Member postings for Nealeb

In the interests of scientific enquiry, I had a careful look at my own lathe under normal working conditions last night. I have LED strip lights fed off a cheap smps nominally designed for the job, plus the lathe worklight which is a cheap eBay car spotlight in the original worklight shade fed off a repurposed "wall wart" 12V PSU rescued from a defunct internet router. My lathe has a mechanical continuously variable speed arrangement; over the range 300-1500rpm I could see no stroboscopic effects. And even wearing my Bluetooth headphones/ear defenders with the radio on, there's no way I would not hear it running!

So, whatever the theoretical risks, I'm happy that they do not apply in my case. I'm more likely to do something a mischief by turning the wrong handle...

Edited By Nealeb on 26/08/2023 08:00:57

Certainly not arguing that stroboscopic effects do not exist; although it's a long time since I had occasion to use one, the strobe lamp I remember used a xenon discharge tube (I think) that gave well-defined pulses of light directly from the discharge rather than via a phosphor which must at least soften the rise and fall times of the light pulses.

I am querying whether, in practice, in a home workshop, it's a significant effect. After all, not many of us are going to install a three-phase inverter just to run a set of fluorescents! And definitely not when a set of replacement LEDs would do the job at lower cost. Rather than install "LED tubes", I removed my workshop's fluorescents altogether and installed low-profile LED strips. Like most commercial LED lighting, I suspect, these are fed from a cheap SMPS that is not likely to produce any meaningful flicker. I accept that those who reuse a surplus 12V or 24V transformer plus bridge rectifier might generate some 100Hz flicker but the SMPS boxes intended to drive LEDs are probably a lot cheaper than buying a transformer and get over any worries about overvoltage with or without smoothing caps. Even with my 12V LED lathe light replacement, the cheapest PSU was a repurposed "wall wart" which might not be the cleanest DC but must be better than transformer+rectifier.

Is all this worrying about strobe effects justified? Apart from the neon light illuminating the strobe marks around the outside of a record turntable and a specialised strobe gun, I can't say that I have ever seen this in practice. It has to be a neon as this is one of the few lighting technologies easily available that does not have any significant "inertia" in the light output. Well, maybe LEDs driven off rectified but not smoothed AC, but even then I'm not sure how powerful an effect it is. Filament bulbs surely have so much thermal inertia in the filament that they can't flicker to any meaningful extent.

Even fluorescents do not give out light directly from the internal discharge which presumably is at mains frequency but stimulate a phosphor which glows - and I'm pretty sure that the phosphor will continue to glow for the few miiliseconds between excitations. Certainly when my workshop overhead lighting was fluorescent I never came close to seeing stroboscopic effects on my lathe chuck. That wasn't the reason I swapped over to LED striplights fed by DC from a SMPS.

Is the whole thing really a bit of an old wives' tale?

...but it says "cosmetic wear only"...

I needed to replace the 48V light on my lathe - not so easy to find bulbs for it these days, and in any case as I was going to run the machine off a 240-415V inverter anyway the light would only have come on with the motor. I bought a couple of small 12V LED spotlights intended for car bumper mounting. I was able to modify the mountings to fit one of them in the existing shade and reuse the wiring and the support arms. I then removed the 415V-48V transformer from the LoVoLight switchbox and replaced it with a surplus 12V power supply that I think had run an old broadband modem/router. Adequate power for the LED. Gives a decent amount of light. I plan to mount the other one (I think they were about £15 for the pair) to do a similar job on my mill. Cheap but effective.

I also mounted the light on the splash guard - could you do something similar with yours? Looks like you already have a small clip-on light on the splash guard?

Edited By Nealeb on 18/08/2023 22:31:20

A couple of other factors:

- start/stop - if you are frequently starting and stopping the machine, 3-phase is generally better than single-phase.

- speed control - possible with 3-phase but not single-phase.

