Not Exactly a Benchtop Mill

[dcprecision]

I thought you subscribers might enjoy a few pictures of a Burgmaster drill and tap machine I recently finished. These machines are excellent when used manually for accurate drilling,reaming, countersinking and tapping. With a skilled operator they can outwork a $ 200,000.00 machining center on certail types of parts. I now have the best of both worlds in not having to pay for operator time. Using a fixture base and pallets, we can load 6 to 8 parts at a time and let it run.

The tool change takes 1 1/2 seconds. Top X-Y-Z is 55 IPM. It uses the IMserve servos and boards with desk cnc software and controller. Maximum spindle rpm is 5500. Super precision omron home/limit switches allow fast, absolutly accurate start up. The entire head motor assembly moves as a Z axis unit. Indexing is software actuated by a solid state relay and a pneumatic solenoid valve. Spindle and mist coolant are also software driven.

I added a custom built feedback circuit which senses servo fault using a mosfet transistor. The FET opens a relay to force an e-stop. This prevents the secondary functions such as coolant, spindle and index from continuing to run if a servos faults. Lack of a complete system shutdown is a major weakness with most pc driven control products because they do not have servo fault feedback.

[InspirationTool]

Nice!

I assume it is for drilling only, with no side loads?

Did you get the burgmaster unit new or used?

Did you build your own XY table, or did the unit come with it?

Ballscrews or acme?

Great job!

-Jeff

[dcprecision]

The bearing design of these machines allows for drilling thrust loads only, not the side load end mills generate.

These machines have been out of production for at least 35 years. I own 4 of them, 3 OB's and 1 - 1D unit. The oldest was made in 1952 and is now almost 55 years old. How many of todays Chinese mill drills will still be working like new in the year 2060 when used in a continuous production environment.???

The larger 1D size Burgmaster has 1/2 " drill/tap capacity, a 2hp - 2 speed motor and 4010 rpm max spindle speed. I have designed an end milling conversion employing an er collet shaft supported by an extending block with a radial contact bearing for side loads. When I CNC this one, it will have a 9 x 24" machining envelope. 2 spindles will have 1/2" end milling capacity.

If you look closely at the photos, you can see that the XY axis is actually a TAG milling machine table with the original lead screws. I felt my tag cnc was such an incapable machine that I decided to decommission most of it and use the XY table which is just rigid enough for drilling purposes. It supplied most of the auxillary parts and servos.

This whole system was hand built by me, including electronics. . These OB size machines have never been offered as a CNC and to my knowledge this is the only one currently in existence worldwide.

The Z axis is actuated by a 6mm pitch ground ball screw. So far I have not felt the need to change the xy tag lead screws out because +/- .001 or so is plenty accurate for the parts made on this unit.

I have made several thousand parts on this one in the past 2 months and soon will have a good idea just how durable the Tag table & screws are.

Dennis

[S_J_H]

Very impressive looking machine!!
Steve

[Ron111]

Very nice indead!!

I have not done anything with servos yet, who makes your drivers and I notices that your full wave bridge rectifiers look very high tech.

Good work,

Ron

[dcprecision]

The board components are all IMserve / Desk CNC components. The power supply is a combination of parts obtained from Allied electronics and IMserve. Other parts are from Radio Shack...

I control all power functions with solid state and reed relays that use 5v activation. 1N914 diodes isolate the relays from the controller board. I use a 5v auxillary plug in power supply to act as a relay activation voltage with a momentary reset push button switch.

When the servo box is first turned on,, only the fans and 5v auxillary power supply are powered up. pushing the reset button activates the main DC power relay. once the power supply is on line, a separate 5v tap keeps the relays latched. The reset voltage is isolated from the controller by diodes.

The latch voltage is supplied through a FET transistor which is held closed by the servo fault circuit. If a servo axis faults, main DC power is cut. This shuts down the controller and all other functions. This type of circuit can also be wired to activate an Estop if desired.

There are override switches to bypass spindle, coolant and index to aid in machine setup and program testing.

