GeckoSub Mirage Evo - And Adventures in 3D Printing Speargun Parts

Voodoo Prints...
A few posts back, I said 3D printing is not really black magic any longer. Well... sometimes it still is and I think it's only fair that I share when that happens, too. Some filament types are easy to control and since I had had good luck so far, I got carried away and thought I was a master printer now;-)

But then, I tried to print with HIPS and it has humbled me...
I like the idea of HIPS as it is supposed to be dissolvable in d-Limonene, a chemical that allegedly wont fry your brain, eat your skin or give you cancer.
In this pic, you see two HIPS prints, from the same brand, printed at the same settings with the same g-code. The only thing that was supposed to be different was the color of the filament but obviously, while both prints have issues, one is a lot worse than the other:


It's not unheard of that filaments are different between batches or even between colors and I have no clue what causes these differences in my case.

But in regards to the overall issues that both prints exhibit, I think I know what's going on. HIPS needs to be printed at quite high temps for the layers to bond well and it becomes quite soft and runny at those temps. To make matters worse, HIPS doesn't like sudden cooling so you can't run the fan much and the printer needs an enclosure. So, on the areas of the part where the circumference is rather small (top and bottom) the last bead hasn't had time to cool down before the 250C hot print head comes around and lays down another layer. The result is that the layers start drooping. Once the travel time is a bit longer, as in the areas with higher circumference, it looks a lot better and the blue one looks very decent in the middle part.
BTW, I am using a bigger nozzle than normal and a rather "drafty" layer height to get faster print speeds.

You could fine tune the print g-code in the slicer software a bit more. E.g. print a bit cooler and slower at top and bottom which I tried - but overall, I think this material may just not be suited for small parts or parts with widely ranging geometry. Luckily, it might still work for me as the parts I am thinking of using it for are going to be quite a lot larger.

Anyways, as I said - I just wanted to balance the scales a bit.

Not much to report, but I did run a few more heat resistance tests - will share the conclusions later. Surprisingly the PLA+ I have been using didn't fare as badly as I thought.
Some real work coming up for a few days, so as mentioned, more indepth thoughts later:
Some HIPS and three different types of PLA:


In the oven:



And then trying to take some notes and measurements on the test parts after the baking:


While this little fat, curvy shape has been good to see how easily the filament prints, I figure I will redo the test with a shape that is easier to get consistent measurements on. Thing is, I need to know how badly the parts swell, shrink or warp. So, perhaps just a simple cylinder shape will better.

More OOOMPH
Also, I mentioned that I had ordered a new motor for the mini lathe and this little beast just turned up:




I have a 500W brushed motor in the lathe now and this new one is a 750W brushless motor - so, I am hoping for a lot more torque at low revs plus a quieter motor. Just from playing with it on the table, I think I will certainly get both.
It will take a while to get this upgrade done. I need to find the time and then make some good measurements and get a mounting plate fabricated. Also, need to order some new timing belt pulleys and bore them out before I actually pull the old motor out.

popgun pete said:

The "Mirage" has been relisted, same starting bid of 169 euros as before. https://www.ebay.com.au/itm/MARES-M...253998560933?_trksid=p2385738.m4383.l4275.c10
Back in the day when "Mirage" guns leaking copious amounts of oil in the shop were being discounted they were selling brand new for AU$50.00!

Click to expand...

It's often such a personal thing when it comes to pricing. But I got one of my first ones for not that much more than this one. It was a later generation and it came with two really nice Sigal shafts as well as an STC vacuum muzzle. I now think of the Mirages as project guns, so they actually have to be both cheap and the internals should be in pretty good order for me to consider buying one. My second Mirage was quite a bit cheaper than this Ebay one, if I recall correctly, and the seller was great - he took the gun apart and took pics of everything so we could establish which generation it was and if there were any obvious issues with the parts. The third one was the cheapest - it was gifted to me by a friend and fellow member here;-)

More OOOMPH - Part 2
I have been busy with real, paid work recently but finally found a bit of time to work a bit on the motor upgrade for the lathe - so not not really gun related...
But let me elaborate on that a bit. I still plan on getting the Mirage with 3D printed parts done, but the more I kept thinking about the nose cone warping in the sun, the more nervous I got. So, I started testing various materials and how they behave when heated and that work is still ongoing. Right now, it looks like I needn't have been that worried, but a few more tests and I will make up my mind on whether to stick to the filament I was using or trying something else.

