[Joey5226]

I may have missed it, but did you recheck table flattness after milling it?

I have a bowed table like yours (Google search to a YouTube video to a link to this thread brought me here.) I'm thinking of taking it to a grinder to have a few thousand taken off. Mine is 0 front and rear and +.004 on the two land edges adjacent to the center T slot.

Hearing your saddle is out gives me fear that mine may be out too.

[atomarc]

Joey,

Our table was bowed before we milled the bottom for the linear rails...that's why we installed them. The rails are now flat but the bow in the top still exists. The plan is to have the table either Blanchard ground or surface ground when registered off the flat surface of the linear rails. This 'should' make the top flat and parallel with the rails..we hope. There are no shops locally that can accommodate our table length on their surface grinder so it will have to be shipped out of the area.

While these little mills are cost effective and do have the ability to do some work, I think one of their downfalls is very poor manufacturing tolerances..plus using green castings that move all over the place as they're machined. My dove tails looked like they may have been scraped in with a dull cold chisel..but remember, the machine was only $1200 dollars.

Because all these components stack on top of each other..the saddle on the base and the table on the saddle, they all must be flat/parallel to each other or the table top won't be 90 degrees to the column. Lots of stuff has to be just right for the thing to be accurate, so it all depends on how much accuracy you expect or need out of the machine. That's my take on it.

Stuart

[Joey5226]

Thanks for the response. I read the earlier post of you mapping it out on a surface plate before machining for the linear rails. I'm curious if you checked if the machining has moved it more or if it is still the same.

Awesome project by the way.

[atomarc]

Hello!!!

a couple things..........


(a viideeeooo!! to start things off... subscribe for updates on the build!!! )

https://youtu.be/0aVEDB_BYTc


(received the replacement hinges, "all metal construction" = ) they are much stiffer than the original ones. (slightly modified with countersinks) )






(a magnet latch.... for positive latching action.... was installed behind the main switch "box")




(the last loc-line valve....)





(some chamfers for the switch(s) face plate)









(the tool steel "rubber clip holder" thing post "heat treat" lol .... I was pleasantly surprised when it "clicked" into position! not sure if it will be used or not... ...."experimental" stage.... (chair)





:cheers:

[atomarc]

Thanks for the response. I read the earlier post of you mapping it out on a surface plate before machining for the linear rails. I'm curious if you checked if the machining has moved it more or if it is still the same.

Awesome project by the way.

Joey,

I didn't check them after machining...sort of afraid to do that. Seriously, I'll do that tomorrow and report back here. The important thing is that the rails are flat and parallel to each other..we shall see.

Stuart

[Joey5226]

Joey,

I didn't check them after machining...sort of afraid to do that. Seriously, I'll do that tomorrow and report back here. The important thing is that the rails are flat and parallel to each other..we shall see.

Stuart

I would be willing to bet that machining probably didn't change the geometry beyond a thenth. It may have moved ten millionths. And the way you machined the pockets and with the rib, It likely increased the redgidity. However, depending on the datum you clamped the table to the mill, it may be slightly more out of parallelism that surface grinding can fix..

My table, like yours has the curvature of a potato chip. About .003/.004 high in the middle. I just checked the parallel from the two outer lands to the way datum and mine is parallel. I also checked the table outer lands drop to the ways on the base and they two are parallel.

I just ordered new spindle bearings and once I get those on (my angular contact bearings have too much run-out to check at the moment) I'll take the table off and check the saddle to table way parallelism from a trued spindle. If all of those are good I'm going to send my table out to grind. If they are not, I may consider your route.

[atomarc]

Its Friday!!!!



https://youtu.be/kREMQckW1xY



and few "camera roll" photos....because photos are the shizz... = )
















:cheers:

[atomarc]

I would be willing to bet that machining probably didn't change the geometry beyond a thenth. It may have moved ten millionths. And the way you machined the pockets and with the rib, It likely increased the redgidity. However, depending on the datum you clamped the table to the mill, it may be slightly more out of parallelism that surface grinding can fix..

My table, like yours has the curvature of a potato chip. About .003/.004 high in the middle. I just checked the parallel from the two outer lands to the way datum and mine is parallel. I also checked the table outer lands drop to the ways on the base and they two are parallel.

I just ordered new spindle bearings and once I get those on (my angular contact bearings have too much run-out to check at the moment) I'll take the table off and check the saddle to table way parallelism from a trued spindle. If all of those are good I'm going to send my table out to grind. If they are not, I may consider your route.

I set the table on the surface plate today, rail side down of course. It was suspended near the four corners by the Hiwin bearing blocks and all 4 corners were dead nuts..but the middle of the table had a little over a .0005 sag in it, consistent on both side..that's the way the cookie crumbles I guess.
Nick and I took the table to the machine shop at the mill I worked at for years to have the top Blanchard ground flat and true with the linear rails. Lo and behold, they have a giant surface grinder too..just the ticket. So the Blanchard is out and the surface grinder is in. They do very precise work so it should be pretty good when it's finished. We'll take some pictures, because we all know, without pictures it didn't happen.

Stuart

[atomarc]

I'll post this question here although it may be more appropriate elsewhere or in a new thread. On the ESS smooth stepper and the MX3660 integral break out board there are inputs and outputs for control of any real world component..is that correct? Are those outputs merely switch (not powered) closures that can control a remote relay or relay board that has its own power. This is a basic question but I'm sort of a dough head.

So..are those outputs powered or just switched, that's my question.

