Member postings for SillyOldDuffer

I started an earlier post by saying Tony and I were at cross-purposes! Perhaps the issue is confused?

No-one is telling Tony he's doing it all wrong for not working to drawings. Rather, the gripe is Tony expresses a particular point of view strongly, saying things like :

'You need a drawing to build a bike? I don't know anybody who does that. I guess it's down to visualising what you're after, and dare I say it - imagination!'

Now that's a challenge, not least because Tony implies anyone who needs a drawing lacks imagination.

It happens Tony's opinion isn't 'the truth, the whole truth, and nothing but the truth'. Tony is being rebutted because the whole truth is that most engineering isn't done Tony's way. His is only one of many techniques that engineers can apply to problem solving, and objectionable because it doesn't scale up. Teams can't work like that.

The idea one shouldn't need drawings or other help is wrong. Engineers use whatever is necessary to get the job done without imposing artificial boundaries on themselves or others. They look for alternatives, and don't believe ignorance is bliss!

Nothing personal, even though I think Tony's results are excellent, I don't care for purist approaches. If I want a plan, drawing, specification, risk assessment, bill of materials, CAD software, bought parts, external help or anything else, I just get on with it. My time is valuable.

We all know jobs often go wrong and different approaches are needed, therefore the best workers are flexible. Tony might do even better if he learned a few tricks of the trade - even drawing.

Dave

I suspect implementation problems would make this difficult to build.

• If the edges weren't perfectly horizontal, the roller would tend to walk downhill
• If the edges were perfectly horizontal but not perfectly parallel, the roller would tend to walk towards the narrow end
• If the edges weren't perfectly horizontal and parallel, the pendulum would swing in an ellipse
• As above unless the roller is a perfect cylinder - circular, with no taper

Got to be worth trying though. Perfect is the enemy of good!

Dave

This is the problem with becoming an old dog! If you know Python, this, and similar problems are trivial. Much harder to implement in 'C' starting from scratch with a recursive algorithm. Unless of course, you already know C and understand recursion! It saves having to learn Python.

C/C++ and Python are my all time favourite languages. They complement each other delightfully. In this particular case Python easily wins on rapid development grounds. Unless high performance is essential, or the code has to be low-level for a system reason, Python gets the code written faster and performance is usually more than acceptable.

I strongly recommend Python to anyone looking for a general purpose computing language. (Not so useful on microcontrollers: MicroPython is sawn-off compared with big brother. Not bad on a powerful microcontoller like the Pi Pico though. )

Dave

Yes it does, if you know how. Here's a test screenshot of a post on the new forum, you might recognise the code example:

It's done by switching the editor into Text mode, which allows HTML to be inserted. <code> is supposed to change a single line to a monospace font, and it doesn't. However text pasted between <pre> and </pre> tags does the needful. (pre is short for pre-formatted, which in HTML means "leave this alone".

Other parts of this test post were less happy. An advanced version of the editor was enabled, which does headings, colour, special characters and other gizmos. They all work in edit mode, but not in the post. <h2> is meant to switch to Heading2 format, but the post part of the forum doesn't understand formats. So we're back to the basic editor, at least until the developer gets back Monday week.

A related feature request is for the new forum to host PDF and other documents: this is under discussion.

Dave

Best not to make any assumptions! There are many motor variations, and it will be lucky if anyone knows for sure what should be on a 70 year old grinder.

Some photos of the beast would help. Especially the wiring and terminal marks inside the connector box.

Nothing to be lost by trying more capacitance, fingers crossed. Otherwise it will be necessary to confirm the motor hasn't got a centrifugal switch by looking inside, and making sure the contacts and mechanism are both OK.

If the thing has been in storage, the windings may be damp, and keeping it warm for a few days might help. Worst case, the insulation has failed and some turns are shorted out inside: if so it will get hot quickly - the cure is a rewind or replace the motor,

Above all make sure it's properly earthed! Ancient electrical insulation is often iffy.

Dave

Yeah I vote for that, it will be interesting to see how the approach works out.

Your Brass Bob is gorgeous but I happened to read this morning that the best bob material is annealed cast-iron. It seems Invar isn't the only internally unstable metal! My rod is made of Silver Steel, the Bob is Bright Mild-Steel, and I wouldn't be surprised to find both of them are unstable. Might be the Brass Bob moving about if you get poor results from your proper Invar Rod.

It's a nightmare - every time something is improved, other smaller problems become obvious enough to be worth fixing. Having GPS and GPSDO make it much too easy to detect and chase tiny pendulum issues. I'd be happy if my clock only had a minute hand!

Dave

To be technical, I haven't improved DC31K's program. All I've done is correct the source code that was mangled by the forum in his earlier post. That's all I've done - his program works and I see no reason why it wouldn't run on Windows or Apple. It requires Python and a couple of modules.