My vertical mill has a motor with electronic speed control of something like 200-10000 rpm, nominal 2HP, which I believe was designed for washing machine use. It has one useful thing for heavily-loaded slow speed motors - a separate cooling fan. Even with 3-phase motors with a VFD for low-speed control, a built-in fan might not be enough at those lower speeds.

I seem to remember a children's programme on television years back where they demonstrated this powder-based explosion problem by blowing a puff of ordinary cooking flour into a transparent vessel continuing a lit candle. Very much a "don't try this at home" experiment - wonder if that sort of thing would be allowed on television these days?

All of this talk of explosive custard is a wicked urban myth spread by militant "ice cream at all costs" North American dessert eaters. I can confidently say that in all my years of making custard, I have never had a pan of it do so much as burst into flames even when cooking on a gas ring.

Now, the old "treacle tin" experiment in the chemistry lab was quite a different story...

One reason the 2D to 3D conversion is tricky is that you have to remember that in 2D typically we are preparing engineering/manufacturing drawings; in 3D a sketch is representing some design aspect without any regard for manufacturing. Yes, it has to be "manufacturable" but we don't do things like always dimension from a common reference and we do use constraints like "align with centre" and "make symmetrical".

One of these differences alluded to in the last few posts is an example. When creating a sketch, I often do not try to create a feature exactly where I want it. I create it close to the required position and then add constraints to get it into place and make it stay there. For example, this evening I was drawing a rectangle with a hole in the centre. I drew the rectangle roughly the right size, then edited the dimensions of the sides to make it exact. Then I drew a circle somewhere near the centre, added a dimension to specify the diameter, then applied constraints to align its centre with the centres of each side. My sketches are a bit back-of-envelope while I'm creating them, then I apply constraints (where I can) and dimensions (where I have to) to get it exact. It's a different approach to the way I drew in TurboCAD, for example. In Nigel's example, I would have drawn my line somewhere near the circle, then used a tangent constraint to force it into place. Then a coincident constraint to lock the other end of the line to the line representing the end of the crosshead, and a dimension to get it the right distance from the edge of the crosshead. Quicker to do than to write! But it does need a shift in approach compared to drawing-board techniques.

These village show competitions started going downhill when they allowed entrants in the baking competitions to use digital scales instead of weighing everything by eye, like proper cooks.

I was wondering about the technology that is in Voyager. Launched in '77, it looks to me as if the latest microprocessor around would have been the Intel 8080 so the processing power on board can't have been anything like as high as we take for granted today. I think the 8080 had a 2MHz clock speed. The 8086 16-bit chip was launched about the same time as Voyager...

Around that time I was using 74 series TTL for my degree project and discrete transistors were still in widespread use - FETs were fairly exotic, I seem to remember. Maybe the larger semiconductor geometry makes it more reliable, in fact, in regions of space subject to cosmic rays? A pretty good effort, whatever it's using!

Andrew - I didn't mention the "create a new part in the context of an assembly" mainly to simplify what was turning into a full-on essay rather than a quick pointer! I think it has been adequately covered in subsequent posts but I realise that it might not be obvious as a technique. I use it a lot. Clearly, once you have that separate part in its own file you can create 2D drawings in the usual way which will include all the dimensions you choose including those "copied" from the parent assembly. As I often go straight to CNC, I often don't bother with drawings unless it's a turning job. Keep thinking about CNC lathe conversion...

I didn't want to add precise details anyway as I'm not an Alibre user - mainly design in Solid Edge and move to F360 for CAM toolpath generation. I think all the general ideas are common to them all but the details differ.

A friend of mine has one of these, although I think it might be the flat rather than 3D version. However, he is rather a fine clockmaker and uses it for crossing out clock wheels rather than wood, which is perhaps a little more demanding. His comment is that it is a bit of a luxury, but it is very nice to use and does allow rather more blade tension than a more conventional frame.

