What's new
What's new

Best way to go from Bridgeport to Low End CNC in 2024

Clive603

Titanium
Joined
Aug 2, 2008
Location
Sussex, England
A device I've been back of the enveloping for myself in a fairly desultory manner over the last 20 years or so might have a niche market. Maybe 50 to 100 a year at £100 to £150 net profit per item which is adequately interesting for even part time "retired guy" with a shop in the back yard. Turning the handles on my Bridgeport will take far too much time and trouble to be viable and proper CNC VMC will be far too expensive. Especially as I need knee mill versatility, not to mention space for oversize components, to handle what I normally do.

So I need to look into what's affordable in terms of either converting the Bridgeport or buying a three axis CNC mill to replace it. This years mad money stash is about £5,000 - £6000 which will have to cover the investment, helped by selling the Bridgeport if need be. So call it £8,000 tops.

I'm in the UK so options are probably more limited than in the USA.

Historically there was a certain craze for converting Bridgeports but the economics have always been dubious due to ball screw and motor costs. My Bridgeport is in pretty good shape and I have new nuts for both axes and a new Y-axis screw in stock so in principle simply buying a new X axis screw might make conversion marginally viable using something like Leadshine "servo-stepper style" motors. I can't see me doing enough work for screw wear to become a major problem. I have Mach 3 on a computer hooked up to my white elephant Taig which could do the control bit although I'd much rather something more modern able to interface with CAD files. Significant work not helped by much of the readily findable, internet, information being more of "project" than "sorted" nature. The quill drive seems to be the tricky bit.

E-Bay currently shows an assortment of Prototracks, a three axis Acra with older Anilam control, a couple or three Interacts and rather attractive Gate with rotary axis all on older Heidenhain controls ranging from £750 to £4,000 so the budget doesn't seem impossibly unreasonable. The Acra at £2,000 is said to have had a light duty life so looks a good buy. I imagine the Interacts will be rather long in the tooth and somewhat worn by now as being proper industrial machines that have worked hard for their living. One of the Prototracks is only £750, poor cosmetics and pre-Edge control isn't good for showroom appeal but if I'm going to be replacing the control anyway the price is attractive.

Whatever I get will almost certainly need the control replaced if I want to use modern CAD generated files with the machine. Prototracks seem remarkably limited when it comes to accepting CAD based data. I've no desire to become an old style G-Code jockey as would be pretty essential to exploit the older Heidenhain and Anilam controls so I'm pretty much committed to changing the control.

But what for?

Linux CNC is cheap and said to be effective but the learning curve looks horrible. Centroid Acorn is said to be decent but is there a UK source? There are various direct from China Fanuc clones about at attractive prices that look suitably industrial but support is probably limited and its said code implementation can be iffy. The Siemens Sinumerik 808D is said to be well suited to knee mills and appears to be a modern industrial control that may be within budget with some support in the UK.

From my position of ignorance buying the aforementioned Acra and retrofitting a Siemens Sinumerik 808D looks to be closest to a "just works" and in budget option. OK its a retrofit so expecting "just works" is somewhat optimistic but I can hope.

What advice does the PM team have?

Clive
 
I've done a Bridgeport Interact with Centroid as well as as number of different other machines / controllers.

Ive done Mach 3, 4, CS Labs, most of them. I've also done the SZGH Fanuc clone controllers on my Hardinge Lathe.

It all depends on what you need.

I've just bought a 5 Axis Gate Mill (Bridgeport type clone) with the original controller and also exploring options.

How many of your questions are related to the mechanicals (new ballscrews etc) versus motors / controllers etc?
 
Thanks for the response.

Fundamentally it's about what control to choose. The Heidenhain, Anilam et al controls on a machine old enough to be affordable aren't going to do what I need to match the way I'd like to work. Changes to the mechatronics may also be needed. If I do a conversion I need to choose a control and mechatronics from scratch and hope that new screws and nut will be accurate for long enough as ball screws will bust the budget.

Whatever I do has to be pretty much plug and play, not a project.

Basically I think I need the machine to do two things.

1) Retain capabilities similar to driving the Bridgeport manually albeit via an MPG or similar pendant. Also being able to set really accurate stops, cut circles and do things like hole positioning in both rectilinear arrays circles et al in a similar manner to using a DRO on a manual machine. Albeit quicker. Probably want to retain manual hole drilling ability too as its easier for simple stuff.

