Musicbox and UPS of Superpower

One week of work/summer left, then back to school.  I’m excited; I miss school.  OK that isn’t exactly accurate... what I miss is State College.  I miss J4 and beeffalo.  I miss hanging out with people.

But I’m going to miss Unisys during the school year.  I’ll miss working with Mark.  That’s one thing that made this summer way better than all the previous ones working there -- I had a kid my age working on the same stuff as me, in the next office as a matter of fact... and he’s into the same music I am, and we’re constantly messing around.  I’ve never IMed someone so much who was so close to me.

[img src="00329--dsc06315small--20030820-2356.jpg" width=271 height=400 align=right clear=left] I’ll definitely miss being able to get gigantic 35 amp diodes like this one, from the power guys.

My project this week was to finally get my musicbox to the point where I could turn the car off and re-ignite the engine without it rebooting.  It turned out there were 2 problems in one here.  The first is that during ignition, the car battery’s 12V becomes more like 9 or 10 volts, and the computer can’t run on that, so it reboots.  The second problem, which I only learned about after fixing the first one, is that the car’s accessory power line actually switches off for a second while the engine is turning over.  Since I use the accessory line to control a relay that turns the computer on, the computer turns off when the accessory line isn’t at 12V.

To fix the dropping-voltage problem, I took the small sealed lead-acid battery out of my UPS to put in the car.  (I also bought a car battery to replace it with; my UPS is now running on that.  It’s supposed to keep a single computer with 15" monitor running for 18 minutes when the power fails; with the car battery, it kept my 2 computers, 19" monitor, and stereo running for about 40 minutes the other night, before I got tired of waiting around and went to bed.  But it would almost certainly keep them running for a day or so.  If this had become its own post, it would have been titled "What the UPS companies don’t want you to know.")  I planned to hook that battery up in parallel with the car battery, so that when the engine is turning over, there’s extra voltage to keep the line up for the computer.  However, the cranking engine would steal that voltage too, unless I put a diode between the batteries.  So current can flow from the car battery to the smaller battery, but not vice versa.

To fix the second problem, I needed a way to keep the accessory line on even when it’s off.  Or at least, a way to make the computer *think* the accessory line is on.  So I built a simple RC circuit.  I googled and came up with this (the circuit on the right).  My version is slightly modified: I used a generic NPN transistor, not specifically a 2N3053, and I didn’t put a diode around the relay coil.  I used a 47F capacitor and a 100-kohm resistor, which gives me a time constant of about 5 seconds.  So the computer will run for up to 5 seconds without the accessory line actually being up... because it takes the capacitor 5 seconds to discharge through the resistor, and that current hits the gate of the transistor, which switches the transistor "on" thus completing the circuit from +12V to ground through the coil of the relay.

Exciting, no?

Posted by Anthony on 1 reply


01. Aug 22, 2003 at 12:16am by Anthony:

Well, it turns out I didn’t need that circuit after all.  The accessory line *doesn’t* actually go low for a moment during ignition.  The reason I had thought it did was because my head unit turns off for a moment.  But I realized that that could be because of the voltage sag, too, and it turns out that that is the case.  So, I’m now using the accessory line to power the computer on, instead of using the "remote turn-on" aka "amp turn-on" line from the head unit.

Reply to this message here:

Your name
Website (optional)

HomeCreate PostArchivesLoginCMS by Encodable ]