[Adobe Machine]

Nice unit, again shows what skill ,knowledge and determination can do.
Question: you said if a servo faults, the main dc power is cut. Does your Desk CNC screen still show activity on the axis ? ? My only gripe ( if you can call it that ) is if there is a servo fault, Desk CNC will not see the fault and will keep going..( Axis not moving, but screen showing they are) ? My DRO has saved the day a couple of times when I first started using the program. (found most of the servo faults were due to my not setting the accel ramps correctly ).Just wondered if you found some magic there.

Adobe (old as dirt )

[dcprecision]

Hi Adobe,

With most pc servo systems, the servos boards are linked together with a fault circuit. I specifically use the IMServe boards so this info is based on them. When an axis count error value exceeds a certain number, the board shuts down. At the same time, the voltage at the interconnect points sinks to zero. This shuts down the other axis drivers boards. However the problem is that there is no feedback to the PC. It keeps going along its merry way sending step and direction down line as if nothing happened. I don't know the behavior of other servo board makes, but a simple test can confirm fault voltage behavior. My problem with this machine is that not only are step and direction being sent but also all relay commands as well. So my indexing and spindle are still operating. Potential injury and machine damage exists...

My solution was to switch a 5v reed relay with a FET. This relay passes voltage to a solid state relay which in turn powers the DC power supply. The momentary voltage sink cuts power to the reed relay.Once the relay drops out, the dc power is cut to servos and controller board. A small auxillary 5v power supply is used as a momentary start voltage source to get things running again.

This system does not react fast enough to keep position info intact if used at the board estop terminal. Since the servo faulted,,, the info is wrong anyway.. The DRO is the best way to confirm actual position short of homing.

Dennis

[Adobe Machine]

Thanks for your in depth reply..I have not had a "servo incident" since I got my acceleration curve and SPS in synch..but I sure screwed up some parts prior to that . I have no safety issues , as I can see you did, as the 3 axis just stopped moving.

Yes , the DRO was the answer, but I'm suprised that as advanced as Desk CNC is, there has not been another solution..Hopefully there will be an answer for this one and only problem I have found.

The Lathe program CLGG ( thats turn, thread, drill and grove) put together by Fred Smith @ imserv is a good economical lathe program that puts out an instant g code program that works ! Really neat in that it supports Desk CNC ,you might down load the free one and see if it helps your many operations..

Again thanks for your reply, and it does the old heart good to see skill, craftmanship and some intense planning and thought to build a productive machine (s).


Adobe (old as dirt )

[dcprecision]

It would be really cool if the DeskCNC software/Hardware had an encoder counting feature that could display a virtual dro popup independent of the cnc software functions. This would provide a constant position monitoring system independent of a servo crash and could also help with setup and possibly manual moves.

When that servo faults, the step/direction are out of sync with the program, but the encoder continues to send accurate quadrature data down line that could be used to reset the machine coordinates.

I am exploring the possibility of connecting a dro readout box in parallel with the encoders on a ball screw actuated mill using the dro to power the encoders so that they operate independent of the cnc hardware but are still usable by it but electronically isolated.

[Adobe Machine]

Cool, keep us informed. That solution would make this CNC program a really pro type..Both of my DRO's ( Mill and lathe in progress) have quadrature signals ( I think that is what I read)Being able to have that independant true position would help a lot in case of servo fault.The DRO has saved the day a couple of times in long programs.

Adobe (old as dirt)

[dcprecision]

Adobe,

I can't say enough about how important precision homing switches are. Many times I feel that it falls on deaf ears with the benchtop hobby guys. An investment in a few high precision switches on ebay will save the day every time a servo faults or program crashes. Home then run from line #.

The DRO is a great alternative to them but having both is the best possible situation.. So far as using a dro with rotary encoders, it would have to be a software/hardware programmable virtual dro. I don't think a dedicated box dro has the ability to be programmed for # steps per inch so if you dont build your machine to it's exact scale formulas, it won't work properly.

Dennis