OLLY, my mate !! I'm so glad you ran with this idea ! From a verbal where we bounced ideas off each other, you came up with a working model. You are right, it didn't work. But I can see so much more now you have something physical not just an image in my head. Again after seeing this wonderful work, I am thinking banding or timing belt material instead of cable. Think you are onto a winner, I really do ! Don't quit on this, I honestly think this will work. Sorry for not being around. Health is a issue and haven't been in the workshop this year. After surgery I'm hoping to be able to spend time making instead of thinking! All the very best Nell 😊😊😊
Hi Neil, hope you're doing well. Good to hear you're still kicking! I've had loads of thoughts, hydraulic transfer, solid wire, even a recirculating ball system. This was the most practical, solution, but needs a lot of tuning to get it to work. Don't worry, the old brain doesn't give up that easily, so perhaps I'll stumble upon the silver bullet at some stage. 😁👍
@@ollysworkshop see! Hydraulic transfer was and is an idea I'm working on for tattoo machines ! That's another story for another day 😊. I really like what you have done Olly. Backlash in directional changes is the issue. The 90⁰ transition is the issue. I'm still thinking metal style packing strap and pinch rollers or speedo cable ???? One of your subscribers will pull the bunny out of the hat. Wait and see! I planted the seed, you grew the idea. Someone will harvest 🤔🤔🤔😊😊😊
Thanks so much for this - you've saved a lot of us a lot of time. I often think up "creative" solutions that generations of real engineers have long since abandoned for very good reason!
It's a flawed design, but it was drawn up in a few hours. I'm sure if man years were spent on it, a solution could be found, but in engineering the compromise usually wins.
Hi Olly, this was a very interesting exercise and maybe there is still scope to make it work. I think your efforts would be better focussed on traditional positions for the DRO though. That said, it did remind me of the chart recorders that I was involved in making before they made a jump into graphic display with printed traces. These models had a paper tape indicator, coloured on one half such that the colour divide coincided with zero at one extreme and full scale at the other. Behind this you had cord which was wrapped around a bobbin and worked consistently, of course using this kind of system in a lathe environment would be too harsh for it I think. Not too dissimilar to the radio as already mentioned. Definitely no time wasted, well done for giving it a go. All the best!!
I like the hydraulic transfer idea, although not sure how to make that work, the I thought about a geared system and then thought there must be some sort of radial encoder too!!
Brilliant 👍. I'd never have thought of that in a million years. A similar drive cord system works perfectly between the scale and capacitor in old radios, so I guess on the lathe it just needs tuning (no pun intended). Cheers Olly 👍
Gday Olly, bloody shame this didn’t work out, it would’ve been a perfect solution for many lathes, really it’s a brilliant idea, Neil definitely had a creative mind and I hope to see more videos pop up over time on his channel, great video mate, cheers
Thanks Matty. It was certainly something I wanted to try, I knew it would have backlash, but was eye opening to see quite how much. I could spend a long time refining it, but every step in the mechanical chain adds some inaccuracy which is unavoidable.
G'day Olly. Kudos mate, that was a bloody good idea mate. I have a similar problem with my colchester as I need a very slim scale to fit it. Hence why it still doesn't have DRO just yet. Cheers, Aaron.
Hi Aaron, it's a really "nice in theory" idea, and I wanted to do it. It could be made much better with a stronger spring (or possibly a gas strut as they have constant spring rate), or even with a continuous loop of wire that both pushes and pulls. There's been loads of great ideas in the comments, so maybe one day I'll get round to trying them.
I like the idea. It must be able to be solved or the reel type scales wouldn't work at all. Better wire options may be the stranded stainless wire used as gear cable replacement on mountain bikes, or, pike trace , even thinner stranded stainless wire used for the first few feet of a fishing line for fish with sharp teeth!.
Hi Peter, thanks for the ideas. The original wire I used was gear cable from a mountain bike, but I'll look into the stainless fishing wire, that sounds promising. There are many areas of possible improvement on this, not least is a stronger spring, but I just built it with what I had lying around. Ultimately, if the spring is strong enough it would overcome any stiction and slop in the system, but it may make it hard to turn the handwheel!
I think it wloul be easier to just read the dials instead of installing a DRO. But, of course, that's not a challenge :) This is a brilliant idea and with a bit more thinking, it would work just fine.
I encountered an application that had a push/pull hydraulic system to create an accurate and positive, backlash-free tool feed feed movement. No cable, no screws. You could still use your feed screw, the hydraulic would only move the measuring parts.
I understand the principle, makes perfect sense. But in practice I don't see how it could ever be guaranteed 100% bled free of air. Air would cause hysteresis in the measurement. You say this was actually implemented somewhere, what kind of machine was it? Was it accurate enough for this application? Like within .00001"?
