I’m feeling amazingly frustrated with the project I’m working on.

I think I know why the data coming in isn’t working, but the data going out has suddenly stopped working coming out corrupt (test is received as QsT.), and I really don’t know why.

I’m considering taking a bit of a break from it, or maybe stepping way back and taking a wholly different look at it, what it does, and how I could do it better, end to end.

The problems all seem to stem from what a hack the board itself is that I’m trying to clone. Maybe I shouldn’t try to keep approximately one half of it original.

RS232 to Eaton Leonard Test Harness
(Click for full resolution image…)

I spent a lot of time in front of this test harness tonight. Most of it was spent thinking and reading and talking with about it. And yes, because I forgot to grab some “Phoenix Connectors” from my friend’s shop, I had to hack together those power connectors seen there. And I had to be VERY careful with them. However, they work. :)

At this point I’m pretty sure I know what needs to be done to make it work. I can send data (and set RTS) out through it, so half of the circuit is working. I’m fairly certain the other half will be done soon.

See, the old board has a MC1488/1489 pair used for RS232/TTL level conversion. It turns out that there was another two hidden traces beneath the MC1488. These traces connected the two inputs which were being used with their counterpart on the gate, causing the MC1488 to be used as both a level converter and an inverter. As the MAX232 which I replaced it with can’t invert things, I need to come up with an inverter.

There is an inverter on the CTS line, but not the RX one; the two lines which feed data to the input on the MAX232. Since there’s only one inverter, I’m pretty certain I’ll be able to do this by swapping the circuits which handle CTS and RX, and I’ll test this tomorrow.

Thankfully I was able to borrow the receiving board from inside of one of the machines today, and now I can test the product end to end, as opposed to just making it, taking it all the way to the shop and trying it, watching it fail, thinking about it, making a change, lathering, rinsing, and repeating.

Now I have no reason not to get it working on the bench, and I’m very confidant I will do so, in very short order.

I don’t think it’ll be tomorrow, as I’m heading straight to my parents house after work in order to set up my Mom’s new iMac, get the old one back from her, and other things like that.

So, the board I mentioned yesterday? It’s not working.

I have both the part I’m cloning and the part it needs to talk to now, though, so I can test things end to end.

I know what I’m working on tonight now.

So, that RS232 to Eaton Leonard (photo gallery retired) level shifter thing I’ve been working on? Well, it wasn’t working. I took a bit of a break from it to get some other stuff done, and tonight I started back in on it. I’m fairly certain I found the problem, and hopefully that’ll be tested tomorrow.

If so, I might have my first for-pay electronics job. Yay!

(This would have been done loads sooner if I had the test environment at home, or a matching board for testing, or something. But, I didn’t, so it was a little more complicated. I don’t mean to blame anyone for this, it’s just the reality of working with one small piece of a puzzle and needing to take it to the customer’s site to test it.)