Looooongshot...
Godfrey has done well to suggest Siemens and Allen-Bradley. I agree, but for different reasons.
First...
I think new programmers need to learn at least two PLC's (three would probably be better).
AND, those PLC's should be as far apart from each other in their programming styles as possible.
The idea is, to learn that...
If you can program those two or three PLC's, then you should be able to program any PLC.
I am not suggesting that being an A-B (or Siemans) expert means you can be an expert in (or even program) other PLC's.
The lesson you have to learn is that, fundamentally (no matter how much you might LOVE(TI)/HATE(AB) ( <--bias-based sub-liminal message!) one brand over another), they are all the same in one unifying respect - LOGIC.
To get to the point where you can program any PLC, you first have to get to the point where you can understand and develop LOGIC.
Once you know LOGIC, it then becomes a relatively simple matter of learning the secret codes, winks and hand-shakes for any given PLC.
Once you think in terms of logical concepts, you simply look for the secret code that delivers on the concept - in any PLC!
Now, a history lesson (or, maybe it should be called a Brief Auto-Biography)...
My very first PLC experiences were with Allen-Bradley.
WOW! PLC's ARE COOL!!!
I worked with Allen-Bradley for about 3-years. Then I went to college for Electrical Engineering. When I got out of school, I went to work for a prominent (at the time) specialized computer manufacturer.
All of my programming experience was based on direct interaction (or at least, relatively more so than with PLC's) with the CPU (Assembly, FORTRAN, C... and a few of the other weird ones like COBOL).
The point is, I was dealing with the processor. There were no high-level interface layers - such as those that are typically presented by all PLC's. (Yes, there are now some PLC's that let you into that level, but that is so now, it wasn't so then.)
In dealing with the processor, I learned to look at things from the processors' point of view. From the processor-point-of-view, things are a lot different! (more later)
So, economic bust came in, Hi-Tech went out...
I went back to work on my Electricians License in a lo-tech manufacturing plant. They didn't have a slot for a Process Engineer, but they did have a job for an Electrician/PLC Programmer. So, I took it.
They had some Allen-Bradley PLC's that I thought I would slip right into. I did... sorta... My previous experience was hindering me a bit. I didn't know why.
Then they brought in a few machines that had the GE Series-One. I was expected to pick up on them. I read the books a couple of times, played games with the spare CPU we ordered and, frankly, I was astounded!
The difference between the Allen-Bradley PLC and the GE PLC was like Day-vs.-Night! The GE was simply more intuitive! I had a ball with the GE and became less enthralled with the Allen-Bradley.
Time goes on, jobs change, and I find myself doing Process Controls at a place that has a Texas Instrument PLC (TI-505) running the big show. (It's fairly well understood by Human Resource-types that good PLC programmers can program any PLC - so I get, and take, the job).
It takes me 6-months to figure out the system (it's 4 stories tall and 2 city-blocks long). Meanwhile, I'm reading the manuals - 2 or 3 times.
I couldn't believe it... this TI stuff is even more intuitive than the GE stuff! Not just more intuitive, but even closer to the CPU-talk that I experienced a few years back.
Like Day-vs.-Night-vs.???
And therefore, my bias against AB.
Others might have gone through the same experience... maybe not.
The point is, learn PLC's that are as far removed, in style, from each other as possible.
So, I suggest...
Allen-Bradley and Siemens Simatic 505 (TI-505) or the Siemens S7-2xx series (a TI-Spawn) - not the S7-3xx, -4xx.
And possibly, the Honeywell System - it's not quite like either the AB or the Siemens.
For all of the potential Flame-Throwers out there... It's just my opinion... find a way to get over it.