Member postings for Andrew Johnston

There is a factor of pi involved. When going from tpi to DP if one value is an integer, then the other must be irrational. For instance if the tpi (of a rack) is an integer then the DP of any mating gear will be irrational. Conversely if a gear has an integer value of DP then the tpi of the mating rack will be irrational.

Andrew

Totally missed the point. I was trying to explain why VFDs do not generally have built-in filters. The manufacturers don't give two hoots about home users, but design for the requirements of industrial installations.

I apologise for mentioning engineering on a modelling forum.

Andrew

Some aircraft I fly have similar red bands on the rev counter signifying ranges where the engine shouldn't be operated continuously, normally due to propeller/engine resonance. However I don't think that is the case here. I can't think why the VFD/motor combination would have speeds to avoid.

I agree with NigelB, I think it's a warning to be careful at low speeds due to reduced cooling. The increasing width of the red bands indicates that the problem is greater as the speed gets lower. Unfortunately it's not done terribly well, as the graduations from more green to more red are too coarse.

Andrew

Industrially the VFD is considered to be a component, which is used within a larger installation. The need for EMI filters will be considered taking into account the whole installation. That may mean a VFD has it's own filter, or a filter is shared, or one isn't needed. Consequently most VFDs do not have a built-in filter.

EMI might be lowered by reducing carrier frequency, but it's not usually clear cut. The carrier frequency determines how far apart the harmonics are spaced. How far the harmonics extend, at a significant amplitude, is determined by the speed of the waveform edges, which probably won't change much with changes in carrier frequency.

Andrew

The OP really needs one of these......

....set up to screwcut RH 19tpi BSPT threads:

the resultant 3/8" BSPT blowdown plugs, and a couple of embryo fusible plugs:

Andrew

I'm not sure why it is assumed that the problem is due to conducted emissions down the mains? It's possibly more likely that the issue is due to radiated emissions. Is the cable from the VFD to the motor shielded? If so, is the shield grounded at one end or both ends?

Does the problem occur when the VFD is on, but in standby, or only when the motor is being powered?

How often does the comms dropout occur and how long is each dropout?

As stated above mains filters are readily available, and are a good thing to fit, even though they may not solve the current problem.

Andrew

That's a bit harsh.

The ER collet system was developed by Regoflex in 1972 for professional use. Standard collets should have a runout of better than 0.01mm, precision ones better than 0.005mm and better than 0.002mm if you have deep pockets. i use ER11 and ER20 collets on my CNC mill, and ER20 collets on my Bridgeport, for small cutters and sizes I don't use very often. Can't remember exactly what the runout is but small, around 0.015mm on the cutter shank. I use cutters down to 0.5mm on the CNC mill and any significant runout would fudge the cutter.

It's possible the holder/collets that were bought are not up to scratch. I bought my holders from Tormach (they made my CNC mill) and the collets came from professional tool shops.

I've never used finger collets on any of my mills for holding cutters.

Without a numerical specification the word 'precision' in an advert is simply marketing sitting bulls and can be ignored.

Andrew

Never knew that; I've always bought silver steel and gauge plate in standard lengths, as it is more economical. I buy both from Coventry Grinders.

I also buy key steel (EN8) from Coventry Grinders. It's bright drawn, not plated, and is to the appropriate British Standard, meaning that it is slightly oversize, so can be drawfiled to be a snug fit in the keyway.

Andrew

Agreed, but not had a problem with gauge plate rusting, unlike plain low carbon steel such as EN1A which can rust overnight, even in the house.

Andrew

Silver steel comes in 333mm (aka 13" lengths, as well as 1m and 2m lengths. But standard gauge plate length is 500mm (aka 18".

Andrew

Bring it up to the club next time I/we are on duty and I'll attempt to identify the accessories.

Andrew

If I was going to reproduce the handle I'd use a hydraulic copying unit and template:

Final parts:

But I wouldn't reproduce the handle. Instead I'd design, or buy, something simpler and less dangerous. I took one of the table handles off my Bridgeport. It's the right height to hit one where it hurts if it catches you while under power feed.

Andrew

Conversely it's a pain having to second guess about parts with which we're not familiar.

Glad it's been fixed, but I suspect the shmming has hidden the original problem rather than solved it. Is the backlash consistent or does it vary as the bevel gear is rotated?

Andrew

Some pictures of the parts would useful.

Andrew

I'm not familiar with the detail of the Allchin design, but it seems like one of the bevel gears is not concentric with it's axis. How were the bevel gears manufactured?

Andrew

With both bevel gears fitted does the centre run smoothly at some points and bind at others? Or does it bind everywhere? If the former then I'd suspect an axial misalignment of the bevel gears rather than radial misalignment of the pinions.

Andrew

Whale!

It's the lowest grade of tufnol that uses fabric, as opposed to cheaper versions that use paper.

Andrew

It can rebound when the originator is wrong.

For unified threads below 1/4" the number is related to OD, for numbers from 0 to 12:

OD = thread_number x 0.013 + 0.06

Andrew

Ketan beat me to it!

Debugging a system is a matter of making measurements, thinking logically and not jumping to unjustified assumptions. We could assume that the resistor is the correct value, although to be sure we could measure it. The next stage is to do as Ketan suggests and load the motor when it is running on it's own. Can be done by hand, or hold a couple of pieces of wood either side of the spindle. If the motor slows more than it should then either the motor and/or the board is faulty. If the motor doesn't slow then the problem is the lathe spindle.

If the motor/board has a problem I'd start by making measurements. However I assume the OP doesn't have appropriate test gear? i agree with Les, don't fiddle with settings on the board. That is simply attempting to paper over cracks, rather than addressing the root cause of the cracks.

Andrew

.....run from a 3-phase supply.

The cause in this case might have been an external short caused by swarf, In which case the reliablilty of the electronics is not so relevant. The issue with power electronics is that it is easy to create a design that works, but more difficult to decide what fault conditions need to be protected against, and to modifiy the design accordingly.

Andrew