Terminal boards for WebControl

[rossw]

I'd been debating using stubs into the outputs but was worried about stability. I'm not using the analog currently and probably not in the foreseeable future, although that could change. I'm more interested in  the temps and humidity right now. I was mainly interested in powering the board from 1 source instead of having multiple. I was wondering about whether the 5v from your board would power up the W/C via the 24v.
My thought was to put terminals above the outputs and use loops up into the inputs for the 2987. I stumbled onto the 2987 and liked it because the inputs on my plc are sinking.
Now I'm going to have to look at something compact for a switching ps to 12v.
I already have a separate board right now with the ULN2003's but it requires pull ups and got unsightly rather fast. It also requires inverting the inputs on the plc. Not a big deal to NOT the inputs in the program. I usually put all the inputs into internal bits, so I could do the inversion then.

 
Part of the reason for the regulator on my V2 terminal boards was specifically to address the issue of multiple power supplies, and yes, it can run the WC board directly without needing to even connect anything to the two-way power screw connector beside the output terminals.
 
The regulator is happy with 24V input, but to bear in mind the power dissipation. With the modest heatsink I provided, and running 24V into the board I was measuring 130mA with no temperature sensors connected. After about 10 minutes with no added cooling airflow, the heatsink had stabilised at 91 deg C. Within safe operating parameters for the regulator, but far hotter than I would be comfortable leaving it run. The regulator was blowing 2.5 watts out as heat. It would need a larger heatsink, and/or forced air cooling.
 
Rather than stubs, use 90-degree male header pins. Spaced right, you'd run a strip of 10 of them straight into the output connector (and optionally a couple more for power) and do up all the screws. It'd be quite stable and solid.
 
There is a drop-in replacement for the 7805 which is a switching version. Ebay item 121026036348 claims 5.5v to 36V input and up to 95% efficiency.

 

[todster]

The regulator is on order. That should be perfect. Now I just have to wait 2 weeks. In the mean time I can start my programming with the board I have already. Now if I could just have the outputs as 1 word........and a free running clock or timer accessible at the bit level.

 

[rossw]

and a free running clock or timer accessible at the bit level.

 
what timer granularity do you need? Since the program execution is asynchronous, you already have a "free-running timer" with a 1 second granularity.

 

[todster]

I'm not real keen on the waits. I'd like to have something a little more stable to use for delays, pulses.

 

[rossw]

I'm not real keen on the waits. I'd like to have something a little more stable to use for delays, pulses.

 
It's rude. It's crude. It's unattractive.... but you could always make a simple free-running oscillator at whatever frequency you like - lets say 1KHz for a 1ms nominal interval.
Feed that into TTL Input 1 set as a counter.
That would give you a 32-bit counter running at 1KHz.
4294967 seconds (1193 hours, nearly 50 days) before it overflows.

 

[todster]

That's exactly what I was thinking. I had ordered some parts from tayda electronics and they sent me 8.0 mhz crystals instead of triacs. I have these now and some 555's as well. The crystals are too fast though. While we're here, how would you extract the sign of an integer? ANDB (2^31) isn't the answer?
Whats interesting is that when you ANDB -1024 with 2^31and put the results into another var it shows as a double negative ie.. --214...........

 

[CAI_Support]

What do you want to do with that free-running timer at bit level?  Maybe we can think up a solution for you.

ANDB -1024 and 2^31 are getting -1 correctly. However, when we convert the negative number to string for display, we displayed an extra negative sign. On the earlier compiler we used, it did not convert the sign bit correct, so we had to add that in our code when display the string.  Now, the compiler convert that correctly, we should take out the extra '-'.   Internally, the VAR is stored as 32 bit signed number, that is always correct.

 

[todster]

If I need a flip flop or flasher I don't have to use extra lines to create it. The flip flop would be something I might use as a timing signal. It's not a big deal. Just a convenience I'm used to having on a plc.

 

[CAI_Support]

I see, I guess this is some kind of communication protocol over IP and OP that you and Ross working on.....

 

[rossw]

I guess this is some kind of communication protocol over IP and OP that you and Ross working on.....

 
No, nothing like that at all... my only contribution to this has been the suggestion of a freerunning osc to the externally clocked counter.

 

[todster]

It was just a crazy idea. I need to learn to live within my limitations.

 

[bbrendon]

I'm looking to measure the amount of light outside with this. How do you go about purchasing one? I have a photocell, I'm not sure how to wire it up yet. I'm sure its possible though.
Thanks.

 

[rossw]

I'm looking to measure the amount of light outside with this. How do you go about purchasing one? I have a photocell, I'm not sure how to wire it up yet. I'm sure its possible though.
Thanks.

 
If you're referring to the terminal board and amplifier module, yes it would let you measure light levels.
Depending on your requirement for precision, accuracy, stability and temperature sensitivity, there are a number of ways to do it.
 
A small PV cell would be easy, with a gain of around 20 would give you 0-10V out.
If you already have a Cds cell (photoresistor), you can connect it in a voltage-divider circuit and use the amplifier to give you a greater voltage swing, but it's difficult to get them linear over a wide light range, and Cds cells have poor temperature stability.
A PIN diode or phototransistor would work too. Perhaps you can give us more information about your needs and intended application?

 

[qwiksilver96]

You're Jim? First to use the online purchase - didn't know who you were, had to hope it all worked ok!
Thanks for the update. There's a few others should have got theirs by now that I haven't heard from.

 
Ross, mine were shipped on 2/18 according to an email I received from you on 2/17 and I have not received them just yet. FYI.
 
Thanks,
Frank Mueller (In USA - NC)

 

[rossw]

Ross, mine were shipped on 2/18 according to an email I received from you on 2/17 and I have not received them just yet. FYI.

 
That sucks. The ones I sent to Atlanta, Ga were posted the day before yours and he got his on 28/Feb. Yours were sent the same way from the same post office.
A set I sent to Iowa took a month to be delivered - the longest any have ever taken. I'll email you shortly.