When does it start violating the rules, IF I were to....

[TorontoBuilder]

Yes. It was a kit designed to save the end user money by sourcing many parts locally in their country. I sold very few in Canada. Others went to Australia, UK, Europe, South Africa etc. To ship motors to Canada and then reship to another country was by consensus way too expensive.

So users bought the following themselves.
1. 12V AC adaptor for the ELS
2. Appropriate power supply for their motor systems
3. Appropriate motor, couplers or belts and pulleys etc for their unique system.
4. Either optical or hall sensor for the 1 PPR spindle.
5. A box of some sort for the ELS and for their motor hardware.
6. I was selling keypad overlays. A friend on Saltspring made a silk screen for the white mask. I can now provide that here since I have the laminater and laser printer.
7. Also recommended was some sort of break out board to simplify swapping back and forth between a PC for MACH3 or LinuxCNC and the ELS. This way you can have an quadrature encoder connected to the lathe but use a 1 PPR index pulse for the ELS and MACH3 but full quadrature for LinuxCNC if wanted.
Note that to combat electrical noise the ELS requires a long index pulse rather than that short one from full quadrature encoders. There are ways to make that work though.

I now supply a 3D printed knob for the MPG. It's a detented 16 pulse per rev (64 units per rev in quadrature) with a momentary button when you press down on it. You can select 00.001 or 0.010 etc for distance per click. I bought enough of those to make all 200 boards. A few years ago that model of encoder was discontinued so a new PC board would be required anyway.

Or, since there's a spot on the board for a keypad connector one could built the board without the buttons and make a separate board that has more than 35 buttons and a panel mounted MPG. The LCD pinout is also designed for up to a 4 line unit but that does require other software.

I built just this one to demo the concept. Haven't looked further into LCD displays with 4 lines. Between the ENTER and the L-JOG button is the keyboard expansion strip. Wires or header pins could be soldered where the MPG goes for an external encoder from China.

It's no longer worth the money to put the stepper motor driver right on the PC board.
View attachment 28949

One project I did with it added this relay and optically isolated input board. Also had A/D and stationary battery current sensor. It plugged into the jumper block on the bottom of the ELS.
View attachment 28950
And was designed to also run stand alone with it's own processor and RS232 + CAN bus connection.

View attachment 28951

IMO one of the biggest barriers is leaving so much to the end user.

Many people who would otherwise be interested dont have or want to develop soldering skills. Nor do they want to figure out break out boards etc. Plug and play is a necessity for wider adoption rates. The german fellows seem to get that pretty well

 

[jcdammeyer]

IMO one of the biggest barriers is leaving so much to the end user.

Many people who would otherwise be interested dont have or want to develop soldering skills. Nor do they want to figure out break out boards etc. Plug and play is a necessity for wider adoption rates. The german fellows seem to get that pretty well

Problem is you have to go back in time to see why things were done this way.
In 2006 a stepper motor driver was over $100US. The motors were also in the $80 to $120 range.

If I wanted a hi res encoder on my South Bend it had to have a 1.65" bore to be able to mount onto the outside of the spindle hub. Pulleys and belts large enough to do that were also in the high end dollars.
Now I did find a large encoder disk with I think 360 lines for only $75 US. Very exciting until I found I had to order 100 of them... Standard panel mount encoders were running from $100 surplus to $800 new.

And the PIC family of processors I was using back then didn't handle quadrature. The ELS has the software and did have the chips to have a microstepper driver built in giving 1600 steps per rev and getting rid of stepper motor resonance. But now the chips cost more than a far east driver.

If next year China decides to invade Taiwan and the US decides to defend the ramifications of cheap hardware from the far east goes down the toilet.

One of the products I manufacture now has a full 1 year lead time on the M9S12 processor. My client is willing to let me remove them from other hardware and build onto theirs. When you make a piece of equipment worth $1 million per unit and it's held up by what used to be a $18 processor that now costs $65 and takes a year to get attitudes change.

 

[gerritv]

IMO one of the biggest barriers is leaving so much to the end user.

