Automated Master Caster and Star run report

[Tazza]

Now, that would be fancy

[No_1]

The high pitched whine does make them less enjoyable, especially when running for long periods of time.

The switch you have in place for the lead pour could be used as a limit switch, i'd like to have one that activated when the mold is fully down too.

Did you modify the ram or set it is fully extended and retracted every stroke as to not put extra force on the MC frame? Or did you just wind the pressure right down to do this? I work on hydraulic gear in my spare time and know just how much force a hydraulic ram, even a small one has. I built an argon transfer pump that was powered by a hydraulic ram, i underestimated for force it could apply, and it tore welds off the support brackets.

Your setup is great that you can adjust on the fly, no coding to alter for any delay you want, yet the timing relays for most of us are very expensive. Hatch was talking $80 or so each, and you needed quite a few of them to do the job.

Every move is based on time, the timers control the hydraulic servo valve, the servo valve controls all fluid coming and going, I have set the timers so the ram never reaches the end of the stroke on either end. There have been (many) times when a boolit has hung the system up but I have seen no adverse affect most likely due to the presence of the relief valve being set somewhat low. As a side note, the cylinder is not a commercial unit but custom built. The original design was a bit big (2") so another cylinder of much smaller diameter (1") was fabricated then pressed into the original cylinder.

[No_1]

Yes the limit switch I gave you for your lead pour safety would work fine. I think it only has a single set of contacts so you would need to use a relay if you wanted to keep the lead pour safety.

The problem with relays is the cost and amount of wiring needed. My cost on the relays you used is @$150 each or was several years ago when I looked them up.
So that would put the cost on relays alone at $1200. Factor in another $150 for a enclosure and small parts. That's a pile of money that most people wouldn't want to spend.
The PLC setup doesn't allow for on the fly adjustments without the use of a expensive display or hooking it up to a PC, however the PLC can do the functions of a 100 relays or more.
Prime example is that the current PLC program uses less then 1% of the available program space.

I was able to duplicate the basic automation functions using just 3 relays. It runs fine but doesn't have a double tap feature. It allows you to control the three basic functions. Top timer (lead cool), lead pour amount and bottom timer (mold cool).
It runs off a limit switch and will run only one cycle per limit switch closure.

I am not saying that any design is better then any other design. What I am saying is there are different ways to do it and with each way there are limitations.

Switching over to a double acting cylinder on the Wyman design would solve the only problem that I have come across. Which is the sprue plate not closing 100% sometimes.

The system you and Wyman have designed (in my eyes) is superior to mine. I considered a PLC as well as proximity switches and then timers when designing mine. Timers was the choice and the cost of them was a non-issue since a close friend had a HUGE box and said I could take whatever I needed. The original plan was to use proximity switches to "locate" the mould in the machine but it was determined that was not needed because the timers would determine the location. What was not considered was the need to bring the hydraulic system to operating temp before setting the works into motion because not doing so causes the carriage to stop short of its destination. If I leave out the stop in the middle (under the fan) I could reduce the amount of timers by 2 but that is a stop I like when the outside temperature gets above 95 degrees, the boolit weights get above 200 grs and the total cycle times get close to 10 seconds. Adding a proximity switch at the bottom would not help me due to the design of my system.

I did not mean to hijack your thread, I only added information because I thought you had indicated you were going to "2 timers" for control of your system (vs the PLC) when in fact it appears you are going to 2 separate programs which consist of different sets of times but using the same PLC. For my type of system 8 timers is the requirement. I still am considering the PLC but I need to do some heavy research on the control panel you are talking about because I would like to retain the ability of making quick changes on the fly related to different bullet weight, temperatures and dew point.

On the same note, the addition of a tapper at the bottom would greatly help both of our designs. I think I will purchase one designed for the commercial machines then adapt it to my system.

[HATCH]

The program for the PLC is ladder logic.
There is a rung that contains the timer for lead cool. I put a normally open 'switch' block at the beginning that says if the heavy boolit switch is on then skip the rung.
I copied the rung and used the normally closed block and said if it is on then run the rung.
I did the same thing to the rung that contains the movement timer.
So its not two programs really.
I have two more inputs that I can use so I could do two more sets of timers very easy. In fact I could do a lot more because I can do multiple switches on run. If switch 1 is on and switch two on then run....
I have learned a lot since 2011 when I saw you machine running.
To be totally honest it was your machine that got wanting to automate mine. I just couldn't afford to do it the way you did.
It wasn't until Wyman came up with the pneumatic design that got my mind thinking.
If I were to build one from scratch I would still use a PLC but would go with a double acting cylinder. I would also use a touch screen display and have the lead pour operated by the PLC as well.