@@charlesstaton8104 The original question as I understood it was, is there another way other than cables. I'm saying there is another way. I believe I did not say the method was developed. The systems were used on a wide variety of Wickes Crankshaft lathes (circa 1940? - 1970?) to rough out , and semi finish forged steel crankshafts prior to heat treat and finishing. The system is a platform for smooth control of motion by using opposed hydraulic devices. The idea is in control of motion and precision is in the measuring devices (scale, pick-up head). To answer the questions that come next, I would begin with: push a thin fluid in a small diameter piston (attached to the tool slide) in a small diameter cylinder through a tube into a matching cylinder that is attached to a head and scale to see what happens.Measure travel at the slide with an indicator compared to the the DRO. Going forward, I require recognition for the design and compensation where appropriate. Additionally, a small tank is required to supply fluid to both cylinders to accept fluid from one cylinder being pushed by the other, check valves to prevent back flow and added tube. The supply tank would be vented to room air, no pump is required. Movement of the tool slide would push to fluid back and forth resulting in travel of the reader head. Feed the tool in, head moves the read head in. Return the tool slide out, the read head moves back. All displayed on the DRO.
@@charlesstaton8104 The supply tank allows all air to rise to the surface and dissipate. There is no chance of cavitation to form vacuum fracture in the fluid because of the very low speed of the fluid. Once air remaining after assembly is bled, it will be bubble free. No pump, slow fluid movement.
Perhaps you could do a conversion like they do for a myford ML7 or Super seven. Mill a slot in your cross slide and glue a magnetic strip there. I bought a kit fro M-DRO ,but I'm sure you could make one yourself.
I think every moving part adds some backlash, the fundamental physics can't be changed. Even the optical scale readers have backlash specified in the data sheets +/- 5um for the decent ones.
Your idea to convert to push/pull (pull/pull) does not have to run up inside the cross slide. As it is only a wire (2 wires), the interference with the tailstock would be insignificant compared to the scale.
needs to be push-pull... two cables tensioned against each other. as you will see in applications like exhaust servo valves, throttle cables, or ailerons etc. as it isnt under load, mere fishing line will do... braided filament? lot more flexible. the cable has to pull on the slide in parallel, in line as it comes off the pulley. the angle gives a trig error... and places excessive load on the slide. using a rotary encoder, and a timing belt, with the same "stationary belt moving pulley" setup you see on larger gantry style machines... resolution will be dictated by the pulley diameter. as well as encoder... a small geared stepper will possibly achieve better resolution, but will need signal conditioning to give decent logic levels. how many pulses can you get per turn and how does that equate to travel?
I covered the push-pull idea, used Kevlar thread, and pointed out the cosine error in the video. A rotary encoder is certainly one way to go with this. I'm not sure it's worth struggling with stepper motors though.
I think you gave up too soon. I know, easy for me to say, you obviously put in many many hours already, but hey just doing my job as a spectator. If you pull the cable straight out of the pulley and eliminate the cosine error as you mentioned, i think that is half the problem. The other half i think is not enough tension. I would try a MUCH stiffer spring, something strong enough to actually preload the nut against the leadscrew. Might have to double up or triple up on your aramid yarn. If that doesnt work then yeah, you will need to close the loop like you metioned. But I like the spring because (if stiff enough) it turns your acme nut into a zero backlash nut. If you're going to close the loop and use timibg belt you might as well ditch the linear scale and just use a rotary encoder. It would take up a lot less room. And the belts could be routed through those places where you didn't want a linear scale because it might get damaged. It would just need a low profile shield to keep chips off it.
Hi, thanks for the comment. I reckon I'll revisit this, but for the moment I've "given up" because I want to get the DRO fitted, and the lathe back in service. I agree a much stronger spring would be good, I was just working with what I had, that'll be an easy one to try. I also like the hydraulic idea that's been talked about, more just because I like it rather than it being the right solution. I would also like to try a rotary encoder. I'm really good at starting projects, less good at finishing them, so we'll see what happens.
OLLY, my mate !!
I'm so glad you ran with this idea !
From a verbal where we bounced ideas off each other, you came up with a working model.
You are right, it didn't work. But I can see so much more now you have something physical not just an image in my head.
Again after seeing this wonderful work, I am thinking banding or timing belt material instead of cable.
Think you are onto a winner, I really do !
Don't quit on this, I honestly think this will work.
Sorry for not being around. Health is a issue and haven't been in the workshop this year.
After surgery I'm hoping to be able to spend time making instead of thinking!
All the very best
Nell 😊😊😊
Hi Neil, hope you're doing well. Good to hear you're still kicking! I've had loads of thoughts, hydraulic transfer, solid wire, even a recirculating ball system. This was the most practical, solution, but needs a lot of tuning to get it to work. Don't worry, the old brain doesn't give up that easily, so perhaps I'll stumble upon the silver bullet at some stage. 😁👍
@@ollysworkshop see! Hydraulic transfer was and is an idea I'm working on for tattoo machines !
