Harrison L5 Metalworking Lathe (1950) Teardown & Overhaul

[Sideways]

These machinery overhaul threads can take a while if posted up in real time. Life often gets in the way and plays havoc with our good intentions.

This thread will hopefully be of interest to a few folk and has the advantage that the work is essentially done as I start to post so it will only take as long as it takes me to post pictures and tell the story. The action takes place in the workshop of my good friend Deema who generously shared his space and know-how along the way. He was doing a similar overhaul on his own J&S 540 surface grinder while I tackled this rather simpler project. We tried not to mix up the parts ....

The machine in question is an early Harrison metalworking lathe. Made in Yorkshire in 1950 based on the serial number, I'm told it spent most of it's life in a training school of some sort before I became the second private owner. It's a gap bed lathe, turns 9" diameter not including the gap. It has an 8 speed spindle from 21 to 480 rpm and cuts threads courtesy of a heap of change gears and a three speed gearbox.

I've used lathes several times over the years but never owned one before. The L5 is old and slow, but strongly built and heavy. It's also, just, small enough to shoehorn into my garage.

The L5 includes a one piece cabinet base made from heavy plate. The manual says they weigh 560Kg all up and it felt like it. It was a monster to load into a Luton van even with two hydraulic pallet lifters to support a tail lift which was nowhere near strong enough to raise it.

Back in Deema's workshop, it looked very promising. Relatively clean, simple and strong.

We had seen the machine working and the immediate priority was to sort out the electrics so that the lathe could be properly checked.. A massively oversize 4kW 3ph cast iron motor had been fitted at some point in it's life driven by a 1.5kW VFD which was literally kicking around the floor on a metal plate. These needed to come off, motor stripped, cleaned, bearings renewed as routine and a new, properly sized VFD fitted.

Life was to prove much more complicated than that ....

 

[Sideways]

The motor that came with the lathe was a fine, British made, item but hadn't been properly connected. Thanks to the cables being taken into the top of the terminal box with no gland to seal it, I was able to tip half an eggcup full of brass turnings out of the inside of the motor. Surprisingly this hadn't damaged the windings though both the armature and stator carried a few scratches from brass between the moving surfaces. I dislike painting over a painted motor so a deal of time was spent with a wire wheel stripping it back for repaint, then new bearings and reassembly. Sadly after all that work the motor ran well but with a little more vibration than I wanted. I decided to replace the motor so it was primed, boxed up and sent to a new home. In excess of Hermes's 20Kg limit the postage cost rather a lot.

The replacement was a little used top quality ABB 4 pole 2.2kW motor.

Many Harrison lathes hang their motors simply on the back of the machine using sliding brackets. These are a cast part and not exactly flat.
They went on the mill and had the slots, top surfaces and feet milled flat straight and square.

The motor mounted nicely but the spindle only just cleared the end cover so two lengths of tube were drilled and tapped to serve as stand offs

Much better. A new motor with a metric shaft required a new pulley. This is a taper lock type by Dunlop. Three rather long thin belts are used. The upper pulley is original and imperial, but I found that a metric SPZ profile is a close enough fit that the motor pulley and belts in that profile would work happily with the original upper pulley.

Ebay provided a good 2.2kW VFD to match the motor and we were ready to spin it up.

 

[Sideways]

The headstock has a filler and oil level gauge on the back side of the head. The oil level looked to be well below minimum so the headstock gears needed checking.

4 obvious capscrews and the alloy cover is easily lifted off from in front of the speed selection levers on top of the head.

Oil was thick and dark in the bottom, barely enough to actually drain out of the plug on the sight gauge.
I consulted with my buddy and we decided the best way to flush it would be to pour in a litre of diesel and run the motor gently.

This was repeated a few times, draining and replacing the diesel with rounds of brushing around all the gears, the inside of the head and copious amounts of good blue paper to remove the metal fragments that had probably accumulated inside since the machine was made.

As it came cleaner, I put several rare earth magnets around the inside of the head to collect swarf from the gears and this worked well to pick up the fine black stuff.

