Member postings for Andrew Johnston

In 20 odd years of having my current centre lathe I've made very little tooling; mainly haven't had the need. What I have made would be irrelevant to the OP anyway. I've bought plenty of tooling, often secondhand, and also made many fixtures that get used for the job in hand and then put aside, or re-purposed.

For the mills I have made simple tooling that get used on a regular basis, but that wasn't the original question.

Andrew

A very simplistic calculation says the battery is better per unit mass. But in reality that's nonsense since I don't believe the battery is anywhere near 900Ah.

Battery chemistries have made big strides over the last few years, but the fundamental energy capacity hasn't kept pace. Lithium batteries compare well on Wh/kg since lithium is so light. A lot of the work on batteries has gone into reliability and charge/discharge cycles. It's no good having a sooper dooper electric vehicle if the battery dies after a few hundred cycles, or even several thousand.

Andrew

A carbide cutter is a must. General rule of thumb is run at the same speeds as ordinary steel, but with small DOC and high feedrate. Ideally the shear zone needs to be red hot. Just to prove it works here's a toolbit being milled:

In this particular case the milling was to achieve an accurate involute profile rather than bulk material removal, but the principle is the same.

Andrew

There's no need to waste time grinding to shape. If a lot of material needs to be removed (like the tool shown) use a mill to shape and then grind to add reliefs.

Andrew

That's what happens when formulae are quoted without understanding how they are derived. The formula given relates to undercutting when a gear is hobbed. The undercutting is a consequence of the hobbing, not a neccessity for the resulting gears to mate.

I have made 13 tooth 5DP pinions with an involute cutter that mesh properly without needing undercutting. On the other hand these 6DP bevel gears are 10 teeth, and do need an undercut:

In both cases the gears were 20 degrees PA.

Andrew

m = micro = marketing b*****ks

The spec, for what it's worth, states the voltage is 5V.

Andrew

That's interesting, I had a factory tour in 1971. At the time it looked like my school career was going to be a major car crash so my parents were looking for an alternative to the academic path. My father knew the Allens apprentice master through the IMechE. So we got a factory tour and I informally took, and passed, the apprentice entrance exam. Three things stick in my mind, watching a turner working on a rotor forging for a steam turbine and wondering if he was more important to the company than the MD, large wooden patterns for pump bodies in the woodwork shop and a massive 25+ foot planer machining engine blocks for the diesel engines.

Andrew

Postscript: My school career was a car crash but fortunately not too serious a one as it turned out.

I normally use nominal size; but I'm not a fan of split dies and avoid using them where possible. Obtaining a rattle free fit is a bit of a crap shoot. I prefer to screwcut or use Coventry die heads, where it is easy to finely adjust the finished thread diameter.

Andrew

Any association with W H Allen in Bedford who made steam turbines, diesel engines and pumps?

Andrew

Addendum: Apparently the answer is yes, a management buy out in the 90s, now seems to have been sold on

I haven't bothered to read the helicon method but hobbing and planing are approximations. They 'generate' the tooth form in the sense of a creating a series of straight lines rather than relying upon the cutter profile.

For many applications the tooth form as hobbed is fine. But for precision and/or high speed gears (as in a lathe headstock) the gears are often shaved and ground after hobbing to refine the profile.

Andrew

Austenitic stainless steels are paramagnetic, ie, weakly attracted to a magnet but cannot be magnetised. Cold working austenitic steels can induce martensite, due to deformation, which is ferromagnetic. The effect can be reduced, or eliminated, by using a high nickel steel and by full annealing after manufacture.

Personally I wouldn't use stainless steel, but if one must use it then 316 will be better than 304 as the nickel content is higher.

Andrew

You could try Wiseman Threading Tools:

**LINK**

I've bought spares and chaser sets from them. They will grind specials although they don't mention standard resharpening. I've got the wherewithal to sharpen some sizes of chaser, but have never had the need to try it.

Andrew

I'd agree with that.

One of the problems when designing switch-mode power supplies is that magnetic components have all sorts of limitations compared to the ideal. When designing a buck converter in continuous mode the key parameters for the inductor are the delta current and the peak current. The delta current is a compromise between output ripple and inductor value as well as determining at what output current the converter will go into discontinuous operation. Values are commonly in the range 20-50%. Peak current is what determines the choice of inductor. Many small inductor datasheets quote a current rating based on an inductance decrease of 20%, ie, starting to go into saturation. Given the wide manufacturing tolerances I'd normally keep the peak current to about 70-80% of rated current.