2) Easily program the bunch of relatively simple components that I'll need for the putative project to take the laser cut parts that will make it all potentially worthwhile. I'd want to do that from CAD drawings. Hopefully the whole CAD to machine process would be easy enough to do other reasonably simple things without too much effort. ProtoTrak style standing at the machine conversational programming isn't going to fly. I understand the modern ProtoTrak is much better when it comes to CAD input but upgrading an old systems is £15,000 over here.

Around 2005 /06 I got myself a Taig CNC mill and Mach 3 to play with and learn CNC with. The machine still sits there but I never really bothered with it because it was clear that learning to be serious with G code was never going to happen and, at that time the sort of CAD integration I'd have been happy with just wasn't out there. I'd have learned G Code et al pretty quick if there was need but one never occurred. Quite enough manual machine jobs to keep me busy.

No issues with learning a new 3D CAD system if need be but right now 2D does me fine. Have a nice 2D to 3D converter in my head.

Clive
 
Understood. I'm happy to share my number if you want to chat (also UK based, Oxfordshire).

In your shoes (and likely on my 5 Axis) ill likely go Centroid Acorn with AC Servos - happy to chat through the pros / cons that Ive seen with the various options over the years.
 
What about sub contracting it out to somebody else? It sounds like the volume isn't quite there to justify a machine purchase. Would you be willing to make say, £60/unit, and basically just assemble what you have contracted out? Depending upon the work, some smaller shops or one man bands with machines already would probably love the quantities you are looking at.
 
I take it you bought the tempting green Gate with the 5 tha axis off E-Bay then.

Chat would be nice, I'll share my number via private message soon, but things have gotten pushed back due to something else blowing up in my face. Friends network thing I couldn't justfiably say no I won't help to.

Clive
 
Yep, it was a few miles down the road from me so transport was a smaller cost and the seller had offered to help me get it into my garage and reinstall the head which we have done. I'm now working out how to power it as I have no 3 phase but some options ... Ill PM you my number.
 
Mmmmm - struggling to see how to PM, maybe my post count is too low?
 
Last edited:
What about sub contracting it out to somebody else? It sounds like the volume isn't quite there to justify a machine purchase. Would you be willing to make say, £60/unit, and basically just assemble what you have contracted out? Depending upon the work, some smaller shops or one man bands with machines already would probably love the quantities you are looking at.
The economics of contracting out don't work at that sort of volume.

Basically I need to be able to make to order, if its only 1 or 2 a month thats what it is. Frankly I've no idea how many I'd sell. Very much a "Gosh thats wonderful I'd love one." thing right up to when it comes to handing over money!

To be honest the whole product thing is bit of an excuse anyway to make life easier on the usual stuff I do for folks as retired guy therapy. Hence the mad-money comment. Bitter experience has shown Clive doesn't do well buying stuff without a proper requirement to hang the choice on.

Clive
 
At work The Prototrak controller died on one of our mills, and they decided to get rid of it. So it went home to my garage. After doing a whole lot of research I finally settled on the Centroid all-in-one DC system. The reason for going with that is is that it's set up to handle the DC servo motors that are already on the '90s-early 2000's vintage CNC knee milling machines. It's a lot more money than going with the Acorn system, but then you don't need any individual motor controllers and you can use the motors that are already there. The added advantage, and key selling point for me was that when you go to the all-in-one, the machine has feedback that comes from the encoder. This allows you to turn off your motors and crank the handles on the mill as if you're running a standard knee mill. Since most of what I do is prototyping onesie twosies maintaining the use of a manual mill was still a priority of mine. It's my understanding that if you go to the Acorn system any movement of the table must be done with a with an encoder wheel, and through the control. There is no just cranking the table and taking a simple cut when that's all you want to do. I had also considered Mach, and a few of the other homebrew systems but it did seem like the centroid gave you the most professional level, least buggy and well supported hardware.

As far as a conversion it was very straightforward. My father is an electrical engineer and him and I wired it all up together, the instructions were very easy to follow and it took the two of us working together about 4 days of hard work. I work for an industrial machinery builder so I was able to scrounge most of the wire, cabinet etc that I needed out of our scrap bins which helped out a lot, but it comes with most of the parts you need. You probably will need to also buy new encoders which will run another $200 each, and you will need one per axis. So factor that into your overall costs what your time is worth.

Centroid's tech support via their discussion forms has been very good. My biggest complaint is that while it seems they keep up on their hardware, and have very good support, the software is very old. There is a great YouTube tutorial on how to run the software that was done back in 2011 or 12, but in there you see a 2004 date show up on something. Meanwhile the software I have on my mill at least in their CAM system is more or less the exact same as it was 20 years ago. It would be nice for them to innovate and add a few more features to make it even more intuitive and easy to use.

