I succumbed to a new LG smart TV this week, which is a vast improvement visually on the steam-powered one that has served for 10 or more years. It was quite easy to set up apart from one major issue: getting the LG Magic Remote to pair with a BT Youview Ultra HD set-top box. The internet really wasn’t much help other than acknowledging this is an issue, but I sorted it within a day.
I first enabled HDMI-CEC on the LG TV (or SimpLink as LG calls it, located in the general settings). I’m not sure whether this is actually needed, but it happened as part of the general attempt to solve the problem after reading about it on the internet. I also reasoned it would help with linking to my Blu-ray player.
Then the trick, rather than selecting BT as the service provider box in the connections settings for the LG TV, is to instead choose not in the list and then Humax from the list that follows (the box itself is made by Humax). In the following button-testing screen for the remote options, “remote option 1” then worked fine.
The downside is that the Magic Remote does not have obvious play, stop and skip buttons, so it’s easier using the BT remote to control video or to pause live TV. And I find it easier to use the specific remotes to turn on devices rather than going through the connections menu. But that may just be an issue of habit.
One topic that recently came up on a discussion group was whether to pre-heat spin-casting moulds for wargames figures before running them. Nowadays, I always pre-heat, leaving moulds in an oven at 100 deg C for at least 30 minutes before casting. Moulds that have thin items such as horses’ reins are often heated to 120 deg C for the same time before casting.
The main reason for doing so is to improve metal flow and surface finish. Better flow helps cast fine parts of a figure. And the better surface finish results by eliminating the bubbling on the surface caused by the temperature difference between molten metal and rubber. Some casters will tell you this bubbling is caused by metal being too hot; it’s also caused by a mould being too cold. More accurately, it’s a matter of reducing the temperature difference between rubber and metal to improve surface finish.
Of course running hot moulds requires gloves to be worn at all times.
How much a mould has to be heated also depends on the casting metal. I find lead-free pewter is trickier to cast in organic black rubber moulds than the leadier alloys I’ve previously used. It is lower density and therefore lower mass, producing less force behind it when poured. Leadier alloys may only need a mould to be gently warmed.
Running a hot mould will also allow a lower metal temperature to be used. This stands to reason because the metal will lose less heat flowing through a hot mould.
The other benefit of heating the mould is that it reduces figure shrinkage. A cold mould is smaller than a hot mould, and you can demonstrate this by simply dropping a cold mould back into the moulding can and looking at the difference in diameter. Depending on rubber, a mould may be 2mm to 5mm less in diameter when cold, translating to a 1% to 2.5% shrinkage. This shrinkage naturally affects the size of mould cavities.
I heated up natural black rubber and silicone moulds to compare them. The black rubber shows more of a difference, i.e. there is more shrinkage on a cold black rubber mould.
Shrinkage from silicone moulds also appears to be dependent on the vulcanization temperature. I use low temperature cream silicone from J Coker of Faversham vulcanized at 90 deg C for two hours. Vulcanizing it at the recommended higher temperature for just an hour appears to result in more shrinkage, and I switched to lower temperatures and longer times years ago.
This is backed by the data sheet for Nicem’s low temperature silicones. Nicem’s light green or orange low temperature silicones boast a 0.3% shrinkage when vulcanized at a very low temperature, but shrinkage increases as curing temperature increases. I’m waiting on samples to check what is possible.
Shrinkage caused by mould material is in addition to compression shrinkage caused by pressure on the mould. Casting with a hot mould will by and large reduce shrinkage to that caused only by compression. By comparison, a metal figure cast in a cold mould will lose height, length and thickness because the mould is smaller, and thickness because the mould is also compressed by the casting machine.
Well, here’s a surprise: the article on casting tips for a Saunders spin-casting machine is put back for a while thanks to the arrival of a new baby in my life: a virtually mint Saunders Mark 5. I am now the proud owner of three of the beasts: two working, and one for spares, repairs and ultimately refurbishment.
The idea was that my spares machine would be machine number two, but refurbishment is currently awaiting welding work on the casing and lid, and it really needs a new funnel.
