JSR's workshop chronicles

[jackssmirkingrevenge]

double post

[jackssmirkingrevenge]

Here's my latest toy. Making a 1000x200 mm surface plate, needed some precision measurement

Neat!

Still some way away from that level (pun?), I'm here impressed by automatic feed... look at it go!



I think I've machined more steel in one evening than I did with my Sherline since I owned it

[mrfoo]

I kinda wish mine had automatic feed and thread cutting :)

How's accuracy? Have you done all the alignment checks? If you don't have a machinist's level to level the ways, it's worth doing a quick eccentricity / alignment check using "Rollie's Dad's Method", i.e:

Take a reasonably long piece of rigid round stock, a very good candidate is a suspension strut piston or similar, and chuck it up. Don't worry about the fact you're in a 3 jaw and it will be eccentric, it doesn't matter. Mount a dial indicator on your cross slide, at centre height, and move the carriage to the far end of the rod. Bring the indicator into contact with the rod, lock off the cross slide. now rotate the chuck by hand, and note the furthest and nearest readings on the indicator. Take an average of these, and you know where the axis of rotation is (relative to an arbitrary point on your cross slide, obviously). Now, without unlocking the cross slide or moving the indicator, move the carriage to the nearest point to the headstock, and take the same readings. In a perfect world, with everything set up correctly, the headstock end and tailstock end averages will be identical. They won't be, though - there's some slop in the headstock bearings which will allow the rod to sag, and the rod itself will bend under its own weight. If you're within 0.01mm or so, though, you're as close to perfect as you're likely to get.

[jackssmirkingrevenge]

How's accuracy? Have you done all the alignment checks? If you don't have a machinist's level to level the ways, it's worth doing a quick eccentricity / alignment check using "Rollie's Dad's Method"

I still don't have a dial indicator... but I did play with some test cuts and everything seems to be within acceptable limits.

Just finished the mill DROs:



Shopped for a bit more tooling for the lathe too, it's clear that I'll be needing a quick change tool post, plus some tooling for thread cutting and boring.

[mrfoo]

I'd suggest you need metrology stuff more than you need a QCTP. At the very least a plunger type comparator and a finger type DTI, plus a decent mag base. The indicators you can get new from SHAHE on AliExpress, but avoid chinese mag bases like the plague - get yourself a good quality second hand one. While you're shopping for metrology, you definitely want a set of small bore gauges, and a set of telescoping bore gauges (that's pretty cheap), and eventually a set of micrometers (mint condition, good quality second hand micrometers in the sub-100mm range run about 25 euros each over here). Your digital guesstimators might *indicate* a precision of 0.01mm, but they are only actually *accurate* to around 0.05 at best (and that's taken from the Mitutoyo documentation, chinesium may well vary). I'm not even sure how accurate DROs are that use digital caliper technology.

I'm not really a fan of QCTPs. Most of your turning can be done with a very limited set of tools, you need a facing tool, a turning tool, a cutoff tool if you're feeling brave, and probably a threading tool. That's 4, and your toolpost has 4 stations. As you need to mount a cutoff blade in a toolholder, you can make up a boring bar holder (or multiple boring bar holders) and a grooving tool holder that can take its place, if you make these to sit at centre height you won't need to shim them so swapping them is relatively painless.

Personally, I do all my work with HSS, but that requires grinding your own tools. For making small o-ring grooves, for example, you're almost always going to need to grind your own tooling, you won't find carbide thats exactly the right size..

If you're going to buy a QCTP, don't even /think/ about the super cheap ones. They're made of aluminium, but might as well be made of dogshit. Prefer a wedge style over a piston style, or splurge on a multifix clone. If you're not spending over 300$ for a toolpost plus a few holders, you're buying something less solid and capable than a decent 4-station holder. And remember, a QCTP means having toolholders, they run to 40 or 50 bucks a piece (although you have a mill, so you could probably make them yourself)

[jackssmirkingrevenge]

If you're going to buy a QCTP, don't even /think/ about the super cheap ones. They're made of aluminium, but might as well be made of dogshit. Prefer a wedge style over a piston style, or splurge on a multifix clone. If you're not spending over 300$ for a toolpost plus a few holders, you're buying something less solid and capable than a decent 4-station holder. And remember, a QCTP means having toolholders, they run to 40 or 50 bucks a piece (although you have a mill, so you could probably make them yourself)

Oooh you're not going to be happy with my recent purchases then...

QCTP

spare holders

We'll see how that pans out.

Your digital guesstimators might *indicate* a precision of 0.01mm, but they are only actually *accurate* to around 0.05 at best (and that's taken from the Mitutoyo documentation, chinesium may well vary). I'm not even sure how accurate DROs are that use digital caliper technology.

This is a very fair point that I was aware of, my reasoning is that for what I'm doing I think it will be good enough. I have no ambitions of series production or making parts with +/- 0.01mm tolerance.

I am treating this tooling as basically a larger and more powerful version of the Sherline and so far it seems to be good enough.

[mrfoo]

Hey, that's not totally cheap shit throwaway stuff, it'll probably be more than OK. Wedge style is a good call. Watch out for chatter, a QCTP brings the tool out away from the topslide, effectively it's hanging out in the void, any slop in your gibs will show itself here, particularly if you're doing cutoffs, grooving, or other form tools. You can help this by having the absolute minimum tool stick-out you can possibly get away with (i.e. not what your steel turning gif shows *grin* ). A lot of people replace the topslide with a solid block (except when thread cutting) for exactly this reason - less gibs = less chatter.