So far I've only programmed up one part. As I've only had it working for the past week and a half. Their cutter comp and entry into the parts is a little interesting, but I don't have much experience on any other CNC programming to really compare the differences. I really just wanted a conversational machine that would be able to do contours, arcs and basic functions better than my manual machine can do. I build model steam engines and most of my stuff replicates pre-CNC era parts, thus for low volume, I don't really need too much of a mill.

One more thing to consider in all this that I went through and making the decision as to what machine to retrofit, is the height of your garage door or how you're moving your mill into your shop. Around here in the USA CNC iron with blown controls really isn't worth that much money, so you really have your pic iron to start working with. If you want to go the retrofit route. I could have probably traded/sold my machine which is a standard knee mill for the prototrak KMX style bedmill. A local used machinery dealer got a few of these in, that had all been scavenged as the controllers died, but they basically were 3 axes with ball screws, and DC servo motors. I wanted to maybe get rid of my machine and go for one of those but then I realized it wouldn't clear my garage door. I came across some other really great 3-axis, early '90s late '80s machinery with dead controls on it, DC servo motors, and many of these machines since their controllers weren't that good to begin with had little wear.

However in the end of the day, the standard Bridgeport style knee mill was all I could fit in my garage, and get in under the door, even if the third axis options should I ever go that route are suboptimal.
 
Thanks very much for that thoughtful and detailed response.

Looks like you've worked through most of the queries and thoughts I have and come out the other side with a working system. The Centroid Acorn certainly seems a very good way to go when trying to organise a conversion without too much individual development. Plug and play is lovely idea but it's never going to happen. Centroid appear to make it relatively easy to sort out matching sets of components so the wheels go round reliably once its all wired in together.

Unfortunately no representation in the UK so I've got to arrange my own import. I'm told its pretty easy but dealing with customs is another step.

The age of many software components does worry me a bit. The potentials have changed drastically in the last decade, mostly driven by imaging and 3D printing. Be nice to have a modern interface with modern code behind the screen rather than just "bitt better than Mach 3, less clunky than a Chinee Fanuc clone.".

Time to carve out some research time I think.

Clive
 
Thanks very much for that thoughtful and detailed response.

Looks like you've worked through most of the queries and thoughts I have and come out the other side with a working system. The Centroid Acorn certainly seems a very good way to go when trying to organise a conversion without too much individual development. Plug and play is lovely idea but it's never going to happen. Centroid appear to make it relatively easy to sort out matching sets of components so the wheels go round reliably once its all wired in together.

Unfortunately no representation in the UK so I've got to arrange my own import. I'm told its pretty easy but dealing with customs is another step.

The age of many software components does worry me a bit. The potentials have changed drastically in the last decade, mostly driven by imaging and 3D printing. Be nice to have a modern interface with modern code behind the screen rather than just "bitt better than Mach 3, less clunky than a Chinee Fanuc clone.".

Time to carve out some research time I think.

Clive
I buy a decent amount from the UK and there's a few great parcel shipping services that offer very cheap rates from the USA. The packages actually come almost overnight to me as I'm near the JFK hub for DHL. Centroid is out in Pennsylvania which means they're pretty close to JFK too.

I wouldn't worry too much about centroid support once you get the basic system. Their web support has been very very good, I post something at night and usually hear back from them by noon time the next day here in the USA. They have their applications engineering department answering their web questions on their own form, and some of the issues I've had have already been answered there.

The other thing I like about centroid is that it supports a lot of third-party stuff very easily. That's something I could not stand about the prototrak, is that they want you to upgrade to their $15K system and up charge on everything. You can buy most of your parts you need off of AliExpress, or other sources if you decide not to go with centroid

If you're going to go with the Acorn system start pricing things up and you likely will find yourself just deciding to go up to the all-in-one DC as it's essentially made to be the plug and play retrofit for any DC servo motor machine. Once you start to realize what all the parts cost, you wind up getting pretty close to the all-in-one controller and then you're putting in questionable Chineseium motors versus the well built industrial quality stuff that you'll probably find on your old vintage CNC. The other nice advantage of using that system is you don't need to worry about changing your motor mounts they're already there and correct. The only thing I had to do was have some cover plates laser cut up at work when I stripped out the old prototrak controllers. They also have a similar version that is the same thing except for for AC servo motors if your machines have those.

