Master caster electronic automation - some questions

Hello Everyone

I’m at the very start of automating my Master caster. I decided I would program a Single Board Computer (Arduino or Raspberry - cost effective) to operate a step motor for replacing the handle and an electric actuator for regulating the pour. The SBC could read out potentiometers or a dip switch to set the required delays at the mold position and how I want it to shake out the boolits at the end.

Big part of the costs is selecting the torque of the step motor and max current associated with it for feeding it.

I removed the handle, made a kind of 1/2” connection so I could plug a borrowed torque wrench on the axle. Only to find out it is a counter clockwise torque measurement, and most torque wrenches only gives a clockwise torque for fastening stuff.

Here is my question: Can anyone give me an educated guess of a worst case torque that normally is applied on the handle? I assume that mechanics who are familiar with torque measurements can give a rough estimation just by feeling it. Biggest mold I cast is 45acp, dual cavity.

I need an answer on this question to avoid any of the following situations:
- buying a left handed torque wrench (200-300usd, only need it once)
- buying a step motor too small
- buying a step motor too big (can cost over 500-1000usd)
- buying a H-bridge driver too small or too big along with the motor and the power supply (same problem)

Thanks for the input
Kind regards

I’ve thought about doing the same thing with a Master Caster. If you look at the AmmoBot, they use multiple chain driven sprockets to increase the torque of their stepper motors. I think this would be key as once you go over 2200 in-lbs in a stepper motor the prices increase dramatically.

I’m curious what you plans are for actuating the lead pour? Like you, I want to stay away from air cylinders but in this case the speed of an air cylinder is exactly what you need. Most linear actuations are far too slow so I would assume a high speed Servo motor or powerful solenoid would be required.

Anyway, Keep us updated as you venture into this project.

Ammobot? Do you mean this?
https://www.fishpond.com.au/Toys/Wor.../0018859813009
Sorry just joking ;-)

Depending on the needed torque and available budget, I may need to adjust the ratio of the transmission if I use a V-belt. Anyway, I thought using a V-belt may be safe if the step motor operates the mold further than it’s supposed to be, then at least the belt can slip before some part gets broken.

Razorfish, to answer your question: I wrote down ‘linear actuator’ but probably a solenoid would be a better idea. Other stuff comes to my mind, but before I want to try this out and share my experiences, I need to solve the biggest issue:

—> What torque do we approximately exert ont the handle of a master caster? Say, worst case, a .45 double cavity bullet mold with a normal alloy?

As written above, I made a whole test setup only to find out my torque wrench doesn’t measure counter clockwise momentum. Feel a bit silly.

If anyone can give a rough estimation of the torque we need to apply to operate the Master caster, that would be really helpful!

Looking forward to seeing how this project goes. I rescued some stepper motors on the weekend to use on a bullet collator, still need to work out how to wire them up....

Keep in mind that an Arduino is not industrial and potentially will glitch on you. My master caster is automated with one and it does glitch on me, it uses air cylinders for pour and dump. The dump really does need a lot of torque, if the sprue has hardened too much, it can be a job to cut it with a motor running at a linear speed. With the handle, you can at least get a run up to shear it off. I liked the idea of using a gear box of some sort to give more torque.

Be careful of what type of electric solenoids you use, i used a 240v solenoid for my pour, the EMF given off when it released messed with my arduino, it took me MONTHS to work out what was causing the glitch, but when i swapped it out to an air cylinder, it glitches one in a thousand or so.

A good source of stepper motors are photocopiers. I got about 10 out of one on the weekend, a fairly large one, 3 medium and a bunch of small ones.Attachment 227801

Hopefully that attachment worked, those tiles are 400mm square for perspective.

Hi Tazza,
Thanks for the hints. If I use large coils like a solenoid then I will think of a diode snubber to contain the back EMF when the current through the circuit is cut. Also separating galvanically the power supply of the Arduino and the power electronics for the solenoid + step motor would help. Maybe a little more filtering like soldering rod core chokes in series with the 5V or 3V power supply of the Arduino will help cleaning up the supply voltage, and maybe an input capacitor. This should reduce the glitching regardless of which single board computer is used, being it a Beaglebone, Raspberry or Arduino.

