Single Phase To Three Phase Rewind - Lathe Uprate

Home Model Engine Machinist Forum

Help Support Home Model Engine Machinist Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
As I said a great article. I have an under powered 1940 Atlas F model lathe that needs the ways scraped more then a power increase. But as an engineer that design parts on paper I know my machining skills are poor so I consider the low power as a way of keeping things safer. I have made an upgrade to it however; purchase a quick change gearbox for an Atlas lathe built by another company but basically the same design ~ late 1950 vintage. The used gearboxes for the 40's models are worn out so I added mounting problem but eliminated gear problems. Motor would be the next upgrade on may list. Also have a Fung Fu RF40 bench top milling machine, used, that has a 3phase 3/4 Hp motor. I added a VFD drive.
Life's too short to want to make something it's not into something you want, unless the effort floats your boat. Me, I'd rather be making chips...

I know it's off topic, and I know what you meant, but I'm sorry, I can't help myself...

You do realize that you literally said that you'd rather be doing, exactly what you said life is too short for?
Whilst being off-topic - To Err Is Human - To AARGGH is Pirate ! (sorry couldn't resist that).

A Small Off Topic Diversion

As I mentioned my tailstock bush stripped so I had to make taps for M12 x 2.0 ACME Left Hand vis :-


I measured up the quill screw using the 3 wire method and found it to be a few microns under nominal but worn more towards the front - which is used more - no surprise there.
So I turned up an oversize blank Ø12.05 but cut the thread spot on size by the wires and gashed with zero rake flutes.

Whilst at it I made a second tap 0.05mm over by the wires - however the on-size ended up fitting perfectly.
There is obviously no relief by this method - but I could have backed off the lands on the lead in - but as it turns out it cut just fine on brass.

I put the bush failure down to my heavy handed and all too frequent driving the quill off the end of the thread - every time you do that, you break another piece of feathered edge off the bush - ultimately having the whole thing strip.

To avoid this problem, I have added a warning line to the quill - at which point there is only 12mm of thread remaining engaged - so Stop !

Now I can finish the rest of the modifications - belt tensioner and Synchroflex drive belt.

Regards, Ken
Last edited:
Changing The Gearbox Lubricant.

I have no Idea what the manufacturers put in but it doesn’t feel or smell like a high pressure gear oil. (Over the years I have seen several examples of Chinese manufacturers simply using whatever they have to hand as long as it fits :- oil, rectifiers, transformers, fuses, switches and so on………………………) build standards appear to be absent.

As a thin film the oil is a very strange yellow colour and it also has an unusual smell – like it’s part linseed oil.

Whatever it is I don’t like it one bit.

So I drained the gearbox, flushed it with paraffin and then flushed it with motor oil (for a few days worth of running) and finally filled the gearbox up with some very expensive BMW diff oil that I have left over from that last time I changed the oil in my M3’s LS Diff.

Whilst changing the gearbox oil I noticed a number of problems all of which have been placed in, or are in, the photo below:-

No gaskets or oil seals anywhere – metal to metal lid, bearings are capped (blind) but no seals.

So it weeps oil everywhere – but having said that surprisingly little.

There are 4 redundant holes in the rear wall of the gearbox – not visible from the outside as they were filled with body filler (Bondo) prior to painting.

There is a monstrously large burr left on one of the gear selector forks.

The oil level sight glass – from the lower front of the headstock (removed, cleaned & placed in view here) has an optical target which partially traps oil between its thick outer wall and the window (it can’t drain back) making it look like there is still some oil in the headstock when in fact there is none. Not that it would have helped as the oil in the sight glass was congealed into some gelatinous mass up to the original fill level which I noticed after I had drained the headstock and it still appeared “full”.

Cleaned out the sight glass and removed the optical background, replaced it with a homemade yellow Perspex optical background the oil can get past.

I also found an accumulated layer of “sludge” and metal sparkles in the bottom of the gearbox which I obviously cleaned out.

Since apart from the bronze selector forks, everything in the gearbox is steel – I added a pair of NiB magnets to the cover plate – directly in the throw path from the gears to act as magnetic scavengers. They are also easy to get to for periodic inspections and cleanups.


Note: Oil “weeping” marks below the “A” lever.

Next: Changing the drivebelt to a Synchroflex.

Regards, Ken
Diff oil is perhaps a trifle heavy for normal spur gears. Ordinary engine oil (as used in the venerable mini engine/gearbox ) might be a better choice.
Regards, Philip.
I have an 1985 8" CNC with a 2 speed headstock, it uses ISO 32 hydraulic oil so I'm not sure I'd be using a gear oil either. Spur gears typically only need a relatively light oil and some gear oils have extreme pressure additives that, depending on the additive, may attack brass and bronze bushings. I have no idea of the viscosity of the expensive BMW diff oil but usually that kind of oil is significantly more viscous and not needed for spur gears and maybe not even preferred.
Thanks for the heads up. I was thinking the same thing myself - as I said I ran it on engine oil for a few days.
The BMW diff oil MTF-LT-2 is remarkably thin (feel only) it's listed as a fully synthetic SAE 75W-80 - its used as a "lifetime" oil for the manual transmission but is changed on the LS diff - probably because of friction plate contamination.

