Light Machine Corp. Benchman XTr (retrofit)

I've decided to go ahead and retrofit my XT using the Mesa 5i25 + 7i77 combo and LinuxCnc. The Nextmove cards I have don't seem to be the correct version and I'd like to have a more robust machine going ahead. I would like to keep the original power supply with the smoothing circuit on the main board, but has I've read over this thread I see that the power to the servo amps is enable by the OEM software. My thought was that I could enable that power by sending the proper signal to the correct pin on the J1 connector for the Nextmove card. The wiring schematics that were posted here earlier have a pinout of the Nextmove's 100 pin connector, but some of the descriptions for each pin function are difficult to decipher (at least for me). If Don is still monitoring this thread I would really appreciate some help with this.

Here's the wiring schematic I was referring to. The Nextmove pinout is on the last page.

If you only need to supply DC power to the servo amplifiers, the circuit on page 2 in the upper right needs to be powered.
This is done on page 1. You will need 240VAC on J2-5 to J2-6. You will also need +24V on J1-98 to turn on K4 which will bring power out to the transformer on J3. It comes back from the transformer/bridge/inductor to J9 on page 2. How do you intend to wire to the J1 connector? Will you just cut off the Nextmove side of the main cable and rewire as needed? Let me know if you need any advice/comments.

Thanks for the quick reply, Don! That was the plan, to just hack up the Nextmove cable to deliver that signal. So the Nextmove card was able to supply 24V? I was expecting it to be something like 5V.

Actually the 24V is generated on the PCB. The Nextmove card has a relay that is used to switch the 24V. The 24V is only present on the Nextmove card if the ESTOP button is out. See schematic (page 7D2). J1-100 is the relay NO contact, J1-98 (page 1A3) is the relay COM contact. So 24V is generated on the board and is conected to the ESTOP switch. If the switch is out then 24V is applied to the relay contact. When the relay is switched on through the software, 24V is present on J1-98 and powers K4 bringing 240VAC to J3. It is not even close to meeting any current safety standards, but back in 2000, it was sufficient.

So I just need to add a relay between J1-100 and J1-98. What would you recommend to make the ESTOP circuit safer?

With the liability issues that it naturally entails, ESTOP safety is a major undertaking. I cannot make any recommendations on specific changes to make, but I can comment that to meet current commercial/industrial safety requirements, you will most likely need a "safety-rated" control to monitor various inputs to make sure that a pre-defined set of parameters are met before spindle operation is allowed and/or high speed motion is initiated. I'm not sure of the exact vintage of your machine, but Intelitek did do an upgrade to the safety circuit as part of the Benchman MX series and even that doesn't meet current requirements. Current standards require "fail-safe" operation (single-point failure cannot compromise the safety circuit) and redundancy, that is where it gets sticky. Nobody wants to NEED the ESTOP switch but have it not work. Sorry, I can't be more specific.

I understand your hesitance in providing that kind of advice. Thanks again for all your help. This would be much more difficult without it!

A little more HW information for these machines.

The ballscrews are unusual. Whilst preventative maintenance is always preferred to a replacement, it is especially so for these machines. Since I was replacing the X-axis linear rail system I decided to take a good look at the ballscrew too. It's a major undertaking to replace the X linear rail bearing blocks, you need to remove the Y blocks, and the X and Y ballscrews from the saddle plate in order to get to them. The X ballscrew showed signs of corrosion, much as the X rails did. One side of the screw itself has signs of minor pitting corrosion, mainly in the root of the minor diameter where the balls don't run. Lash and notchiness was evident, and upon disassembly it was discovered that about a third of the balls had flats, pitting corrosion or fractures.

I cannot stress this enough - keep your rails and ballscrews lubricated, especially the X axis, and doubly so if you have coolant in the sump.

Anyway after a couple of attempts I managed to get through to a parts guy at Intelitek. They do not carry any spare ballscrews for these machines, and his estimate of the replacement cost was $2200 in single quantities. They were originally bought from Danaher.
The markings on the ballnut are 31-0802-0026 (which I believe is an Intelitek part number), R16-05-209, a pair of serial numbers, and a symbol that I couldn't read but eventually matched online as IBL.

Some internet sleuthing: International Ballscrews Limited were a manufacturer in Devon, UK. They got bought by Danaher. Danaher later bought Thomson, and rebranded under the Thomson name. For a while IBL were known as Thomson-IBL. Now they are just Thomson.

So I called Thomson UK looking for information on the ballscrew. They only seemed interested in selling me a replacement and since I am in the USA bounced me to the US sales team. Several weeks and phone calls later they finally got back to me with a quote for replacement - I never got to speak to anyone about a repair. Anyway, Thomson no longer sell IBL designs as stock. Ballscrew designs (and in particular the nut) have standardized around a DIN standard. They wanted about $2400 for a replacement, custom made in Germany by a different subsidiary (Thomson-Neff) , and I'd have to pay shipping and duties on top of that.

The ballscrews are rolled to a P5 (positioning) standard, which is high precision for a rolled screw. Most vendors top out rolled ballscrews at T7 (transport). The only other two vendors that roll a 16mm ballscrew to P5 or better are Gten and TBI Motion (both Taiwan). Unfortunately they use DIN standard ballnuts too.

One problem with switching to a DIN standard ballnut is that as far as I can tell, none could easily be made to fit. The centerline to saddle distance is critical, as is centerline to table. A taller ballnut would need the bearing block at the end of the table to be replaced with one bored at the new centerline height. The X rail blocks might need to be shimmed to provide clearance too (at the cost of Z clearance. I don't think that it's impossible, but it isn't just a simple drop-in replacement.

The other challenge is that these ballscrews are 16mm nominal diameter, but the lead is 5.08mm (0.2"). That is also unusual. One of the Thomson companies sold a 0.631"x0.2" ballscrew, but I don't know whether it would work with the nut or not. (Ball size is also a unknown critical factor here).

I decided to try a simple repair for now - replacing the balls. I bought 4 different sizes of balls from Bal-tec spanning a total of 0.0075mm with the range centered on the size found in the Y ballscrew (about 0.01mm larger than the "good" balls in the X ballscrew).

I tested one of the larger sizes first, but it felt too stiff, so I backed off to 0.124425" which feels to provide about the same preload as the Y ballscrew, albeit not quite as smooth. It turns out that ball size is pretty much exactly as those in Y anyway, as far as I can accurately measure.

Keep those ballscrews lubricated. They appear to be the weak link in the supply chain needed to keep these machines working.

Anyone needing a replacement seal for a motor box cover, theoringstore.com part number VN70.103-012 ordered with a 6.5" ID is a perfect replacement. Intelitek don't have any stock, and the original vendor doesn't list them on their website anymore. The replacement was selected using Intelitek's original part description as guidance, and cost under $5.

If you replace any rails, replacement hole plugs can be ordered from an IKO bearing distributor for about $1 ea. Kinda outrageous for a 9.8mm diameter 2mm thick injection molded bit of plastic, but you can't get them any other way.