MiniSport Laptop Hacker - Vol #21. April 1994 To discourage pecuniary interests, Copyright (c) 1994 Brian Mork >>> ADMIN The MLHacker series is formatted to print out on 8.5"x11" sheets manual- ly fed into an Epson LQ-850. I print at 8 lpi, which gives 82 lines per page. My goal is to make each MLHacker one to two pages, to fit front and back on one page. Back issues are available via Internet e-mail and ftp archives and direct from my data line listed below. File download privilege is given on the first call. Use the F)ile option and L)og into the \public\computer directory. No cost 'cause I just like doing it. Do something you're interested in for free and pass it on! The Air Force has me committed to some massive amounts of overseas flights during April and May. For this reason, I expect MLHacker will take a hiatus during that time. >>> AUTOMOBILE POWER SUPPLY Something has been bothering me ever since issue #12 of MLHacker. I mentioned how you could tap into your car cigarette lighter, step the voltage down to 9 volts and use that to power your Minisport. So far, true. My bad recommendation was the placement of a noise-killing capac- itor and inductor. Yes, put a big inductor in series (old lamp ballast works fine) to kill ignition glitches. But don't put the capacitor on the car side of the inductor. Put it on the *computer* side of the inductor. The problem is this: When you switch off the computer, the magnetic field sustained by the inductor collapses, forcing continued current flow. But the com- puter is off! Being off is a high resistance. As the charge collects on this high resistance, the voltage peaks very high. To some extent, the internal Minisport battery, which is effectively in parallel across the power line, begins to conduct (i.e. charge up) as the voltage rises, so there is some limit on the voltage. But I was wrong to suggest put- ting the capacitor where the inductor could feed off from it. It's much better to put it on the computer side, where it will dampen the inbound current. An additional improvement would be to put a diode across the inductor (cathode, banded end toward the car). It will then short out the inductor spike upon power-down. Some folks have indicated they run the Minisport straight from 12 volts. I'm not sure I'd recommend that, but I'd be interested in anybody's suc- cess doing this. Battery? Direct from car? One MLH reader is working to modify a drill battery pack to power his Minisport. I hope to talk him into writing up notes on the project. >>> INSIDE THE BATTERY PACK Jim sent me a dead battery pack to rip apart. Aha! The final link to the puzzle. MLH #12 discusses the wall adapter/battery pack connection. MLH #14 discusses the battery pack/computer connection. Issues #17 and #18 discuss the switching power supply board inside the Minisport. Fi- nally, here's the stuff from inside the battery pack. In the process of documenting this, I discovered another error from issue #12--my descrip- tion of the small pin on the 3-pin charging connector. I believe the information provided here to be the correct interpretation. BATTERY PACK INTERNALS Round Power Supply Connector Tab Connectors ------------------------------ ---------------- 9V, 2A BOTM --------------------------------------------- CTR (red) 6V, 1A TOP -------------------------------------------+-- AFT (yellow) | 8.3V O.C. LITL ----TC1---+ +---:|:|:|--TC2--:|:|---+ | | GND SHIELD --------------+----+---------------------------FORW (black) There is also a tiny chip capacitor between the 9V supply rail and ground. A parallel surface mount spot (for a resistor?) is empty on the circuit board inside the one battery pack I've inspected. The charger adapter has four connections: the shield, two similar pins, and a third pin smaller than the other two, thus the titles BOTM, TOP, and LITL. The tab connections on the bottom of the battery pack corre- spond to the three slide connectors visible inside the minisport when you remove the battery pack. The colors correspond to the wiring color visible inside the computer when the bottom is removed. TC1 is a thermal cutoff unit, tightly mounted to the battery cell clos- est to ground. TC2 appears physically similar. While charging this battery pack, I monitored the voltages on the power supply connector. BOTM was 9.2V, TOP was 7.6V, and LITL was zero volts. I applied a heat gun onto TC1, and at 0:12 the voltage at LITL jumped up to 8.3 volts. At 0:40, TC1 closed and the voltage at LITL went back to 0 volts. It seems clear that the design of this system is to overvoltage the bat- teries (7.2v onto a 5x1.2 => 6.0v battery pack) and when the batteries get warm, the thermal cutoff ungrounds a signal back to the charging unit. To confirm that the charger responds to the signal, I loaded the battery pack with a 20 ohm resistor and hooked up the charger, giving 7.2 volts across the resistor (about 350 mA drain, 2.5 watts). I would expect that if the charger shut off, the voltage would drop a little bit as the battery pack picked up the entire load and started discharging. Sure enough, after about 20 seconds of heating TC1, the voltage across the resistor dropped to 6.8 volts. Heating TC2 has expected results. While charging, I measured 7.0 volts across an unloaded, charging battery. At 0:20 of heating, an audible click occurred and the voltage dropped to 2.9 volts. I turned off the heat gun and at 1:14, I heard another click and the voltage comes back up. While *not* charging, the voltage goes from 6.5 to 0.0 with nearly identical timing. I think the 2.9 volts observed while charging was residual output from the charging unit. >>> BAYCOM VS POOR MAN'S PACKET Poor Man's Packet software works fine to receive packets on the Minis- port using the hardware described back in MLHacker #9. It does not work on transmit. If you try to send a packet, the transmit tones just lock up at 1200 Hz or 2200 Hz. This was confirmed using another ham's (thanks Ron) PMP hardware rigged to run off COM1 (normal DB-9 out the back). PMP is written in C, including the AX.25 bit packing routines. I think these are just not fast enough when run on the 8 MHz Minisport. Baycom uses a separately loaded driver to talk to the hardware. It's small and efficient, probably written in assembly. I got a copy of Bay- com software and using Ron's Baycom hardware, which was rigged to oper- ate on a parallel port, everything worked fine. Conclusion? The Minisport needs bit packing routines written in assem- bly. Baycom offers this; PMP does not. I thought the PMP hardware was designed to be bit compatible with Baycom (e.g. RTS does the PTT func- tion, and so on). Apparently, this is not so. When I redirect the Bay- com to COM2, it does not work with the PMP hardware I built and de- scribed in MLHacker #9. I don't have the time right now to do an assembly rewrite of the PMP software, so I want to figure out how to use the Baycom software. Any- body out there more familiar with Baycom and what bit lines it uses for what? Eventually I'll poke around and find out what's going on, but maybe somebody could save me the trouble. >>> FUTURE PROJECTS I saved up money for several months and purchased a Fluke 97 digital storage scope / DMM "do everything" box. It's roll mode data monitor worked great for probing the battery pack. Expensive, but a *superior* bench top accessory. I'm interested in anybody using one of these. Mine has an optical data interface and that will, I'm sure, eventually be hooked to my Minisport. I also picked up a 670 nm laser diode assembly, including a photodetec- tor and stepper motor. The light beam comes up through the shaft of the stepper motor, allowing the outbound beam to be aimed onto one of eight mirrored segments, and reflected forward. Inbound light is collected by all eight segments simultaneously and focused onto the on-axis photode- tector. I wonder if I can hook this up to the computer to do a laser light show. Or maybe a proof of concept optical radar. Hmm... :-) Please provide feedback! * Direct data 1-509-244-9260 * ARO Net KA9SNF@wb7nnf.#ewa.wa.usa * Internet bmork@opus-ovh.spk.wa.us 73, Brian * 6006-B Eaker, Fairchild, WA 99011