The new machine has arrived thanks to a chance conversation with Paul Thompson of Early War Miniatures at the Bovington wargames show. It was his spare, but he has now acquired another compressor machine. According to Paul the Saunders had been languishing in a basement for 10 years before he acquired it, having originally been bought to cast model railway parts. But the then owner couldn’t get it to cast and abandoned the project.
I spent the weekend clearing out the workshop to make space for the new beast, and a morning cleaning the machine and freeing-up seized parts. I have never seen such a pristine example of a Saunders, however, it is showing all the signs of long storage. Most of the internal arms and threads still have a golden tinge of what looks like zinc plating, and slight oxidation of this means that the threads on the weight arms needed re-stripping using the 5/8 BSF die nut I’d bought to refurbish machine number two (see part 3). I am fortunate in that I have a sturdy bench vice to hold parts securely while I turn nuts with a wrench after applying universal engineering solution no 1 for parts that don’t move but should – WD40.
The machine had some slight metal build up on screw threads, all of which was possible to flick off with the help of a screwdriver. I also scraped the inside of the casing to retrieve a few hundred grammes of unknown metal.
It should be obvious from the picture above that although both machines are Mark V Saunders separated by only a couple of hundred by serial number, they are both different. The most noticeable differences are the height of the motor mounting, the nature of the belt guard – which accounts for the different height of the motor, and the start-stop switch and lid-lift break switch.
The belt guard on the new machine is a fully enclosed type – the old machine has a guard only at the front. After fitting it, I have to say it’s the worst designed bit of safety kit out, as it catches the parts it is supposed to enclose. In the end I left the back part of the guard off, and ultimately I may replace the front part with the half-guard from the spares and repairs machine. The enclosed guard necessitates a lower mounting position for the motor assembly.
The start-stop switch is a soft, NVR type; the older machine has a less-safe on-off type switch which in my experience causes RSI through repeated usage because it takes a good prod with a finger to turn it on. The lid has a complex arrangement that sounds as if it operates a relay to turn off the machine when the lid is lifted; it takes some time for the turntable to come to a halt, so safe operation is best accomplished using the start-stop switch. As this is soft-touch it shouldn’t cause RSI.
My original machine has a simple push-rod system that operates a switch under the machine when the lid is lifted or closed, and I therefore tend to use the lid to start and stop the turntable – saving me a fortune in chiropractor’s bills. The spares machine has no lid switch at all.
Inside, the bottom platter of the turntable is different: it is recessed by several millimetres to accommodate 11-inch moulds. This also lowers the typical height of the top platter in the machine, in theory allowing slightly thicker moulds to be accommodated. One of the weaknesses of the design of the Saunders is that it cannot cope with really thick moulds because the adjusting screws on the top of the weight arms can only be adjusted upwards so far (a modern compressor/ram machine is far more flexible when it comes to mould thickness). The recess on the platter unfortunately catches the first knuckle of the fingers when using a nine-inch moulds, making protective gloves essential.
Update: The recess has the unfortunate side-effect of increasing the distance between the end of the funnel in the lid and the mould. I found this led to more metal spray across the top plate, and so I put in a 5mm packing sheet of rubber (I vulcanise a few layers from an organic rubber mould to make these) when casting to make it behave like my other machine.
Owning two machines that show no signs of being able to adjust the motor speed – the knob to adjust the motor distance from the turntable belt wheel and therefore alter tension on the variable speed pulley on the motor spindle is just too stiff – it was a surprise that this aspect of the new machine works as intended, with the adjustment knob turning freely and allowing a speed selection of, in theory, 500 to 900rpm (see part 1). However, the machine doesn’t run sweetly until set to 700rpm or faster; increased speed is certainly noticeable when cranked up to 900rpm, which should make this machine better for dealing with problem castings.