Now, on the motor upgrade for the lathe. First of all, why am I doing this?
I am getting by alright with the brushed 500w motor that I have now - but the thing is, this lathe was designed to run that motor with a gearbox and mine doesn't have a gearbox... Yup, the manufacturer decided to save a few bucks and just leave it out. They still kept the face plate to remind me to not change the non-existent gears while the motor is running. So, there's that...


But why all this talk about a gearbox? Well, while the controller lets the brushed motor do RPMs below 50 (which is really low and great for threading) it has no torque at those low RPMs. I can stall it with a light grip on the chuck. So, basically it is worthless. I need to run it at much higher RPMs for it to not bog down and it makes threading frustrating and stressful. Perhaps, if it had come with a gearbox and thus, the low speed/higher torque option, I would have been fine.
But it didn't and that's why I wanted to try my luck with the brushless motor.

The brushless motor I chose is 750W and physically about the same size more or less. It's a tad shorter and instead of being 80mm in diameter as the brushed motor, it is 80x80mm squarish:


It's supposed to have more torque on the low end but I also wanted to lower the gearing as much as possible - even more than in the stock setup. Furthermore, the belt and pulleys used by the factory are a very rare standard. It's called 1.5 Module and is a square teeth with a pitch that is close to 5mm but not exactly. Those belts and pulleys are extremely hard to come by, so I decided to change over to a more common standard which is HDT5M.

First I had to do some measurements to see how much room there was for a bigger pulley on the spindle. I went practical with some rubber tape and a large o-ring:


It's the housing which is part of the gearbox for the leadscrew (not to be confused with the missing gearing for the spindle speed) which dictates how big I can go on the pulley spindle:


I found I could go quite a bit bigger and ordered a 37 teeth pulley whereas the original had 31 teeth (both pitches are close enough to 5mm that we can do an apples to apples comparison).
Some work had to be done on it, though and here I am just setting it up as close to concentric as my three-jaw chuck will let me. I liked the mounting boss(?) on the end for work holding, but it had to be turned down to fit in the inner jaws:



Once that was done, I could change out the jaws and flip the part. I blued the face and marked the approximate bore diameter I needed to get to - just so I didn't have to take measurements the whole time (my DRO has been pulled off the lathe for the time being, so we are back to manual measurements).


I had made a gauge of sorts out of Delrin which I knew was a perfect match for the spindle diameter, so once that could just be pushed into the bore, I knew I had hit the correct bore diameter:


After I parted off the pulley, I needed to cut a key way in it. With a bit of youtubing, I found that you could use a parting blade mounted axially on the toolpost and scrape your way there. My lathe is not that sturdy, but with the pulley being alu and me taking my time, it worked out surprisingly well:



Here, just for a comparison, I mounted the old and new pulley next to each other. You can clearly see the difference in teeth shape and of course also the difference in size - this is how I am getting a lower gearing out of this new setup:


Number wise, the old gearing on the brushed motor was 1:1.8 whereas on the new one, I will be getting 1:2.5. Not too bad an improvement, actually.
The brushless motor controller will only go as low as 200RPM on the motor, but with the gearing I will be at 80RPM on the lathe spindle. If the torque is there for threading, I will be a very happy camper. Even if I have to bump it a bit, it will be alright.
The way the old, brushed motor behaves when loaded is that it's controller just sends it a higher voltage (I think) to overcome the load, but it always overshoots and speeds up way too much - and then your threading runs too fast all of a sudden which is not cool if you are threading close to a shoulder (yes, you could also thread away from the shoulder, but that's another story).
On the other hand, the new, brushless motor has a hall sensor on its spindle so at all times, the controller knows how fast the motor is running. If it bogs down, I guess it also beefs up the voltage but it should happen in a much more controlled way - we will see once it's all mounted together.