On the whole, I subscribe to the argument that paying a bit too much for a tool that you will use a lot is better than paying too little - although there must be a limit to "too much"!

I think that you might be missing one of the great features of the kind of 3D parametric CAD we are talking about here. You are in excellent company - I have done a certain amount of tutoring in 3D CAD and have always had the biggest problem working with very experienced draughtsmen/designers coming from a 2D background. There are certain principles you learn in 2D that are better forgotten when working in 3D. One of the mantras that I tell people is "never use a dimension if you can use a constraint." Like all such things, it is not an absolute hard-and-fast rule but it's always worth considering.

In this case (and I'm simplifying the geometry a bit) let's say you want two gears that the supplier's catalogue gives as 6.35mm thick. You want a shaft through them that sticks out by 10mm either side. So, I would model the two gears (as plain cylinders based on their pitch circle, quite possibly) and assemble them in their correct relative positions. For the shaft, I draw a circle on the outer face of one gear, then extrude that by the design dimension of 10mm from that face. I then extrude the same circle in the opposite direction using a constraint so that it finishes 10mm from the outer face of the other gear. Net result - a shaft that exactly matches my design goal. At no point do I calculate the length of the shaft - I shall find that later from the manufacturing drawing that my software produces. A variation might be that the gears sit on a shaft that runs between bearings in a fixed position, and you need a spacer to suit. In this case, the spacer is constrained to have one face on the face of one gear, and the other to be some clearance dimension short of the face of the bearing. I never need to actually calculate the spacer length.

Why do it this way? Couple of reasons: one is that coupling individual parts together in this way is likely to lead to fewer drawing errors. If you have tied the parts together like this then you can't accidentally draw them with different dimensions. A second reason is that by implementing the design in terms of "design goals" rather than calculating dependent dimensions and hard-coding them into the model, it is much easier when things change. In this case, what if your supplier actually says, "Sorry - no 6.35mm thick gears; they are now 6mm." You go to your drawing, change two dimensions (gear thicknesses) and everything that depends on these dimensions automatically gets updated. This actually did happen to me when I was designing my CNC router. A hole spacing in one bracket depended on the combined thickness of several other components. I had designed this around the 50x12mm bar I had ordered, only to be told that the supplier only had imperial dimension bar - 2"x1/2". So I changed the bar thickness from 12mm to 1/2" in the model and reprinted the bracket drawing. Job done. Actually, one of the biggest gripes I had with TurboCAD which I used to use regularly was that it was so difficult to make these kinds of changes without scrapping the whole drawing and starting again. 3D CAD came as such a breath of fresh air.

So, in response to your comment - not only can you draw parts of your assembly without knowing in advance their dimensions, but very often you should!

Edited By Nealeb on 31/07/2023 22:13:10

Is it possible to look at the machine parameters to check? Maybe compare mill and turn mode settings?

Quite plausible that there are limit switches at both ends of travel, though, or even one switch with two triggers. My own mill only has switches at one end for homing and I rely on soft limits at the other end.

CNC control software includes a set of machine-dependent parameters that define things like "steps per mm" for the different axes, max speeds and acceleration, etc. One parameter is direction to move on homing. Don't know which generation of Tormach control software you are using or whether these parameters are easily accessible or locked away in some hidden area but that's where I would be looking. I presume that you load a different set of parameters for milling or lathe modes?

I have seen comments here on different bolt heads due to access restrictions, but I don't think I have seen any informed comment on the torque-handling capabilities of the different head forms. Particularly important when you are taking the d****d things out when you have to fight both the original tighening torque plus the corrosion that's occurred in the years since it was put in in the factory...

