The LNW-80 Model II, by LNW Research, was an enhanced clone of the TRS-80 Model I. It was said to be software and hardware compatible with the TRS-80 Model I, but had many additional features. The Model II was an improvement over the original LNW-80, and put it on a level with the TRS-80 Model 4.

The LNW-80 II came standard with the LNW Expansion Board, LNW 5/8 Doubler, and a bank switching board to allow 62K CP/M to run. The CPU clock was switchable between 1.77 and 4MHz. Program RAM was 96KB, divided into banks, with 64Kx6-bit RAM for hi-res video RAM. 80x16 and 80x24 text modes are available. Hi-res mono (480x192) and color video modes, with 8-colors.

This particular unit was very dirty, and required a fair amount of work to get it running. To start, the +12V rail was down, which I found was due to a shorted tantalum capacitor. After I replaced it, another tantalum started to smoke, and then a third. At that point, I ordered a batch of new capacitors and replaced 25 6.8uf 15V tantalums with 35V versions.

With +12V stable, the machine came up with the usual random character display you'd expect to find on a machine with no disk inserted. However, doing a reset, even with the BREAK key, had no effect. I eventually found that the Z80A was dead, and replaced it. With the Z80A now working, ROM BASIC still wouldn't load. I removed and tested the program RAM on the main board, and found and replaced two bad 4164 chips. Reset-BREAK now got into ROM BASIC and I hooked up a dual-Gotek drive with some bootable disk images.

From the Gotek, I was able to boot LDOS 5.3.1 and verified the machine was mostly working. But, it would not boot any double-density disks. I figured the Western Digital 1791 FDC on the Doubler had a problem. I was able to find one at Unicorn Electronics, and could now boot double-density disks.

While I continued to work on the system, the display suddenly reverted to random characters again. After more testing, I realized that it was actually booting and responding to commands, but nothing was written to video memory. With the help of the logic analyzer, I narrowed it down to the area where the video RAM (a pair of 2114s) were enabled for write. Unfortunately, everything besides RAM was soldered to the board, and I ending up replacing 4 chips before finding a defective gate on a 74LS32.

Finally, I found that although double-density disks would boot, any attempt to write to them resulted in a corrupt disk. There's an alignment procedure for the 5/8 Doubler in the technical manual, and I found that the write pre-compensation setting was way off. I was able to adjust the pulse width to the 200ns setting in the manual. Writing to double-density disks now works reliably.