Help needed with new project.

[jetmotorcycle]

I will start from the beginning.
What I have in my home workshop is an old relay and timer driven Band saw, its seen better days yet is mechanically sound.
It has an Auto feed and cutting head both driven by hydraulics and controlled by relays, contacter's and timers 110v.
The machine runs and cuts very well but needs constant attention to the electrics. Mostly due to neglect and interference by the previous owners.

My plan and a serious one at that is to fully re-wire and upgrade all the electrical system. It was wired for 3 phase 415v but I,ve change one motor to 240 and added an inverter for the other. It also had a 240v pump motor so thats been left as is.

So what I need is a spec on the 240v supply gear, would be running about 20amps at present.

Recommendation on what plc would be best for this application

help with the programing.

Function:

Mains in

auto feed mode/ manual mode switch, various switches already fitted for start and stop positions(EOS) of each action, control action is made by 110v two way hydraulic solenoids. Various warning lights, start and stop buttons.

Wish list and final goal.

The length of bar feed in on each cycle is governed by a movable hard stop, this I would like to change for something that is both accurate and changeable from an HMI or even a touch screen.

I'm just building an addition to the machine that will place a large drill in line with the machines bed, I'm wanting to use the bar feed but switch function to operate the drill head and not the saw blade. The mechanics is simply but the hardware and programing will need to added.

So if your still reading and can offer any advice I would be most grateful.

Regards
Paul

[evanh]

The simplest solution is a smart relay.  However, smart relays are not much fun to make regular adjustments with.  The T100's will do a great job, especially with the measuring side of things.

Attaching a rotary encoder to the end of the feed shaft will give the required feedback for cut-to-length function.   Go for a low resolution encoder of say 100 lines per revolution.  You don't need better than that and a low performance controller like this often can't handle pulse rates above a few kiloHertz.

To eliminate backlash as a cause of error between encoder and material you should always reference the machine in the same direction of movement as when performing the measuring stroke.

Stay with ON/OFF switching for everything else.  The T100 will have no trouble handling all the logic and timer functions.

The general approach for wiring is to have all limits and buttons and sensors go to inputs on the PLC and all contactors, valves and drives come from outputs of the PLC.  This way you eliminate all logic from the electro-mechanical sections.  Two benefits arise from this approach, one is ease of sequence modification and the other is clarity when fault finding.  Exception is the human safety wiring if any is needed, this must be hard wired and verified to function correctly.


Evan

[jetmotorcycle]

Thanks for posting a reply.

Taking into account the wish list, which one of the T100's will perform all the functions I need and should I be looking for other useful features, making the system a little future proof.
Also do you have a spec on a rotary encoder so I can find the correct one ( assuming there are different ones ). I.e voltage etc.

[evanh]

MD2424 to be safe on the I/O count and the MD-HMI for entering the lengths.


Evan

[evanh]

opps ... best encoder would be a 12-24 volt unit.  Make sure it's either a push-pull or NPN.  PNP won't work with these PLCs, this goes for other sensors also.


Evan

[jetmotorcycle]

Been looking at encoders and it all seems a bit of a mine field? Would I be correct in thinking a linear encoder is what I'm looking for as the action of the feed arm is linear. I most likely won't understand but can you give me a quick lesson on encoders for dummies or even a few part numbers to look up.

The saw is hard at work and so am I but as soon as I get a minute I will begin the task of detailing spec on all the component's that need to be controlled, solenoid valves etc. I will also attempt a drawing of the electrical system for members hopefully to edit.

At present the motor contacters and solenoid valves are all 110v, powered from a transformer. Would this be a good time to be thinking about changing them to 24v. I already have two 24v power supply's that I think may come in useful. Again i will post spec on all items soon.

Many Thanks

[evanh]

No, don't use a linear encoder, they are way more expensive and use 5 volt differential signalling designed for servo controllers and precision readouts.  You want a rotary encoder designed for PLC use.  Attach it to the end of the screw adjustment shaft with a flexable coupling and lightweight bracket.

Contactors are ten times the price of relays so unless they are warn out I suggest you have an intermediate bank of din-rail mounted relays.  I prefer this setup instead of having the relays on the PCB where they can be frustrating to fault find and replace.