I spoke to a coworker today and he verified that the PLC we use would do exactly what I wanted. So I may build one of those just to try out.

[Tazza]

The ram did look pretty big for the job you had for it, even 1" would provide more than enough power.

I never thought about the fact that when it was cool that the time taken for the ram to extend and retract would be different, but it does make sense as things heat up the oil becomes thinner.

[No_1]

It does not take long to warm the equipment up. I turn the pump and let it run against the pressure relief for a few minutes then cycle the equipment for 2 to 3 minutes. When the lead has melted I turn the timers on and away we go.

[Tazza]

Can't get much easier than that.

Are there any fail safes just in case a timer failed? you don't want the ram pulling or pushing too far out of the desired range.

[No_1]

The only fail safe installed is the proximity switch which keeps the lead pour circuit open (spout closed) unless the mould is under the spout. Charles came up with this idea after witnessing a lead pour failure because the mould was not where it should have been. I have been looking over my diagram to determine if a simple rewire will be all that is needed to implement the new idea presented by Charles (the switch being the start control) or if it will be a total rewiring nightmare. I upgraded to a PID controller when I replaced the heating band some time back which consumed all remaining space in the box so I have considered upgrading to a larger box that will allow the addition of another relay, a couple terminal boards and house all of the control switches internally. Hopefully by time I have decided to do that I will have determined whether I am going to modify my current system or move to the PLC set up.

[HATCH]

Relays are normally spring return. So if a relay fails it will return to its unpowered state.
I would assume that it would continue to run just skipping the function that relay would provide.
With the lead safety relay and the relief valve no damage or mess would occur.
If I remember correctly there is a master relay that everything is timed off of. If this relay fails then I suspect the machine would stop running completely.

There is NO way for you to have a runaway lead pour situation unless something gets in the pot and jams the valve open (that has happened recently to someone else).


As far as switching over to PLC, I am not sure if I would do that since you have a working relay setup.
The only feature advantage at this time would be if you wanted to do double tap and having the limit switch start the process.
You have cast tens of thousands of boolits and the only major flaw in your design has been fixed with the addition of the lead pour safety switch.

One feature that I somewhat like is that if a boolit jams the mold open, it continues to run and in the next cycle it frees the boolit and operates properly.
My only fear I had initially was a breakage because the cylinder tries to move the carrier. But with the relief valve you have that would prevent the pressures from getting high enough to produce enough force to break something.
If you were to switch over to a limit switch cycle control then you would need to also have a way to activate the movement cylinder so you can clear a jam and reset the mold carrier back to the top. This could be down with a momentary toggle switch with center being off.

I am thinking of a feature in the PLC which would allow this to happen.
Basically after the "relay" for stroke stops it would have a delay timer of a sec or two then if the limit switch isn't closed it would run the stroke again.
Kinda like if after two seconds the limit switch isn't closed then run double tap.

Robert the major advantage the PLC has over a relay based design is that ability to "rewire" every function using nothing but software.
Prime example is the double tap feature. With the exception of adding the input for the on/off double tap switch, this feature is all software. The setup that Wyman has (which is running V2.0 of the software) could be upgraded to all the latest features by just downloading the project into the PLC and adding a single wire to the PLC. The wire is only needed if you want to switch doubletap on/off. You can enable double tap using the software so its always on.

At my work we deal with DDC modules. They are just PLCs. When we are dealing with a new piece of equiptment (like a big AC roof top unit), we look at the points we want to control and the points we want to monitor.
On the MC automation process you only have to look at the following points
Inputs
Start
Stop
Limit

Outputs
Lead cool
Lead pour
Stroke

The start/stop could be one input as its either on or its off.
So you really only have 5 points to deal with. As long as all 5 of these points are wired into the PLC you can control the machine

On my design we have more inputs and outputs only because of features

Inputs
Start
Stop
Limit switch
Double Tap
Heavy Boolit (new feature in beta)


Outputs
run light (green)
Stop light (red)
Lead pour
Movement cylinder

The lights aren't really needed for function but I include them just so you can see what is happening.

[Tazza]

I had a tinker with the double acting ram last night. My dad thought it had a longer stroke than it actually did, so i had to make a few modifications to the cylinder to get an extra 15mm of stroke. Long story short, the ram now has the required stroke for the task at hand. The end was re-threaded to fit the clevis for the mold. I am still machining a sleeve to mount this ram where the old one was. The old one had a slightly larger OD, so a sleeve needed making to fit the mount i have. A little longer at the lathe and that will be finished, then i will need to put the cylinder mount in the mill and cut a slot to accommodate the air coupling.