As a lot of members downloading in bulk at the same time could crash the server, I've asked Morton's to confirm they're happy with it being run. Net etiquette and all that...

Dave

I thought 'chipboard' straight away. Then wood work.

Unlikely - the router and network don't trust Microsoft, or anyone else.

More likely bing has cached some pages locally, and isn't using the network. It fails when a page can't be found locally, and the computer has to get it from the web.

As Michael says 'it cannot see any other device on my wi-fi network, and it cannot access the internet' I suspect the XP network configuration is wrong. Assuming a physical cable to the router, in network settings:

First thing to check is the XP sub-net mask, which controls how IP addresses are interpreted. Home networks are usually 255.255.255.0 All devices on the sub-network have to share the same mask; if not their IP addresses are incompatible.

Second thing is that XP has a valid IP address. Could be static, where an administrator allocated and set the address in both computer and router. More likely, XP should be set to ask the router for a dynamic address at boot time (DHCP).

XP can't connect to the network without a valid IP address.

After that we're into driver and hardware problems. Wifi works the same way, with an additional layer of complexity - it's encrypted, requiring passwords, and protocol and other settings. Fault finding, always get it working with a cable first.

Dave

I don't know if it's right or not because maths makes my head hurt, but I dislike the idea of compensating in the bob. Not that I know the idea is wrong, but the approach with Brass is unusual.

The main temperature error is caused by expansion of the long pendulum rod, with a component coming from expansion of the bob.

Expansion of the bob can be eliminated by hanging it from the middle. Then it balances out by expanding equally up and down. a useful trick. Then only the rod need be compensated, whilst a simple nut allows rate adjustment.

Mechanical compensation is usually achieved by arranging two metals such that expansion of one cancels expansion of the other. For convenience, the rod is low expansion and the compensator is high expansion.

The only way of compensating in the bob I know of is by making it a jar part filled with Mercury. Mercury has a high coefficient of expansion, and it rises in jar as it warms up. The rod expands down, and the Mercury expands up. The maths is intimidating, at least for me.

Other methods compensate in the rod.

The classic method is Harrison's Grid Iron in which the rod ends in a horizontal bar, on which are mounted two vertical columns. On top of the columns, are a cross-piece from which two rods drop down to a cross-bar to which the bob is attached. When temperature rises the rod comes down, but this is counter-balanced by the columns expanding up. The maths is much easier.

Later it was realised that the grid-iron can be implemented with tubes. The rod runs loosely through and supports a tube of high expansion metal. A second tube, low expansion, sits on top of the expansion tune and drops down over it to the bob. The original high expansion metal was Zinc, which tends to very slowly compress under the weight of the bob. I'd be inclined to try Aluminium alloy, stronger than Zinc, similar coefficient of expansion.

The maths can be simplified by using the ratio between the two temperature coefficients of expansion to keep dimensions relative rather than starting with absolutes. If the ratio between rod expansion and compensator expansion is 1:20, the compensator tube is 1/20 rod length. (I think!!!)

All done above the bob, which only has to hang from it's mid-point.

Dave

Python supports permutations with the 'itertools' module. Example is a snip from the program I used to calculate all the ratios available from my lathe's change gears. All combinations of 4 gears from a list of 11 gears.

import itertools # Load the module that does permutations

metric = [ 20, 30, 45, 50, 60, 60, 65, 70, 75, 80, 85 ] # list the gears available by tooth count

# print all combinations by looping through the list of permutations

for a,b,c,d in itertools.permutations( gears, 4 ):
print(a,b,c,d)

Dave

Really good work, but welcome to the Great Grippen Mire, where nothing is solid.

When John says 'a good wall' he means something truly medieval. To minimise vibration, I think the French kept their standard pendulum clocks in an underground vault cut deep into the bed rock.

Many accounts of pendula synchronising when clocks are mounted well apart on the same wall. The tiny forces exerted by the pendula must be enough to upset the clocks by vibrating massive walls.

Much better than mine though, which is free-standing. Not good. Far too sensitive to run on a dining table, almost a seismograph.

Dave

Could be. The gun-trade is notorious for non-standard threads and somebody had to make their weird taps and dies. Even government arsenals misbehaved - the Enfield Inch was 4 tenths shorter than the inch used by the rest of British Industry. And the US inch was bigger than the biggest British inch.

Dave

Good advice, and also, if the letter is a complaint, start the letter with a heading containing the word COMPLAINT.

Regulated companies have to report the number of complaints received but letters, no matter what they say, don't count as complaints unless the word complaint is used. Means that large numbers of customers can be dissatisfied and the regulator doesn't do anything about it.

I advise writing a COMPLAINT letter to OVO saying that, despite having had the situation explained several times, they are still not managing a customer with dementia appropriately. Ask OVO to explain in writing what they are doing to fix their mistakes. Tell them that OFGEN have been informed. Copy the letter to OFGEN and ask them to investigate.