That's another story for another day 😊.
I really like what you have done Olly. Backlash in directional changes is the issue. The 90⁰ transition is the issue. I'm still thinking metal style packing strap and pinch rollers or speedo cable ????
One of your subscribers will pull the bunny out of the hat. Wait and see!
I planted the seed, you grew the idea. Someone will harvest 🤔🤔🤔😊😊😊
Thanks so much for this - you've saved a lot of us a lot of time. I often think up "creative" solutions that generations of real engineers have long since abandoned for very good reason!
It's a flawed design, but it was drawn up in a few hours. I'm sure if man years were spent on it, a solution could be found, but in engineering the compromise usually wins.
Hi Olly, this was a very interesting exercise and maybe there is still scope to make it work. I think your efforts would be better focussed on traditional positions for the DRO though.
That said, it did remind me of the chart recorders that I was involved in making before they made a jump into graphic display with printed traces. These models had a paper tape indicator, coloured on one half such that the colour divide coincided with zero at one extreme and full scale at the other. Behind this you had cord which was wrapped around a bobbin and worked consistently, of course using this kind of system in a lathe environment would be too harsh for it I think. Not too dissimilar to the radio as already mentioned.
Definitely no time wasted, well done for giving it a go.
All the best!!
I like the hydraulic transfer idea, although not sure how to make that work, the I thought about a geared system and then thought there must be some sort of radial encoder too!!
Brilliant 👍.
I'd never have thought of that in a million years. A similar drive cord system works perfectly between the scale and capacitor in old radios, so I guess on the lathe it just needs tuning (no pun intended). Cheers Olly 👍
I was definitely channeling classic AM radio tuner on this one! Unfortunately this one has tuning drift 😁
Gday Olly, bloody shame this didn’t work out, it would’ve been a perfect solution for many lathes, really it’s a brilliant idea, Neil definitely had a creative mind and I hope to see more videos pop up over time on his channel, great video mate, cheers
Thanks Matty. It was certainly something I wanted to try, I knew it would have backlash, but was eye opening to see quite how much. I could spend a long time refining it, but every step in the mechanical chain adds some inaccuracy which is unavoidable.
Just catching up with your videos after a busy year. This was a fantastic idea for the scale
Thanks. It was a fun experiment.
Hi Olly. That was an interesting and enjoyable fail. Overcoming backlash is indeed the curse of that design. I hope you find a solution. Good luck.
Hi Andrew, definitely more backlash than I was expecting, but I didn't really know how to quantify it without building it.
bloody marvellous idea Olly and Neil, unlucky, back to the drawing board
new ideas always need to be tried, hard luck
see you next time
atb
kev
I knew it would have backlash issues, but wasn't expecting quite so much! I needed to do it to it to find out. Cheers Kev
@@ollysworkshop keep working on it, there's usually a solution👍👍👍👍
G'day Olly. Kudos mate, that was a bloody good idea mate. I have a similar problem with my colchester as I need a very slim scale to fit it. Hence why it still doesn't have DRO just yet. Cheers, Aaron.
Hi Aaron, it's a really "nice in theory" idea, and I wanted to do it. It could be made much better with a stronger spring (or possibly a gas strut as they have constant spring rate), or even with a continuous loop of wire that both pushes and pulls. There's been loads of great ideas in the comments, so maybe one day I'll get round to trying them.
I like the idea. It must be able to be solved or the reel type scales wouldn't work at all. Better wire options may be the stranded stainless wire used as gear cable replacement on mountain bikes, or, pike trace , even thinner stranded stainless wire used for the first few feet of a fishing line for fish with sharp teeth!.
Hi Peter, thanks for the ideas. The original wire I used was gear cable from a mountain bike, but I'll look into the stainless fishing wire, that sounds promising. There are many areas of possible improvement on this, not least is a stronger spring, but I just built it with what I had lying around. Ultimately, if the spring is strong enough it would overcome any stiction and slop in the system, but it may make it hard to turn the handwheel!
@@ollysworkshop The cable you used looked like the normal galvanised hi tensile stuff. the stainless is more flexible. cheers
Great! I’ve the same issue with my lathe and had purchased a similar dro! I look forward to your solution!
The brain will keep thinking about it. In the meantime I'm just going to push the splashback further back!
@@ollysworkshop I really like the out of box thinking. I was resigned to putting the dro sensor on the chip side. You’ve given me hope!
Very creative idea
Thanks!
I think it wloul be easier to just read the dials instead of installing a DRO. But, of course, that's not a challenge :)
This is a brilliant idea and with a bit more thinking, it would work just fine.
The thinking will probably continue, even if the project does not.
