My Techno/ISEL 4 foot square CNC router table build

travis_farmer

Administrator
Staff member
Thought Club
This project has been in the works for quite a while now. it does work, last i checked, but i tried to sell it... without any luck. or perhaps it was luck, as my Music Studio failed, so here i am back to this... i will try and update as i go and make progress. :D
 
right now, i need money, so i have a little work to complete on the machine first before it can make money.
currently, the Z axis limit/home switch locates on the down-most stroke. very less than ideal, but it works for now. so i need to change that to signal the top of the travel. given the way Techno/ISEL designed the machine, it is very hard to mount the limit external. so i think the best option is to dismount the Z axis and open up the box that encloses the movement to first see what is wrong with the factory limits, and to repair or replace those, with something.

really, over-all, the machine needs to mildly taken apart to make sure all points are able to be lubed. Y i can do under the table, but X and Z need to be taken apart to lube. i plan on installing some well placed Zerk fittings at some point. or remove the covers that hide the movement entirely, and just keep them clean somehow.

the controls are fairly situated, and the power, while it could use some improvement, does work.

my plans for the Shed are to use my large tent-shed, but build a platform under it, and build actual walls for the front and back. that way i can better keep the critters out from chewing on wires. i just have to figure out how to keep the platform from rotting out. maybe just build the joist frame, and lay plastic over that, then the Advantec, or whatever over that. but this time also, use some crawl-space vents to let the moisture out.
 
7i80hdcolo-500x500.png20210723_160813.jpg20210723_160824.jpg20210723_160919.jpg20210723_160930.jpg20210723_160941.jpg20210723_160958.jpg
 
All i wanted to do was fix the limit switch inside the Z axis, and lube it up while i was in there... now the machine is all apart, screw heads are stripped, and a window is broken (slipped T-handle hex wrench). i wish the machine was put together with anti-sieze...
long story short, i called it quits for the day, as things were devolving very fast.
my plan of solution is to use an EZ-out to remove the screws that are stuck, if possible. then use anti-sieze compound to prevent them from locking tight again, and fusing (dissimilar metal fusing).

once i am able to get the Z axis off, and opened up, it is my hope to be able to either mount a new home/limit switch inside, or make one somehow. i would like to use optical sensors, throughout, but that would be a major overhaul. though with the machine not working, it may be a good time... something to think about.

so, not dead in the water, but i am not swimming very well...

if i ever get around to building the platform and improving the shed, then having the machine apart will make it lighter, but over that amount of time, i will likely forget how to put it together. i think my best bet is to just fix it, and put it back together. if the time ever comes to build the platform, then i worry about it then.
 
Currently thinking i will go with Hall-vane interrupt limits. similar to a optical break-beam, but easier to keep clean. i have digital hall-effect sensors... if i place a magnet opposite the sensor face (printed side), then a steel vane can interrupt the magnetic signal to trigger it. should be as precise as optical, but without the dirty-sensor issues.
 
machine is back together and working. still need to replace fasteners, but i shouldn't lose parts this way.
 
current connections from the Mesa 7i80HD-16 to the machine.
 

Attachments

  • connections.pdf
    31.7 KB · Views: 196
Last edited:
been trying to think of a way to automate laser ride height... i think the easiest would be to mount a 3D probe in the laser attachment. but i would like to somehow keep a laser on the machine, and simply put a dust cover over it when not in use, or perhaps a glass "shield" to protect it. what i need is a quick-change way of taking the spindle off, and mounting the laser, or, leave the spindle on, and mount the laser with it, and take it off when not in use.
Now, i plan on having dust collection for the spindle operation. so i need a way to mount the dust-shoe as well. so an easier method of removing the whole spindle sounds like the best option right now. i could mount a steel plate to the Z axis, and drill and tap it for bolts, and then use bolts or cap screws to mount the spindle and laser, plus accessories.
 
I have been pondering the idea of converting my machine to an ATC spindle. it will be expensive, but i think it will be worth it in the long run.
not yet sure how i will do the tool carousel yet, but the idea is an interesting one...
 
I now have 180 IPM rates!!!
Upgraded the X and Y steppers, and now have more speed! i also boosted the Z axis up to 120 IPM, and it works just fine. very pleased :D
 
Almost forgot, i also have 3 solid-state AC relays now, for the CNC. will take 5Vdc input, and up to 240Vac @ 25A of load. this will be for my vacuum control, and either 3 spares, or 2 other controls.