Many people who would otherwise be interested dont have or want to develop soldering skills. Nor do they want to figure out break out boards etc. Plug and play is a necessity for wider adoption rates. The german fellows seem to get that pretty well

Yep, Rocketronics have thought through the connection methods for the most part. But the Rocketronics, like all the other options available, still requires you to decide how to mount the motors and connect to machine axis. And you still need to decide which motors and drivers to buy. Bigger motors is not alwyas better so a bit of decision making, guessing and hoping are required. And you won't know until you try it and have tuned the paramaters whether you got it close enough. And until you try 10tpi or coarser threads you still won't be sure in my experience with 2 different controllers.
And chances are you need to drill holes in lathe or machine leadscrew extentions etc.
I can see coming up with a kit for a specific style of machine, e.g the ubiquitous G0602/G0752/PM/King 1022 lathes as there mostly the same. But that setup won't work the same for e.g. a Standard Modern 2000.

 

[jcdammeyer]

Yep, Rocketronics have thought through the connection methods for the most part. But the Rocketronics, like all the other options available, still requires you to decide how to mount the motors and connect to machine axis. And you still need to decide which motors and drivers to buy. Bigger motors is not alwyas better so a bit of decision making, guessing and hoping are required. And you won't know until you try it and have tuned the paramaters whether you got it close enough. And until you try 10tpi or coarser threads you still won't be sure in my experience with 2 different controllers.
And chances are you need to drill holes in lathe or machine leadscrew extentions etc.
I can see coming up with a kit for a specific style of machine, e.g the ubiquitous G0602/G0752/PM/King 1022 lathes as there mostly the same. But that setup won't work the same for e.g. a Standard Modern 2000.

I even went into BusyBee Tools and tried to talk to them about making a kit for I think their 8x24 metal lathe. They weren't interested or they just didn't have the authority to even attempt/promote something like that.

Thing is, the Busy Bee, KMS and KBC lathes all come with gears to do metric and imperial threading so they don't really need an ELS because a DRO is just as good.

The original purpose of my ELS was to add metric threading to imperial lathes or add thread pitches for which there were no gears; my Gingery home made lathe for example. The cross slide control and tapering came in the second revision after many of the users on the group wanted that feature. And that's something I've not yet added to my South Bend.

 

[DPittman]

Thing is, the Busy Bee, KMS and KBC lathes all come with gears to do metric and imperial threading so they don't really need an ELS because a DRO is just as good.

My lathe has the ability to thread both metric and imperial threads with the change gears and I also have a DRO but I'd still love an ELS. How is a DRO as good as a els? Not trying to be arrogant here I'm just learning.

 

[jcdammeyer]

That's a really good question. Really good!

Another good one is if I have an ELS on the South Bend Carriage why not on the cross slide? The South Bend has a mechanical Taper Attachment and power cross feed driven from the same slotted lead screw.

This hub was within the reach of the compound slide travel but I think the carriage and taper attachment left a nicer finish.


So most operations can be done with existing hardware. And if you have one of these:

Blind holes to exact depth only require the setup and then release the half nut or clutch on time and turn the last bit with the hand wheel up to the micrometer end stop.

What I like about my ELS is that when I bore a blind hole it stops. I can turn up to the collet or chuck jaws and have it stop instantly. When I thread it stops without me having to be lightening fast on the half nut.

When I face the end and then want to turn to a shoulder at 1.25" I set the Z=0.000 where I faced and set the END position at -1.250". The BEGIN just in front of the face say 0.2" to give the system time to get up to speed and take out backlash.

But on my lathe it's unlikely it's exactly 1.250" and without a DRO I wouldn't know because my lead screw and ways are worn. Cutting the 10 TPI ACME thread for my Gingery Lathe Lead screw I had to tweak the cross slide as the carriage traveled along the 16" or so. Just to try and keep the depth of cut constant.

Yes, my ELS has DRO output but it's based on how far the motor has moved the lead screw. Not how far the lead screw has mechanically moved the carriage. I'm finding that issue on my mill too. My Shumatech DRO tells me when LinuxCNC has not taken out the backlash correctly. Would I be without LinuxCNC now?
Nope.

 

[Susquatch]

How is a DRO as good as a els?

I "THINK" one of the many things that @jcdammeyer was trying to say is that you can do the same things with both - it might take a bit longer manually though.

I am a hobbiest. Ya, my farm and even my neighbours farms need my machining and other skills, but they are not my bread and butter. Unlike many others, my time is not an issue either. I enjoy machining and I don't do production runs so I really don't care how long it takes me to make something. And if I enjoy doing it, why would I let a computer do it for me and take my fun away?