Speaking of which, the motor needs a bracket. I wanted to have one fabricated in steel or alu, but then thought I would 3D print one first, just to get the measurements right. But I had this nice, new roll of carbon fiber reinforced nylon filament looking at me... So, for the fun of it, I beefed up the bracket design and will now have a go at making a printed working one. If it fails, I can still have a metal one made.
It is being printed as I write this:


As for the printer itself, I added a cable chain (can be seen behind the carriage) and two strips of LED lights (visible to the left in the pic). I think I will add a bit more LEDs - it's a relief being able to really see what's going on and not having to rely on a flashlight or a video light like I have until now. The LEDs were easy, they plug straight into the 12V out of the main power supply inside the printer and have sticky back tape on the strips.

The leadscrew gearbox is certainly there - that's how I can cut all the different thread pitches, I need. I don't use it as a power feed for normal cuts that often anymore. On bigger lathes than these, the leadscrew gears are easily selected by pulling levers or turning knobs but I have to change the gearwheels themselves by hand, unscrewing them and swapping them from a box of extras. So, these days, the gearbox is either set up for threading M1 or M1.25 as that is what I most often use for gun parts - and then for all the normal cuts, I just hand feed which works fine for me.
Down the line, I might make another modification I have seen online which is to add a motor to the leadscrew and use that for an independent powerfeed. The gearbox would be uncoupled, but it would still be there for coupling to the spindle for threading ops.

As for the "missing gearbox" on my lathe it normally sits between the motor and spindle and then the gearbox for the leadscrew gets powered from the spindle, so whether there's a spindle gearbox or not, the leadscrew gearing in relation to the spindle is the same (if that makes sense).

The mini-lathes generally come in two models power-wise, one with the spindle gearbox and one without. But the one without has a stronger brushless motor and it is this setup, I have copied (except for going with an even stronger motor and lower gearing). The version with gearbox is cheaper but runs off of a normal brushed motor (the gearbox is to make up for that motor's lack of torque at low speeds).
What my lathe has is kinda the worst of both worlds; no gearbox and the older, less torquey brushed motor.
But for USD 350, I guess they had to cut some corners somewhere. The new motor including the controller was USD 93 plus a bit for pulleys and belts, so I think it's a very worthwhile upgrade and it's ending up cheaper than had I bought the brushless version in the first place. If I had to do it again, I would probably buy the proper brushless version from the get go.

I just ran a quick test and so far the 3D printed motor bracket is holding up. Better yet, I can get sub-100 RPMs with a whole lot more torque. E.g. at 94 RPMs I tried stopping the chuck with my hands, which would have been very easy with the older setup, but I couldn't. It felt like it beefed up the torque when it sensed the load but the RPMs stayed completely constant so that's great news. It's also quieter but I was hoping for some pretty substantial motor braking (brushless motors can do that) but the controller seems to be programmed to a much milder one. It's supposedly healthier for the controller the less braking it has to do, so I might just keep it like this and anyways, I would have to send the controller back to the shop to have it reprogrammed if I wanted it changed.

@popgun pete I took a pic for you so you can see how these mini lathes look behind the scenes:

popgun pete said:

This is my lathe inherited from my late Dad who was a toolmaker earlier in his working life.
View attachment 53106
Here I was using the chuck jaws as a powerful clamp with the back drive simultaneously engaged to lock the chuck from rotating. Believe me it needed such treatment as the part was super hard to unscrew as powder paint "overspray" made the inner tube non-smooth and twin seals had to be dragged out! Plenty of oil prevented the "O" rings from being trashed.
View attachment 53107

Click to expand...