A word of caution. Because the actual speed depends on the tension of the belt on the variable speed pulley, the speed on the gauge may not be accurate. If links have been taken out of the belt since its factory set-up then there will be more tension on the variable speed pulley and it will run slower than indicated. Trial casting seems to indicate that turning the new machine at the indicated 800rpm seems to produce much the same result for casting success as my other machine set to 600rpm – but that has a fan-belt rather than a multi-link V-belt between motor and turntable. I haven’t changed that to a V-belt because the machine is running fine and I work on the principle that if it works, don’t touch it.
Update: Inspecting underneath the machine, I eventually twigged that the main V-belt pulley on the new machine and spares machine is 6 inches in diameter; the one on my original machine is 8 inches. I don’t know if was subsituted by a previous owner, but it explains why the maximum speed of the original is capped at about 640rpm, whereas the others will get up to 850-900rpm. There again, different Saunders machines may have been built to different specifications. I’ve ordered a new 6in 5/8in bore V-belt pulley for the old machine – ideally I’d like both machines to behave in exactly the same way.
The new machine has only a set of heavy weights, which were set slightly more than half-way down the weight arms. It’s no surprise that the original owner couldn’t get anything to cast. I tried it, and even my AB horse moulds, which require more pressure than others to seal, would not cast successfully at any motor speed. I then decided to adjust the arm weights, which is when I discovered they were locked pretty solidly onto the shaft, which meant I reached for the WD40 and took everything apart. However, it was interesting to try a machine that looked factory fresh on its set-up.
Once I’d freed up the weights and the adjustment screws on the top of the arms, I set the weights to my usual number of turns down the shaft (10 complete turns from the top for horses). To do this, wind the top nut on each arm to the top of the arm (i.e. nearest the platter), then wind the weight up to touch the nut. Mark the weight with a felt pen – I draw a line outwards from the centre – and use this line to count the number of turns back down the shaft. Then tighten the top and bottom nuts lightly with a spanner to lock the weight in place on the shaft.
With a mould and the top platter in place, hold each arm in turn so that the top adjustment screw just touches the platter (there’s a circular steel disc where each arm falls). Turn the screw so that it falls roughly in the centre of this disc (see part 2, for how to accomplish this easily using record turntable spirit levels).
Update: The adjustment screws on this machine are tight and don’t move freely, despite cleaning, and I have ordered a 5/16in BSF die nut to re-cut the threads.
I set the speed to what was supposed to be 700rpm, but eventually moved up to 800rpm to replicate the results from the older machine.
I will at some point use a non-contact laser photo tachometer to accurately measure the speeds of all the machines. When I last borrowed one to measure the speed of my original Saunders, it was running at 620rpm or so regardless of what the machine’s speed gauge said.
Update: I fitted reflective strips for the laser photo tachometer to the belt wheels on the underside of both machines so I could read the turntable’s revolutions per minute without opening the lid. The new machine, set to 750 on the gauge, runs at 620 rpm; the original machine, set to 600 on the gauge also runs at 620 rpm. So both machines are now set to run at the same speed. The new machine when set at 800 on the gauge was actually running at nearer 700 rpm. Just so you don’t go around believing anything a machine tells you.
Encouraged by the fact that the motor speed was clearly adjustable, I applied engineering solution no 1 to the screw elements of the motor rack on my original machine, and within a few minutes the WD40 had penetrated and I was able to adjust the motor position. Again, there was no discernible change in speed, but the original machine uses a fan belt which I suspect when only just taut is too long for the distance between turntable belt wheel and the variable drive pulley to make a difference. When it eventually fails, I’ll replace it with a proper multi-link V-belt.
WD40, however, has failed to help on the spares machine, so I will be disassembling the motor mount and adjustable rack when I have time. I suspect the screw threads must corrode slightly and they just need cleaning and freeing up.
Anyway, with two machines now set up to cast properly, I should be more efficient in what I turn out. I’ll be able to have one machine running while I empty the other. I work by optimising the machine for a mould and then running that one mould repeatedly, rather than working through a stack of five or more moulds, to avoid having to adjust the machine each time between moulds. This is because I am an insane perfectionist who would rather work slower to achieve the optimum result for a mould. 🙂 I should add that even this trait does not guarantee an end to mould lines; it just minimises them. If I ran compressor machines that automatically adjust for different mould thicknesses, where pressure can be changed at the touch of a button, and motor speed can be changed at the turn of a dial, then I would work through a stack of moulds like any normal person.