Sounds as if OVO are in financial poo - many energy retailers have gone bankrupt over the last 18 months. Could be the firm has horrible staff shortages because everyone who can has jumped ship.

Dave

The tachometer magnets on my lathe spindle sit in a flat bottomed shallow recess. Whilst the magnets are more than powerful enough to stay on the spindle at 2500 rpm. the recess lip is needed to stop them sliding. No adhesive needed.

As Magnets and glue are both weak in shear, a physical restraint is best, not glue. Only needs a shallow lip.

Dave

Not exactly. Libre outputs Open Document Format, which is non-proprietary. MS-Word should be able to read it. By default MS-Word uses proprietary files, which other software can usually read, but not necessarily perfectly.

For ordinary purposes, not doing anything advanced, Libre/Word incompatibility doesn't seem to a major problem. Your mileage may vary!!!

Libre (and MS-Word) can both work with a long list of other file formats.

Dave

As I read the spec, 601 can cope with total gap (both sides) of up to 0.1mm. So the side clearance of a round rod in a round hole should be no more than 0.05mm (about 2 thou)

In general the fit should be kept tighter than the 'up to' figure, tighter the better.

My experience:

• fresh glue essential because old glue gradually weakens with age.
• cleanliness is usually vital
• gap-fillers are a compromise. Thick bonds are weaker than thin bonds.
• Whether a key is needed or not depends on the type of glue. I don't think 601 needs a key.
• glue not a good idea if the joint gets hot

Dave

Edited By SillyOldDuffer on 23/08/2023 10:37:20

We're at cross-purposes Tony!

All designers do trial and error to some extent, and it's normal to select, modify and combine existing parts. Though we could, no-one on the forum makes drawings of standard nuts and bolts; we just use them. What you're doing is good as far as it goes, but not the whole story.

The unplanned approach is OK up to a point, but it breaks when complex objects are made from scratch because no-one can hold all the details in their head. For example, engines are full of moving parts that have to fit together and rotate without hitting anything, and the engine won't work unless suck, squeeze, bang, blow are timed correctly, and the fuel mix is right, and the spark is hot and long enough to ignite the compressed mix. This type of design problem is only practical with research, planning, drawings and maths. Only a BF would cast and machine a cylinder block without being certain the pistons and cranks will fit into it.

Trial and error might be OK in a custom bike business, but on its own it would bankrupt a manufacturer. They always need a portfolio of design information that can used by others to build the item.

It's OK to be proud of results, but what you're doing isn't a better way of doing engineering. It's only a small part of what engineers do, and only suitable for simple stuff. What engineers mean by 'planning' covers a blurred range of ever increasing complexity from:

• ideas simple enough to hold in your head
• slightly more complicated ideas described on the back of an envelope
• complicated enough to require formal drawings
• So complicated that multiple formal drawings with version control are required. This level becomes mandatory as soon as more than one person is involved. Multidisciplinary teams, purchasing departments, subcontractors, production engineers, machine operators, accountants etc.
• Computer support essential - high-tech, CAD/CAM etc.

In a home workshop it's not unusual to find everything from bonehead to computers being applied more-or-less on the same project. In the right circumstances they're all useful, but 'planning on the run' is only a small part of a much bigger story. Might be a good way of customising a motorbike, but wouldn't work on a Boeing 787 or an internet router.

And because an engineer is a bloke who can do for a pound what any fool can do for a guinea, it's best practice to reuse other people's plans and methods. Believing originality is essential isn't engineering!

Dave

Only because the electrician's estimate was wrong...

In the power station:

The National Grid Control Room:

And the consumer...

Dave

Both tools Margaret is using should be OK on Brass, though the sharp silver inserts are much better.

Rule of thumb, an unsupported rod shouldn't protrude more than 5 times it's own diameter. Beyond that the cutter pushes the work away, and chatters. Note that a blunt insert has to be pushed harder into the job before it will start cutting.

The Brass is a suspect. Though most Brasses machine well I have a rod that doesn't!

Centre height is important, especially on small diameters. Too low and the cutter tends to dig in and chatter.

Here's a photo of tool height being checked with a steel rule. The cutter is wound gently to nip the rule in place. When the cutter is too high the top of the rule tilts away from the operator. If too low, the top of the rule tips towards the operator. When the rule is vertical the cutter is exactly right.

I suspect all these factors have come together to spoil Margaret's day. In order, I'd:

1. Set Centre height
2. Reduce how far the rod protrudes from the chuck to much less than 5x diameter
3. Try a sharper insert (or sharpen the HSS)
4. Try a different Brass - though it's probably OK

There are other possibilities, such as bell-mouthed chuck jaws, but deal with the most likely first. It's possible that steps one and two will fix it.

Dave

Edited By SillyOldDuffer on 22/08/2023 18:46:14