Member postings for Andrew Johnston

Standard endmills and slotdrills are not flat on the end. They're slightly hollow ground, so you end up with a barrel shaped crankpin not a cylindrical one. The problem can be ameliorated, but not eliminated, by offsetting the cutter.

Andrew

In theory there is some resistance to upward movement of the work due to friction. But it would be best to orient the work so that it sits without rocking before clamping.

Andrew

Forget parallels, flycutters and machine vices. A tolerance of +/-0.02mm is doable but not simple. First I'd ensure that the material is stress relieved. We don't know what the material is, so It's diffcult to suggest temperatures and times.

I'd the put the material on the milling machine table against a thin sheet as a stop. Then hold the material using side clamps with eccentric screws. This is what I mean by side clamps, the small gold hexagons:

If the clamps are too thick use a piece of thinner gauge plate as a filler. Don't tap the material down, just let it sit in it's natural position. Then I'd take a series of shallow cuts with an end mill, say 10mm diameter. It may not look pretty but if the mill is trammed properly it'll be flatter than using a flycutter. Once one side is machined turn the material over, clamp and machine the second side.

When clamping the material you need to be clean room obsessive to ensure there is no dirt, swarf or dust between the material and table. Brush it and then run over with a kitchen towel soaked in acetone or similar. It would be a good idea to stone the table before starting to remove any small burrs.

Of course the above assumes that the top of the table is parallel to the dovetails in both X and Y to better than the tolerance wanted. If not then all bets are off.

Andrew

How flat and how parallel? There's no point in discussing possible methods until we know what is acceptable.

Andrew

The 300 series stainless steels bend cold without a problem, if the fold is straight. On a curve is more complex. As the flange is formed the metal needs to compress, or more likely stretch, to accomodate the surplus metal. Of course there are methods involving spinning or forming dies that can be done cold. But to do it by hand would need heat. Here's a flange hot formed in 3mm cold rolled mild steel to give an idea. Note the increased depth of flange on the curve:

However, one needs to control the temperature. When hot stainless steel tends to absorb oxygen from the air. To some extent the chromium reacts with the oxygen leaving a scale that slows the absorption. Another issue is that if one melts the stainless steel it forms a black crystallised "sugar". When TIG welding it is normal to back purge the weld with argon to prevent this.

Andrew

We could add:

Injectors

Fast Fourier transform

Star versus delta connection

Bridgeport mills

Kurt vices

All topics that have lead to me being insulted over the years.

Andrew

Good grief! Is nothing sacrosanct? I'd put you in the always worth reading category, even if I don't always agree with you.

Andrew

Well, if you're a baddy I must be off the scale, but I don't care.

Andrew

When I used my Clarkson tap grinder I think I set the tap orientation by eye so that the cutting edge was more or less on the horizontal plane along with the tap axis. I don't think it's that critical. Not a great picture but this shows the general arrangement:

Andrew

Possibly less stringent testing for the requirement to fail open, but it would seem unlikely to design for failure to be a short. What happens if the shorted capacitor doesn't blow the fuse? You burn to death rather than get a shock.

Andrew

It would seem odd to have Y-capacitor suppressors when there's no earth connection?

Andrew

None of the T-slots on my three milling machines are painted and I have no issues with rust. On two of the mills I use flood coolant most of the time and on the third I give an occasional spray of WD40.

I use gauge plate and close fitting cylinders in the T-slots for alignment so it'd be a right royal PITA if they were painted.

Andrew

Toolpaths are selected by the programmer, but the program sorts out the movement details. The programmer also selects speeds, feeds, DOC and WOC. Although of course tool manufacturers give guidelines. I've never found the online speeds and feeds calculators very useful, I prefer to use a mix of experience and calculation.

When I started using my CNC mill (same horsepower as my Bridfgeport) I eventually learnt that it was happier running smaller cutters at higher speeds and feeds to achieve the same metal removal rate. I rarely use a cutter bigger than 10mm on the CNC mill in contrast to the Bridgeport.

Using CNC is a slightly different mindset. I run mine at much faster spindle speeds than I run the Bridgeport and I often run a roughing pass and a finish pass with different cutters. Another useful feature of the more advanced CAM programs is the ability to re-machine. Let''s say we need a recess 70mm by 50mm by 10mm deep with 1.8mm radii in the corner. The 'simple' approach would be to machine the whole recess with a 3mm cutter. But using re-machining one can machine the whole recess with, say, a 10mm cutter. Then one swaps to a 3mm cutter and the CAM program knows that only the radii in the corners need tidying up.

Andrew

I keep my part built engines (which are mostly steel and cast iron) in the house.

Andrew

NAWCC

**LINK**

Andrew

It's not just temperature but the amount of heat. I'd guess that one is 250W so plenty big enough! I've used one that size once to solder copper busbars, carrying several hundred amps, in an inverter.

I've recently acquired it's baby brother, a Weller 100W iron from an ex-colleague who was closing down his acoustics and vibration consultancy.

Andrew

Simply eighths, sixteenths, thirty seconds and sixty fourths in decimal format. No idea what 4 is, but probably off some sort of scientific instrument.

Andrew

Like many things in engineering desoldering is counter-intuitive.

To start with use a proper soldering iron at a good temperature. My Weller iron is 80W and I run at 360°C. Forget solder suckers and braid for desoldering. It doesn't look like the switch leads can be cut to start with. First add solder in two beads down both sides, each covering three pins. Then melt one bead and push on the switch, it should tilt slightly. By melting alternating sides you can work the switch out far enough to get one side completely out, or at least cut the leads on one side. A high wattage iron helps as you want to go in, melt the solder and be out again in a few seconds. A low power iron that takes ages to heat up is useless. Although not relevant in this case a multilayer board with internal planes will need even more heat to melt the solder quickly. Once the leads are seperated it is simple to remove them one by one.

After the switches are removed you can clean up the board. Solder braid is great for surface mount pads, but isn't that useful for thru hole. To clean up with a solder sucker add solder until the hole is full. Then melt the solder and suck! Solder suckers are ineffective at cleaning partially filled holes as it isn't always easy to get the solder properly melted. Filling the hole first makes it easy to melt all the solder.

Note to SoD: Come on! Meet the challenge and get those SM devices soldered. For ICs down to around 0.05" pitch it's simple to solder individual pins. Below that we need to be counter-intuitive again. I use a liquid flux and add a big blob of solder covering several pins. If you then melt the blob of solder and slowly move the iron along the row of pins the solder will follow and amazingly it will leave the pins soldered while surface tension ensures there is no solder bridging between pins.

Andrew

A single word will do - pettifogging.

Now back in vogue having recently been used in the US Supreme Court.

Andrew

I drilled four 2mm holes 2mm apart with a 4-facet carbide stub drill. Then joined the dots with a coarse round needle file and finally cleaned up the scalloped sides with a fine flat needle file.

Andrew