I’ve managed to operate Fighting 15s for 10 years using just one casting machine with no guarantee of finding spare parts or being able to get it repaired if key parts go wrong. The nightmares associated with its possible failure – the one week I spent without it was problematic to say the least – mean that although it may seem excessive to have two running machines and a spare, I am at last in the position where I can survive the failure of one machine and quickly repair it or slot in the spare until the failed machine can be repaired.
Next: Casting tips or the Saunders instruction manual (if it arrives!)
My second Saunders casting machine arrived in a state. To be fair, I knew it would. It was the machine I learned on back when I went up to Gladiator Games with Mike Lewis of Black Hat Miniatures for a crash course in casting and mould making with Bill and Elaine Lucas. Mike and I have known Bill since our early days of roleplaying. Mike was buying Gladiator Games and Coat d’arms paints from Bill and asked if I’d like to come along for the day.
Back then I was still importing everything from Eureka and AB, and being shown the casting process was a revelation. I’d imagined it would be difficult and involve lots of snipping of figures from sprues. Instead, I learned about kiss gates and how they made figures easy to snap off a sprue, how to use packing to bring a thinner mould up to height, and had a stab – literally and painfully when using a Stanley knife – at cutting a mould. Bill showed us his tricks at getting his machine to cast with his moulds and it seemed so easy.
I arrived home fired with enthusiasm to buy a casting machine, and was fortunate enough to find my Mark V Saunders on eBay. No one else bid for it, and it cost just £450. The seller was coming down to Southampton for the cricket and kindly offered to deliver it there free, and Mark Rowsell of Flashing Blade kindly stored it in his garage until I could get friends to pick it up as back in 2007 I couldn’t drive. It fitted in their people carrier with about a centimetre to spare at the top. eBay also provided a second-hand 20kg Tiranti/SEBA melt pot for £100. Nic Robson at Eureka sent me an old mould to use and to prove that I could cast. I just used the machine as it was, melted some scrap Eureka figures (accumulated breakages in transit), and miraculously turned out my first AB French light infantry. It has led to me casting more AB Figures under licence and, of course, Eureka’s 18mm SYW range and my own figures.
The day that machine broke was the day I realized I needed a spare, even if Nelson Engineering of Bembridge did fix the machine in a week. It had been put together without a keyway on the spindle of the main platen, and the belt wheel was held in place simply with a grub screw. A keyway is a groove cut into the shaft of a motor or spindle; there is a corresponding groove (the keyseat) in the belt wheel, with another piece of metal (the key) cut to go into both grooves: this keyed joint prevents the belt wheel from slipping on the shaft. Nelson cut a new keyway, and welded the belt wheel that they’d had to use an engineering solution to remove (they hit it with a hammer). It’s an experience I didn’t want to repeat, so checking the new arrival and discovering it had a keyed joint was a relief. I don’t know whether Saunders machines have never had a keyway or whether my machine was a one-off, but it’s something that any Saunders owner should check.
Mike of Black Hat has gradually been moving over to compressor machines, first an MCP machine that was Calpe’s spare spare (Peter has two of everything as a contingency plan, and at the time had three casting machines) and most recently a secondhand SEBA machine. When Mike said he was selling his Saunders, I immediately said I wanted to buy it. I knew it was in a state, caked with casting metal and the victim of many over-enthusiastic engineering solutions (it had been hit repeatedly with hammer and cold chisel), but it was mechanically and electrically reliable and importantly came with a full set of Saunders weights that would allow me more control over my main machine’s pressure – I was already borrowing one set from Mike anyway.