Diodes would help for sure, mine was a 240v ac coil, so diodes weren't an option :( Possibly a capacitor across the wires might have helped.

Any extra filtering will help, i may look at adding that to mine. Mostly it does run right, but every so often......

Is there really no one able to help me with a rough torque estimation of the Master Caster?

You can have someone measure the air pressure they use on a cylinder and the piston diameter. That will give you the force

Then measure the distance between the pivot point and cylinder.

That will give you the torque needed.

Quote:

---

Originally Posted by Pebbles

Is there really no one able to help me with a rough torque estimation of the Master Caster?

---

I would like to know the answer too but I think the answer is going to be more complicated. Notice that there are a couple of popular auto drives that use stepper motors but they use gearing to increase the torque. So, from a power standpoint a large stepper motor should have enough power if geared properly.

The question might be if the stepper motor can gain enough speed (accelerate) to knock the bullets/sprue out of the mold. If not, you “may” need tappers. (And if they have enough holding power to stop immediately without skipping steps)

You can probably calculate the force that the more popular air cylinders create but they have an advantage of accelerating very quickly so it may not be a fair comparison. There is also designs that use electronic motors and a horsepower rating could be helpful.

Again, I’m not much help but if you look closely at some of the auto drives for progressive presses, you’ll probably be closer to an answer.

RazorFish,
I agree with you the pressure of the cylinder on a Wyman/Hatch setup is not all you need. I would recommend not accelerating to knock bullets out like the air system.

Pebbles,
Why don't you take your master caster out and pull on the handle a few thousand times. (You will get bullets out of it and they can be used so no time wasted)
Then take your Torque wrench put the end in a vice and pull on the handle and try to duplicate the pressure you used to shear a cold bullet that sat in the mold too long (worst case scenario)
Then convert the Nm or ft-lb back to the shorter handle.

Sorry i missed this thread.

There are too many variables.
The amount of clearance between the rails and the mold carrier.
The amount of lube on the bearing surfaces on the shaft and rails.
The temperature and alloy of the lead.
The timing of the machine itself.

Now your gonna ask why does the temp and alloy of the lead matter? Why does the timing of the machine matter??
Simple, softer lead requires less force to cut.
Shorter cooling time before sprue is cut means it’s easier to cut.

The movement cylinder provides about 30 lbs of force based on the average pressure used by most of the customers that use my panel (70psi)
But some have had to run 100psi which is is 40lbs of force.

Like I said this all is dependent on the variables listed above.

On the air setups we use flow regulators to slow the speed of movement while still keeping the pressure(force) up.
The ‘slamming’ of the mold carrier to release the boolits is due to the molds not being polished and additional force needed to release the boolits from the mold.
Even when i operate it manually I have had to slam the handle down to release stuck boolits.
I polish all my molds now to make the drop easier.

Look at this video -> https://youtu.be/B0TWNmgEMnM

Look at the 3:30 mark. He uses tappers to release the boolits.
If your using a gear motor your gonna need rappers for your setup too.

Steppers and torque motors slip when torque is exceeded but they overheat if drive is not reduced.

Tappers
Autocorrect got me

Quote:

---

Originally Posted by HATCH

Tappers
Autocorrect got me

---

Has anyone ever sourced these tappers? They look like some sort of solenoid with a pretty long travel.


Sent from my iPhone using Tapatalk

I never located them and the German guy is useless when it came to information about them.
He paid someone to automate his machine.

Thanks Hatch for chiming in.

In the mean time I nearly started a small machine shop for converting all my moulds, bought an X2 type mill. I acquired a Left hand torque wrench and found a source of Step motors and appropriate drivers. I believe the step motors can be programmed to do a back-and-forth shaking at the end of the stroke. I assume this would be sufficient if combined with polished molds, time will tell.