Running the motor on a variable drive is ideal for this type of experiment. You can gradually increase speed and watch how the splash lube works.

Although the gears showed signs of burring at the edges (ham fisted selection and possibly attempts to change gear before the spindle had stopped) nothing looked or felt too bad for a lathe that's older than I am.

I ordered up some oil for the head and the slides while the diesel was having a good soak and 3 days after first opening it up, I drained the head and put in the oil.

Filled to the top line, the oil isn't very deep in the box. All the lubrication relies on just a few larger gear wheels dipping down into the oil and throwing it up for everything else. Oil thrown onto the inside of the head drips down onto a ledge around the top cover, runs around and tops up a hollow above the main spindle bearing. From here it simply pours down a hole into the two big angular roller bearings at the front.
One clear lesson from this is that when starting the lathe from cold, it's important to get that splash lubrication running to get oil into the main bearings. Don't just start the lathe and put it straight to work in low gear. Run it for a couple of minutes in one of the top ranges first to get the gears and bearings properly lubricated.

Gear selection on the L5
Back in the first post, there's a picture of the speed selection chart on the top of the lathe. It's confusing until you know how it works.
Spindle speed is done with 3 levers.
On the FRONT of the head is a big lever than shifts between pointing at 10 o'clock and 2 o'clock. This selects high or low range represented by the yellow and red numbers on the speed table.
On the top there are two small levers pointing towards the operator. Each has 3 positions. Left and right select actual speeds. Middle is a neutral. There's an interlock so that you can never try and select a speed with both levers at once. In order to move a speed lever left or right to choose a speed. The other lever must first be put in the middle position. Left lever selects between the two low speeds in each range. Right lever selects between the two high speeds.
It's actually easy once you get used to it.
The gemlin waiting to bite you is that the combined clutch brake lever that applies power is on top with the gear selectors. The lever closest to the operator and the one you instinctively want to grab in a panic is the high / low range selector which mustn't be moved when the lathe is running. I wonder how many apprentices got that wrong at least once

The standard L5 came fitted with a 6 pole motor (just under 1000 rpm) that gave a speed range from 21 to 480 rpm.
It was also available with a 4 pole motor and a twin speed (2 or 4 pole motor). Depending on the motor that gave top speeds up to 720 or 1440 rpm. very low by the standards of today where lathes may hit 3000 rpm or more to turn smaller parts with replaceable insert tooling.

With a 2.2kW 4 pole motor I have plenty of power available and the variable speed option. I've configured the VFD to spin the motor faster than it's design 1400 rpm. 4 pole motors are physically almost identical to 2 pole, so there will be no mechanical reason that it can't take 2800 rpm if I wanted, but when the lathe was sold with a 2800 rpm motor, it was also given a second bearing on the tail of the shaft. I won't try to go that fast. My setup will allow maximum power upto 720 rpm but thanks to the VFD can be increased to 1000 rpm at reduced torque for light cuts.

21 rpm low speed feels SO slow !

 

[Sideways]

Having seen inside the headstock and actually cut metal, I decided that this really was a lathe that I'd like to keep for a while.
I decided that I wasn't brave enough to strip down the head when it didn't seem to need it, but peering around and underneath the saddle it was clear that most everything else would need to be stripped for a serious clean.

The previous owner hadn't used the lathe much I suspect and when he had, it had been to make small parts, turned dry from brass. The lathe was full of brass shavings - they get everywhere - and the whole machine had ben wrongly lubed with thick grease. The L5 and several of its successors have several grease nipples but they aren't. In fact they are oiling points and should be regularly lubed with a push type oiler.

Here are some images to show what was lurking inside the saddle and apron once we got them off ...

The apron seen from above and behind

Looking down into the apron - front control handwheels at the top edge of the photo

The split nut that opens and closes on the screw for screw cutting

This mechanism is the surfacing feed.