So the first task was to get rid of all the casting metal that had built up around the pillars and arm adjustment bolts. No matter how careful you are, inevitably metal gets flung around the inside of the machine, either because insufficient pressure has been applied to the mould and it flashes, or because too much metal is poured into the mould and it is flung out of the hole in the top plate. And that’s why you never cast without the protective casing of the machine in place and why you ensure the lid is securely down.
I haven’t got pictures of how to remove metal, but basically it involves knocking off as much excess metal as possible with a heavy hammer and cold chisel – carefully, to avoid damaging any further the platen, pillars or adjustment bolts – and then unscrewing everything, attaching each bit to a wire, and dunking it in a pot of molten scrap metal to take off most of the remaining metal non-destructively. Of course, these parts get very hot, and once out of the metal they need to be left for some time on a fireproof surface – an old silicone mould is fine – until cool enough to handle. There will still be bits of metal in the threads, but these can largely be flicked out with a fine, pointed modelling tool (one of those dental/wax modelling picks sold for sculpting in green stuff) and a fine wire brush.
Another consequence of hitting the machine to remove metal is that the threads of the adjustment bolts and the arm weight shaft may be damaged. They need recutting using an appropriate die from a tap and die set, or a die nut.
In essence you lubricate the thread with WD40 or PTFE spray and then just use a spanner to wind the appropriate die nut up the thread and it re-cuts the thread, taking off distorted metal. Alternatively, use a die in its special handle. It’ll still leave a dent, but this won’t impede the movement of a nut or arm weight up and down the shaft. The aim is simply to get everything free-moving again. BTW, wear gloves to avoid getting particles of swarf embedded in your hands.
The pillars have two holes through them that allow a tommy bar to be inserted so that they can be turned in a die nut, itself held in a spanner, to restore the thread. The securing nuts for the pillars were badly damaged, and I have replaced them with 3/4in 10tpi UNC stainless steel nuts, which I ordered from eBay. They’re not going to be subject to great pressure, so the stainless nuts should be OK. I don’t intend taking an engineering solution to them.
One of the arm adjustment bolts was damaged beyond fixing, so I replaced the lot with new stainless steel ones from eBay. Again, used in compression and not whacked with a hammer, I think they’ll stand up. I will look for a non-stainless alternative. These are 1 3/4in fully threaded bolts – i.e. the threaded length is 1 3/4 inches; the head is extra. It’s the maximum size possible that will keep the head clear of the casting machine lid in normal usage. The bolts and corresponding nuts are 5/16in BSF (22 tpi).
The bolts that provide the pivot point for the weight arm didn’t need replacing. The look to be 3/8in UNF (16 tpi) 3 inch partly threaded bolts, threaded to about 1 inch so the weight arm pivots on a smooth surface. These really are out of harm’s way, which is no doubt why they are OK. I did, however, undo them to take off the arm to rethread the weight shaft, cleaned them up with a fine wire brush and lubricated them with PTFE spray before reassembling. I used a spanner to lightly tighten the nut beyond hand-tight. All the lock nuts on a Saunders need tightening slightly beyond hand-tight because the vibrations of the machine will shake them loose. This is particularly relevant to the lock-nuts for the arm weights in order to stop the weights travelling along the arm when in use and changing the pressure.
I’ve taken the protective casing off the machine. It was held on with only three screws, but it should use more. With only the three, the base flange of the casing lifts from the wooden base, allowing metal particles to get underneath. I took about half a kilo of metal out of the machine. The casing has been damaged on one corner and needs welding, which a local garage has said it can do. I’ll add details of this when it happens.
When the casing comes back, I’ll also sort out the mounting of the new NVR magnetic switch to turn the machine on and off.
The top plate is also “worn” and I’ll be investigating using high-temperature silicone to provide a liner for the top hole to make it easier to remove excess metal that’s escaped from the top of the mould.
This is the ultra nerdy part about using a Saunders spin casting machine. It will get absolutely the best results and allow easy setting up for different moulds while running a typical cycle of five or more moulds at a time.
The key to getting the most mould-line free castings on a Saunders is where the arm adjustment bolt falls on the top plate. It needs to fall over the centre of the pillar that both separates the two casting machine plates and locks the top plate into position to stop it spinning off when the machine starts up.
