Help on GX developer programme

[bob1371]

Here is a small program with comments to give ya an idea.

 

[Bobtheclock]

Dear Bob

I have attached a file with all the info I have re Inputs/Outputs and descriptions also the setting instructions I have been given plus the results of me following the instructions today.
Basically the Hour dial continued to index , Y2.
I noted the line numbers in monitor mode that were flashing and I have screen dumps of the program when its was not performing how it should have been. If you need them I can download them also.

Cheers

Bob

 

[Bobtheclock]

Input/Outputs for Kyoto

not sure what happened, I posted a reply but it doesnt appear to have gone , so this is a repeat.

Attached is a word document with the In/Outs for X and Y's
In addition my Instructions and Results.

Just woke up and back on uk time.

I have now got another file from UK, Iss 14, along with some comments and more instructions, they look a bit complicated and apparently I have just been told I might not have access to the clock tonight till 2100 - so it may be while before I am back.

Back in a bit.

Bob

 

[Bobtheclock]

Thanks Bob

Just like the buses , expecting 1 reply and 2 come at once - see what you mean about the comments - makes it a bit easier.

Is the info I sent ok in that format ?

Bob

 

[Bobtheclock]

Iss 14 tests

Well back again after test of Iss 14.

Found an error in the instructions Bob , X11 should read X13 - ah well - suppose I should have expected it.

X11 is Reset soldier sensor count not Hour dial count.

Am going to write up my results - basically Hour dial did not index at all on set up (10:15) and at 24:00 hrs when X0 triggered.
It also will not index manually when triggering X14 - X17 - X14.

Qu :
Can someone give me an honest opinion - am I wasting my time, surely theres some sort of methodology (is that a word) to PLC programming.

If I say "stop" now what would be a structured way forward, to some extent I feel each change is adding to the problem.

For example if I wrote down what I expected to happen each 15 minutes could the program be tested offline as it were.

I very much appreciate your help so far but feel as though I might be asking too much.

Best regards

Bob

 

[Bobtheclock]

Iss 14 Results

Onwards and upwards

I made some notes in the results about the hands at midnight, doesnt concern the PLC its to do with the GPS system. It just means the PLC timer is set a few minutes out - I can fix this.

Goodnight

Bob

PS : When I come out of the monitor mode it asks if you want to save the project - does this store any of the register values and should it be saved as a new file ?

 

[Goody]

To be honest, I think the program is now getting more and more unessesarily complicated.
There seems to not have been much collusion between the programer and the clock designer.
I have studied the program at lengh and while I can understand it (or most of it) there is too much data moving of the RTC (real time clock) values.

It should be fairly easy to set up a 24 hour clock and have functions run at pre-determined times from it.

The PLC clock could even have been set in England for Japanese time and it would not be out much (even without power) when it got there.

You might have to ask him to completely re-write it giving him a 24 hour chart of all the events and the times of these events.

 

[Bobtheclock]

Many thanks Goody thats what I was thinking. The reason behind setting the plc time each day was so that over time it wouldnt get out of sync with the GPS system. Cant comment on why the data is moved about so much but I think I know what the real problem is.
On Iss 14 there is a M141 - in line 325 (my man originally said this would be line 326 - this has relevance and bear with me)
I think this should be M140 - from

line 363 CMP D8015 (HRS) C1 (DIAL) M140

I also cant see where M101 comes from ?

My point is this - my man is a big bloke with big fingers and a small laptop.

Hence instruction X11 should have been X13 and all other Typo's.

If this is right am changing my name to sherlock.

Can you confirm if my thoughts are correct - I have only used this tutorial this morning - well I think it was yesterday for you guys.

Off to bed for an hour

Bob

 

[Bobtheclock]

Sorry last message was tosh I know S1=S2 MXXX + 1 - long day/long night but nearly there now.

Soldiers working,lights working, just got hours to sort.

Bob

 

[Bobtheclock]

Sorted - Thanks for all your help.

If anyone wants to know what the problem was let me know.

Cheers

Bob

PS ; Next big clock I do will have a bloody big weight, pendulum and loads of ratchets/cams/pawls/levers - much simpler than a PLC.
Progress - who needs it.

 

[geniusintraining]

If anyone wants to know what the problem was let me know

Yes I want to know...I was off yesterday, but yes that is the reasoning for the website Bob, so we all can learn

Next big clock I do will have a bloody big weight, pendulum and loads of ratchets/cams/pawls/levers - much simpler than a PLC.

I'd go with a sundial...

 

[Bobtheclock]

Sun dial

Well apparently this was it :

"The reason that its not indexing is because of a little line of code I put in to try to ensure that the count C1 reset ay midnight. Unfortunatley it counts the pulses of Y2 not the movement.
So, in short, the count C1 is always equal to the hours without having to move.

However, I've taken this out now and replaced it with a different falling edge trigger from the 2400 sensor"

Simple eh.

Now get this - The GPS movement I bought in as a package, included gearbox (which was the worst ever machining I have had the pleasure to see), GPS box of tricks and receiver and a Chiming unit - oh and a speaker.

After re-making the gearbox it was just case of plug in and go. Now we were on a very tight lead time for this clock - hence the mess (customer pulled delivery forward 3 months after he placed the order and I am always up for a challenge), and so I never really listened to it striking.
Last night, sat in the back of the clock, freezing my nuts off, midnight came and -- no strike.

Am now told -

"I have had &%** (%$£"%% check out the chimer and you were correct it is chiming '0' at 12 midnight. If this is a problem he can email the files to you and if you can have access to a card reader it is simple to copy the additional files onto the removable card".

You can imagine my reply.

Sun dial sounds a good shout, especially as global warming increases - couldnt be anything to do with all the billions of PLC's on stand by could it !

Have a great day, I was serious when I said you have been a great help, it helped me put some structure into the tests so my man was able to solve the problem and I now know how and what a PLC is and does.

Bob

 

[Goody]

Bob, glad you got it sorted

When you get home tell your programmer on that CPU you can do inline compares (lol not online but inline)

With the normal compare 'CMP' you use 3 bits;;; less than ,equal to and greater than. Often you are only wanting one condition, maybe 'equal to'

eg -[CMP D0 K10 M10}- this uses 3 M bits

-[= D0 K0]---(M10) uses only the M10 bit/coil

 

[bob1371]

I recieved a email from Bob yesterday. He is back in the UK. He did send a picture of the project I thought I would share here.
I checked out some of the mantle clocks they make and even went as far as asking a price. 1/2 a years wages.
www.clockmakers.com


Bob