I know that I'm a minority here, but I even like changing gears to cut threads. Cutting a new thread pitch is just another opportunity to change my gears! Woooo Hoooo!

As an old man and a hobbiest, I look at life a bit differently than some. CNC & ELS are both wonderful tools. But I've been able to do everything they could have done for me without them so far. I have simply learned to enjoy the manual journey as much as those who have CNC & ELS enjoy the time these electronic wonders give them to do something else.

Maybe that's oversimplifying, but that's how I feel about it and I think that's sorta what @jcdammeyer was getting at too.

Of course, some smart ass is gunna come up with a huge giant list of all the things you can do with CNC & ELS and can't do manually. Like a 16.6 tpi thread pitch, etc etc. Until the day comes when I need that, I really don't care. More importantly, I don't want to care. The truth is that I don't expect to live long enough to care either.

 

[6.5 Fan]

I must disagree Susquatch, i do not like mucking about with change gears. My big lathe has a gear box for threads and feed rates, i don't have a need for speed when i'm machining but 'tis a whole lot more convenient than trying to read a chart then fitting gears to greasy shafts. Just my opinion and worth what i was paid for it. Things would be very boring if we all had the same thought process.

 

[gerritv]

Some good points. Indeed most hobby-sized Asian lathes can do m and i threadng. But once set up for threading, it is useless for feeding and vice versa. As a result, due to my 'work' pattern, I never used power feed. Tapers were limited to short ones, if I took the time to put my compund back on. And the coarser the pitch, the more passes and too often opportunity to mess up.
Even with the not-so-totally-great Russian ELS I could move from one op method to another with a few buttons. so now power feed was in my arsenal. Now with the Rocketronics I have boring, threading, parting, tapers, ball (convex and concave). carriage drilling, keyway cutting (drilling cycle with spindle locked), and all of these to a set diameter and distance with control over DOC and passes. Internal threading to a shoulder is now possible, my reaction time was too often too short.

All things I could do with just my DRO or even dials and a carriage stop but as I get to making smaller and smaller parts, the less I can emery cloth things to dimension
To keep my interest in full manual I have 2 watchmaker lathes and a Taig so I won't forget those skills I only started machining in earnest 6 years ago, and at my age it is too late to develop full manual skills, and my arthritic hands make it a bigger challenge every year. This is for my why ELS of any kind or brand keeps things possible.

The only downside so far with the Rocketronics is metric only, but they confirmed after asking that the inferial feature is being finalized over the Christmas break.

 

[historicalarms]

The feed speed transmission on my lathe is one of the features that gets used often in a session, I will change feed rates on the go, sometimes multiple times in one pass if i require different quality finish on different parts of a shaft ...change gears would take me forever on some jobs.
I realize every operators manual published for every lathe built says "don't change gears under power" but it can be done easily without any gear clash or jamming on the cuts i make with my lathe, if your cutting 50 thou at a pass the cut pressure may make it a bit more difficult but for me just pull the pin, slide over and lift until pin falls into hole. firm precise movements will result in no gear clash at all.....if we can shift an 18 sp driven by 600 hp, pulling a 90,000 lb load 150 times a day without using the clutch you damn sure can shift a lathe feed transmission LOL

 

[Dabbler]

Like @gerritv I primarily use manual feed, even though my lathes have very easy options for feeding. For me it is the tactile feedback and the focus of machining that make it so much fun. Taking longer isn't a downside, but... In that regard I'm like @Susquatch.

However I'm interested in ELS, because of eye-hand coordination, and some concerns about concentration as I age. I've never crashed a lathe, and really don't want the initiation.

Besides I'd love to do the project, purely because it is cool project... A useful cool project.

 

[jcdammeyer]

I agree with what @gerritv is saying. Especially if the lathe in question requires changing gears to go from turning to threading.
I didn't want to go through the process of first making the Gingery shaper and then the Gingery mill in order to create the Gingery dividing head to make gears for the Gingery lathe. Hence my ELS since the only other one out there was the FROG which had been discontinued due to lack of sales.
You can see in this photo there is one feed speed using the sewing machine O-Ring belts with pulleys right from the Gingery lathe book. And even there I modified the original with a parallel set of screws to prevent the flat plate from twisting.

http://www.autoartisans.com/gingery.htm

Hence the ELS was a pretty well must have. As yet I haven't put one on my Unimat DB200 which I only use for tiny things and never for threading. It doesn't even have power feed.