I later learnt that removing the saddle and apron of the lathe should be easy. The bearing block at the right hand end of the two long shafts (one slotted, one threaded) is unbolted, some pins removed and then the two shafts can be drawn out through the apron, to the right of the lathe.
This worked great for reassembly once everything was cleaned. During the teardown, everything was so dirty that this wasn't obvious and I just removed bolts and tapped out pins wherever I could get at them. Eventually parts becoming loose enough to wriggle free of the bed.

 

[Sideways]

The L5 turned out to be easy to disassemble. Shafts were mostly steel running directly in the cast iron or occasionally in a bronze sleeve in the casting. Shafts were held in place either by grub screws that only needed an imperial allen key to remove, or in many cases they were pinned using a gentle tapered bore and a tapered pin. These lock securely enough but are easy to tap out. They have the advantage for maintenance that there is a right and wrong way to line parts up. It is more difficult with a sprung roll pin because unless carefully checked beforehand you may get one of the two parts being pinned 180 degrees out and start driving the pin only to have it stick and need driving out again.

Despite the grime, only one significant problem turned up. One of the large gears in the apron involved in the power feed had a tooth completely missing. Presumably a user crashed the tool into the work or chuck under power at some point and the tooth was the thing to shear off.

The next stage of the overhaul was a massive clean up taking several weeks off and on.

Weapons for this included Gunk degreaser (cheap) and Jizer (expensive but much better, about as good as diesel but without the stink and 5x the price).
Wire brushes, scotchbrite pads, electolytic rust removal and most useful, an improvised tank to do the cleaning uin.
I bought some large army surplus stainless steel catering trays and these were very helpful - it was like having a large sink to do the dirty work iin and the holes in the trays allowed the grit to drop through into a heavy bin bag sandwiched between two of these nested trays.

Eventually, a clean kit of parts emerged

These had to be kept clean while the adjacent lathe and mill were put to use repairing the broken gear

 

[Sideways]

Gear repair:

There are a number of ways of repairing or making a new gear. The gear I needed to fix or replace was about 4" diameter and 42 teeth. It was an imperial size and after reading around and taking measurements, I was confident I knew what the tooth profile was and bought a milling cutter capable of replicating it for about £25 from RDG.
I also bought two cast iron blanks in case I needed to machine a brand new gear from scratch. One to use. One to mess up. A 5 inch cast iron blank, one inch thick, cost about a fiver plus £5 postage last year.

Nonetheless the gear was in decent condition apart from the missing tooth and I wanted to try out a repair where a new piece of steel is silver soldered into a closely fitting pocket milled in the gear, then the tooth is recut by horizontal milling.

A tightly fitting silver soldered joint is reputed to be nearly as strong as a brazed joint but needs less heat to achieve it.

The process worked well and I'm really pleased with the end result.

First, a robust arbor needs to be turned from a bar end to hold the gear for milling. This has to be a tight fit in the bore (zero play) and threaded so that it can be clamped tight with a nut.
The spindle is then mounted on a mill between a dividing head with horizontal chuck and a fixed centre, and a pocket cut out where the broken tooth was.

A piece of steel was prepared that was a tight fit in the cutout.
The gear was setup on edge between fire bricks to try and keep the heat in the gear with as little loss as possible

The joint was fluxed and silver solder placed on both sides where it could wick into the joint. 2 torches with mapp gas provided the heat and it took less than a minute to reach the 700 degrees C needed to melt the 55% Ag silver solder

After cooling, the gear goes back on the horizontal mill, this time with the gear cutter mounted.
It took much longer to position the job than it did to cut it. Two cuts need to be made, one taking away each flank of the tooth blank, with the aim of getting the correct shape, depth, height of the new tooth, perfectly spaced between the adjacent pair.

Several passes are made cutting deeper until the correct depth is reached, then index the gear around and cut the other flank to the same depth.
Guessing the depth is a challenge as the new tooth is too tall and looks narrower than the others. The tooth is taken down afterwards on the lathe but meantime, it's all about judging the bottom of the cut.

