AUSTIN ATARI COMPUTER ENTHUSIASTS
                           Objective:ATR
                            Larry Short
           (originally published in AACE Neuron, Oct 1985)

I thought Bernie Bennett's article on the ATR 8000 was the most objective 
review of the machine I have seen. Maybe some people will he interested 
in my experience with it. I had two double density drives, a Percom and 
a Rana, and an 850, so the advantages the ATR offered were CP/M, a 
printer buffer, and standard drives. I am a believer in standard drives 
because, unlike Atari drives, just about anyone will work on them. That's 
why I had the Percom, which is also a standard drive, and when it broke 
I had it repaired by Computer Doctor, who will not work on Atari drives 
because of bad experiences with the company. Since I also have a Tandy 
1000 with standard drives I decided I could use the drives for both 
machines and save money and space. I needed CP/M to maintain a payroll 
program I had written for a Kaypro. Since the Diskdef program supplied 
with ATR CP/M allows the ATR to read and write Kaypro format disks, and 
numerous other formats, as well, the ATR seemed like a reasonable 
purchase. So I made a deal with Ron Hawkins at Computer Resales; he got 
the Rana, Percom, and B50, and 1 got a 64K ATR 8000.

Ron and I did the upgrade from 16K to 64K. It was simple: swap eight 
chips and change six or eight jumpers. The instructions are very clear. 
I removed the drives (Teac's) from the Tandy, and mounted them in a 
dual half-height case with power supply,

There's nothing like building your own cables to make you feel like a 
hacker. I built two drive cables, one for the Tandy and one for the 
ATR. Neither cable worked the first time. Since these are 34 conductor 
cables, that makes 68 tiny connections on each cable. I used and reused 
the press fit connectors that puncture each conductor when you press 
them down, and some of the connections didn't make. I finally tested 
each connection for continuity and pressed each suspect connection with 
needle nose pliers and the drive cables were in order.

I also built the cable that adapted the ATR RS232 part to my Anchor Mark 
II modem. This involves a joystick type plug an one end and 8 or so 
connections to a card-edge connector on the ATR end. I had to match the 
ATR RS232 pinout with the 850 pinout, for which the Anchor plug was 
designed. This didn't work the first time either, because I made those 
8 connections exactly backwards. Luckily, the printer cable from the 
Tandy plugged right into the printer part on the ATR. Once the cables 
were working the disk drives booted right up and the modem worked the 
first time I tried it with a variety of terminal programs, including 
Chameleon. The Print Shop even worked with the printer on the first try. 
I will admit that it is very pleasant to catalog your drives and see 
1400+ Free (double density) Sectors. I tried the ATR with some heavily 
protected programs, including some from Electronic Arts, Blue Max, etc. 
and all of them loaded and ran.

Next I bought a half-price used copy of CP/M. I took about three weeks 
to get this running. The basic problem was that CP/M would just hang, 
frequently. After about a week I called SWP for help. In my job I deal 
with technical support people often, and the folks at SWP were the most 
courteous and professional I have ever talked to. Usually before I call 
a technical support department, I observe the symptoms so that I can 
relate them as clearly as possible.	I also try to have a list of 
questions. Well, before I had even finished reciting the symptoms, the 
guy at SWP was asking me my own questions. I was impressed. We decided 
that some of the the system files, in particular the Memtest, were 
missing from my copy of CP/M. He sent me a new system disk, and when I 
ran the Memtest, sure enough there were two bad chips. I replaced the 
chips with good ones, but the CP/M still didn't work.

I isolated the problem further and it was a weird one: the system would 
boot and as long as you pressed a key more often than every 30 seconds 
everything was fine, but if you went longer than 30 seconds without 
pressing a key, the system would hang, every time. Finally, I discovered 
that the 64K chips I used were 120 nanosecond chips; when I replaced 
them with 200 nanosecond chips (as specified in the ATR manual, pg.68, 
editor), everything worked fine.

I was surprised when I examined the CP/M system disk, even the upgrade, 
to find that MBASIC was not on it. I was used to Kaypro and other CP/M 
machines that came bundled with MBASIC, CBASIC, SBASIC, and other stuff, 
sometimes including Wordstar, Perfect Writer, and so on. Honestly, I 
can't imagine paying $250 for MBASIC, and I don't think anyone ever has, 
at least not for the last 5 years.

It might sound like the ATR has been a lot of trouble, and it has.  It 
has been worth it though; my old system was working so well that it got 
boring. The ATR makes me feel like a hobbyist rather than a consumer.