My South Bend Heavy 10L has spoiled me even if the ways and lead screw are worn. A single lever gearbox and both clutch for turning speeds and half nut for threading and power cross slide with taper attachment and that micrometer stop means the only thing it was lacking was metric threading. Hence my ELS on the Z axis for it.

But as @gerritv mentioned, having to change gears for turning verses threading or even just for different turning pitches will head most people over to an electronic gearing system of some sort. And at some point cost verses replacing the lathe with one more capable enters the picture too.

His ELS can likely with the press of a button and turn of a knob set 0.003" per rev or 0.020" per rev for finish and rough turning. If you cannot easily do that without physically changing gears then an Electronic gearing or full blown ELS does become more interesting.

My only word of caution is verify the maximum spindle speed the electronic gearing system can handle. For old iron limited to 1200 RPM it's not a problem. For a more modern lathe capable of 6000 RPM it may well be.

 

[gerritv]

The Rocketronics is good to 6000 rpm depending on the encoder. At 400ppr you can go to 3000 rpm, needs fewer ppr as the spindle speed goes up. They want you to keep the (ppr * rpm) below 1,200,000 else the control gets over run. Personally I get nervous even with my 5C chuck at > 1600rpm so went with the 400ppr encoder.

It does a finish pass at settable feed rate vs. roughing feed. Still getting my head wrapped around all the features, like John's it can do more than meets the eye.
I have it set to prevent changing params during operation, might turn that off and see what I can fiddle with during cutting. Also haven't enable the spindle speed control yet, need to verify how well that works with my KBCC125R dc controller. that would give me constant SFM on parting.

I'll be trying out the pulley grooving mode soon to make jack shaft pulleys for one of my watchmaker lathes.

 

[jcdammeyer]

I'll be the first to state the failings of mine.
With 1 PPR it's limited to the low end for spindle speed at about 25 RPM.
The X axis is slaved to the Z axis in that it can't do a higher step rate than the Z axis. So that limits the taper to 45 degrees and therefore can't do ball turning.
Maximum step rate is 20,000 steps per second. That's only a limitation with a servo motor. A stepper motor torque falls way off well before that. I've played around with the closed loop stepper motors and even they are effectively useless at anything above about 700 RPM due to torque fall off.

The math? 720 RPM is 12 RPS. At 400 steps per rev with half stepping, which isn't enough to get rid of the resonance dip, that's 4800 steps/sec. Go to 800 steps per rev or 4 micro-steps it's 9600 steps per second. The resonance dip is still there but use a big enough motor and it may not slip and lose steps.

In reality the best stepper drivers use a dual micro-step method. The Gecko's were the first that published that approach. They use a 10 steps per step rate for 2000 steps per rev. Using feedback they detect the resonance point and alter the phase of the output relative to the step inputs just enough to mess up resonance so it doesn't happen. Then once the motor is turning faster than the resonance point they change to I think half stepping or 400 steps per rev up to a higher RPM. I never did add that feature to my ELS micro-stepper module since the hardware drivers became too expensive.

But the advantage of the Geckos was silky smooth stepping at low speeds and more motor torque at the high speeds once they past the resonance point. With a 2000 step per rev my ELS can only run the motor at 600 RPM.

To put that into perspective that's
1. lead screw pitch threading at 600 spindle RPM
2. 2x lead screw pitch threading at 300 spindle RPM
3. 4x lead screw pitch threading at 150 spindle RPM.

The other major disadvantage of my ELS over directly connected gearing is that once you start a thread you cannot change the spindle RPM. That's because the ELS can easily do 2,3 of 4 start threads due to the required index pulse (the 1 PPR sensor).

It works like this: Once the index pulse is detected the assumption is that motor speed is steady and that threading cuts don't overload the motor and change the speed significantly. History has shown this to be true.

At that index event the Z axis motor is accelerated up to the desired threading speed. This acceleration is constant and and part of the motor configuration setup.

Simple physics says to reach that speed a certain distance is covered over a specific period of time. As long as the tool bit is away from the work it will start cutting at the correct speed and enter the same point because the time it takes to get there results in the spindle turning a certain amount from that index position.

Change that starting position away from the work by 1/2 the thread pitch and the spindle will have turned further before the tool touches it. Exactly half the pitch. Now you have a 2 start thread.