Good thing i had a look at this, the ram I was using before was not as good as when i first installed it. It was re-purposed for the task and the spring inside had lost quite a bit of tension. You can grab the end and there is no spring tension for at least 1" of travel. It has done pretty well as it started it's life out as an air operated valve spring compressor for cylinder heads. It has done a good 20 thousand strokes without any real issues, i may still convert it to a double acting cylinder just in case the new cylinder failed down the track.

[HATCH]

Finished working out the bugs last night on Chloe's control panel and the new software.

It now has two sets of toggle switch selectable timers.
One for standard boolits and one for heavy boolits.
Double tap works the same for both settings and also can be set independently for light or heavy boolits.
Had to basically rewrite 1/2 the program last night to get it to function like I wanted to.
Good news is that all you have to do is jumper a wire at the PLC and install a wire from a switch to activate it.
No additional wiring needed.
This is good for those with older setups

I have two more mods in the works but they will wait till after vacation.
(1) tappers
(2) touchscreen interface (adjust all settings)

With the touchscreen I would move lead pour over to the PLC.

Found the touchscreen I can use.
$200.
Requires a programming cable then a interface cable so its gonna be about $260 or so to add it.

It requires no changes in the PLC. All programming is in the display

[Tazza]

Good to hear you got the new panel done with the modifications that you had planned. Handy that the modification if members want it just requires a switch and a jumper.

[Tazza]

I got the double acting cylinder installed on Saturday, it took a lot longer than i had hoped with other issues i came up against, but nothing that couldn't be fixed. Mainly cylinder stroke and mounting another air solenoid/mufflers.

It worked very well, no hang ups. It still needs fine tuning as when the ram retracts it does come back faster than I like. The cylinder does have buffers which help, but i think a restrictor may give it just that little bit more control.

The gas strut is still fitted as when the ram came back, it would bounce a little at the end, It could be the buffer, but i'd rather it come back softer than just a big thud hitting the stopper on the MC

[HATCH]

Use a flow regulator on the exhaust port on the valve to slow the return.

[Tazza]

That is what i was going to do actually. My only concern is that there may be too much air hose between the cylinder and control valve to be as effective as it should be. Time will tell though, i'll need to get some restrictors. The one i have is 1/4 bsp not 1/8. They are cheap out of China, so i can risk a few bucks on a test.

[ProfGAB101]

Somehow I read this when Wyman first posted, then it fell off my radar.

Tazza I feel your grief as I was doing much the same fighting with an Arduino to control my Star luber. I used a 6:1 GT5 timing belt driven by a Nema 34 640 oz-in stepper.

Phantom problems - every once in a while it would want to start up going the wrong direction or some other absurd behavior. I fought it for 3 months. Never could get it to run reliable and with out adding feed back switches all over I don't think it ever would. Don't know if cell phone proximity or other signals were botching up the Arduino's internal logic. Bottom line is one day I just gave up and grabbed a "dumpster grade PC" and loaded up Linux CNC. I made a cable to go from the parallel port to the stepper driver and the whole wiring, software install and config took under 2 hours.

It worked good enough to find out that my stepper motor did not have enough torque at the existing drive ratio to size my extra hard 45 acp slugs. It functioned perfect with slugs that had been sized before, or run without slugs. I have since ditched the timing belt and am switching to a #35 chain drive much as I hate using that type of dirt magnet. I have some of the items on hand, some I have as only blueprints until I can get in some quality shop time.

I think I am done with Arduino's below the "Mega" series. Lets be honest they were intended to be used as an educational tool, not an industrial tool.

Yeah driving this with a PC takes more space but its free - as in cheaper than a direct from China Arduino even, plus I can pull up a different "CNC" program for each different bullet if needed.

edit - not #320 chain its #35 chain

[HATCH]

The arduino has a place but most people I have talked to have issues.
$80 for a PLC from automation direct is cheap and you can do a lot with one

[Tazza]

I am still at a loss as to why i had so many issues with mine. It was a good learning project, a dedicated PLC would have been better as it was designed for industrial situations. I wonder if shielding is what is needed to make the arduino more reliable. Mine runs well when no other inductive loads are run off the same power lines, filters may be needed? who knows.

It would be interesting to see if the issues were fixed in the mega version, but i'd have my doubts.

I like the cheap options, if to works and free, you can't get better than that. Especially if the pc was destined to become land fill, it got a new job.

Hatch is right, the click PLC he uses are quite powerful, reliable and a good price. Looking back, i possibly should have gone for one, but being in Australia, it would have added quite a bit to my build.

[HATCH]

I am gonna have to buy another output expansion block.
I don't have any extra room for tapper control.
Its $35

[Tazza]

Not too expensive at least. And it will do what you want it to.