Mill off the excess length, shorten the new tooth to the same length as the others on the lathe then carefully hand file the ends of the new tooth to match the others

It's not perfect, I should have milled 0.1mm or so deeper, but I'm very happy for the first time of trying a number of these steps. The joint is very sound.

 

[Adam W.]

It's very tidy work and well executed. Will the small voids in the silver solded present a problem at all ?

 

[Sideways]

Reassembly is the reverse of taking it apart so skipping many photos which I'm happy to share if anyone has questions. Here's the cleaned and reassembled apron with the repaired gear installed. The apron manages the left right travel of the cutter both by hand and under power for surfacing and threading.
This is the back and the apron is upside down, so that you can see the split nut that engages the feed screw (towards the R), and the small pinion gear (lower L) that engages with a rack along the side of the bed for manual feed and for surfacing under power.

 

[Sideways]

It's very tidy work and well executed. Will the small voids in the silver solder present a problem at all ?

I don't think so. Certainly not so far. The other end of the tooth is a tighter fit so I hope that porosity doesn't go too deep. The ones that you see in the gullets of the teeth are just on the surface because I didn't mill down quite far enough to remove them completely.

Fingers crossed ....

 

[wallace]

Thoroughly enjoyed that, I think when I get a metal lathe it will be a harrison, I've been watching doubleboost that long I feel I know the lathe quite well. I have a dinky little drummond that I started but got sidetracked by wadkin lumps.

Deema's place looks like a cracking place to play, with some useful bits of kit.

 

[Sideways]

Saddle, slides and gearbox still to come Wallace. Hope you enjoy a long read
I don't know if pictures of the insides of things count as helpful or should come with a spoiler alert !

There's huge advantage from having a like minded buddy to work with. One part of that is you tend to encourage each other to "just have a go". Once we start, we can usually fix it or maybe break it and have to make another

 

[Tony Works Wood]

Great job on the overhaul, very thorough. That gear repair is superb. They are great lathes, I fitted a faster 2HP single phase motor to mine as well. It was my first lathe and was also dated 1949 I checked it with Harrisons in about 1985 with the serial number after I bought it when I rang them up to enquire about a manual for it and a drawing for spindle nose dimensions. They sent me both free of charge. I sold it about eight years ago when I bought a later model Harrison 140. Looks like I got one of the last of the Mk2 design and you got one of the first of the Mk3 design sold in the same year lol. The pic is when I sold it. Looking forward to the rest of the overhaul.

 

[Sideways]

Excellent Tony. Thanks
Yes, I believe mine is an L5 mk3 and yours is slightly different though v similar.
Hope you are enjoying the 140. The L5 is a proper lathe with a useful capacity but it's strange to have the same 1.5" 6tpi threaded spindle on a metal lathe as they used on the graduate wood lathe !
Your newer 140 is nice step up the range !

 

[Fergie 307]

Nice job. Are you a member of the Harrison lathe owners group? I and other have posted extensive details on there regarding the stripping down and repair of every aspect of this machine, which you will find useful. The big old motor fitted to yours was a n option when new. A common problem with these is the clutch, or specifically wear in the pins, worth checking and an easy repair if needed. You also want to carefully check the state of the friction surfaces on the brake, they can sometimes wear right down. Mine is a 1961 5A. The second picture is of the magnetic sump plug I added to mine to catch any particles. In the trough over the main bearing feed hole you should hopefully have a very fine mesh gauze, intended to trap anything that might otherwise find it's way into the bearings. Good luck with the rest of it, if you get stuck with anything then only too pleased to help. The owners group is well worthwhile, and a good source of spares too like your gear, although you have made a nice job of repairing it.