But at the end of the day I just didn't like that acorn much. The whole system relies on telling something to go 1 in to the left and then assuming it went there. If you have a knee mill being able to turn off the motors, and just crank like you're used to is really nice. On the other hand I can also quickly tell the thing to rapid to positions which is pretty cool without having to program anything so I kind of like having it both ways. I think you would be foolish to try to save a few $100 and use the acorn. That thing is really set up for the guys with the cheap routers or DIY plasma cutters. You on the other handle probably be buying an industrial quality machine with industrial quality motors. So why not just stick with the industrial quality controller.

Once you build out the centroid control box (most of my electronics came from AliExpress) it pretty much actually was plug and play. We did have 8 fun man days of work between my dad and I wiring the whole thing up. I seem to recall somewhere on centroid's website they will even sell you the whole thing wired and in a box and that will be pretty damn close to plug and play if you want that but I think it was another 2 to 3,000. On the other hand what's your time worth? For me it was just a fun activity with my father which was actually priceless.
 
I hope this is not considered hijacking of your post, however I was thinking about this some more. I haven't fully had the time to play with the Intercon CAM system that's embedded in the centroid but so far as I wrote above it's the one part that I'm less than satisfied with.

On the hardware side I think Centroid put together a very good easy to use user friendly system, and I like that it has so many options for future expansion all built into the base cost. From the homework I did I didn't find anyone else who offered similar level of support, and hardware integration. Factory support through their website has been excellent, and since the hardware is all designed to drop onto my DC motors it all worked almost flawlessly.

The few hiccups I ran into we're all a result of me misreading the instructions, and we're quickly cleared up on their form. On the other side as I wrote above they kind of leave you hanging when it comes to their conversational programming. It seems to have the best that 2004 had to offer, and at some point they just gave up on innovating it.

I am still on the free trial of their premium software but I may soonfind myself having to either upgrade to their mill pro software or find something else.

Can anyone here recommend a good intuitive easy to use CAM system, with conversational?

I have access to the Autodesk software as someone who is a mentor on a first robotics team, ( they allow you to download anything in their family for free) but when I started playing with their inventor cam plug in it felt like it was way overkill for what I'm trying to do. The tool changer is me, the z-axis is me, the drill cycle is me on the quill, the spindle speeds are me. I really just want something where I can either draw out a simple contour to go cut, call out some bolt circles, call out some pockets etc... Basically I'm trying to use the machine just like a prototrak mill is used in a prototype environment.

I sense though that this is really backwards to how modern CNC is going, which puts our needs into a little niche that it seems like prototrak likes to service for about $15k

I'm only 2 weeks into learning curve, and I don't get that much time to play with it, so maybe some of this is a little too preemptive, but I'm curious what others think is the best way to go when it comes to actually programming these machines?

Anyone ever seen this software? https://kentechinc.com/kipwarem/ is it worth it? Centroid limits you on the number of lines they give in their intercon program. Time will tell if it's enough for me. So far I've been only doing very simple shapes, but if I have to upgrade to their mill pro software I am tempted to consider other options that would generate g code, while possibly being more intuitive to learn.
 
Interesting post. FWIW, Clive and myself had a 30-45 min call a few days ago on my experiences.

I'm doing a small Denford Triac retrofit as we speak and have opted for Centroid Acorn powering Panasonic Servos. Included in the purchase was the Mill Pro license. I dont see a feasible way to run Centroid without this. For the cost of a tank of fuel (in the UK, I know, horrible), gives you the license, its affordable. Paying £250-300 for an MPG that is a rebranded Chinese unit isnt great but it is what it is when you consider the larger eco system.

I'm taking a look at the Kipware, not seen before, thanks. I used to use Cambam before, lightweight, good for '2D' type stuff which might suit you?
 
Adammil1

Interesting and thoughtful post.

I'm glad you are getting to grips with your system but it seems that the objections I had when taking a run at Mach 3 on the Taig horror show still haven't been fully answered. I'm inclined to agree that modern CNC & CAD is evolving the wrong way for folk like us. I can see a fair number of instructional videos, manual downloads et al in my future before I decide whether or not its worth pursuing the project given that, at 70, I don't have too many good years left. Might even have to fire up the Taig thing to remind myself what I didn't like. Yuck!

I draw everything as a matter of course, so Prototrak style conversational isn't really a great need. Certainly not at £15,000 minimum upgrade price over here. Being able to run in crypto manual mode would be nice for some things but I'd probably not miss turning handles so long as I can simply set the end stop position and hit the go button. But handles would be nice for setting up with a wiggler. I think my Huffams are great but adapting to an MPG pendant might take a while. I'm surprised you can turn the handles with a servo machine though.

Possibly the hardest thing will be getting my brain aligned with the way things are supposed to work. 3D CAD is going to be tricky after so many years of 2D. I've got an excellent 2D to 3D converter in my head! But carving 3D from a block is just plain backwards.

