Member postings for Andrew Johnston

Be careful - datasheet surge ratings are often listed under maximum ratings and can be non-repetitive. So you shouldn't be subjecting the diodes to them every cycle.

Andrew

The first reply seemed pretty straightforward to me?

Andrew

Oh dear, on some points I disagree with Bob and SoD.

For an induction motor at no load the current is low and so is the power factor (around 0.1 to 0.3), ie, the motor looks pretty much like an inductor. As the load increases the current increases and the power factor improves to around 0.7 to 0.9 (still inductive) at full load.

With a VFD all the above is isolated from the mains input. The classic rectifier/capacitor circuit has a short conduction angle. Once the capacitors are charged the rectifier diodes only conduct when the mains input voltage is higher than the voltage on the capacitors. So current only flows for a short period near the peak voltage of the input. This is bad for several reasons. One, the rectifier diodes need to carry a much larger current (for a short period) to supply the nominal power rating of the device. That cause more heating and the need for bigger heatsinks. Two, current is taken from the mains as a series of short spikes, which have a high harmonic content. That upsets the energy suppliers. To comply with regulations a cheaper VFD (with a simple front end rectifier) should be fed from a filter so that the assembly looks like a resistive load from the mains.

Larger and/or better quality (more expensive ) VFDs may have a power factor corrector (PFC) at the front end. The PFC is basically a DC-DC boost converter, but with a control loop that forces the current draw to be proportional to the input voltage, ie, resistive.

Andrew

Low power design is all about attention to detail, and a detailed understanding of the hardware. Some pointers are:

Remove any external hardware that is not needed

Run the processor at a low clock frequency, preferably using an internal oscillator if it has one - no external crystals or oscillators

Disable all processor peripherals that are not being used - especially analogue functions

Shut the processor down (not necessarily the same as a sleep mode) when it is not being used, including the clock oscillator - use a true hardware interrupt that will bring the processor up even if the clock isn't running to start with

Use small low quiescent linear regulators; they're much more efficient than switchers at low output currents

All external components need to be sized for the design; excess current/drive capability often comes with higher base current draw

Look at the current draw of every external component, including passives. It's suprising how many microamps can disappear in pullup and pulldown resistors

Andrew

Thanks; it's not that common to have single phase in and 380V 3-phase out. I'm surprised the listings don't make more of it.

Andrew

I can't see anywhere in the listing, or on the specification plate, where it says the output voltage is 380VAC phase to phase?

Andrew

As blanket rules both of the above statements are incorrect. I do a lot of tapping under power on both the lathe and mill at up to 1000rpm depending upon the size of tap. Although I screwcut many external threads I cut some under power with dies, admittedly often with a Coventry diehead, but also using normal split dies. Similarly I tend to use flood coolant on the lathes, but not the mill as it gets sprayed everywhere.

Andrew

Likewise; I make them from EN1A or EN3B, use them once or twice and then recycle.

Andrew

A possible problem is that the PLL will not exactly follow speed changes within each revolution of the spindle. The bandwidth of the PPL cannot be greater than the data input rate. It's a fundamental axiom of signal processing to gather as much data as possible before extracting information. You can't magic information out of limited data.

A 4000 pulse encoder running at 2000rpm is 7.5us between pulses. Plenty of time for a decent processor to do the calculations. An ARM Cortex processor with onboard floating point and running at 100MHz can do hundreds of instructions in the time available. It will also have on board timer/counters that automate pulse counting or timing. These type of processors are a few dollars. Of course one needs to junk any operating system, but a simple interrupt driven loop should be fine, and largely deterministic.

I've just had a quick skim of the hobbing articles in MEW. A lot about building but nothing about design or specification.

I take no responsibility for any shock and awe suffered by SoD. It's probably just as well he only saw pictures rather than seeing it in reality!

