MiniSport Laptop Hacker - Vol 18, Dec 1993 To discourage pecuniary interests, Copyright (c) 1993 Brian Mork >>> ADMIN Remember, you can get copies on disk of any software or text file I re- fer to (including the MLHacker series) by sending me a disk and SASE. I'll put a variety of other MLHacker related goodies on there, too. MLHacker is available on the KA6ETB Internet HAM-Server. Send a message to ham-server@grafex.cupertino.ca.us with the single line of text: HELP to get access information. Check the \hamradio\newsletters directory. If you're comfortable with late-night long distance phone calls, you may download MLHacker related stuff directly from me at 509-244-9260. Use the FILES command, and L)og into the \public\computer directory. Stand- ard XMODEM and YMODEM protocols are available. ARO related CD-ROM, In- ternet e-mail, and Usenet newsgroups are also available from this free service. >>> POWER SUPPLY Nothing, I repeat nothing, beats a schematic diagram when fixing elec- trical equipment. No fancy test equipment can compare. Thank you big time, Brian, for 1) recognizing my reverse engineered attempt (see Vol 17), and 2) sending me the real schematic! Two hours later... I have a dead n-channel enhancement-mode MOSFET ("K612" in case anybody can offer a cross reference). There's a 7V p-p, 500 KHz square wave arriving at the gate. The intent is that the drain pulls current through a switch- ing power supply transformer primary. Wasn't happ'ning. Drain stays at the +V supply voltage. Comparing with another K612 that handles the electroluminescent screen voltages confirms the first one is not work- ing. Several parts houses couldn't identify the K612 part, so I made a trip to Radio Shack and picked up a 276-2072 IFR510 n-Channel MOSFET. It's about twice as big as the K612, and its specifications are an over-kill, but it's all I had available. It cost just over $2.00. Instead of mounting it as a surface mount on the bottom side of the circuit board, I layed it sideways between some capacitors on the top of the circuit board and ran hookup wire to the bottom side of the board for connec- tions. Everything works fine! A representation of the power supply pinout given in Vol 5 is in order. This time, I know functionality, too. With the bottom of the computer removed, the motherboard still in place, pin 1 is toward the right and front of the computer, numbered like this: 15 13 11 9 7 5 3 1 16 14 12 10 8 6 4 2 1 5 V backup supply #1, 10-300 mA 2 BACKUP-ON request from computer, TTL levels 3 Ground (backup power supplies) 4 5 V backup supply #2, 10-45 mA 5 VEE-ON request from computer, TTL levels 6 -22 V (Vee), 10mA 7 +10 V, 10 mA 8 -10 V, 10 mA 9 EL-ON (LCD screen) request from computer, TTL 10 Li Battery +V input to power supply (~6.3 V) 11 Ground (Li battery circuits) 12 POWER-ON request from computer, TTL levels 13 Ground (hi current return from pins 15/16) 14 Duplicates Pin 13 15 5 V (Vcc), 0.25-0.6 A) 16 Duplicates Pin 15 A note of caution is in order. The computer will automatically shut down if certain voltages don't come up to specs within a few seconds (this appears as a 2 second blink of the power and numlock LEDs and the screen). Requests from the computer can be deceiving unless you know what you're looking at. For instance, I repeatedly measured Pin 9 at zero volts during my debugging. I thought the computer was erroneously not asking for the screen voltages. In fact, right when I pushed the power button, it came high for a few seconds, and then shut back off. It !was! working ok. Pin 12, on the other hand, went high to 5+ volts and then dropped back to 4.7 volts as the main 5 V supplies were shut down. This pin kept the request (4.7 volts is TTL high) active, and only dropped back to zero when I pressed the power switch once again, telling the computer to shut "off". Realize if there are good NiCads, good Li batteries, or the external charger is plugged in, this computer never really shuts totally off. There are three main sections to the power supply. The first handles all the 5 Volt supplies. Both batteries are brought in to this section and the normal computer supply and the backup voltages are provided from this circuitry. If pins 1, 4, 15, or 16 are bad at the power supply header, this section is malfunctioning. The second section is where my trouble occurred. An integrated circuit regulator provides oscillation signals for the switching converter and appropriate voltage feedback control. A three-tapped transformer fol- lowed by diode rectification provides +10v, -10v and -22v. If pins 6, 7, or 8 are not at correct voltages, suspect problems in this area. The last section is associated with the electroluminescent screen volt- ages, and has its own 6-pin header. To see it, open up your Minisport using the directions in MLH Vol 5 and find the 6-pin jumper going from the power supply board into the display pivot joint. Only 4 wires are used. One handles ground. Two carry current to/from the brightness control, and one is the high voltage drive for the display. Specifical- ly: 6 50-150 VAC p-p, nominally 138 VAC 5 n/c 4 GND 3 n/c 2 return from 20 Kohm brightness control 1 diode rectified 10 VDC going to brightness control If pins 1 or 6 !of the screen connector! are at incorrect voltages, sus- pect problems here. Incidentally, Pin 9 of the motherboard power supply connector (see MLH Vol 5 and Vol 17) is a TTL level request from the computer to turn the electroluminescent display on. It goes into Pin 14 of the MB3778. The MB3778 oscillator output is buffered through three parallel CMOS gates and toggles the gate of a K612 MOSFET, which in turn pulls current through a transformer. Subsequently, a high voltage step up transformer (looks like a telephone line isolation transformer) gen- erates the high VAC for the screen. Packing tape applied on the front and sides of the keyboard worked well to bundle the keyboard to the motherboard as I worked on the assembly. Without this, the keyboard connector ribbons get severely strained as the board is flipped this way then that way. I am growing to dislike auto-routed circuit boards. The computer logic generates through-the-board vias that terribly confuse a person trying to trace the circuitry. In a lot of cases, casual inspection shows sin- gle side alternative routes were available. >>> BAD ROM DISK I received a message over Internet from a Minisport user that indicated his ROM disk (C:) has gone dead. He's getting checksum errors on boot- up. Do you have the equipment and willingness to pull your ROM, make a copy and send it on to Ron? I ask this not just for one user, but ever- yone else this will eventually happen to. The source for original ROMs has dried up, and if we're going to keep Minisports working, it would be good to know who could do this for others. >>> MEMORY SWAPS The first MB of memory on a ZL-2 is eight discrete chips soldered on the motherboard. The second MB is a SIPP package. I know from experience the 2nd MB can be unsoldered and swapped into a ZL-1. Thomas asks if the lower and upper MB are interchangeable. Unfortunately, no. Howev- er, I would be interested in soldering a standard PC SIPP into my ZL-1 and see if the graft takes. Do you have an unused 1 MB SIPP you'd be willing to let me use? Please provide feedback: * BBS 1-509-244-9260 * AX.25 KA9SNF@wb7nnf.#spokn.wa.usa * Internet bmork@opus-ovh.spk.wa.us 73, Brian * 6006-B Eaker, Fairchild, WA 99011