I know I advance very slowly but I am committed to convert this machine - family and work occasionaly get in the way. I definitely will not pay an external company.

Thanks all for your ideas.

Please do keep us updated with how you go with this project, i'm interested in seeing how you go about this. I hope to one day complete a project i started of building my own version of a bullet master, thses systems may help do it.

Quote:

---

Originally Posted by Pebbles

Thanks Hatch for chiming in.

In the mean time I nearly started a small machine shop for converting all my moulds, bought an X2 type mill. I acquired a Left hand torque wrench and found a source of Step motors and appropriate drivers. I believe the step motors can be programmed to do a back-and-forth shaking at the end of the stroke. I assume this would be sufficient if combined with polished molds, time will tell.

I know I advance very slowly but I am committed to convert this machine - family and work occasionaly get in the way. I definitely will not pay an external company.

Thanks all for your ideas.

---

so its been a little while, where are you at with your progress?
I did sell a panel to a guy in Belgium last year or was it the year before.

Not so far, I had an urgent renovation at the house. But in the mean time I used the left handed torque wrench on one of my bigger diameter molds and found out the torque that would stall the motor is at +/- 14Nm when operating properly and a bit more if I let the mold cool down too much, like when I start up the casting process and find out I have an adjustment to make while the cold mold cools even further down.

So now I found out what size stepper motor I need to buy. And I think I will use another way instead of the solenoid for operating the pour.

Urgent house work is never fun.... I'm trying to finish mine so i can put it up for sale

Glad you have managed to calculate the torque needed, i wonder if there is a figure you should work with, as you don't want to spec the motor too low, or too close to that figure, just in case. Do you add 50% to the required torque? Different lead mixes will give different torque requirements to cut the sprue, bigger is generally better in these situations, to an extent. You don't want so much power that if it was to jamb up, that it pulls the machine apart.

I saw another way that they did lead pour, you can use a stepper motor for that too, set it up to a cam, so you move it fast, it drops for a short time, turn it slow, it pours longer, it's an option?

A small step motor and a cam is exactly what I had in mind. Will create a less noisy circuit.

For the main step motor I found earlier a 30Nm step motor for 200USD and its driver is about 100USD. I plan on watching the current and switching the thing off if the current is becoming bigger than a certain adjustable setting that I can tune with a potentiometer. That way I dont need a limiter switch to protect my master caster.

Actually the 14Nm were not calculated but measured with a counter-clockwise torque wrench.

So it looks I have answered my own question that I started this thread with.

I like the idea of limiting current, when it detects a current spike, it knows something is stuck and stops.

Double the torque sounds like a good idea, running a motor at it;s limit all day surely wouldn't be good. Do steppers have issues with over heating, as they don't have cooling fans?

Small update - I purchased the step motors and am currently trying to write software for it. I settled for an arduino. C code programming is like 16 years ago but the internet came since then.

It is just going very slow as I can program for something like 2 hours per week with my current obligations. I settled for a Nema 42 step motor for the main drive and a Nema 23 for the pouring. I decided to use a wheel instead of a cam.

The motors get hot if they stand still. But they can handle 100°C+. May install a cooling fan if needed.

I dropped the idea of a V-groove transmission and two pulley wheels. I will use a direct transmission with a sacrificial clutch made out of a kind of hard plastic. The clutch should break before my master caster does, as the Nema 42 looks quite impressive.

Good to hear you're working on it.

How did you come to the conclusion that 30nm that the nema 42 will put out will be enough? are you running this motor as direct drive? If it's done with a belt, you can increase the torque by changing pulley/gear sizes.

As far as i was aware, you can run the motor so it will not "hold" that is the main cause of heat build up. It means that even when it's not moving, it's still powered. Ideally, you only need the motor powered when it's moving for this setup. They will need to hold for CNC work so it can't be moved around by a load on the work piece/spindle, but for this project i think the hold feature isn't required.

