Floating point ADD

[Ben 1]

Hi everybody, I am using +F instruction to move my floating point value from DM1022,1023 to another DM1028,1029. (Using CQM-1H CPU61) Instruction is, +F DM1022 0 DM1028 My question is, It gives me "operand2 memory area not valid" when I try to use #0. I know that 0 means IR000 channel. On this I have some digital inputs. In above case, is it adding the value of channel 1 to my dm1022, I don't want that to happen. I tried with #00000000, #0.00 but not working! How can I add value 0 to my floating point?? (I saw in a code addition of -0, what does it mean?) Thanks, BEN

[PdL 71]

I don't think you can add a decimal number as the addend word has to be of 32-bit floating type. Although the instruction help says so. A workaround would be to fill in a DM of which the value of the DM and the DM+1 is always 0. Additionally you could move 0 in DM and in DM+1 to be sure. I don't work with floating data that much so there must be a cleaner trick.

[Ben 1]

Thanks Pdl, Which means the instruction help is TOTALLY HOPELESS!! and misleading. For the beginners like me it proves to be very costly. Anyways, I was already thinking in that direaction as last resort and did change my code to move 0 to DM, DM+1, and then use those DMs. However, this seems to be very long solution for a small problem. I'm not sure but MOVL should work. However, I don't have space to add this in my expansion instruction! However, i'm not clear whether add 0 means addding the contents (0-65535) of 16 bit IR0 or not. And what is this add -0? is it equivalent symbol for #0 or 0 or something else??

[PdL 71]

Well if you can accomplish what you want with the instruction although it has another purpose, I don't think it is hopeless right ? Anyways, I never used the MOVL instruction but I think it will do what you want, move 2 consecutive words from S and S+1 to D and D+1. You are right if you use 0 as addend word with the +F instruction, that the word value that the inputs of IR0 reflect at that time wil be used as addend value. So no good. No idea what -0 could be used for. But if you are short on expansion slots, use the XFER(70) instruction. Should have thought of it in the first place. XFER(70) #2 DM1022 DM1028

[Ben 1]

"Well if you can accomplish what you want with the instruction although it has another purpose, I don't think it is hopeless right ?" No offence! I was talking about the option of data types it shows for +F instruction. If # can not be used with floating point, why is it shown there! Anyways, thanks for hinting Xfer instruction.

[PdL 71]

Thought you we're saying the instruction itsself is hopeless.. Out to get glasses...

[BobB 58]

I have just had a look at the help in CX-Programmer fo +F and # is not shown as an option. In fact # means BCD and & means binary/HEX. The instruction help clearly states that all words have to be floating point to IEEE754 standard.

[PdL 71]

Instruction help for +F in CXP V6.00

[BobB 58]

Can't read Pim.