Archive for November, 2010

Version 2010.11.15 of SoapBox Snap is now available for download. It includes 11 new Math and Comparison operators for the ladder logic module. This fills in most of the missing functionality you need to make good use of the NUMBER type in your applications.

Let’s step back and talk about types for a minute. SoapBox Snap only has 3 types:

• BOOLEAN
• NUMBER
• STRING

That’s a really lean type system compared to most PLCs (but of course, “Snap is not a PLC”). If you’ve done much traditional PLC programming, you would expect to see a dozen different numeric types, like BYTE, INT, UINT, DINT, UDINT, not to mention SINGLE, and DOUBLE precision floating point numbers, and various BCD numbers if you’re really lucky. At first when I was writing SoapBox Snap, I planned to support every data type in the IEC-61131-3 standard, but I realized all those types only existed because traditional PLCs are so memory constrained that you had to take memory usage and low level memory management into account when you use them. One of the goals of SoapBox Snap is to remove the burden of memory management from the programmer. That’s why there’s only one NUMBER type and it can represent any number.

SoapBox Snap already used the NUMBER type in the Counter and Timer instructions that shipped with the first version. The counters expose a Count output that you can access just like any other signal. Likewise, the timers expose an Elapsed output (in milliseconds). Of course, all analog inputs in your Device Configuration also expose NUMBER type signals, and all analog outputs consume NUMBER type signals, so now you can manipulate all of these signals with the new instructions below.

Math

The new math instructions carry out your four basic mathematical operations: add, subtract, multiply, and divide. Each takes two NUMBER inputs, and produces a single NUMBER output. Since a result is generated, all of these instructions are considered “right hand” instructions (that is, they show up on the right hand side of the rung). They all use the rung-in condition to enable the operation. If the rung condition feeding the instruction is false, the result stays the same as it was during the previous scan. If the rung condition is true, the result is overwritten with the new result of applying the operation to the two inputs.

There is one exception: if the second input to the divide instruction is zero, it disables the instruction from executing, so the result will stay the same as the last scan. There are various alternative ways it could handle a divide-by-zero condition (throwing an error, setting the result to zero or infinity, etc.), but I thought that holding the last “good” value was the most reasonable thing to do.

Comparison

These are your basic comparison operators. They are all “left hand” instructions, so they go on the left side of the rung, where you normally use contacts. Each one takes two NUMBER inputs, just like a Math instruction, but doesn’t generate an output, other than the rung-out condition. If the rung-in condition is false, the rung-out condition will always be false. If the rung-in condition is true, the instruction will read both inputs, apply the comparison, and set the rung-out condition to the result of the comparison. Therefore you can chain these comparison operators together on the same rung (for instance to say A > 250 and A < 750).

Choose Number

Similar to the Math instructions (in that it has two NUMBER inputs and one NUMBER result) is the new Choose Number instruction. This instruction copies the first input to the result if the rung-in condition is true, and otherwise it copies the second input to the result. If you use this instruction in combination with a Greater Than instruction, then you can easily implement a Maximum instruction. That is, if A > B, return A, else return B.

What’s Next?

There are some small features we’re going to be adding to SoapBox Core shortly. After that we’ll come back to adding some new features to SoapBox Snap, so keep watching this space for updates. Remember, contact us if you have any questions, or post a question on the support site. Most of you like to email me directly, which is just fine. 🙂