Do keep us updated

Hi Tazza,
I used a left-hand torque wrench to measure the force required to cut the sprue of my largest diameter mold I currently have. I measured 14Nm max. I am aware that there are a lot of variables that affect this result. The amount of time the bullet can cool down is a big factor, together with size and alloy.

The hold feature for the step motor is important to make the machine operate at a known position. The step motor I use has no feedback as to where the spindle is pointing. That’s what encoders are for but then again here is not much choice in 30Nm motors, this pushed me a bit in the direction of a step motor (and I admit they are easier for me to understand and operate).

Once I set up the machine, the step motor may not slip or the arm of the master caster may hit its metal frame and break something. The motor I use has no dynamic current limitation. The max current it draws is set by a dip switch so i can not program a safety in that way. Therefor I use a weak machined plastic link between the motor and the caster.

A V-belt would also offer this safety feature. But I must also ensure the position is sure, I mean: no slippage. If I program the motor to do 100 steps to the left then the result should be exactly that.

If the 30Nm are not enough then you are absolutely right about your suggestion of changing the transmission ratio. an advantage of the direct drive are the lower costs, the motors were expensive enough for me.

If I would need two pulleys, the mounting plates, the ball bearings, one or more V-belts and a way to tension it so I get no slippage below 30Nm I would at least spend 100-200 Euro more I guess.

Thanks for thinking along

Ahh, that's right, i remember you mentioning the torque wrench earlier, sorry i forgot about that.

Another thing you can do is to add a limit switch so you can "zero" the machine every time it starts up. That way there will be one point that is the same every time, even if something binds and it does a step but does not move, no need for a rotary encoder that way, those are way too expensive.

Mine uses air, so it needs to know when the arm reaches set points, i have one under the pour spout and one for when the mould hits the stopper. There is no reason why you can't set switches on yours like this, so you still get the strike when it hits the end of travel, unless you wanted smoother operation without the impact at the end to shock the projectiles out.

It has been a while so here is a small update
I started programming and exploring examples of code beginning of april 2020, by summer I was programming a few days per month on this project. By november I got it all set up and mostly software bug free to a point where I could start testing. Unfortunately the automation failed because of the following points:
- The 30Nm torque, even if huge for a step motor, is marginal. Sometimes it slips when the sprue is hard, and then I have to re-calibrate which is a pain. Recalibrating means that the sprue of the next bullet is cooling too much.
- the direct clutch is not precise enough, the motor can get out of alignment if the sprue is somewhat hard, because of this there is a degree or so of angular uncertainty of the mold’s position.
- the above uncertaintity trips my microswitches that limit the motors’ movement;
- I don’t like to go to full speed for bullet extraction given this set-up, which causes sometimes a bullet stuck in the mold. However the motor is capable of much more speed than required. I’m aware that max torque is at low speed. Microstepping helped to smoothen the movement but did worsen the torque issue.

I’m not ready yet to give up. I think the solution lies in a pulley drive to double the torque and a tight timing belt. Should do away with the flimsy microswitches and put an encoder directly on the axis of the master caster (12 bit resolution = accurate up to 0,08° = more than enough).

Doing this electronically gives in theory a lot of possible refinement options (I’m not there yet), however I believe the air pressure concept is more robust, but may give more wear.

Addendum
I found it interesting to see the step motor is capable of accelerating almost instantly and stopping in such a sudden way that no impacting of the frame is necessary. I just need to tackle the stiff sprue torque issue and be more precise in the drive pulley/encoder.

Nice work, love to see pics of your progress.

As for air giving more wear, i'm not so sure. I have cast about 800,000 with my master caster driven by an air cylinder. It does thump at the end of the stroke to eject projectiles, honestly, i see no wear on these points of impact or anywhere else really. Even the pivot points for the mould carrier, there is zero wear and i bought the main casting machine used, and it looked well used.

I like the idea of a rotary encoder, but they are pricey and for what it's going to be used for, it seems to be an over kill. Can you not just use limit switches for it to know it's home position? I had plans of making an 8 station machine and i was going to use magnets or bolts for a proximity switch to detect when it was in the correct position. The arduino would look for this signal every cycle, so it essentially calibrates itself each cycle.