Specs :-

Any comments or recommendations ?

Regards, Ken
Last edited:
I can't be specific about your BMW oil, but we changed the oil in a 50+ yr old Smart and Brown recently at work. Light use in an education workshop - I don't think it had ever been changed :) We guessed at an 'EP' type gear oil and the clutch dragged horribly - it took a lot of reading about but to our surprise the modern equivalent of whatever was in the manual, turned out to be our thinnest hydraulic oil which we use in jacks and power packs. It seems happy enough now. Most importantly the oil now circulates, which it didn't for years before the recent sort out!
Changing The Drive Pulley System & Ratios.

Subsequent to performing the motor/VFD uprate, the lathe ran just fine on the original “V” belt and this entire change could have been eliminated – but WTH I’ve bought the belt, pulleys etc. and I’m going to fit them anyway.

I wanted to change the primary drive ratio to effectively gear down my now higher revving motor to better make use of its newfound power.

I can’t change the ratio of the “V” belt drive as the primary pulley can’t go smaller (too tight a radius for the belt) and the secondary can’t go bigger – it will get in the way of the spindle bore.

So I will be changing to a Synchroflex T5 toothed belt drive which can support a smaller primary pulley.

The centerline distance between the motor and the driven shaft gave me some problems.

I wanted to use a T5 Synchroflex belt but the nearest belt lengths of 720mm and 750mm were size #too ( too short or too long ) – forcing me to install an idler / tensioning wheel which I mounted on the M10 drain plug hole at the back of the gearbox (the existing motor position adjuster couldn’t manage this either – the motor adjuster is also a PITB to adjust in any case).

Using a 28T driving a 60T via a 150T = 750 long x 16 wide drive belt.

Permissible continuous belt tension is 540N – so driving 1.5kW @ 3800rpm off the motor using a Ø44.58 pitch circle would be capable of handling 540 x π x 44.58 ÷ 1000 x 3800 ÷ 60 = 4789 Nm/sec.

≈ 4.8kW a factor of safety of more than 3:1.

The belt is a steel wire cored polyurethane belt. Very flexible, efficient and strong.

This is the revised belt layout using a pair of 6004 2RS ball bearings (Ø20xØ42X12) as an idler pulley.

Above the unassembled parts – below as fitted.

As planned this now produces the original speed range at ≈67.7Hz on the motor – since the range set on my VFD is 30Hz. To 125Hz will give me a range of approximately 2:1 either up or down – again consider my favourite gear of 736rpm now allows me to go from 326 rpm to 1359 rpm just by turning a dial – and still have more power and torque than I used to.

For heavier work, simply gear down and rev up – like driving a car.

Regards, Ken
An Unexpected Improvement - In Part-Off Performance.

The V1000 VFD has two features :- “slip correction” and “torque correction”, the purpose of these is to maintain the rpm’s of the motor at the “reference frequency” rpm.

A normal squirrel cage motor responds to increased load by increasing its “slip” – it runs slower – this increases the voltage induced in the squirrel cage bars thus increasing the current in the bars and the reaction torque to the rotating field – the downside is the increased slip also increases the impedance of the squirrel cage which diminishes the current and at some “tipping point” the increasing impedance defeats the increasing current and stall occurs suddenly.

Typically slip is 5% - this is no small issue – take for instance when you are parting off – as you apply load (feed) the spindle rpm’s will slow down slightly – which in turn actually increases the force being applied to the part-off blade – so the process is made inherently unstable. Sure the inertia of the system comes to your aid – but “slip” is not your friend.

With the VFD the speed is maintained by reacting to the current draw and Hall-Effect feedback caused by the out of sync squirrel cage current reaction.

The VFD actually increases the output frequency (but not the reference frequency display) to maintain speed.

Example if I dial in 50Hz. The unloaded 4-Pole motor will turn at 1500rpm but as the load increases it will slow down to 1425 rpm – what the inverter does is increase the output frequency to ±52.6Hz to keep the motor turning at the indicated 50Hz/1500rpm.

Now I hadn’t given this much thought but I found that parting off was much improved by both the increase in deliverable power, torque uniformity and speed stability provided by the VFD.

I’m not sure how much I can ascribe to speed stability on its own but I believe it’s significant.

Regards, Ken
I may be wrong but it looks like the belt tensioner is on the wrong side for a forward spindle drive. On the other hand there may be a perfectly sound reason.
Mauro - Well spotted - you are correct.