What you need is three 15mm turntable spirit levels, typically available from eBay. These are intended to get a turntable and turntable arm set up on a record deck, which will mean nothing to anyone under the age of 30… apart from hipsters with new vinyl record collections. Plus you’ll need some green stuff (epoxy putty) or five-minute epoxy adhesive – green stuff takes longer to set but will allow the spirit levels to be seated better.
First, ensure the casting machine is level. Use a conventional spirit level on the wooden base and if necessary pack under each foot of the machine’s tubular frame until the casting machine is level.
Take off the top platter. For each weight arm and pillar, turn the adjusting bolt in the top of the weight arm and raise the pillar until they touch and by eye align centre to centre. It doesn’t matter about how high the pillar is or how much through the top of the arm the adjusting bolt projects – it simply matters that they align. Link all the arms using rubber bands to pull them together and keep them in place for the next step.
Mix up some epoxy putty and put enough into the base of each spirit level so that any cavity is filled and there is some surplus to provide a bed into which the level can, if necessary, be pushed. Stick one level to the top of each weight arm, being sure to leave enough space between the level and the nut on the adjusting bolt. Carefully press the level into its bed of epoxy putty so that the bubble in the spirit level is centred. Leave to set. (You can use five-minute epoxy adhesive instead, which sets quickly but is fluid and needs continuous adjustment of the level until the epoxy sets firm.)
If you’re thinking why not attach the level to the top of the adjusting bolt, which will be horizontal, it’s because there may not be enough clearance between a level on top of the bolt and the lid of the machine when in use.
When the epoxy has set, remove the bands and wind the adjusting bolts and pillars to approximately their usual positions.
Place a mould on the bottom plate, and then put the top plate on, locating the tops of the pillars in the appropriate holes to hold it in position. Adjust the pillar height so there is a gap between the shoulder of the pillar and the top plate even when the assembly is compressed slightly.
While holding the weight arm so that the adjusting bolt is lightly in contact with the top plate, turn each adjusting bolt so that the bubble is the spirit level shows that the arm is level.
This adjustment is the key to good castings. It ensures that each weight arm applies the same pressure at the same distance from the edge of the top plate, without tipping the top plate in any one direction. Combined with correctly positioned weights on the weight arm to give enough pressure to seal a mould, this will minimise mould lines.
Update: Using this arm position on the top plate is a starting point. Moulds may benefit from having the arm shown slightly off-level – provided that all arms are the same degree off-level. I get better results on some moulds by having the contact point slightly towards the edge of the top plate, and adjust the bolt so that the bubble in the level is up on the top of the marked circle in the level – i.e. nearer the bolt (bubble nearer the bolt equals contact point nearer the edge; bubble nearer the edge equals contact point nearer the centre).
Weights can be positioned evenly down the weight arms simply by measuring from a fixed point. When setting up for the first time, I take the weight arm off a machine and measure using a ruler from the pivot point (actually the top of the hole in the weight arm) to the top securing nut of the weight. When in use, I typically use an item of fixed width, such as a metal bracket, to set each weight the same distance along each arm. I then fine-tune the distance by making test castings and increasing or reducing the pressure by moving each weight the same number of turns respectively down or up the arm until the mould stops flashing or starts filling all the cavities. Once I’ve identified the right setting for, say, a 15mm infantry mould, I note which gauge I’ve used and the number of adjusting turns, and then use that setup as the basis for all other 15mm infantry moulds.
Importantly, this is just a starting point. Individual arm weights may vary very slightly in weight thanks to the accumulation of casting metal from inside the machine (hence it is important to clean them). And the three threads making up the weight arms may vary very slightly in length, adding marginally to the mass and therefore causing a slight difference in pressure on the mould. Even after careful setting up, you may find that one weight may need a quarter or half turn up or down the arm to balance the machine: exactly what adjustment is needed is part of the black art that is spin casting.