That said, contactors are quite modular and if your ones are  modern then you could replace just the coils with 24 volt ones.   This may also apply to valves.

Indicators should be changed to solid state units for reliability.  Limit switches are usually clean contact so will happily run on 24 volt.


Evan

[evanh]

Another common way of attaching rotary encoders is with a toothed belt and pullies.

[jetmotorcycle]

Thanx for the advice, I will look into it and see if i can make a toothed belt and pulley work. It seems like a good option.
Just putting together a drawing detailing how it will all look, Thanx again.

[Dogface]

Just use a NPN prox switch with a toothed gear.  Or if it is a high speed application, use a HESS (Hall Effect Speed Sensor)  Unless rotary encoders have come down in price dramaticaly, this solution is both economical and simple.

[jetmotorcycle]

Got a good deal on a pair of encoders. One to use and one for spare.

Hollow End - Shaft      
Incremental encoder with integrated OPTO-ASIC, providing high interference immunity  
Speed up to 6,000 RPM  
Signal outputs A, A inv., B, B inv., N, N inv  
Temperature range for 5 VDC up to +100  
Power Supply 10 to 30 V  
Current consumption  
Operating frequency 150kH max  
Output signals Channel A, B, N + inverted  
Output stages  
10 - 30Vdc Push-pull output,short-circuit-protected  
RPM Value 6,000 rpm max  
Hollow end-shaft 12 mm / 22 mm deep  
Material Flange / Housing Aluminium  
Operating temperature -20 ?C to 70 ?C  
 

Voltage   10 - 30 Vdc
PPR 1024
GI330.0604123 349-0038  

I trust this will do the job ?

[evanh]

Yep, those encoders look fine, you'll need to watch you don't go too fast though.  The T100MD is rated for 10kHz, normally this means the pulse rate of each channel so the max count rate is 40000 counts/s.  I'll take a stab and say that's optimal so you should stay well clear of it by limiting yourself to 5 kHz absolute max pulse frequency which for a 1024 line encoder is about 300 rpm (5000 Hz x 60 sec / 1024 lines = 293 rpm).  A good target for the encoder would be around 100 rpm.

Also use the command SETSYSTEM 4,1 to give proper counting on channel one - I'm not sure if this provides full resolution or not but this is not a concern for you as the resolution is already plenty high enough.

Use an aux contact of the M2 contactor as the supply for all other contactors and valves.  This then becomes your safety enable as well as powering the pump.  The coil of M2 is powered through a series of contacts:  SW1, SW2, PLC_OK, INVERTER_OK, M2_OVERLOAD,  M3_OVERLOAD, M4_OVERLOAD.

The PLC can turn the hydraulic pump on and off with the PLC_OK output.  All other outputs control individual items and all other limit-switches and push-buttons are individual PLC inputs.


Evan

[plc_user]

Evan, can you expand on the first paragraph of your last post.  I like to understand your thoughts on the frequency range of the MD Series PLC when using encoder's.  I understand what you are saying about the 5kHz range being a more reseasonable range than the posted 10kHz, but if you would be so kind as to expand on that paragraph it would be deeply appreciated.  Also, would a interrupts going on in the RS232 and 485 ports cause the interrupt driven encoder counter to be even slower than the advertised 10kHz range on the encoder inputs?  You see very knowlegdable about the PLC and its features and any info is appreciated.

[evanh]

Yeah, sorry about that, when I said "pulse rate of each channel" I ment each of the two phases of the quadrature encoding, often refered to as A and B channels of an encoder.  The other point that may need explaining is the 40000 counts/s, this is related because each quadrant of the combined A and B phases is a usable position thereby producing 4x the resolution compared to the pulse frequency.


Evan

[evanh]

Your final question is a valid concern.  Counters are a very demanding job without support harware and when the processor is doing many jobs it can't afford to spend too much time on the one function.

There is, however, more than one way to skin this cat.  Yes you can use a direct interrupt from the rising edge of each of the phases, which I assume is roughly how the basic method is done.  The enhanced method will likely use a timer based interrupt that observes all quadrature transistions and therefore could achieve the 4x resolution feature.  In both cases the IRQ should be a higher priority than the comms IRQ.

PS:  My musing are, on the most part, general knowledge.


Evan