Andrew

It's a non-trivial problem to sync one sampled system with another at a different rate and maintain short term synchronism. There's also the added complexity of maintaining the sync as the spindle accelerates and deccelerates. It's essentially what a CNC mill does when it rigid taps. You don't need to be in the wrong position by very much to jam/break the tap.

The problem is basically one of signal processing and realtime software and is the key to making a CNC hobber. In comparison the issue of driving force and missed steps is fairly simple.

Andrew

John, thanks for the confirmation. Here are the parts:

You are correct that the detent and spring are housed in the hole in the block at top right. Note the block is missing a 4mm dowel. At least the serial number on it agrees with that on the main body.

I don't think the 4mm balls are spring loaded directly. When the Allen key is fully home (as above) the ball is slightly proud of its hole. I think there must be a tophat shaped detent in the hole in the block. When the ball is proud of the hole the detent can't engage with anything. As the Allen key is withdrawn the ball drops to the bottom of its hole and when the screw is in the correct place the detent can drop into the 4mm hole, thus locking the screw. The hole in the block is quite short so the spring is going to be short, given that the detent needs to be long enough to stop it canting over.

Andrew

I think they're part of the primary winding snubber?

I can't tie up the layout with the Sanken datasheet, so I don't think the IC is the one shown on the datasheet.

That's the theory; it should provide creepage and clearance between the switching waveform and a low voltage supply that powers the IC. But look at the tracking on the underside of the PCB - completely negates any isolation obtained from the missing pin.

Andrew

Looking at the underside there's no evidence that a pin was ever soldered in the location of the missing pin. I've seen a lot of fudged ICs, but never seen a pin missing without leaving evidence of the failure.

Andrew

Finally got around to disassembling my UPA4:

Fairly straightforward, although it wasn't quite as the instructions detailed. I also fell at the first hurdle and had to buy a set of larger circlip pliers. Once I'd got the correct tool it took 5 seconds to remove the circlip!

It's clear that the locking mechanism for the coarse feed has parts missing. It's missing one ball, a spring and a detent. Sadly the spring and detent are not shown in the picture in the instructions. I can see how it would work, but it would be very helpful if someone could confirm that the lock only works in three places, spaced by 120°, and corresponding to the positions of the three balls. It's simple enough to get a replacement (4mm) ball but I'll need to make the detent and possibly the spring. I also need to replace a missing 4mm dowel pin.

Andrew

Given that the DIL is actually 7 pins I expect it's a custom chip that won't be available, even from the likes of Farnell and RS. I'd agree with Joe, bin it.

Andrew

Depends on what sort of gear and the application, manufacturing method and whether commercial or home made cutters will be used. There's no magic flowchart that can be followed to give an answer - engineering doesn't work like that.

Andrew

Shouldn't be a problem, use normal feeds and speeds and no coolant/lubrication for brass. With brass pecking is less critical; I'd probably withdraw every 1/4" or so.

Use quality makes of drill. Start the hole with a spot drill. Assuming the lathe is being used ensure that the tailstock and headstock are aligned. Make sure everything is rigid; don't put the screw directly in the chuck, use a threaded split collet.

Andrew

I expect 14.5° PA was originally chosen for convenience as sin(14.5) is 0.2504, which makes life easier when you don't have a calculator.

As Martin says there is a compromise between root thickness of the tooth and radial loads as the PA varies; think about the force vector.

However, there are also practical reasons. Larger values of PA reduce undercutting when hobbing gears. Undercutting further weakens the tooth. As an example multistart worms for power transmission may use a PA as high as 60° to avoid undercutting of the mating worm wheel.

Andrew

I'd largely agree with the above. Either use flood coolant or don't bother. My only exception is drilling aluminium alloy on the vertical mill where I'll use a squirt of WD40 down the hole to reduce the swarf sticking to the drill.

For small, low power, machine tools coolant simply isn't needed. When I get my instrument lathe going I won't be equipping it with any coolant, even though the baseplate does have a drain.

Andrew