Member postings for Andrew Johnston

A plate mounted on a small spigot is an open ended invitation to Mr. Chatter, whatever tooling you use. The ratio of OD to finished bore doesn't look too bad.

I'd mount the rough plate in a 4-jaw chuck held on the OD. Face off and drill/bore the centre hole. Turn it round in the 4-jaw and face to width. It wouldn't take long to knock off more of the edges on the rough part. Finally mount on a spigot, with a thick washer (not a thin commercial one) and machine the OD and chamfer. In my experience 316 turns very nicely with insert tooling. I'd mill the peripheral slots after machining to avoid an interrupted cut and allow one to form a better shaped edge for cutting.

Andrew

Was the buck converter unloaded during the measurement? It's common for simple switchmode power supplies to lose regulation on light loads, or no load.

Andrew

Both heatsinks on my easily accessible VFDs are grounded. However, a VFD is classed as a component, intended to be used in a larger assembly. So there's no absolute requirement for the heatsink to be grounded. It could be floating due to the design, or could just be a case of you get what you pay for.

Andrew

None, because with an imperial or metric leadscrew one or other set of threads will be exact and the other an approximation, not both approximations.

Andrew

Therein lies the problem, we don't know what the OP wants to make. So advice can only be offered on the basis of what we make, not what the OP wants to make.

In general gear cutting is easier on a dividing head:

Although a rotary table can be used; in this case the gear was too big to fit under the spindle:

Of course it helps if the number of teeth is an integer divisor of 360, as it was in the above case (72 teeth).

However, as Mike says, there are some things for which one definitely needs a dividing head:

My dividing head has a reversible index plate behind the chuck, so I also use it for simple indexing work such as squares and hexagons.

Conversely a rotary table is best when radial cuts are needed:

Or for large parts that will not easily fit on a dividing head:

The centre hole on my rotary table is 1" diameter and parallel. In my view that's more useful that a Morse taper as it makes it dead easy to make mandrels or alignment pegs. After buying my rotary table I also bought a cheap 3-jaw chuck, but have never fitted it and never will. I simply don't need it, especially as I use the dividing head for indexing. In the past I've used the rotary table for rounding the ends of rods, but it's a right royal PITA. It takes care to set up the part on the axis of the rotary table. If an arc is needed rather than a full circle it is only too easy to go a gnats whatsit past the desired point. It might only leave a shallow witness mark, but it is surprisingly difficult to remove it. Lots of filing and an end that no longer looks semi-circular. I prefer filing buttons for rounding the ends of rods and clevises.

In summary the OP needs to decide what he wants to make and can then decide which is better, rotary table or dividing head.

Andrew

Agreed - Andrew

Martin: Thanks for the exposition. You've certainly had an interesting career. I remember a few of the names and companies. Pi worked primarily with Indy teams. In F1 we worked with smaller teams, as bigger teams like Mclaren and Ferrari had their own electronics and data analysis groups. Teams I recall were Tyrell, Footwork, Larrousse, and Sauber during their initial entry into F1.

It's a shame that Paul Morgan killed himself in what was a rather unfortunate accident. People in motor racing have a poor record of fatal aircraft crashes. Motor racing can be dangerous, but aviation is particularly unforgiving of over-estimating ones ability.

I'm not sure what you mean by a compound dividing head, presumably one that can do compound indexing? While the movements are different I thought compound indexing was mathematically equivalent to differential indexing, which any universal dividing head can do.

Andrew

No need to worry about finding the photo. I've made bevel gears the old school way on a manual mill with a dividing head. I wouldn't regard CNC as cheating. If nothing else it taught me a lot about the design of bevel gears, which was needed in order to create the 3D CAD model.

For a 32bit ADC reading a 5V signal (assuming that's full scale) I make the LSB 1.1nV? That's the resolution. It will only be precise if the ADC is monotonic and only as accurate as the reference voltage.

The Moog valves I used were ±10mA and had no sawtooth as far as I recall, although it was 30+ years ago. The valves were aerospace ones. The gearboxes were used on Indy cars and F1. I worked for Pi Research, who had strong links with the original Ilmor Engineering, and Roger Penske, before Paul Morgan was killed flying a Sea Fury. Who did you work for?

Andrew

Unfortunately the question has been posed backwards. It would have been better to state what needed to be made and then ask which one would be better.

Andrew

Shocking, I can't imagine any self-respecting forum member having the gall to be so equipped.

As a kid I built a small oscillating engine and boiler using hand tools. Although, to be fair, I think the boiler was silver soldered at school. I then moved onto a 3.5" gauge steam loco with the help of a £10 lathe and free pillar drill, courtesy of the local model engineering society. In the early 1970s, while at secondary school, I used to cycle across town on Saturday morning to use the workshop of a fellow member of the society. He had a small lathe and home built milling machine. On Sunday mornings I went round the corner from my parents house to an amateur radio enthusiast to learn about radio and electronics. Those were different times - my parents never met, or even spoke to, either of the people I visited.

I have a leaflet (a reprint probably from the 1940s or 50s) detailing how to build an oscillating engine using only hand tools. There is a caveat; I suspect the boiler was soft soldered and wouldn't meet modern requirements. But if the OP sends me an email address via a PM I'm happy to scan the leaflet and email it.