Stuart

[gd.marsh]

I can't speak about the MX3660 .. but I do use an ESS on my machine.
It's always been my understanding that the inputs & outputs on all such boards are just logic switches that pull the terminal high or low to trigger other devices (such as relays, etc.) which must have their own power source.

gd.marsh

[atomarc]

I can't speak about the MX3660 .. but I do use an ESS on my machine.
It's always been my understanding that the inputs & outputs on all such boards are just logic switches that pull the terminal high or low to trigger other devices (such as relays, etc.) which must have their own power source.

gd.marsh

Gary,

Thanks, here is what I think your post said. The outputs on the ESS or a BOB are solid state switches that can be programmed high or low (open or closed). These powered outputs can switch relays that have coil voltages matching the supply of the ESS or BOB. If I wire a 5 volt SSR to a ESS output, I can switch that relay ON or OFF through ESS programming. Is that correct?

The other scenario would be that the outputs are not powered, but only open or closed contacts that can switch a device that has its own coil power...maybe that's what you said. This is my "kornfusion"

Thanks,

Stuart

[SCzEngrgGroup]

The outputs on the ESS or a BOB are solid state switches that can be programmed high or low (open or closed). These powered outputs can switch relays that have coil voltages matching the supply of the ESS or BOB.

Absolutely NOT! The ESS outputs, as with nearly ALL BOB outputs, are logic level signals, with VERY limited drive capability (typically 5-20mA MAX). They absolutely will NOT directly drive relay coils. They will drive the control inputs of an SSR, but NOT a mechanical relay. To drive a relay, you need relay boards with logic-level control inputs.

Something like this: https://www.sainsmart.com/8-channel-...ttl-logic.html

Even so, all the outputs of the ESS are NOT the same. Different pins have different drive capabilities, IIRC a few are open-drain, some may be only 3.3V instead of 5V, etc. So, care must be taken in choosing which pins to use for which functions.

Regards,
Ray L.

[atomarc]

Ray,

Thank you. I get a little loose with my 'relay' definitions. The relay board you linked as an example is exactly what I have, although only a 4 channel version. I feel like I may be starting to understand this stuff a bit more.

Stuart

[photomankc]

The MX4660 has Opto-Isolated I/O. In general, for outputs they are **pulled** high when it is inactive. So a large value resistor connects it to the positive side of the I/O voltage you provide presenting a digital high. When the output is active a phototransistor is switched on by the MX4660 electronics and grounds the output pin thus presenting a digital low to the outside world. Outputs like this are not suited to driving a load in the inactive (high) state. They can sink a load when active (low) by providing that load's path to ground. So for instance to light an LED when the I/O is active you place the cathode (neg) side of the LED onto the I/O pin so when the path to ground is complete, the LED lights. This arrangement can usually drive a pretty decent load. I did not see a clear indication of the transistor's current limit but I'd say probably 100ma to 200ma would be a safe bet.

Inputs on the MX4660 are similar. The outside world input will be expected to drive an internal LED inside the MX4660 that activates the internal phototransistor to signal your input to the electronics and parallel port. This input arrangement usually has pretty wide input voltage tolerance but some attention to detail is needed if you get higher than 12V. You can use the supplied 12V or your own supply voltage to do this (say 5V if needed). If you go over 12V however you need to add more resistance to stay within the current limits of that internal LED.


Can't speak for the Ethernet Smooth Stepper, but most interface / motion control cards provide Microcontroller / IC type I/O. Usually that I/O is like a SPDP switch inside a chip. The output is in the middle and the IC switches that between it's positive and negative supply to drive the output. Such I/O can directly drive small loads in either high or low output. However they often have very restrictive limits on the current they can provide doing that. Usually 5ma to maybe 50ma per output. They are dangerous to connect up directly to external loads because they can be easily destroyed by either too much current running through them or even a couple volts above the chip's logic voltage. These are usually expected to be paired to opto-isolators as described above to protect the chip from damage.

If the device is described as having a "breakout board" then often (but not always) they provide the opto-isolators or other electronic protection to the I/O pins they expose but you MUST read carefully to understand if they do or don't.

Hope that helps somewhat.

[atomarc]

Photoman,

Thanks..that helps a bit, although much of the terminology is foreign to me so I'll need to study it to fully understand the concept.

Stuart

[photomankc]

Yeah, it gets a bit involved.


This is a good explanation of the transistor as a switch. This is a very common circuit used to interface a small IC output to switch a bigger load. In this circuit replace their representation of "VIN" for light and you have the operation of an optoisolator.

Transistor as a Switch - Using Transistor Switching




You'll see this circuit mimics closely that of the transistor switch show above.





Anyway the real take-away is that the MX4660 uses optoisolators on all it's provided I/O but that is not the same as contact closure type switching that a relay would provide. Looks to me like the SmoothStepper puts out microcontroller type outputs. If that is the case then only the MX4660 can directly drive electromechanical relays. Either one could drive solid state relays.

I'll not hijack anymore of the discussion. Good luck!

[atomarc]

No hijacking...all good basic info for a electronics boob like me, thank you.

Stuart

[photomankc]

Ok that's funny. I was following the pictures and didn't even notice you are the guy posting them! Duh. I guess it's not hijacking if it's your thread!

[atomarc]

The table is back from being ground!!!! yeee! Other than that..... working on breaking down the machine in preparation for paint! ..... Sherwin Williams blessed us with a gallon of "grey" and "white" DTM enamel. Started work on making a new X ball-nut holder.........uhhhhhh... how about a video = )



:cheers:



(I spent all week making this video for you... )

https://youtu.be/bGhMDIIpQjg




( (awesome iron!) Here are some photos i took when we went to pick up the table.... some beef cake machines no doubt! (surface grinder pictured below did the "grizz" table) )














(paint...."gunna be uh two tone" )






(the start of a new (steel & longer) X ball-"nurt" )




...... more to come

[atomarc]

Happy Thursday! :cheers:





(a special treat.... )

https://youtu.be/-NkipMCbbg4