I was aware it was technically incorrect but being a toothed belt it has no effect on the potential torque transmission - as would be the case for a flat belt - but it does increase the fatigue of the belt in that the reverse bend happens under tension and will therefore shorten belt life - but I don't expect that to be a big issue.

But then it is the right way around for reverse - not that I am likely to transmit much power in that direction during the belts' lifespan.

So it was a choice of simplicity and expediency over technical correctness. Let's call it perfectly unsound reasoning.

I'm glad someone is reading the post.

Regards, Ken
After posting my comment I considered that if the tensioner is locked, as opposed to be a spring loaded type, then the objection looses much of the merit since variable torque is not going to result in movement of the tensioning arm.
Richard, you often see motors rated for 440V @ 60Hz and 380V @ 50Hz - because they will pull the same current because of the impedance change matching the voltage. The 440V run will output 20% more power - same torque, more revs.
380 x 60 ÷ 50 =456 (allowing for resistance they are typically made as equivalents).
Other than that you see the ratio 1.732 (Square Root of 3) pop up a lot in three phase equations such as 380 ÷ 1.732 = 219 ie 220V so a 380V motor connected in Star is only 220V when connected in Delta.
As regards 440 it could only have a 220 option with a double set of windings - doable but unlikely.
The number of "possibilities" is large but generally not done for reasons of cost.
You can always run a motor under-volt - with concomitant loss of torque/power - but overvolt can generate massive current and burnout in short order.

Example: Let's imagine a 220V motor that pulls 1A therefore the impedance is 220 Ohm - but that 220 is made up of 210 Ohms of inductive impedance and 10 Ohms of resistance.
If saturation occurs at 220V (the limit of the induction portion of the impedance) then every 10V over will apply only to the 10 Ohm resistive part and add another 1A to the load - ie 5% overvolt gives 100% overcurrent. Whereas 5% undervolt would result in 5% undercurrent.

Since the resistive component of AC devices is generally very small, once you exceed the saturation voltage the current rise is very steep indeed.

All hell lets loose if you venture past saturation.

Regards, Ken
Last edited:
Is that a tachometer sensor you have installed on the main spindle? ( I read the first page of this post and find that yes you installed this)
Has that shown to display accurate RPM at low spindle RPM?
I use the VFD with a conversion constant on my 10x24 bench top lathe to display spindle RPM on the VFD LED readout. It's V-belt driven, and not a gear head machine.
I allow my vector drive VFD to run down to 2Hz, and have had all the torque needed for tap and die threading of steel up to 1/2" threads.
The only bad thing about my lathe is the threaded lathe chuck as I typically can't reverse it under high torque requirements. A D1 cam lock, pin driven spindle would be nice.
Thanks for these posts, I always thought that the stator slots in a single phase motor would not accommodate rewiring it to 3 phase.
Last edited:
I saw an advert for a 3 phase electric motor on Aussie Ebay the other day and it quoted 440/720 volts on the motor plate. I'm wondering if this a special motor for running on a different line voltage than the normal domestic line voltage of 220/230 volts. Can anyone explain to me what this rating actually means and whether a motor of this kind could be run by a single phase input 3 phase 220 volt output VFD?
I saw an advert for a 3 phase electric motor on Aussie Ebay the other day and it quoted 440/720 volts on the motor plate. I'm wondering if this a special motor for running on a different line voltage than the normal domestic line voltage of 220/230 volts. Can anyone explain to me what this rating actually means and whether a motor of this kind could be run by a single phase input 3 phase 220 volt output VFD?
No, it is for industrial power that is most likely not available for a home shop. You need to find a low voltage (220-240 VAC) 3 phase motor.
Also if you purchase a used VFD, make sure it is for the lower line voltage. I see many that are called the 400volt class for 440VAC, and you want the 200 volt class for 220-240. The 400 volt class will error out on the low input voltage. I have had good luck with used motors and VFDs from eBay sellers, but make sure you can download the manual for the VFD model they are selling. Some manufactures remove obsolete manuals from their web site, so make sure you can get that, as the used VFD, typically does not come with a paper manual, unless it is sitting on the shelf in the original box and listed to be included.
John, I agree with Ignator above.

FYI - the 440/720 rating is its Star/Delta rating ie 440V connected in Delta and 720V connected in Star.
star-delta connection.jpg

Same three sets of windings connected differently as per above.

(Ignore the star point ground - not done that way for motors - only supply transformers)

It's a square root of 3 issue 440 x 1.732 = 760 (when the internal resistance is taken into account 720.

If you wanted to run that motor off a 200 (220) Volt VFD - you could, but you would only get half the torque (and power) out of it - you could rewind it to 110V (use 1/4 the number of turns of twice thickness wire - a simple request even for a dumb rewinder) and run it at double its rated frequency for twice the power etc. etc.

Regards, Ken
Last edited:

Latest posts