Andrew

Use a CNC mill:

Unlike with a dividing head and involute cutters the above method produces a true bevel gear.

Rotary tables and dividing heads are complementary, although there is overlap in use. The well equipped workshop has both.

Accuracy, precision and resolution are not the same thing, so one can't say that analogue is more accurate than digital or vice versa. It all depends.

It's been some years since I used Moog valves, but I don't recall them having dither? The ones I used were about the size of a small matchbox. We used them as part of a crash gearbox on a racing car. They were current driven, so I designed a voltage driven bi-directional current source to control the valve.

Andrew

Looking at the datasheet it doesn't seem to be intended for general machining. It contains high pressure additives for slow, but potentially high pressure, operations like drilling, reaming and tapping. The fact that it is available in aerosols should tell you that flood coolant for general machining is not it's primary purpose.

The safety sheet mentions toxic fumes from combustion, so if you're getting smoke then you are asking for trouble. Use it properly and look for another product for general machining.

It seems to be an equivalent to the Rocol RTD compound I sometimes use for tapping. No way I'd use RTD for general machining.

Andrew

That's a shame, a missed opportunity. An apposite idiom given the current discussion would be no point in crying over spilt milk.

Andrew

I really like my Kurt vice.

Andrew

As far as I'm aware no inverter measures torque, or directly controls it. So you can't increase it by just altering a parameter. The only thing that can be controlled is current which in simplified terms is proportional to torque. If you want more torque from a motor running at 50Hz then you will need to fit a bigger motor. If you want to boost torque at frequencies below 50Hz then you may be able to select a current profile that allows this, but be aware that it may hasten overheating of the motor.

I'm not familiar with Banggood inverters so I have no idea of the exact parameters that might need changing.

Andrew

Like the title says the graph shows the torque-speed curve of a 3-phase motor. What it doesn't say is that it is at a fixed frequency. Let's assume that we have a 3-phase motor running off a 3-phase supply with no load and no internal losses. The motor will run at the synchronous speed, for a 4-pole motor and 50Hz that is 1500rpm. At this point we are at the bottom right of the graph. Of course a practical motor will have some internal mechanical losses which require torque to overcome. To provide this torque the motor has to slow down below synchronous speed; it slips below synchrnous speed. So we move up the torque curve with a slight decrease in speed. If we now add a mechanical load to the motor then we continue up the curve. The speed drops more and the torque produced rises as the current rises. When we reach the rated power of the motor the current will be the rated current and the speed will have dropped to the rating plate value. For most motors the speed at rated power is a few percent less than the synchronous speed.

If we now continue to increase the load the speed will drop further, the currents will rise above rated current and the torque will also rise. At some point we reach the maximum torque - pull out torque. This is normally 250% or so more than rated torque and interestingly is determined by rotor resistance. A VFD can mimic this behaviour in a controlled manner by trading speed for increased torque during a short term overload. My CNC mill does this.

If the motor load continues to rise the speed drops further and the torque also drops. Of course from full rated power onwards the motor is overloaded and will probably have thermal issues.

To summarise, the graph is nothing to do with how the torque varies with frequency for a motor driven by a VFD and within it's rated power.

Measuring torque on a rotating shaft is involved, and done properly requires expensive inline sensors and modification of the drive train. Less accurate estimates of the torque in a rotating shaft can be made by measuring the twist of the shaft under load.

Andrew

Oh dear, that's very disappointing, I thought SoD was smarter than that. He should go and stand in the corner and contemplate his misdemeanor!

In a simple model of an induction motor torque is proportional to current. As the frequency, and speed, decrease from base speed the VFD will decrease the applied voltage to keep the current constant. So the torque is essentially constant below base speed. For the purposes of this thread we'll take base speed as that when the motor is running from a 50Hz supply. As the motor speed drops the torque stays constant but the output power drops in proportion. Gear and pulley arrangements increase the torque in proportion to the speed reduction so the power stays constant.

Newer VFDs, that use vector control, can increase the current, and hence torque, below base speed. To some extent that ameliorates the drop in output power. But it is normally intended to counteract short term overload, not for continuous operation. The motor is being run beyond its design current and will quickly overheat, even with a fan running at normal speed.

I'd agree with the above that a larger VFD will have no problem running a lower power motor. The only real problem occurs when the VFD is capable of much more power than the motor needs. The internal current sensors will be sized according to the rated motor power. If a smaller motor is used naturally the currents will be lower and the sensor will be operating over a small part of it's range, leading to a loss of resolution. That may be a problem for the more complex control algorithms.

I'm not sure why a belt sander needs variable speed control?

Andrew

I don't turn aluminium very often, as it isn't used on my traction engines, other than for machining fixtures. I find that the CCGT inserts work fine without the need for coolant. Birdsnesting of the swarf is a bigger problem, and I don't think coolant would help with that. I don't use RTD very often, mostly for large taps or tough materials, and only when things get iffy. It works very well but is a PITA to clean off. An advantage of using larger tapping drill sizes than most modellers is that tapping is much easier without the need for messy lubricants in most cases.

Andrew

Correct, but it is pertinent to note that as the frequency decreases below 50Hz the torque will stay roughly constant, unlike a belt or gear change arrangement. So the motor output power will go down in proportion to the frequency decrease.

Andrew

Did you ask them why?

Andrew