily be fed into the MB that way. (plus the support nightmare that would result would kill me... )

I was never suggesting using the PCI signaling.  I was just saying that it could hang off a bracket and just plug into part of the PCI slot for vertical stability but with NO electrical connections.  I've seen some OEM USB brackets designed that way.

I've tested them both now, both on pc's, and they work great.   no problems after around 20 cycles with it getting out of sync. (asus p4c800-e-dx MB). System hibernated on power off, resumed on power on each time.

I was just thinking about the potential "out of sync problem".  Couldn't the device monitor one of the 5V leads off the power supply (not the 5V standby).  If there is 120V power but not 5V power, it could trigger the switch to turn the computer on, and if there is 5V but not 120V, trigger the switch to turn the computer off.  I'm just throwing stuff out there and seeing what sticks.  I have no idea if this is even possible.

[SoundDoc]

It could do that as well I suppose.... Just thinking here.
(It'd be a lot easier to do all this with a micro at this point  )

Trying to do it with discrete components, you

[krick]

I was looking around on the web and I saw an interesting way to tap into the ATX wires without actually modifying the power supply.

The trick is to use a short ATX extension cable.  You can solder additional wires into that cable where necessary and then insert it between the ATX connector on your PS and the ATX connector on your motherboard.

[SoundDoc]

Okey, I put a couple more ideas into it, and I've got something that should work as I think you want...  

Its not 100% polished, or tested, but they should both work and do the same thing. I just got my order from Bob Roberts (YAY!!) and I think I'm going to be busy for a while building my CP and finishing my cab, so I'm not going to have much time to tweak these out.

Anyway, basically it ties into 5Vcc, 5Vsb, and 120Vac, compares the cab status On/off with the computer status On/Off.
If the turns on, and the computers off, it sends a adjustable 0-2 second "ON" pulse, every 0-20 seconds (adjustable) until it gets 5vcc from the computer then it stops. Same the other way as well, if the Cab switches off, and the computers on, it sends a 0-10 second (adjustable) pulse every 0-60 seconds (adjustable until 5vcc drops, showing the computer as off.  That last delay may need to be adjusted longer, depends on how long your machine would normally take to shut down. Each of the pulse legths, and repeat delays are individually adjustable for your pleasure.

The jpg is almost unreadable, but the PDF is up on my site:
http://dacab.datplace.com/Design.htm about 2/3 the way down.

Theres 2 circuits that do the same thing, one uses TTL logic for the comparitor, and 555 timers for the pulse control. The other uses discrete transistors for the comparitor (easier to find and cheper) and uses only 2 556 timers for the pulse control. Mix and match, and enjoy!

If anyone really wants one, I could be pursuaded/bribed to build it up and test, I just don't have the timeto work on this anymore if theres no interest, I've got some funky circuits to build for my cab.....

I GOTTA GO BUILD A CAB!!! (ok still 3 hours of work, but then I'm going to... Three people have allready walked by and "fondled" my new trackball. I might try to sneak out of here early)

sd

[danny_galaga]

holy...! i think ill stick to my lonesome, works 98% of the time, relay!

[Lilwolf]

btw, the 3rd post was stating that my old emplorers had to get a way to do this.  And they had an expert build something... they couldn't use a relay for some reasons...

And that they had a solution for about $25..

After doing all this... Do you think that was a realistic price?  (seriously... I've considered trying to get them to sell a few here or at least mention it... just didn't know if it was worth it)

[SoundDoc]

$25 to build, or to sell for $25?
It would me MUCH easier to do this with a small Micro. a pic with a reed relay, and maybe 3 or 4 other components would do it, and be fully programmable. parts would probably run around $10-20  if you were making just one, mass production woud be much less expensive.

If you've just got the parts kicking around, I know theres a cost, but I've got  easily 5000+ IC's and all the other components around to experiment with. So for proto'ing something up for a one off "different" app, this is probably easier. (don't have to code anything etc...)

This started off fairly simple, with a single 556 and a couple of components, and if I ever needed to build this up, thats probably how I'd go.

Just got the first coat of color on my cab! YAY! Another coat tonight, and build up the CP and I should be Rockin' by wednesday!

sd

[Lilwolf]

I think it was 20 - 25 out the door.  Small board so a few bucks more for shipping.

The trouble is the ones my old company are selling had to be certified something (don't remember what) because they are used in manufacturing floors and needed to be OSA? certified.

And they are run on slightly smaller runs.  I think 100 at a time or so.  

But I really haven't dealt with them because my old boss started using them right before I left.  So I heard some of the troubles.  I know why they are being used.  But I never used them.

[independentthread]

I am by no means an electronics expert, but I have a question about the whole thread in general...
The final solution looks extremely comlicated for someone who only has soldering skills and no electronics knowledge.  Is the simple solution of a capacitor strapped to the back of a plug going to hurt the motherboard, or is the cap simply going to wear out after time?

That seems like a bigger question in my book.  Is it worth trying (and probably failing) to make something like the circuit board from the end, or is it more ideal to replace something that I can probaly pull from something old and unworking.

[danny_galaga]

it should be that the cap will last as long as any other cap in the motherboard. so it should go for years!

[SirPeale]

The only thing the cap does is act as a bridge.

By nature, caps start out shorted, and open up as current is applied.  

Since it's such a small cap, and the voltage is tiny (5V) the cap will likely outlast the cab.