What a cool, old beast and great that it has some history.
My brother has one very much like it that he wants to renovate at some point. When I first saw it, I honestly thought it must have been given to him for free or he found it as scrap - I didn't know the first thing about lathes. But he had to pay a fair amount for it, nothing too crazy but not a steal either, as they are incredibly well made and people in the know are always looking for these smaller, beefier machines of which there aren't that many around.

I guess you also have a "gearbox" on the spindle - it looks like you can move over the drive belt on the spindle to get a different gearing?

I have used the lathe chuck as a clamp a few times, too and it certainly beats using a hammer, haha.
I might need to make myself a real spindle lock for my lathe at some point. Putting it in gear doesn't really lock it down tight enough on this lightweight lathe. But for now, it's OK it was just that when cutting the keyway, it would have helped.

More OOOMPH - Part 3
Here's a better pic of the 3D printed motor bracket. BTW, the walls are 6mm thick:


Yesterday, I realized I hadn't managed to think the design fully through as there was no way I could get to the two inner bolts on the motor flange - so for the first trial run, they weren't mounted. But then I remembered another trick, so I pulled the motor again and then locktite'd the bolts in place on the motor and left the glue to cure overnight and then mounted the motor again today:


There's a none of those grip-washers on the bolt (there's no room for a washer and I even had to cut the heads of the bolts down) but the Loctite did the trick as not only could I push the motor in place with the bolts staying put but I could also torque the nuts up quite well without the bolt breaking loose.

Here's a pic of the motor mounted properly with all four bolts in place. The old motor had those two alignment/support screws come down from the frame itself to snug up against the top of the motor and I kept those. Now, the new motor seems to sit firmly in place as well. I still don't actually know if the bracket will hold up or not:


As mentioned previously I wanted to go with biggest possible spindle pulley for a lower gearing but to check if I overdid it, I blued the inside of the gearbox housing in a few critical places:


I already knew that the lower end near the tumbler gear handle needs to be opened up a bit, but I was more interested in the two lugs at the top. They are threaded and it's where the bolts holding the plastic gearbox cover in place screws in.
After running the lathe a bit, I took the gearbox housing off again and I had rubbing on one of these lugs:


Nothing a Dremel can't fix with a few mins of work and it seems I did get the biggest possible pulley in there;-)

Next up is to fix this slight rubbing and then print a box for the controller board. I think it will be something super simple that I'll use for a little while, until I am happy with the layout of the switches and such before I design the final box.

I also have a fancy idea to use an inductive proximity switch as an electrical end stop when threading...;-). Well, I thought it was a great idea until I realized the motor controller doesn't do as powerful motor braking as I thought it would. That said, I just tested it again and at low revs, like the ones I will be threading at, I think the chuck stops spinning fast enough that it might still work. With a cutting tool into the part, it should stop even faster.
I have the switch and I have figured how it should be wired (that took a while...) but I am not sure I will mount it just yet. But it's a fun idea anyhow.

popgun pete said:

Yes, stopping the tool advancing on a thread cut takes some co-ordination to prevent the tool gouging the job at the thread base. On my lathe you wind the cross slide back and simultaneously hit the disengagement lever on the bottom of the saddle. A sharp whack on the mushroom head end and the lever immediately drops free.

Click to expand...

Threading on The Mini Lathe
There's a very good machinist on youtube showing how to run a lathe in reverse and thread away from the shoulder. He points out how even for a pro, it's a lot less stressful to do it that way as you don't have to worry about the timing to disengage the halfnut lever.

