the lack of backlash compensation is one.  If stuff like that is promised for years

Will be interesting to see if something is done in the "very near" future. I have given my 2 cents to Greg on the subject.
Not that i have any influence.........

RICH

Bob: Thank you much for the thorough and well considered response. It is very helpful to me a this time. There seems to be opposing groups of haters and fanboys and it's sometimes difficult to get an objective opinion.

In the midst of my tantrum over the shipping cost, I have continued to research the Kflop. If my butt cools of enough, perhaps I will acquire one to evaluate. We shall see.

Producing the hardware does not seem to be the problem, so I question how helpful Gecko et. all would be as a solution. I will snoop around some more and see what may be out there. If I cannot find what I want, rest assured I will make it myself  . .  and you can quote me on that. There certainly would not be any competition in the marketplace. The question is, how big is the market. My research is showing that many (if not most) systems do not need such a . . . 'helper'  . . like smoothstepper. So long as Mach continues to provide the PP driver, external hardware in optional, and therefor unattractive to develop.  Catch-22. Add to that the very large percentage of hobby users and it becomes more understandable why there is not a line of developers knocking on the door. Just my rambling thoughts on the subject.

Rich: methinks the pile of pennies should be very high by now from the multitude of 2 cents worth that have been thrown in this particular wishing well. I have been hearing 'very near' future since I started on this forum, which has been a while now. The swapAxis board exists because it was faster and cheaper to fix the problem myself rather than wait forever  . . it take far too long buy a feature paying 2 cents at a time . . . 

SIMPSON,
I know what you mean. BTW , i have been following your thread on the rotary and one of these days i will get back into it's use.

I will admit that the SS provided a way to deal with a problem PC that i was using, and i struggled with buying the SS as compared to just putting the money towards a new pc, which i did a year later anyway.  The only thing that kills the use of the SS here is the backlash comp. The only reason for not getting rid of it is that it still can be used with my laptops if i so desire and believe that one day, whenever that may be, you will need an external device like the SS.

Hobby users need BC as it provides an avenue for doing acceptable machining. I have seen some realy crappy systems which use BC and do rather exceptional work. When i say very "near future" i am meaning that maybe Greg will finaly provide something, maybe limited, but accepatable to the hobbiest.

The funny thing about the SS developement is that it agrees with nature.......
"Progress and move forward or wither and die" 

RICH

An expensive SS alternative.
http://www.machsupport.com/forum/index.php/topic,14738.0/topicseen.html

Hobby users need BC as it provides an avenue for doing acceptable machining. I have seen some really crappy systems which use BC and do rather exceptional work.  

As always, you make good points.

I tried Mach's BC and found it to be useful only for spotting type functions . . drilling holes, etc.  The way it is implemented for continuous cutting, if I remember correctly (it's been quite some time) it actually stops with every direction change per axis. This is not useable in my opinion. I think there are solutions to this and I suggested a different approach, but the only result was to raise the ire of a couple of fanboys so it was not a topic that was interesting to discuss, but I'll be brave and just throw out another idea here.

If accuracy is the objective, then another way to spell BC is 'ground ball screw'. If you start with a smoothstepper budget of $180 (assuming $20 shipping . .LOL!) then you can find a pair of used ground ball screws on eBay and fit them to your machine and have money left over for a cold Bud lite. I think the worst used ballscrew will do a much better job than Mach's BC in it's current form.

My need is for a high pulse rate in order to get the RPM I need with the high count encoders that I also need, without resorting to step multipliers. That's a little more difficult to solve if you are going to stay with Mach3. I have quite a time investment in Mach3 at this point and there are a lot of them out there, so it's worth a bit more effort before turing off the porch light.

There are two threads on the Warp9 site were discussions about BC has been going on. Here is Greg's latest response and why i say in the near future.

Thanks for keeping the discussion productive and civil. I think have decided to make it work the best and I will continue on my plans to make it stop whenever there is a reversal. It will be fun to see it working. I will work on it this weekend, but I wouldn't count on me finishing it this weekend. I'll keep you posted however.

There is difficulty in implementing BC to satisfy say CV mode. I would suggest that folks post to the Warp 9 site if they have some constructive  suggestions for Greg.

In a general, good components for the "machine system" are required to eliminate BC, using BC will help in achieving improved machining accuracy
but there will be trade offs in using it.

RICH

2 comments and then I'm moving on.

First, a device like a G540 but with a USB input I think would sell extremely well.  Parallel ports are dying. We can keep haunting eBay for old PC's and buying parallel cards, but at some point that's just not the way.  We need a USB alternative that works and is reliably supported.

Second, RE backlash comp, yes, it can help a marginal machine to produce much better results.  But remember, even VMC's with ground ball screws and all the rest also use backlash comp.  It matters for best results.

Cheers,

BW

Perhaps when Brian gets around to finishing the next version of Mach3, he should get into the hardware business. Or partner up with someone. Moving forward, with the next version, sell Mach as both a software and hardware package. While some people may balk at the added expense, they can always use the last parallel port version. Or EMC. But for the added price, it should be more stable, better integrated, and run on all PC's. Bug  fixes should get done quicker.

But remember, even VMC's with ground ball screws and all the rest also use backlash comp.  It matters for best results.

I suspect this has more to do with marketing than performance. Open loop backlash comp will always have serious drawbacks, no matter what type of machine your trying to use it on. Ask any machinist about watching the table jump on a manual mill as all of the backlash in one axis is suddenly taken up by the cutting forces as your feeding the opposite axis. The same things happens on our CNC machines, so without knowing exactly where the table is there is no way you can accurately compensate for backlash. One benefit of having relatively poor performance lead screws is that the increased friction helps to prevent the table from being pushed around.

Now, if you have glass scales or other form of auxiliary encoders on your table there are some high $$$ motion control cards that can make use of them for closed loop backlash compensation. But then you in a completely different price range, at 10x to 20x the cost of the SmoothStepper just for the motion control card.

Don't get me wrong, I would love to see backlash comp for the SS. My main interest to is to see if there would be any improvement to the circuit boards I make.

Aside from lead screws' increased friction there is also the slides friction and here conventional slides have advantage over linear ball slides.

One thing we usually forget about backlash compensation, and thus underestimate its benefit in producing more accurate parts, by saying "the cutting forces will drag the table in unpredicted directions" is the fact that backlash compensation really matters on the finishing cuts which in their turn produce very small cutting forces which usually wouldn't be able to overcome the friction force of a conventional slide. Not true for every machine and case, but I believe that considering the above, most of the hobby users can benefit backlash compensation, especially when planning well the machining operation.

Also, backlash compensation helps programming a lathe more straight forward and easy, not thinking about the backlash on each move.

Daniel