 

[Fergie 307]

I would urge you not to take the headstock apart. The spindle and associated parts are all dynamically balanced and have to be kept in the correct orientation. It's an easy enough strip down but putting it back together again correctly is another matter, so unless there is something specifically wrong with it I would leave well alone. When you get to the saddle you will find there is no adjustment for the mounting strips which hold it on the bed, one long one at the back and two short ones at the front, the right hand one of which also acts as the saddle lock. If there is play between the saddle and bed this is usually down to wear in the strips. These are deliberately made of fairly soft iron. The fit was adjusted at the factory by scraping the underside mounting area on the saddle casting itself, the strips being completely flat. The beauty of this arrangement is that to restore the original clearance all you usually have to do is to regrind the strips themselves flat to remove any slight step worn in them where they run on the underside of the ways. Have fun.

 

[Sideways]

Excellent insights Fergie. I didn't know those details.
My L5 doesn't have the mesh filter you mention or marks from having had one, but there are now a dozen flat neodymium magnets around the inside so I'm onboard with that method.
My big motor that I removed wasn't an original. It was 1980's, pre owned but not vintage
Your L5a looks great. Very clean, better gearbox and I like the splashback.
I'm registered for the groups.io Harrison mailing list which is mildly helpful but very limited because it doesn't intend to be a forum and has v limited space for image storage.
If there's another forum I'm unaware of it. I don't do facebook in any shape or form so if there is something there, I'll have to do without.
Please continue to advise and point out my mistakes as I post the rest of the cleanup !

 

[Sideways]

Here's the gearbox:

The L5 is available with either this simple 3 speed version or a better Norton gearbox with more ratios. The 3 speed has 1 : 2 : 4 (or 1/2 : 1 : 2 ?) relative speeds and after cleaning up the apron, it seemed lazy not to strip and clean it. It bolts straight through onto the front of the headstock. The drive comes in at the left hand end via the change gears, and it drives the two long shafts from its right hand end.

No synchromesh here. Change gear when the lathe is stopped.
Lube is via oiling nipples on the front, on the two output shaft bearings and importantly, one on the left, inside the changewheel case and obsured by the lower of the 3 changewheels. This is why it can be a good thing to have a list of lubrication points and tick them off as you go. Here's the input end:

Once unbolted, the back of the gearbox is open and the mechanism simple

Points of interest. The lead screw (upper output shaft) is only used for threading. It doesn't need to be turning most of the time so the L5 has a simple sliding dog clutch that can be slid 1/2" left or right along the shaft to engage and disengage the lead screw from the gearbox. Quite neat.

To reduce the damage in case the operator loses attention and crashes the tool while surfacing under power, there is a spring loaded clutch inside the thick stubby shaft on the lower left of the gearbox in the pic above. This should prevent an accident stripping teeth in the gearbox, although as we saw, it didn't prevent a tooth being lost inside the apron.
Here you can see inside that mechanism.

The upper shaft of the gearbox is partially keyed. Gears basically thread on and where necessary are held in place with taper pins through the gear and shaft.
Kindly, the gears were already stamped making assembly a case of pairing the numbers.

To select gear, the cluster of 3 gears on the lower shaft is simply moved between left, centre and right by a fork attached to the gear selector spindle.

All the bearings in the gearbox are plain. No ball bearings. Steel shafts run in bronze bushes.

Here's the output end of the gearbox after refitting. The clutch for the feed shaft is below and the socket for the screw is above, showing the drive dogs ready to mate with the selector.

And from much later in the reassembly, this is what it looks like with the shafts refitted and the sliding dog clutch engaged (this has a spring loaded ball bearing to keep it in the chosen position and the key is PINNED, which makes it quite a challenge to remove !

Overall the gearbox was simple and just needed a good clean. The spring loaded clutch and dog clutch were interesting to figure out. Reassembling the spring clutch involved compressing the spring inside so that one (two ?) tiny countersunk capscrews could be screwed home to hold it together. On the feed screw, the inch long piece of key steel that transfers the drive from the sliding dog to the shaft has a pin through it, and that all needed flattening and filing as this wasn't obvious and I had bent it trying to get it out.
The numbered gears and use of taper pin meant that the reassembly of the gears was very straightforward. But it still makes sense to photograph EVERYTHING as you take it apart ....