A ribbed pulley and gear as a torque multiplier is a smart idea. Will you have an issue with heat from the stepper motor? As they have no fans to cool and if set to hold position, you get heat from the coils even when not moving.

I could post pics but since I’m destroying everything to set up the timing belt pulleys, I need to tinker a bit more. The price for the encoder was ok, I paid something like 40€ and this is a small amount compared to the big Nema 42 step motor and the rest of the hardware.

Considering the heat issue: while testing my software the motor that actuates the pouring with a cam wheel got a bit hot, say 50°C which is not an issue. The drivers for these motors have dip switches to change settings - halving the hold current is one of those options. Less heat get generated in that way.

Also with full holding power the motors don’t heat up that much during the testing. The temperatures in my basement vary from 8°C in winter to 20°C in summer. The motors are allowed to reach 80°C.

If a cooling need arises, then a PC fan and a wooden housing to conduct the air flow over the motor can help.

If I can get everything to run precise enough and with enough torque then the standstill time is well distributed with movement, I have enough parameters in hand to manage the heating issue. Then again it may depend on your selection of motors, to answer on your concern.

Hi Tazza,
I just realize another reason of not having heat issues: I use step motors with a high supply voltage: my small Nema 23 motor (for the lead pour) is fed by a driver fed on 36VDC, the main Nema 42 motor is fed by a driver fed directly on the mains (230Vac 50Hz over here). Somehow the losses are lower with a higher supply voltage.

I know exactly what you mean, with how my setup evolved, i built it up, then disassembled parts that didn't work as planned. I'm not an engineer, so when i built parts, i didn't know if they would work or not, but then you find any parts that are weak or not working as planned. Jmorris helped me with issues i had with my sizer, something so simple, yet i couldn't see it, a different set of eyes or looking at it from a different angle sure can help.

I tried setting up a stepper to run my sizer collator, i have yet to get it to run as it should, first attempt burnt up the controller, i have 4 more to play with and hopefully get it to work. I ran it at 12v, i think it was too nigh and that's why it roasted. 36v seemed really high, but the motor i had was out of a photocopier, so i don't know what voltage it was meant to be. I never knew you could get 230v steppers, we run 230/240vac here too.

I just want to let you guys know that after three years of interrupted development I managed to make the automation work. I had to use a timing belt to increase the torque and a hall sensor over an I2C interface to adjust the motor / shaft position every now and then when the motor slips a few steps and gets out of alignment. The timing belt also helps not to destroy the machine when it stalls when I’m not directly nearby.

Will see if I can post a picture.

Yesterday I did a long test run, and casted 2000 bullets in sessions of 400 (double cavity). Occasionally the machine jams: sometimes a cable disconnects due to high vibrations and it is easily jammed by electrical noise. If I turn on the hood at the wrong moment the machine and/or display stalls.

There are also a few software bugs that I need to fix that accounts to stalling, for instance about negative angle handling (then my machine corrects in the wrong way). Also a great upgrade idea is to have a acceleration sensor notice when the machine is stalling to death and have it turn of the enable button, then I can leave it unattended with slightly more confidence.

But now I know it is worth to further refine the machine and develop a printed circuit board for it, eliminating the cable spaghetti and its unreliability and make a better screening/grounding/filtering against EMC.

Attachment 287129

Picture of big black-silver nema42 step motor (center), also the black and blue drivers are visibleAttachment 287130

Transmission from SKF
Attachment 287131

That is some awesome work there buddy, glad you finally got it running.

Mine took me a good 12 months, if not longer. I also had issues of stalling due to electrical noise. I'm wondering if i install shielded cable to my sensors and enclose the arduino i'm using to prevent most of the EMF would work.

Over time, you'll work out any and all the points that have issues with vibration. My locking screws on my cylinder loosen up over time, but i have not had any parts break off due to vibrations, which does surprise me, i have cast probably 800,000 with mine so far since it was automated.

Love to see a video of it running if you get time.