So for example. Say you want a 4 start thread. Set the ELS to cut 5 TPI which is a pitch of 0.2". The spindle is up to speed in say 1/4 turn. Set the BEGIN position to +0.2" away from the edge of the part but set thread depth for 20 TPI.

Now cut a thread and you get a 5 TPI thread but with a depth of a 20 TPI thread. Change the BEGIN position to +0.25 and cut another 5 TPI thread at 20 TPI depth.

Rinse and repeat for BEGIN set to +0.30" and +0.35".

You now have a piece that looks at first glance like it's 20 TPI because the threads are 0.050" apart as a thread gauge will verify. But once you turn a 4 start nut you will see it behaves like a 5 TPI thread.

So why do this? For small diameter parts it's pretty hard to cut a 5 TPI thread to full depth.

But you can't change the spindle speed during all that because now although the time for the carriage to accelerate up to speed remains the same, the spindle doesn't turn the same distance and you no longer enter the part at the same point.

 

[Susquatch]

I'll be the first to state the failings of mine.
With 1 PPR it's limited to the low end for spindle speed at about 25 RPM.
The X axis is slaved to the Z axis in that it can't do a higher step rate than the Z axis. So that limits the taper to 45 degrees and therefore can't do ball turning.
Maximum step rate is 20,000 steps per second. That's only a limitation with a servo motor. A stepper motor torque falls way off well before that. I've played around with the closed loop stepper motors and even they are effectively useless at anything above about 700 RPM due to torque fall off.

The math? 720 RPM is 12 RPS. At 400 steps per rev with half stepping, which isn't enough to get rid of the resonance dip, that's 4800 steps/sec. Go to 800 steps per rev or 4 micro-steps it's 9600 steps per second. The resonance dip is still there but use a big enough motor and it may not slip and lose steps.

In reality the best stepper drivers use a dual micro-step method. The Gecko's were the first that published that approach. They use a 10 steps per step rate for 2000 steps per rev. Using feedback they detect the resonance point and alter the phase of the output relative to the step inputs just enough to mess up resonance so it doesn't happen. Then once the motor is turning faster than the resonance point they change to I think half stepping or 400 steps per rev up to a higher RPM. I never did add that feature to my ELS micro-stepper module since the hardware drivers became too expensive.

But the advantage of the Geckos was silky smooth stepping at low speeds and more motor torque at the high speeds once they past the resonance point. With a 2000 step per rev my ELS can only run the motor at 600 RPM.

To put that into perspective that's
1. lead screw pitch threading at 600 spindle RPM
2. 2x lead screw pitch threading at 300 spindle RPM
3. 4x lead screw pitch threading at 150 spindle RPM.

The other major disadvantage of my ELS over directly connected gearing is that once you start a thread you cannot change the spindle RPM. That's because the ELS can easily do 2,3 of 4 start threads due to the required index pulse (the 1 PPR sensor).

It works like this: Once the index pulse is detected the assumption is that motor speed is steady and that threading cuts don't overload the motor and change the speed significantly. History has shown this to be true.

At that index event the Z axis motor is accelerated up to the desired threading speed. This acceleration is constant and and part of the motor configuration setup.

Simple physics says to reach that speed a certain distance is covered over a specific period of time. As long as the tool bit is away from the work it will start cutting at the correct speed and enter the same point because the time it takes to get there results in the spindle turning a certain amount from that index position.

Change that starting position away from the work by 1/2 the thread pitch and the spindle will have turned further before the tool touches it. Exactly half the pitch. Now you have a 2 start thread.

So for example. Say you want a 4 start thread. Set the ELS to cut 5 TPI which is a pitch of 0.2". The spindle is up to speed in say 1/4 turn. Set the BEGIN position to +0.2" away from the edge of the part but set thread depth for 20 TPI.

Now cut a thread and you get a 5 TPI thread but with a depth of a 20 TPI thread. Change the BEGIN position to +0.25 and cut another 5 TPI thread at 20 TPI depth.

Rinse and repeat for BEGIN set to +0.30" and +0.35".

You now have a piece that looks at first glance like it's 20 TPI because the threads are 0.050" apart as a thread gauge will verify. But once you turn a 4 start nut you will see it behaves like a 5 TPI thread.

So why do this? For small diameter parts it's pretty hard to cut a 5 TPI thread to full depth.

But you can't change the spindle speed during all that because now although the time for the carriage to accelerate up to speed remains the same, the spindle doesn't turn the same distance and you no longer enter the part at the same point.