I have fair bit of control system design experience but starting out on IR homing for terminally guided sub munitions means that I instinctively think in vectors. Potentially lots of trouble with conventional CNC. Especially as having to fit teh whole homing and control systems into £100 at 1978 values was seriously weird. Actually behaved as an emergent system and essentially impossible to model on a digital computer, i used a hybrid analogue digital one, But it worked just fine, really pissing off the powers that were. Ever since I've never seen much point in knowing every point on a path exactly. So long as the start points and end points are right and the trajectory doesn't exceed the limiting parameters it can do what it likes in between. But judging by some kith things folk report struggling with on here, conventional CNC makes a lot of things difficult that I'd expect to be trivially easy on an emergent vector system. Wish I still had my lab books!

Hafta see how things work out but I'm way further ahead than I was. Clearly I can get something workable up and running within the sort of budget I have. But whether its something I actually want to run is different matter.

Clive
 
Clive, turning the handles on a servo machine is very easy. You just turn the servo motor off. There is a big button right on the centroid control for doing that. When the Servo motor is on it fights to hold its position. If you try to move it you'll feel the motor fight you back. This was the real reason to stay away from Acorn for me, if you try cranking on the acorn it will lose its position and you will have no feedback. When I crank the handles it's like watching the regular old DRO show me the position. I'm still in a transition mode and find it really nice to be able to do both manual and CNC on the same machine. If I didn't have such an awesome manual mill I probably would actually sell it at this point and just stick with the hybrid system that I now have.

What were your Mach3 Taig objections?
 
Clive, turning the handles on a servo machine is very easy. You just turn the servo motor off. There is a big button right on the centroid control for doing that. When the Servo motor is on it fights to hold its position. If you try to move it you'll feel the motor fight you back.
This the kind of thing that makes me not sleep at night.
I turn off the servo (actually I disable the amp).
Now my motor is a generator as I turn the handle.
It is looking at a capacitor bank that will charge fast and the back ass end of a bridge or transistors that will not let power through in this direction.
Does much use of this damage the motor or amp?
 
Look at the Chinese controllers and servos.

GSK comes to mind for the controllers.
Plenty of affordable chinese servos/drives/cables availible.
 
Clive,

What were your Mach3 Taig objections?

20 (ish) years later I'm not sure exactly what they really were! Even after a couple of days contemplation.

I bought the Taig and some software after redundancy to, hopefully, prove out a potential business concept. It was sold to me as being able to do what I wanted. Technically true but involving major league mucking about to generally persuade it to just about handle some things it didn't want to do. Which was where I discovered that normal G code and point to point CNC struggles with some things that are really easy if you can directly run as a vector. Classic example of not properly understanding the difference between concepts.

Didn't help that the man handling Taig CNC at that time was getting out of the business, I may well have been his last customer, and basically disappeared leaving me with machine and software that was never properly set up to behave itself. Not to mention selling me a quite expensive software package that included £1,000 of training without bothering to arrange said training.

So the toys got slung out of the pram.

I briefly looked at writing my own controller.

Then the post redundancy consultancy contract went from maybe 6 months of as and when work to cover any rough edges to 2 1/2 years full time. Hence the whole thing became moot and the Taig was banished to the maybe one day corner. Redundancy money, some manual machining plus the tax bonus my accountant arranged when I shut the VAT registered consultancy business down tided me through the carer years until works pension time quite nicely. So it still gathers dust.

Objectively a blessing in disguise as the business idea would have flown about as well as a ruptured duck.

Clive
 
This the kind of thing that makes me not sleep at night.
I turn off the servo (actually I disable the amp).
Now my motor is a generator as I turn the handle.
It is looking at a capacitor bank that will charge fast and the back ass end of a bridge or transistors that will not let power through in this direction.
Does much use of this damage the motor or amp?
I never thought of it that way. In fact I decided to post a question over on the Centroid CNC form as it is an interesting one to ponder. On the surface I would expect I should be okay but I'm curious to hear what the experts say. I'm thinking that I should be fine. If you think about it if you have any internal load on the system I would think it would want to backfeed a current, thus they must have some way to handle it with the onboard electronics. I know on large servo motors it's not uncommon to see big breaking resistors for the exact reason that when the RPM is supposed to drop rapidly they have to toss the excess current that the motor gives off somewhere. Now in this case I'm not cranking the hell out of it I'm just casually moving things around and so far so good. But once again I'm curious to hear Centroid come back and explain if this is or isn't an issue.
 








 
Back
Top