The spirit levels allow minor adjustments to account for small differences in mould thickness to be made very quickly. If you make sure a casting session involves only moulds of about the same thickness – there is a natural variation in thickness caused by differences in the original mould blanks and by the volume of the originally moulded figures (big figures displace more rubber!) – then each mould may take only a quarter or half turn of the adjusting bolt to set the machine up optimally.
Also, it is important to bear in mind that moulds may be uneven, sloping to a lesser or great degree, because the vulcanising press is not even. What I do is work out the thickest point of the mould (it is, by and large directly opposite the thinnest point) by running the jaws of adjustable wrench round the outside of the mould until the wrench stops (closing up the wrench if necessary!). I mark the point and then run the wrench round the other way until it stops again, and mark that point. Halfway between those two points is usually the thickest point. I mark that point on the mould with paint or a marker pen. I then always align that point with the same pillar on the mould, so that all moulds always have the thickest point aligned with that pillar, meaning I have the fewest turns to make on the adjusting bolts.
Sometimes I just set up the machine perfectly for one mould, securing the adjustment bolts in place by tightening their locking nuts. I then cast only from that mould, leaving slightly longer between spins before opening the mould (i.e. so that the central core is set) compared with operating using a stack of moulds. This keeps the mould hotter, helping the flow of metal – handy for lower density casting metals such as pewter.
I’ll run another blog entry sometime soon on my casting tips, covering metal temperature, mould temperature and talcing.
Next: Cleaning, repairing and refurbishing an old Saunders
Fighting 15s does all its casting on a casting machine made by N Saunders Metal Products Ltd. Saunders appears to have gone out of business in about 2002, but its casting machines live on. A number were apparently used by Games Workshop, and have therefore found their way to other toy soldier businesses: machines are also used to cast items such as fishing weights, and they occasionally crop up on eBay. I recently picked up a second machine that needs attention to get it working properly, and stripping it down provided the opportunity to take pictures and show how it works.
The Saunders is a robustly built, bob-weight machine. It is highly adjustable, and with experience and one simple improvement can reliably turn out high quality castings.
Bob-weight casting machines use weights on the end of three arms to apply pressure to a mould. When the machine is switched on, the platen spins, the weight-ends of the arms rise and the short end of the arm is pushed down, applying pressure to the top plate.
Pressure on the mould is varied by the following means: speed of rotation, the heaviness of the arm weight, and the distance of the weight along the arm. In addition, the adjustment screw at the top of the arm affects where the force is applied on the top plate: the location of this is critical, and in part 2 I’ll look at the improvement that can be made to get the best results.
Speed control is rudimentary. The Saunders uses an AC motor whose speed cannot be controlled by a variable resistor – those only work on DC motors. Instead, it uses a variable speed pulley – the belt wheel on the motor’s shaft is split and has a spring, which has a dust cover, that applies pressure to keep the two halves of the belt wheel together. The belt face of this split wheel is sloped, and the motor can be moved in its frame to increase or decrease the tension on the drive belt: the more tension, the wider the split belt wheel is forced apart and the smaller the diameter of the drive wheel becomes, and hence the slower the main turntable spins (a smaller drive wheel takes more turns to turn the main belt wheel once).
The mechanism to move the motor assembly, however, is basic and difficult to adjust because it is hard to get purchase on the knob used to adjust the distance.
In practice, because of the difficulty of turning the adjustment knob, you learn to work with the speed the machine runs at and change the other variables to cope. It is possible to convert the machine using a DC motor and a variable resistor to provide better speed control, but the cost might not be worth it.
The arm weight and its position on the weight arm, however, give plenty of control. The system may seem arbitrary, but as the pressure applied depends on the principle of levers it’s quite easy to adjust pressure to suit difference moulds. Even on casting machines operated by air pressure and a ram, discovering the initial pressure required to get figures to cast is a matter of guesswork improved by experience.
My initial machine came with only a set of heavy weights (985g or so); the new machine has a complete set of three heavy weights, six mid weights (260g or so) and three light weights (165g). The arrival of these weights has given me more flexibility. It is possible to have some machined, which I will be looking into: the arm the weights screw onto is a 5/8in BSF thread (British Standard Fine, or 14 threads per inch). The weights may have accumulated metal around the milled edges and it’s important to remove this to get all the same-sized weights to the same weight.
