Bringing-Up NorthStar Horizon systems

This document originally edited in 2007. (c) copyright Herbert R. Johnson 2016, updated july 29 2016.
Price and contact-me information can be found in this notice.
For more Northstar information and docs, check this Northstar Web page. To return to my S-100 home page click here.


For more information about NorthStar, and a list of manuals I have, check this Northstar Web page. There's links there to other S-100 pages. Thanks!

Here's a Northstar system I purchased and resurrected in 2006. I bought it at the Vintage Computer Fest - East at Infoage in Wall, New Jersey USA. Here's how I brought it up in 2007.

Here's the chassis view.s/n 10-045XX
Here's the CPU board, a Z80 of course, 4MHz operation
Here's the NS floppy controller.
Here's the view of the motherboard bus and power supply.
Here's a closer view of the back of the motherboard, which contains serial and parallel I/O.
Here's a non-NS memory card, probably by Central Data, 64K of DRAM.
Here's a close view of the I/O card, including a resistor (second from the top) that almost got fried.

The system sat unused for years before I got it, and the floppy drives were not correct or working. I replaced both drives. I cleaned all the board's edge connectors. I tested each card with an ohmmeter to ground, at either side of each board's voltage regulators - that tests for shorted capacitors. I found a few and replaced them.

Next I did some powering tests on the chassis, to make sure the big power supplies were functional. They were. Then I put cards in one at a time, and measured the DC voltages from the regulators. Those were all good, they sometimes fail or shorts appear.

Then I connected a terminal, confirmed with a simple RS-232 LED tester that transmit and recieve were not crossed, and tried to get some activity. It took me a long time to see why I did not. Tracing the wiring of the serial port, I saw that the -12V was wired through a resistor to a pin on the serial connector, which on the cable eventually was set to GROUND. That put a lot of current through that regulator and resistor - which is why there is a "fried resistor" photographed above. Removing the ground brought the negative 12 volts back "down" from ground. I was able to hit reset, a NorthStar DOS disk booted in the drive, and I got a DOS prompt!

I checked my NorthStar disks in my archives, and found some CP/M diskettes. I formatted a few disks, read a few, and found some useful utilities to add to a "working" CP/M disk and backup for this system. Then I mostly put it aside and only brought it up on occasion.

Then, in 2011, I decided to bring up another Northstar Horizon, and I checked this system out. Still working! I used it to confirm my bootable CP/M diskettes for the other Horizon.

Cybertron Horizon

On April 25 2011, I began testing on a "Cybertron" labled, Northstar Horizon system, s/n 10-017XX. I acquired it several years ago with a second S-100 system and some 8-inch drives. Chip and connector date codes on Northstar boards are 1977 to early 1978. The boards are photographed in chassis, and are MDS-AD3 floppy controller. ZPB-A2 Z80 card, Three Problem Solver Systems or PSS RAM16-B SRAM cards. All have 1977-78 chip date codes. One Seattle Computer Products 16K card. Two double-sided 40 track floppy drives, probably original Tandons. The ZPU runs at 4MHz. The previous owner wired a rotary switch to the baud select jumpers in back. Here's another view of the boards.

1) Ohm-ed out all voltage regulators on S-100 cards, inputs and outputs, to look for shorted tantalum caps. Typically 2K to 4K ohms on DVM. A few read as low as 500-700 ohms.

2) Brought up AC power slowly with variac. monitored unregulated voltages at power supply caps, and also on Northstar motherboard +5 at IC's. At 50V AC, read on +8/+18/-18 as 4.5/+9/-9 volts.

3) checked for hot spots on motherboard. Chip at B-1 (1488) was warm, others cold. Pulled chip in case it was shorted.

4) At 70VAC, 4.5/+10.5/-12. Floppy drives spinning constantly. Motherboard +5 is below 1 volt. First two RAM cards (Problem Solver Systems or PSS RAM16) were warm regulators. Others were cold.

5) At 80V AC motherboard +5 is at 3.5 volts. Floppy drives stop rotating. +5/+13/-14 on power caps. At 110V AC, power caps are +7/+18.5/-19.5. Drive tries to boot on reset. Motherboard +5 is now +5 volts. SO it looks like the system will operate at full AC power.

Had trouble establishing baud rate on rotary switch. But after several tries to boot a bootable diskette, found the correct baud rate for both terminal and switch. Was able to boot CP/M 2.2 and run MEMR memory test program. the program indicated good memory from above its run area 0D00 approximately up to the start of BIOS at D300.

Temperature readings of concern

Using IR temperature meter, the PSS #2 board (counting from front) had temps up to 152 degrees F on the topmost regulator. PSS #1 board read 100-113 degrees F. Both boards use TMS 4045-45NL (450 ns) 4K X 1 static RAM, 32 chips equals 16K RAM. There are three of these boards, plus a fourth 16K board by Seattle Computer Products (SCP).

April 27 2011: Took #2 PSS board and checked voltage regulators. LM340T5 and 7805. They were dated 1978 week 42, although the RAM on the board is dated in 1970's. But heatsinks had no thermal compound. Added dabs under each regulator.

Mounted board on extender card and started system. Took thermal temps at short intervals over several minutes, of each heatsink. Hot spot was at the regulator plastic body., several degrees warmer than the heatsink (until thermal equilbrium after maybe 15 minutes).

Temps for four heatsinks within ten degrees F of each other; the lowest heatsink was cooler, the upper three tracked within a few degrees of each other. Temps quickly got up above 100 degrees F in a few minutes, then climbed over several minutes to a stable 125-127 degree F range for the top three, and 115-116 for the lowest heatsink. Reading the DC voltage from each of the regulators, I got about 4.5 volts.

During the 20-minute run The temperature for the other two PSS boards at the top edge of their heatsinks stablized at about 133-134 degrees F. The 8 Volt supply at the caps read 6.75V. This is a little low for these sort of 5-volt regulators, they like more voltage drop and this may account for their 4.5 volt outputs. The apparent effect of adding thermal compound was to reduce temps about 20 degrees F. But I'm more concerned about power consumption and the low +8 volts.

The memory test "crashed" at one point but I was able to reboot and restart. But after running the system for about 20 minutes, I tried to reboot and could not. Letting the system cool for a few minutes, I was able to reboot. after a few trials. I concluded that the "crashes" were more likely do to momentary bad contacts from bumping the boards. But the inability to reboot suggests that heat may be a factor with the other boards such as the floppy controller. So I decided to replace the four 16K cards with one 64K card. That would reduce heat, and make sure there was plenty of +8 to regulate down to +5.

May 6 2011: Pulled a number of 64K static RAMs from my stock of boards. ECT, SMS "Last Memory", and a Seattle Computer Products card. I found a ECT manual on the Web, I had a SMS manual. But I tried to just toggle the switches first and to verify they would not blow up. WIth a little struggle with extended addressing, MWRT and so forth, I got them to all work. And they all worked OK with the Northstar Z80 CPU. They had 200ns or 300ns 2K X 8 static RAMs - 6116, 2016, 8275, etc. The other trick was to disable their memory at E800, where there is some Northstar controller and serial addressing. (THe northstar puts it's I/o in memory space, not in I/O space.) I ran the "MEMR" utility I put on the Northstar disks years ago, and exercised one ECT board for some hours. Ran OK!

Copyright © 2016 Herb Johnson

Herb Johnson
New Jersey, USA
to email @ me follow this link