Very nice discussion John. I enjoyed that very very much. Gives me a much better idea of how the system works.

Also reminds me of my old days developing automotive sensors and control systems. I have not done that in many years. Not sure I could anymore.

 

[jcdammeyer]

Very nice discussion John. I enjoyed that very very much. Gives me a much better idea of how the system works.

Also reminds me of my old days developing automotive sensors and control systems. I have not done that in many years. Not sure I could anymore.

Thanks.
Remember you can do this with a half nut and thread indicator for many of the pitches too. Recall for threading sometimes you must engage the half nut on the threading indicator same number for each pass.
But if you set it up right then choosing # 1 and #3 will give you a 2 start thread. Been years since I did it that way.

 

[jcdammeyer]

Also reminds me of my old days developing automotive sensors and control systems. I have not done that in many years. Not sure I could anymore.

Yeah been ages since I did the ignition system for the Honda VTEC engines used in hovercraft and experimental aircraft.

With respect to stepper motors recall the engine ignition world. Put voltage a cross a coil of wire (the ignition coil or one of the two stepper motor coils) and current doesn't flow right away. The inductance of the coil prevents instant current so it takes time to get to full rated current.

If you think of that in terms of an electromagnet it also takes a bit of time before full current and full magnetism. BTW, change the voltage but still limit the current to the rated value and the current builds up faster. That's why often in older cars the ignition coils were actually 8 volts and had a series resistor. For cold cranking the lowered battery voltage that might well drop to 8V still created a solid spark. But once the ignition key was released to the run position the resistor went in series with the coil to drop some of the now back to normal battery voltage.

And of course that spark is created by the current in the coil being stopped. The collapsing magnetic field is like a generator and creates a much higher voltage in the 12 volt coil. Which is stepped up to the 15Kv and up voltage to the spark plugs. And even that takes time for that to happen.

Now picture that stepper motor winding. It's a 3V winding but the stepper motor driver uses 36V. That higher voltage means the time to full current (say 3A) takes 1 millisecond. To make the magnet on the rotor move the direction of the magnetic field has to reverse. So first you have to take that 3A and reduce it to zero and then change it to -3A. All that takes time because of the inductance.

Pretend for this example it also takes 1mS to do this. That means you can change the direction 1000 times per second.

What happens if you want to change the direction 2000 times per second? It just means you don't get up to the 3A before it's time to change to the target -3A. And 4000 times per second? Even worse for reaching the 3A current.

And what's important about that current? The amount of magnetism. Or more simply said motor torque is based on Amps x Turns. Keep the number of turns in the winding the same but reduce the amps and you get lower torque. Which is one of the reasons why stepper motors lose torque at higher step rates.

There is one other reason too which DC motors also suffer from. Ever wondered why if you took that old Eldon Slot card motor rated for 6V and change the power supply to your dad's 12V battery charger that the cars went faster? (Usually off the track).

Well when a DC motor spins it's also a generator. The voltage created by this generator part of the motor is the opposite polarity of the driving voltage. The top speed of the motor when 6 volts is applied is when that generator part of the motor creates -6V. Change to 12V and now the motor turns faster until again the generator (what's called back emf) equals the applied voltage.

So there's the other issue with all motors but worse with stepper motors. Once it's spinning it's also a generator. So that higher drive voltage is required to push back on the generated voltage to get the current flowing through the winding in the opposite direction as quickly as possible.

Hope that all makes sense.

 

[Susquatch]

Hope that all makes sense.

Perfect sense to me John.

I spent the first 10 years of my career designing electronic control systems and then microprocessor control systems to replace the old electro mechanical devices of the auto industry from the pre-transistor age.

 

[jcdammeyer]

Oh and I realize it's totally off topic but when I mentioned Eldon Slot car set only the really old guys on this list will remember how bad the tracks actually fit together and how bumpy they were at the transitions.\

eldon slot car in All Categories in Canada - Kijiji Canada

Find eldon slot car in All Categories in Canada. Visit Kijiji Classifieds to buy, sell, or trade almost anything! Find new and used items, cars, real estate, jobs, services, vacation rentals and more virtually in Canada.

www.kijiji.ca

 

[Tom O]

Eldon slot cars! Wow that was like 60 years ago I don’t know how many times I took the lap counter apart to fix it.