If you find you're having trouble getting good surface finish, your feeds and speeds are right and your tooling is in good nick and correctly set up, this could be a contributing factor.

[Moonbogg]

From the perspective of a skilled tradesman, I can see how one would prefer quality equipment. But from the perspective of a hobbyist like us, even 0.1mm precision would be fine for anything we do. 0.1mm is about .004", and a standard machined tolerance for most parts is +/- .005". You could just machine the parts to the same dimension and shave off a bit at a time from one part until they fit to your liking. As long as the o-rings seal, the darn thing will work. The benefit of making parts yourself is you can just tweak them until they fit without caring much what the final dimensions are. If making drawings for someone else to make your parts, then of course you need to watch the tolerances and make sure everything will pan out as long as they do their part.

[mrfoo]

It is certainly true that, where what you're making doesn't have to interface with some other precision part, you don't ultimately care about exact dimensions.

However, you do care about fit, and that requires measurement. Take the example of a sliding fit, sub 10mm piston in bore, not uncommon in what we're making here. To get a good sliding fit that isn't going to jam or leak, you need to be bang on the money with your measurements. "skim a bit off, see if it fits, lather, rinse, repeat" will leave you oversize pretty much every time. Ask me how I know that :)

[Moonbogg]

Keep in mind I'm the guy who just puts numbers on paper and then expects everything to somehow work out in the shop. If I tried to machine my last cannon myself, regardless of equipment used, there'd be enough aluminum scrap to form an extinction level-sized asteroid.

[jackssmirkingrevenge]

As long as the o-rings seal, the darn thing will work. The benefit of making parts yourself is you can just tweak them until they fit without caring much what the final dimensions are. If making drawings for someone else to make your parts, then of course you need to watch the tolerances and make sure everything will pan out as long as they do their part.

Nail on the head, there is a massive difference between someone machining parts according to a drawing to be sent across the country to interface with other parts made by someone else in the same boat, and someone working on their own project in one room. Of course you still need to be able to measure with reasonable accuracy, but you are much more flexible as an individual hobbyist. Accuracy costs time and money, so chasing more than you need in practice is counterproductive.

Hey, that's not totally cheap shit throwaway stuff, it'll probably be more than OK. Wedge style is a good call.

Watch out for chatter, a QCTP brings the tool out away from the topslide, effectively it's hanging out in the void, any slop in your gibs will show itself here, particularly if you're doing cutoffs, grooving, or other form tools. You can help this by having the absolute minimum tool stick-out you can possibly get away with (i.e. not what your steel turning gif shows *grin* ). A lot of people replace the topslide with a solid block (except when thread cutting) for exactly this reason - less gibs = less chatter.

The mostly aluminum alloy Sherline has taught me that quite well, but with this new bit of kit I am amazed with how much I can get away with.

Now to build something

[Moonbogg]

As long as the o-rings seal, the darn thing will work. The benefit of making parts yourself is you can just tweak them until they fit without caring much what the final dimensions are. If making drawings for someone else to make your parts, then of course you need to watch the tolerances and make sure everything will pan out as long as they do their part.

Nail on the head, there is a massive difference between someone machining parts according to a drawing to be sent across the country to interface with other parts made by someone else in the same boat, and someone working on their own project in one room. Of course you still need to be able to measure with reasonable accuracy, but you are much more flexible as an individual hobbyist. Accuracy costs time and money, so chasing more than you need in practice is counterproductive.

Hey, that's not totally cheap shit throwaway stuff, it'll probably be more than OK. Wedge style is a good call.

Watch out for chatter, a QCTP brings the tool out away from the topslide, effectively it's hanging out in the void, any slop in your gibs will show itself here, particularly if you're doing cutoffs, grooving, or other form tools. You can help this by having the absolute minimum tool stick-out you can possibly get away with (i.e. not what your steel turning gif shows *grin* ). A lot of people replace the topslide with a solid block (except when thread cutting) for exactly this reason - less gibs = less chatter.

The mostly aluminum alloy Sherline has taught me that quite well, but with this new bit of kit I am amazed with how much I can get away with.

Now to build something

In most cases, chasing more accuracy than you need will probably get you fired as well, lol. Cost is king and time is money. Some industrial repair shops don't have any digital readouts or fancy equipment. The nature of the work doesn't require it, and buying it is literally a waste of money. Some fancy machine shops have nothing but the best equipment because they have to consistently hit tolerances within +/- .0005 or even tighter on high-RPM cylindrical parts. Shooting a potato though? Lol, yeah, it don't matter. My drawings had to be specified thought because I couldn't write them a note saying, "Just make all this stuff fit and work properly, please"

[jackssmirkingrevenge]

My drawings had to be specified thought because I couldn't write them a note saying, "Just make all this stuff fit and work properly, please"

I think this typical JSR working drawing for a Crosman valve illustrates the point quite well D

[Moonbogg]

That's actually pretty good. It actually looks exactly like many sketches I've seen over the years in machine shops. A lot of places work off of sketches just like that. The information content is what matters, and as long as they know what to do, then the drawing is just fine.

[jackssmirkingrevenge]

Quick Change Tool Post installed:



A very solid bit of kit for the price, much better than I expected.

A little taller than I expected so I milled out a 5mm deep square on the compound slide to give a bit more distance before the holder bottoms out and to keep it rigid.

Most setups seem to be free to rotate but I figured I'd want it square with the chuck most of the time and if I needed to turn it I could always turn the whole slide.