I know that security is one reason for the odd heads you sometimes find. Sometimes misplaced ideas of security. I spent some time working in a pretty secure Government establishment. To help with resetting forgotten passwords (fairly common - their technical people were pretty switched-on but the staff they supported dwelt on a planet slightly to one side of our own) there was a "super-secure" PC in a custom cabinet with some kind of specially trusted network connection. To reduce the chance of hardware tampering, there was a clear polycarbonate cover so you could see all the components and which was secured with "security bolts". The guy in charge was very proud of this. I had to tell him that I had bought a set of bits off eBay the previous weekend for a tenner that included the bits necessary to undo his security bolts.

As memory slowly oozes back into my brain, I recollect that actually, some years ago, I solved a problem with lime scale in the boiler frequently clogging a pressure relief valve in this machine by buying a piece of fine brass gauze from one of our usual ME suppliers and making a sleeve to fit over the internal exit pipe from the boiler. Fixed that problem nicely - but it seems that I wasn't worrying too much about brass in the filter then. In any case, given that I seldom use the thing to generate steam, it only gets to around 80C, less than those Turkish coffee-boiling cups.

What I have subsequently found on moving from a very hard to a very soft water area is that I now no longer get lime scale problems. Instead, there seems to be steady corrosion of the inside of the aluminium boiler which I am sure will go pop one day. There is now a slow build-up of, presumably, something like aluminium oxide in the boiler (which is being retained there by my filter, fortunately). On the plus side, being aluminium, my club boiler inspector is not qualified to look at it and therefore can't condemn it!

Thanks for the comments. Yes, PEEK would be ideal - but I have none (used my small stock making a hot end for my first 3D printer!) and it's not cheap, would need to wait for online order to arrive, and I want my coffee machine back working again asap! It's on the limit for ABS, looking at Wikipedia data, but I suspect that the original was indeed ABS and, as the Wikipedia article suggested, that it became so brittle after maybe 10 years or more that it shattered into small fragments when hit by a sudden stream of coffee once the dam broke above it. Just picking up some fragments to try to take measurements showed me just how brittle.

I've just spent a little time making a replacement from an odd stub of brass which seems to work fine. I have had to guess dimensions and precise shape from various online outfits that sell this component as a spare part but I doubt that any of it is super-critical. I have also done what I should have done in the first place which is a quick Google search on brass and coffee. Seems that there are some cultures in which coffee is actually brewed in brassware, and there is even a suggestion that the zinc that leaches out in tiny quantities is good for your health. Well, maybe - I was more concerned about the copper leaching out and leading to heavy metal poisoning!

Spent some time wondering where this question should go but as it will probably turn into a debate on how to grind beans, then "Tearoom" seemed appropriate...

I have a Gaggia Classic espresso machine, much used. The ground coffee goes into an aluminium cup with a double-layer base with a small gap between the two. The inner base has many fine perforations; the lower has one slightly larger hole. I gather this is to do with the production of "crema" - froth, to the rest of us. This cup sits in the portafilter (the removable bit with a handle) which has a double-ended spout to distribute the coffee into one or two cups. Inside the portafilter is a small plastic plug with a blind hole from the top and a cross-hole near the bottom. Apparently this helps to control coffee flow. It is apparently called the jet swirl emulsifier.

From time to time, the single hole in the cup blocks. I let it sit with some special cleaning powder and run the pump occasionally until, usually, the blockage clears. Today, it did clear, almost explosively, and coffee spayed enthusiastically out of the spout. Turned out that for whatever reason, the plastic plug thing broke into many pieces. Likely to cost best part of a fiver to replace plus postage. Can't 3D print in PLA due to temperature (80C or so). The only plastic rod in the workshop is acetal which is not suited for that temperature either. So, plan is to use an odd scrap of brass - fairly simple bit of turning and cross-drilling.

And so - the question (at last, I hear you say). Anyone have any idea of compatibility (taste, corrosion, etc) between hot coffee and brass? I can't think of any examples of seeing brass and coffee in contact; brass components on the coffee machine all seem to be plated. Thoughts?

Edited By Nealeb on 25/07/2023 15:36:50

Edited By Nealeb on 25/07/2023 15:37:57