On the vacuum system... how to have it work, and still be able to use a ATC in the future... ?
i was thinking of having it slide up and down on the spindle body of the ATC spindle, so when it hits something, it just rides up on a spring. hitting something includes when it is hitting the fixture to load a new tool. so it needs to ride up easy. looks like about a 100mm straight-cylinder body, so it may not bee too hard to make a dust shoe. the other option is just a couple vacuum nozzles to either side of the spindle head.
in theory, i could make a rigid, yet movable fixture for a dust shoe. mount it with a small lead screw, and as part of the tool change macro, it would lift the dust shoe up. seems rather over complicated though...
if i have 3 nozzles... two behind, one in front... it should negate the need for the need for nozzle-hair. with enough suction, it would suck up most debris. i suspect i would want a 4" inlet dust collection system. but the 3 nozzles could be mounted high enough to hopefully be mostly out of the way, maybe.
the cheapest idea is to just have little or no dust collection, and just vacuum up after the job is done. but that is just boring. ;)

currently, with no ATC, dust collection is easier... i could just modify the dust shoe i have so it is a single front nozzle. being in front it would be out of the way of the ATC. and if it is high enough, no part collision risk. this could work with my current spindle too, so less to change over. just a matter or mounting it somehow.
I wonder... if i could mount it behind, so the hose comes up behind the gantry... out of the way. the ATC, that i don't have yet, could just as easily change in the front of the machine. a front change ATC could be fun to watch. :unsure:
the dust hose could just hang from suspension springs and a travel track to take up slack. 4" hose, flexible (dryer hose?) and the rest of the hose could be hard-line to reduce vacuum loss.
 
on the dust collection shoe... my current line of thought is to use a pneumatic double-acting piston i have to lift the dust shoe out of the way of the tool change, and then back into place. i mean, i will need shop-air for the ATC spindle anyway, so why not make use of it in other ways.
so, for that to work, i would need to either mount the dust shoe on short slides, or levers to guide the dust shoe up and away. perhaps just a short 45deg angle away from the head will work.

for the ATC carousel, i am thinking a tear-drop shaped rotary system. i will use a large pitch chain, like a motorcycle chain, or better. something easy to buy. i will use two layer of chain for anti-tip support. the tool pocket will be on just one chain-link (both top and bottom chain), so the chain can freely ride around chain sprockets, without binding. planning on 3 sets of sprockets, in a triangle shape. can stretch the length to get more tool pockets on.
pocket sensing... not sure yet how...
 
just the other day, i received shipment of 10 inductive proximity switches, to replace the current mechanical ones. it is my hope that they will be more precise, and have less noise from contact bounce. they work much like a metal detector, but with better distance triggering. they will work with aluminum, but steel/iron works best.
for each axis, i will mount two switches. one combo home/end-limit, and a limit for the other extreme.

Y axis: switches will mount on the base frame, and the target flag (the steel the switch detects) will mount on the gantry support.

X axis: Switches will mount on the Z axis support plate, on tabs welded to the steel support plate, spaced wide apart, to prevent them from interfering with each other. the target flags will mount on top of the gantry box.

Z axis: Switches will mount on the Z axis box, and the single target flag will be welded to the Z axis support plate.

the switches will require 24Vdc to function. i can't use the current limit-switch wires, because for each axis, i need 4 wires, or 6 wires for the Y axis. 4 for X and Z, because i can common the power and ground wires., then the 2 signal wires. 6 for the Y axis because it is just easier, given the connection point will be mounted on the base frame, along with the sensors.

so i will need to pull back the current home switch wires, and source out some 4-wire cables, in a small diameter, and flexible. i can work on mounting the new switches for home, though it will be only temporary. it may even be possible to only need one per axis, and two flags at each end of the travel. i will have to look into it, but it would allow me to use the same wires, i think.
 
i am thinking the best option for target flags, is a bolt, so it can be threaded into a mount, and adjusted up or down to set it up. i can mount the target mount fixed on the Y axis, and make the sensor mount adjustable to get the home/limit switch in the correct position. that would be a bit away from the machine hard-stop, to account for coasting, should the machine develop a problem (unlikely, but just in case).
i am still not sure where i want to mount the sensors on the machine. the easiest would be to mount them on the side frame rails. the targets, being a 3/8" bolt, will mount... somewhere... just thought, that puts the sensor in harms way... better to mount it under the table somehow, and have the target bolt mounted on the gantry drive beam.
the X axis, i would like to mount the sensor inside the gantry box, but i don't know is that is possible. it would be better protected, but if it fails, harder to get to. maybe just mount on top of the gantry box.
Z axis, i would really like to mount inside, but will likely mount outside for ease of access.

i may also use 3 sensors per axis, and have a home and 2 limits. i will lose some working area, but i think i will have less problems... either that or set each axis to offset zero after homing, and set the software limits to stop just before the home/limits sensor.
 