I encountered an application that had a push/pull hydraulic system to create an accurate and positive, backlash-free tool feed feed movement. No cable, no screws. You could still use your feed screw, the hydraulic would only move the measuring parts.
That's definitely something worth trying then, thanks for the comment!👍👍
I understand the principle, makes perfect sense. But in practice I don't see how it could ever be guaranteed 100% bled free of air. Air would cause hysteresis in the measurement. You say this was actually implemented somewhere, what kind of machine was it? Was it accurate enough for this application? Like within .00001"?
@@charlesstaton8104 The original question as I understood it was, is there another way other than cables. I'm saying there is another way. I believe I did not say the method was developed.
The systems were used on a wide variety of Wickes Crankshaft lathes (circa 1940? - 1970?) to rough out , and semi finish forged steel crankshafts prior to heat treat and finishing. The system is a platform for smooth control of motion by using opposed hydraulic devices.
The idea is in control of motion and precision is in the measuring devices (scale, pick-up head).
To answer the questions that come next, I would begin with: push a thin fluid in a small diameter piston (attached to the tool slide) in a small diameter cylinder through a tube into a matching cylinder that is attached to a head and scale to see what happens.Measure travel at the slide with an indicator compared to the the DRO.
Going forward, I require recognition for the design and compensation where appropriate.
Additionally, a small tank is required to supply fluid to both cylinders to accept fluid from one cylinder being pushed by the other, check valves to prevent back flow and added tube. The supply tank would be vented to room air, no pump is required. Movement of the tool slide would push to fluid back and forth resulting in travel of the reader head. Feed the tool in, head moves the read head in. Return the tool slide out, the read head moves back. All displayed on the DRO.
@@charlesstaton8104 The supply tank allows all air to rise to the surface and dissipate. There is no chance of cavitation to form vacuum fracture in the fluid because of the very low speed of the fluid. Once air remaining after assembly is bled, it will be bubble free. No pump, slow fluid movement.
Perhaps you could do a conversion like they do for a myford ML7 or Super seven. Mill a slot in your cross slide and glue a magnetic strip there. I bought a kit fro M-DRO ,but I'm sure you could make one yourself.
Hi, that's certainly something worth looking at, thanks. 👍
I think every moving part adds some backlash, the fundamental physics can't be changed. Even the optical scale readers have backlash specified in the data sheets +/- 5um for the decent ones.
Yes, too many moving parts.
@@ollysworkshop good to try things out nonetheless. If you're chasing microns lookup laser interferometer encoders 💸
Your idea to convert to push/pull (pull/pull) does not have to run up inside the cross slide.
As it is only a wire (2 wires), the interference with the tailstock would be insignificant compared to the scale.
Good point. I'll have a bit of a think on that one. Thanks 👍
needs to be push-pull... two cables tensioned against each other. as you will see in applications like exhaust servo valves, throttle cables, or ailerons etc.
as it isnt under load, mere fishing line will do... braided filament? lot more flexible.
the cable has to pull on the slide in parallel, in line as it comes off the pulley. the angle gives a trig error... and places excessive load on the slide.
using a rotary encoder, and a timing belt, with the same "stationary belt moving pulley" setup you see on larger gantry style machines... resolution will be dictated by the pulley diameter. as well as encoder...
a small geared stepper will possibly achieve better resolution, but will need signal conditioning to give decent logic levels. how many pulses can you get per turn and how does that equate to travel?
I covered the push-pull idea, used Kevlar thread, and pointed out the cosine error in the video. A rotary encoder is certainly one way to go with this. I'm not sure it's worth struggling with stepper motors though.
I think you gave up too soon. I know, easy for me to say, you obviously put in many many hours already, but hey just doing my job as a spectator.
If you pull the cable straight out of the pulley and eliminate the cosine error as you mentioned, i think that is half the problem. The other half i think is not enough tension. I would try a MUCH stiffer spring, something strong enough to actually preload the nut against the leadscrew. Might have to double up or triple up on your aramid yarn.
If that doesnt work then yeah, you will need to close the loop like you metioned. But I like the spring because (if stiff enough) it turns your acme nut into a zero backlash nut.
If you're going to close the loop and use timibg belt you might as well ditch the linear scale and just use a rotary encoder. It would take up a lot less room. And the belts could be routed through those places where you didn't want a linear scale because it might get damaged. It would just need a low profile shield to keep chips off it.
Hi, thanks for the comment. I reckon I'll revisit this, but for the moment I've "given up" because I want to get the DRO fitted, and the lathe back in service. I agree a much stronger spring would be good, I was just working with what I had, that'll be an easy one to try. I also like the hydraulic idea that's been talked about, more just because I like it rather than it being the right solution. I would also like to try a rotary encoder. I'm really good at starting projects, less good at finishing them, so we'll see what happens.