I tried this technique on my mini lathe but again the problem was that at low revs there's as much torque in it as in a rubber band. I have to bump the revs to about 300 for it to not bog down when taking a cut and that speed makes using the thread dial indicator a real nightmare. As you know, if you don't engage it correctly on every single pass, you pretty much ruin the part. Eventually, because of this, I removed the thread dial indicator completely.
Instead, I have been threading by keeping the halfnuts engaged the whole time, reversing the carriage back out by changing the direction of the motor. But disengaging timing is still an issue as it now has to happen by just turning off the lathe motor and hoping the whole thing stops before the cutting tool hits the shoulder (without a thread dial indicator, if I disengage the halfnut I am "screwed"...). It's a tedious process to say the least and I don't think I explained in detail that I have been threading this way, but this is why I started dreaming of a new motor.

If you are wondering why I haven't combined these two techniques - cutting away from the shoulder and never disengaging the halfnuts - again, the motor is to blame. Even with a relief cut between the shoulder and the end of the thread, there's not enough time for the motor to spin up to some revs that will give it enough torque not to bog down as soon as it hits the cut.

The good news is that judging from the very brief early test of the new setup it seems I finally have plenty of low speed torque. I actually think I will be able to thread at about 80-90rpms without any bogging down. This would be amazing and it would be slow enough to use the thread dial indicator with confidence. Even if I use the technique of just turning off the motor, keeping the halfnuts engaged and reversing the motor direction, it should still be a whole lot better.
But it would be cooler to have the thread dial back, so I need to redesign and print a new holder for the leadscrew cover on the apron to make space for it, but that's alright. I'll put that on the to do list right now;-)

More OOOMPH - Part 4
Been busy with work and other real life stuff lately but did find time to finally wire everything together on the new motor and controller.
I mentioned that for now, I just wanted to make a simple box until I figure out where exactly I want the controls to sit. I also decided to split the controller board and the switches into two different boxes, whereas on the mini lathes it's all in the same box on the front.

Now, on the front, I have a small box where I can set the speed, direction and turn the LED lights on and off. Also, I have an emergency switch there:


Overall, I decided to not run any high voltage into this front box at all. So, it's all 5V for the lathe controls and 12V for the LEDs. The small LED on the top is wired to the 5V from the controller board which is in a different box, so I can use that to verify that the control board is powered up if I have any issues.
But not having 220V on this panel means that the E-switch doesn't kill everything, it basically just works as on/off switch through the board - which is not ideal.

The box housing the controller board is as simple as this:


I am not sure where to mount it yet. I think I'll put it just under the lathe on the side of the black tool chest it is sitting on. There I would be able to reach the main on'/off button easily. And then the next version of this will have a proper E-switch wired to the mains in this box.

Honestly, the way the switches and controller works is less safe than the original. E.g. to switch direction on the old board, the direction switch would automatically trigger the e-stop. This meant that you couldn't just switch directions without resetting the power and stopping the lathe completely which was a nice safety touch. Also, the older motor started up a bit more gently, so you could leave it at high RPM and just hit the start button. Mine just wants to go to full power/speed right away which makes for some unneeded forces right from the get go.
So, the workaround is to turn the speed dial down before you restart it. I've already gotten used to it. But I might ask the vendor if he can program a soft start into the board.

On the plus side, the lathe runs smoother, is much quieter and I feel it has all the power I would ever need. I haven't really run it in anger yet, but did do a threading test yesterday and I could easily thread at 90-100rpm which is seriously slooow. Even at 150rpm, I have all the time in the world to disengage at the right spot and the motor doesn't bog down at all, doesn't drop any rpms at all. It's quite amazing, really. This was exactly what I was hoping for and it seems the upgrade has delivered

On Hold...
Not much to report on the Mirage 3D project and a few other things came up in terms of life, work and possible relocation, so it's been a bit hectic here. Also, I am running out of time before my next trip so I have decided to put this project on hold for now. That said, I am still pretty sure that this can be done and I will hopefully get back on it come spring.

