Plc5/80e

Christoph · Aug 18, 2015

All

Merging 2 data systems. Need to know the highest value constant
I can use in a PLC5/80 and what is the largest value you can have
in a Floating point. Thanks in advance.

TWS · Aug 18, 2015

PLC Data Type: IEEE Float (32-bit value)
Range: +/– 1.1754944 E–38 thru +/– 3.4028237 E+38

Steve Bailey · Aug 18, 2015

Unsigned 16-bit integers, 0 - 65535
Signed 16-bit integers, -32768 - + 32767
Unsigned 32-bit integers, 0 - 4,294,967,295
Signed 32-bit integers, -2,147,483,648 - +2,147,483,647

Mickey · Aug 18, 2015

See pdf for Floating point and integer range.

I don't believe a PLC5 supports unsigned integers or long integer ( 32 bit)

Steve Bailey · Aug 18, 2015

I don't believe a PLC5 supports unsigned integers or long integer ( 32 bit)

It's been a while since I worked on a PLC5, but it's hard to imagine that there are things you can do in a Micrologix 1500 that you can't in a PLC5. 16-bit integers are N files and 32-bit integers are L files in the Micrologix.

(8{)} ( .) · Aug 18, 2015

It doesn't. Actually I may be the person responsible for that...

TConnolly · Aug 18, 2015

Christoph said:
All

Merging 2 data systems. Need to know the highest value constant
I can use in a PLC5/80 and what is the largest value you can have
in a Floating point. Thanks in advance.

The PLC/5 uses the IEEE 758 floating point standard. This standard is how most computers represent and use floats. With floating point numbers the largest values are subject to resolution. For a resolution of one count, the largest floating point number you can have in a 32 bit float is 16,777,215 (2²⁴-1, the largest value that can be represented in a 23 bit mantissa plus one implied bit). Although you can have bigger numbers, values larger than 2²⁴ can no longer increment by 1, you start loosing resolution in the 1's column. As a general rule of thumb, you have 7 to 8 significant digits in a float when written as a base 10 number.

(8{)} ( .) · Aug 18, 2015

And while it's a standard the byte order is different than the CLX. The 5 are 680x0 based.

Christoph · Aug 18, 2015

When I store a counter value and I reset on both ends to 0 at 999,999
everything is fine but if I try to go out 1 more to 9,999,999
everything just doesn't work right. this is in the PLC5. data is coming from
a ControlLogix messaged over.

Ken Roach · Aug 18, 2015

Is your "counter" a REAL or a DINT ? Are you "storing" the value with an ADD instruction ?

You're probably encountering a basic feature of IEEE single-precision floating point while trying to add a large number to a small number. TConnolly referred to this, and it's a classic problem when folks are writing accumulators in PLC-5 controllers.

The only sensible approach to large integer accumulation in the PLC-5 is to use cascading accumulators.

nightline · Aug 19, 2015

(8{)} ( .) said:
It doesn't. Actually I may be the person responsible for that...

Can you explain the second part...?

(8{)} ( .) · Aug 20, 2015

Yes, I was the tech support guy in charge of the PLC-5 back in the late 1980's and early 1990's. I was asked by engineering if we could live without the DINT data in the NP5 (New Platform 5) because memory space was really low etc. I said that we can use floating point for large numbers. Hindsight says that I should have fought that one.

Plc5/80e

Christoph

Member

TWS

Member

Steve Bailey

Lifetime Supporting Member + Moderator

Mickey

Lifetime Supporting Member

Steve Bailey

Lifetime Supporting Member + Moderator

(8{)} ( .)

Lifetime Supporting Member

TConnolly

Lifetime Supporting Member

(8{)} ( .)

Lifetime Supporting Member

Christoph

Member

Ken Roach

Lifetime Supporting Member + Moderator

nightline

Member

(8{)} ( .)

Lifetime Supporting Member

Similar Topics