Another consideration is the curent limit within the IC. This is a last resort and is poorly characterised: worst case more than a 2 to 1 variation in the datasheet. It's a moot point as to whether the inductor should be sized to cope with the worst case over current. It would depend upon the inductor and IC capabilities and heatsinking. As far as I can see the LM2596 goes into cycle by cycle current limiting if there's a problem, so both the inductor and IC might get hotter than desirable.

Time to test the lockdown by going for a cycle ride - less than an hour though!

Andrew

Sounds like you've got the taper issue sorted. Could be a lack of currency?

I'm probably the most current pilot at Gransden as I flew my DG200 the Thursday before last (20th) at Milfield where I was retrieving it, fortunately just before the lockdown kicked in.

Andrew

The angle needed on the topslide is about 1.8 degrees. So I suspect a set up problem. A picture would be helpful. Less likely is that the work is deflecting. Is the turning exactly parallel or just close?

Andrew

Change of material? Tearing is common with low carbon steels and insert tooling if you don't get speeds and feeds correct. Cutting on the return is common. It can be caused by the work deflecting away from the tool slightly on the cutting pass and then springing back.

Andrew

Cabin fever must be setting in.

Andrew

That was really annoying; the BBC probably though the peasants would be too thick to understand it. A value of 1050 would indeed be high. Around 1030 to 1040 is a pretty good peak in the UK.

Andrew

Obviously one needs to make full use of the magnetic material one is paying for, but I can't see you'd deliberately run in saturation?

As the current in an inductor increases the associated magnetic field increases, and it's that increase that opposes the current flow, which is the property called inductance. The magnetic field creates a magnetic flux in the core. But the core material is not perfect and can only cope with a certain level of magnetic flux. When that limit is reached the inductor is said to be in saturation. In saturation an increase in current does not cause an increase in magnetic field and magnetic flux, so by definition the component is no longer an inductor. The result of that is that the current can increase until limited by something else, usually resistance.

On all the switch mode converters I've designed I've always been careful to keep the inductors out of saturation.

I've had another read of the LM2596 datasheet. Not a great fan of TI datasheets (the old Nat Semi ones were better), but buried in the applications section is a note saying the IC is perfectly happy running in discontinuous mode, even if designed to run in continuous mode. It's been a long time since I used any of the Simple Switcher IC family, but I don't recall any particular instability problems at low currents. They should operate at tens or hundreds of milliamps without stability problems. Although of course there will be greater output voltage ripple and the unit will be noiser due to ringing on the switching edges.

The only application I can immediately think of that uses saturation in an inductor is a magnetic amplifier.

Andrew

I agree entirely, although the LM317 is a bit old hat. There will be better, and smaller linear regulators available.

The selected DC-DC converter is the wrong device. The LM2596 is rather out-dated, although it was very popular at it's introduction by National Semiconductor. One of the first completely integrated and easy to use buck converters. No need to understand Bode plots and feedback loops!

I'm not going to bore the pants off people with buck converter theory. It basically outputs a PWM waveform which is then filtered to give a DC voltage lower than the input voltage. It is normal to operate buck converters in continuous mode as the design is simpler and output ripple is less. What that means is that the current in the inductor does not fall to zero during the off period of the switch. In this way the current ripple in the inductor, and hence output voltage ripple, can be controlled as part of the design. For low output currents and/or high input voltages, which implies low PWM duty cycles, the converter operates in discontunuous mode, ie, the current in the inductor falls to zero on each cycle. The converter in the link will be operating in continuous mode. I suspect that what is happening is that the converter does a few cycles at a low on time then whoopsie the output voltage is too high. So the converter turns off for a period until the output voltage falls as current is drawn off. Then the converter fires for a few cycles and whoopsie again the output voltage is too high. The result is relatively large output voltage ripple. This is not a mode the LM2596 is really intended for, a lot of the graphs in the datasheet for output current versus something start at several hundred milliamps.

In newer more complex ICs operation at very low currents is catered for and is often known as hiccup or burst mode.

I'm going to get a glass of wine; you may need to do the same after that.

Andrew