First attempt at testing my ind. prox. switches, failed. i forgot the pull-up resistor, so the signalling was wrong. the Z-axis would not pickup the home switch, and during homing, it ran into the hard limit. fortunately it was moving slowly. next time around, i will try adding the resistor, and see if it signals correctly. if so, then i have to hard-mount the switch, as it is currently just taped on for testing.

a few things on my ToDo list, currently:
  • finish mounting home/limit prox switches
  • add a spoilboard
    • particle-board
    • held down with perimeter T-bolts (outside of cutting area, inset)
  • mount tool touch-off sensor
    • hard-mount to aluminum table, back-left corner
  • add a machine wiring cabinet
    • shorten signal wires to remove excess length
  • finalize power cabinet
    • replace 36V PSU with 48V psu
    • add master power switch
    • add arduino power control system
    • install vacuum/coolant SSRs
  • clean up wiring in control cabinet
    • build two layers for mounting parts
    • bring the component layer close to front so the components are easier to get to
    • mount 7i80HD on side layer, access through a side panel in cabinet
      • BoBs are mounted on front layer
    • SSRs could actually be mounted behind other side access panel
    • mount relay board, if still needed, with SSRs
  • install provisions to grease the machines moving parts
  • add provision to plug in 3D probe, near spindle.
    • could hard-wire, with some sort of storage holder.
  • install laser somewhere where it can be of some use
    • make dust cap
  • make dust collection shoe
 
I have pretty much abandoned my LinuxCNC laptop, in favor of a Dell PowerEdge R200 that gets much better real-time kernel performance. my latency jitter was over 200K on the laptop, but only 37K on the R200.

also, my spoilboard will not be particle board, but will either be plywood, or Advantec OSB. whatever i can get, that is good and flat.

for mounting the prox switches, i am thinking i will find some small machine screws that i can get a tap for, and thread the screws directly into the mounting locations of the machine. it will be the most rigid method, though i will have to be careful about what is behind where i am drilling. unfortunately, there are some tight clearances where some of the switches will go. the Z axis switch, for instance, has about 1/8" or less between the sensor and gantry support.
i think the X axis prox switch can be mounted on the gantry, so it detects the Z mounting plate.
still not sure how best to detect the Y axis.

i really need to get access to the X and Z axis mechanicals, so i can lube them. and the Y axis, i just need to climb under the machine and lube it as well. while under there, i can inspect for prox switch locations.
i really want to have two prox switches for each axis. one home/limit, and a limit for the other end.

i am really caught between a rock and a hard place with this machine. it is old, and exact match parts can't be found anymore, i don't think. so i really have to keep it running long enough to be able to buy a new machine, unless i can, by some chance, find parts that fit. my biggest gripe is the aluminum fasteners. they just keep stripping out the heads. i would like to replace them with steel fasteners, and use some anti-sieze to keep them so i can remove them.

right now, my biggest issue with the machine location is that i can't get around it, but just one side. i want so very bad to be able to build my new CNC shed, and move it in. but there is a money issue, and a winter snow issue.
those issues are also why i haven't changed to a ATC spindle, like i want. the old Perske spindle i have works, but i don't know for how long.
 
Thinking my spoil-board will be this... a piece of plywood, surfaced flat, with inset T-track. this will not be for through-cuts. for those, i will place a scrap of wood under the piece to be cut, just to protect the table surface.

thinking of re-doing the control cabinet... i want to bring the main mounting surface forward, so i will have room for additional mounting through the side openings of the enclosure.
 
I have been reviewing the BB Charge Pump activated PSU cabinet…

there will be a master power disconnect for AC power input.

5Vdc, 12Vdc, and 24Vdc will be powered full-time, as long as master power is on. This is to make sure the 7i80HD-16 has the correct signals when LinuxCNC is started up.

The stepper PSU will be what is powered up only when LinuxCNC is active. This will ensure no movement of the steppers when LinuxCNC is not able to command them.

I would like to order a better PSU for the 12Vdc supply. It is currently a camper power-converter. It supplies for the 5Vdc and 24Vdc as well, using DC to DC converters. Works, but less than ideal…
 
I need to check the DC to DC converters, and see if the ground is pass-through, so i can simply connect the grounds together. this is necessary for a star-ground system. that way i can hopefully improve the shielding in the control cabinet.

also, i need some better power-distribution in the machine signal distribution cabinet (to be built, or bought). all signals will be 24Vdc, unless otherwise specified, just to make things easier.
 
Back
Top