Next up will be the saddle and slides

 

[Spectric]

A good old solid machine, they were common in the school metalwork shops and found in many workshops in the days when we had large manufacturing in the UK. The gearboxes were better made than what you find on modern cars.

 

[Sideways]

Saddle and cross slide

Back in 1949 this lathe would have been made in imperial. I received it with metric dials fitted and I assume that in it's life as a training lathe, it would have been converted some time before or after decimalisation. The gearbox and changewheels allow either imperial or metric threads to be cut depaending on the combination of change gears, but converting the dials would mean new screws and nuts in the slides.

Firstly the saddle. The L5 mk3 doesn't have hardened ways. just cast iron on cast iron. The larger section on the L is to the front and the apron with it's gears hangs from this.

A good clean was needed and at some point, wire brushing the scarred paintwork on the top of the saddle turned into a full strip of the paint and re dressing the casting to remove the worst of the grind marks etc left after manufacture. Using a blue Norton vortex blending disk I cleaned up the saddle and although the casting wasn't perfect, I decided that I was going to leave the top surfaces bare.

Lathes always get scratched up after a while from brushing away the swarf so I decided that this was going to be maintained steam punk style from now on. Wiped down with an oily rag and covered when not in use.

In the image above you see that the cross slide slides on dovetail ways and that a tapered gib (cast iron or steel) is used to take up the play and give a good sliding fit. Two large chrome headed screws set the position of the gib strip to keep it snug and to stop it wedging itself even tighter.
The wear is at the limit of what's usable. I need to make a new screw for the back as there isn't enough thread to let me use the last 1/4" of adjustment left in the gib.
These are a carefully matched taper so making a completely new gib will be a challenge best done with Deema's surface grinder once he finishes that rebuild, and we both need to learn to scrape these in. That will be a big month when the time comes.

The cross slide screw looks in good condition.

The brass nut is a nice fit and has a slot so that it can be tightened to take out backlash as it wears. This is interesting. It isn't the original L5 imperial nut. In fact I think it's a nut from a later model L6. The L5 mk3 wasn't made with a split nut and this is a clue to the metric conversion.

The cross slide screw simply slides into a plain bore in the cast iron of the saddle.
A disc is screwed on to provide the index mark for the graduatel dial, then a sleeve slides over the shaft and is secured with a taper pin. This is filed flush with the sleeve.
There are no special baring surfaces at all, just steel on steel or cast iron and some holes above to allow oil to be dripped into some of the sliding surfaces.

This isn't a sophisticated arrangement, so we decided to improve it by machining the index plate and adding a slim roller type thrust bearing.
After some careful measuring and lathe work because there's no facility to adjust clearances to add or remove axial play it turned into this. A satisfying mod !

 

[Sideways]

Saddle and slides 2

Here you see the cross slide nut as the lathe was received. A simple round boss (round tenon) on the nut fits snugly in a blind hole in the underside of the cross slide. A countersunk screw keeps it in place and allows it to turn to align with its lead screw.

Unfortunately, the lathes must have been hand fitted when made, and this wasn't done correctly when the replacement metric screw and nut were installed.
As the cross slide was wound in and out, the nut would run freely then bind as it came closer to the front.
We measured and found the nut was offset to one side and wasn't in line with the screw. As the screw was snugly fitted in a long hole at the handwheel end, it was nicely aligned in the saddle, but the mounting boss on the metric nut was offset about 1.5mm to the side compared to the hole it needed to fit in the cross slide. The closer it was wound forward, the more it tried to bend the screw and eventually jammed.
I think the lathe must have been like this ever since it's metric conversion. The only way to make it work at all was to leave the gib sloppy and use only the far end of it's travel (small parts).

The nut was otherwise good and a replacement would guarantee nothing, so we devised a fix involving a top hat insert with an eccentric bore so that the lateral error could be dialled out. Dimensions allow about 2mm of sideways adjustment. This was enough with a bit to spare and it worked well. You need a lathe and a mill to fix either of these tools but it made another good day

Once done, the cross slide could be wound fully forward and back without binding