I'm loosing it...

[fredtoo]

Comm errors have been an ongoing, since new (3 months), problem with my PRS Alpha and 4 hp single phase spindle. Grounding has solved 'some' of the errors but until yesterday I could not do a continuous cut for more than 15-20 minutes without an error. Prior to my grounding exercise (x-rails tied together, x-rail to dust collector, wire through dust collector hose, x-rail to ground inside control box and x-rail to pc chassis, I often could not get past a few minutes without an error. So what happened yesterday? I decided to ground the spindle control box chassis even though it was grounded via the power cable from the control box. When I opened the spindle control box (first time ever) to see where it was safe to drill a hole for a ground screw I found the little board which the USB plugs into resting on top of the fans inside the control box. The four little foam stickies had let loose at some time. I relocated the board and grounded the control box chassis to the control box heat sink and to the x-rail.

So, since that time I've cut two test files, each over two hours long AND the spindle speed control on the pc now works (it never did!).

No happy dance yet, but I'm warming up......

[bill_lumley]

Greg , thanks for sharing . You made me thinking of something on my Alpha that may be a problem . When I mounted my VFD I did not have room to use the supplied Shobpot brackets and so mounted the VFD to the frame using a piece of wood so in fact my VFD chassis is insulated from the frame . So I called a buddy of mine who is building a Mechmate and he said he was having all sorts of issues with files not running and eventually it came back to the connection from the VFD and Spindle in his case . He said what happens is when the spindle is under load it creates a field that if fed back to the controller will end up overloading the relatively weak signal of the USB connection . His comment was to make sure that the shielding of the Spindle cable is properly connected so that it is not acting like an antenna to drive the power of the Spindle signal back into the control system . This also explains why the inputs are part of the resolutuon for some folks . By disconnection the limit switches you are taking away an antenna from the VFD signal and limiting the power of what gets back to the control box .

In summary he said it is more of a shidling issue than a groudning issue but can manifest itself as one . The explaination also explains why I was never able to get the VFD speed controller working . All I was told on that one was to make sure the VFD cable did not cross the USB cable but I think it is all related . If you have that USB device inside the VFD connected to the same USB HUB as going to the PC of course that VFD signal will get fed back if grounding is not done properly in the VFD .

I hope this makes sense . It is hard to describe but frankly I think Shopbot really needs to get it's act together on this issue ....

[fredtoo]

Bill,

Ditto for me. I mounted my VFD controller on a piece of plywood parallel to the Y-axis on the middle leg, more or less under the table. I 'assumed' the power ground to the SB control box would do the job. Also, in my case the little USB board sitting on top of the fans could not have helped.

I didn't mention in my earlier post that all my added ground connections are sanded so that I have a good connection to bare metal.

But I did mention, no happy dance (no videos allowed) yet but I'm hoping.

[bill_lumley]

Well I talked to Ted Hall yesterday and we 'discovered' a few things while talking . Firstly I had the exhaust from my vacuum motors pointing at the control box on my Alpha and the exhaust can get quite hot especially when it was funnelled through the muffler on the vac so was only about 8" from the control box with it's blast . Aparently the control card will shut down if it detects too much heat . In an upcoming software release there is a facility to monitor the temp. in real time inside the control box .

Another point came up when I mentioned a com error I had seen early on with my machine when I touched the dust collection attachment point to the gantry while the machine was running . I have flex plastic piping connected to a galvanized metal pipe which is attached with band clamps to the gantry . In effect I had my dust collection foot grouded through my gantry instead of back to nuetral where I thought it was going .


I have removed the muffler on my vac to allow the heat to disperse more and put an insulating layer between my dust collection hose and the gantry . I have only run 4 sheets without issue so I am not ready for the happy dance either but I am more hopefull than I was a few days ago . Ted also assured me they are looking at improving the handling of the Comm Error in a future release but I didn't get specifics on that ....

Bill

[paco]

Ted also assured me they are looking at improving the handling of the Comm Error in a future release but I didn't get specifics on that ....

Yeah, if it just wait rather than crash...

Most, if not all, of my latest 'Missing com...' prompt/crash has been while the tool is idling with nothing running like during the night or lunch break, in between files run when I take care of the parts and scraps and also during K mode.

Ted has been emphatic to my problems as well and I'm very thankful for that.

[bleeth]

There seems to be one major difference between my upgraded PRT with Gecko's and what I have been reading about those with regular comm issues on this thread. That is that my control box is hooked up to the operating computer with a serial cable and high speed serial card rather than a USB. It would probably take more technical analysis to see if this is actually a difference that makes a difference or not but it does seem like a place to look.
I can tell you that my operating computer is not all that strong and is not loaded with ultra ram. My machine is grounded, but once again, I would not say that I have gone into extremes in making sure that every bit of it is totally grounded in the most perfect way. Cables are not very isolated from each other but I will say that I am most definately not running the most up to date SBC software. I stopped upgrading just as the first 3.6xx came out.
I have a distinct feeling that some of the more technically edgy developments that have been done to the bot in the last few years had a bit too little field testing under various operational conditions and too quick a release due to an (understandable) desire to please the more and more demanding nature of the professional user. When one looks at the few short years between the PRT and the latest Alphas and even PRS's the difference both mechanically and technically is exponential.
Throw in the R & D on the whole Buddy series, and the pending release of the next generation Bot and you have an awful lot of changes in a short time. To properly test each and every one of these updates to the nth degree and ensure that they all operate flawlessly would tax any operation, let alone one as closely held as SB.
It could be that it is time to slow down on new developments and spend some time verifying all the operating parameters of the current equipment and software. A bit of managerial conservatism could go a long way here. After all, if SB is, as seems clear, going after a larger market share of those who simply wish to plug in their machines and make dust, than the dust needs to be made as intended.
Those are the thoughts of this very supportive SB'er anyway.

[beacon14]

I had similar lost comm errors several years ago over a period of many months. Long story short when the computer finally died, I took it in for repair and was told one of the memory cards was defective. The card was replaced by the manufacturer, and the lost comm problems went away.

I know at least one responder listed a new computer as a troubleshooting step that did not help, but there could be more than one reason for similar problems in different shops. If you have these problems and can swap computers to see anything changes, if might be worth a try.

[boubou]

Hi everyone, if this can be any help...the only time as was getting (missing com) messages was when i was trying to push the xy speed too fast for my prs standard, loosing com going around corners.
4 weeks ago i upgraded to a spindle and i was getting (miss com) error all the time, could'nt get the bot to work properly. So i took a 14ga (normal plastic wrapped house wire) and grounded the spindle mounting plate to the power earth ground. I dont have any more problems and running ecabinet since with big files, 20 sheets and more. i am using window 7 with great success. Boubou

[paco]

Anyone is having disconnect, 'Missing com...' and such with laptop?

[bill_lumley]

Hi Paco , sorry you are still have these ... I may be on to a solution for mine in a place I never expected it . Ted tells me the Controller card in the Alpha (and I expect standard) will shut down or show eractive behaviour when there is not enough cooling . I am using 4 of the central vac motors for hold down and had one pointing directly at my control box . I did a successful run for 1 hour yesterday with no issues and no vac motors pointing at my control box . I am not sure I have solved it yet until I get some runs that go for 3 or 4 hours but it may be a strong possibility . I measuered the output temp of the air on those vac's and it was 180 degree F climibng to 210 degrees F when there is a muffler attached to it !

The electronics are only rated at 180 degrees F but can show functioanl failures at temperatures below that . Good Luck . Hope to see you at th eCanuck camp coming up .

Bill