I still need a gun for the next trip but instead of the 3D Mirage, I will hurry up and work on something much simpler but potentially still quite powerful.
Here's a teaser (I will probably do a thread on it as the project takes form):


P.S. (Esun CF-PA)
Now, a very specific "warning" to the few of you who have shown interest in the 3D materials. I got a sample of some carbon fiber filled filament from Esun and it was pretty cool stuff. But the full spool I have since ordered has been a dud. It absorbs quite a lot of moisture (I think) and goes incredibly soft within a few days. So, for now, I can certainly not recommend that at all. I will see if I can find time to print two identical pieces from the sample material and the bigger order and then talk to eSun about it. Maybe they effed up a batch...

Off Hold...
So, I found a bit of time to revisit this project - I guess Pete might have spurred it on;-)

When I left this off a while back, I had one working nose cone and had started working on the check valves in the bulkhead. When I first drew the check valves based on the ones in pump inlet valves, I had not made a step for the o-rings to sit on as I thought they would stay nicely in place from the pressure from the spring. But it turned out, that the airflow can blow the o-rings into the bore and skew them. (I talked about this all the way back in post 57).

Now, I have added the step for the o-ring to the bulkhead:


When I last worked on the check valves, I just printed some small test pieces as proof of concept, but now I have started printing some bulkheads.

One of the things I learned early on in this endeavor is that I need secondary machining on the lathe to make these 3D printed parts seal properly. But work holding became an issue. And for the bulkhead, the peg that the pumping barrel mounts onto was a real killer, so I have had to make that as a separate part but that's not a big deal, just takes a bit more time:


Here are some parts and while the insert peg can be seen in the "full" bulkhead at the top, I have since decided to split the bulkhead and print it in two parts. This will make access to some of the inner features much easier:


The square part is a jig I use when mounting the part in the four jaw chuck - it makes it a lot easier to line up the part. Without it, the part can easily turn in the jaws and I have to start the setup over.

To help set the part up in the four jaw chuck, I have also made a range of "mandrels" - one for pretty much each bore I have to machine. Here's a pic of the one I use for the main shooting barrel bore:


It's a really tight fit but the Delrin helps me being able to still spin the part when I tighten the jaws. I then use the tailstock to push the part into the chuck:


Finally, I indicate the part using the mandrel. I get to about +/-0.015mm pretty easily this way:


I used a cut off of a barrel to check the fit in the bore and I also added an o-ring with the o-ring seat design I intend to use. I have found that even with a nice chamfer it can sometimes be hard to insert the o-rings when you run as high compression as I ideal would like. This way, I can check on the spot:


I will use the same "slim design' at the nose cone where it is even more needed as I don't have much wall thickness between the pumping barrel and the shooting barrel up there. The o-ring is "only" 1.5mm thick as opposed to the 2-3mm ones often used.

(And yes, my four jaw chuck is just held in the three jaw chuck. I still like the three jaw chuck for the vast majority and it's just too much of a hassle making it run true if I had to put it take it off and put it back on every time I had a need for the four jaw;-))

I mentioned that I had decided to split the bulkhead into two parts. The main reason is that I needed to chamfer the inside of the air transfer bore and it is hard to get to and see what I am doing if I need to make that cut deep into the full length bulkhead. On the split part, it is much more accessible:


There's not a lot of room - or room for mistakes - when making that cut but it seemed to work out really well. I have tried sliding the power regulator in and out and it's quite smooth.
As for gluing the two parts together I don't think it's much a risk. There's mostly compression acting on those two parts once the gun is assembled and even if the glue seam leaks, it's not a problem as the seam is rear of the main o-ring sealing the bulkhead against the reservoir.

The tool of the day was a close competition between the chamfer drill bit I ground down to work for the internal chamfer but the winner has to be my cheap rubber hammer:

The fit in the jig is so tight that this hammer has been a life saver

Honestly, making the check valves internally like this has made it all a lot more difficult. I hope it will all work, but I am not 100% confident yet. I think the basic design is simple and should work but if I have just one spring get caught or one o-ring not sit perfectly, then it will fail. We will have to wait a few more days to see if it works out;-)