As I said, how much pressure is applied depends on the formula for levers. And here’s the science bit… If you assume that the rotation speed is constant (it cannot be easily changed anyway), then lever formulae decide the force applied to the top plate. In short, a heavy weight at the top of the arm (nearest the pivot) can apply the same pressure as a lighter weight at the bottom of the arm (furthest from the pivot). The relationship is quite simple, involving mass and distance from the pivot point (fulcrum): M1 x a = M2 x b, where M1 is the mass of one weight and a is the distance from the fulcrum, and M2 is the mass of the second weight and b its distance from the fulcrum. Very crudely, without converting into appropriate units, a 1,000g weight that is 50mm from the fulcrum is equivalent to a 250g weight that is 200mm from the fulcrum.
Where the Saunders limits this is the length of the weight arm, which is around 110mm, so you can’t use a mid-weight at one end of the arm to be the equivalent to the big weight at the top of the arm, because the arm isn’t long enough. With the locking nuts above and below the weight in position, the maximum distance a weight can be positioned along the threaded arm is about 75mm, or about 105mm from the pivot point allowing for the unthreaded elements of the weight and arm assembly. It means the heaviest weight at the top of the arm (it’s about 50mm from the pivot point) is, crudely, 980 x 50 = 49,000 “force units” – the same as two of the 260g mid-weights positioned 94mm or so along the arm. You also have to allow for the weight of the nuts and where the actual centre of mass is for the total weight if you want to be 100 per cent accurate.
In general, once you find out the correct weight and distance along the arm for one mould, that setting will work for most other moulds of the same type. There may be slight differences, and more or less pressure may be applied by turning the weight down or up the arm respectively. Marking each weight with a line allows you to count the turns up or down, so you know how many are required to restore a weight to its initial position. If nothing has cast on a first spin you may be 10 or more turns off: slight flash takes only one or two turns to cure.
If you’re starting with a Saunders for the first time, having the weight as far away from the pivot point as it can go will reduce the likelihood of spraying metal all around the insides and over all the adjustable elements. Getting metal off the parts may lead to the sort of damage my second machine has suffered.
Using lighter weights gives finer control over pressure. That levers formula means that a light weight has to be moved a greater distance to have the same effect of moving a heavy weight a small distance. So one turn of a light weight is equivalent to a fraction of a turn of a big weight.
Finally, it’s important to get the adjustment bolt at the top of each arm set to come down at the same distance in from the edge of the top plate to ensure that pressure is even. The top platen has three hard metal discs screwed onto it and on which the adjustment bolt lands: the right position is approximately in the centre of each of these discs. The discs are directly above the locking pillars that are screwed into the bottom platen and which have location studs onto which the top platen fits to stop it spinning off when the machine starts up. Basically, to set the adjustment bolts you hold the arm onto the top plate and adjust the bolt so it is central. Doing this ensures that even an uneven mould is evenly clamped by the machine. It’s a bit hit and miss, and the improvement I’ll cover in part 2 deals with how to do this quickly and reliably.
Lastly, a look at some of the basics of the machine. The Saunders should have an industrial machine stop-start switch. However, it’s what is known as a “dangerous start” switch, because it’s possible to start the machine simply by inserting the plug if the stop-start switch has been left on. However, my Mark V also has a lid-operated safety switch that means it only operates when the lid is down, so replacing the stop-start switch is not a concern.
The machine I am stripping and rebuilding, however, has lost its original stop-start switch, and has been fitted with an immersion heater switch. It also does not have a lid safety interlock switch. I will be replacing the main switch with a no voltage release (NVR) switch as a priority, and adding a safety interlock switch to the lid later.
Next: modding the Saunders to make life easier
Update 22 July 2017: Included the term “variable speed pulley” under speed control as that is the technical name for the sprung, split belt wheel.