| | |
Since I can't seem to post in the old thread I'll repost here:
Hi all,
Sorry for the temporary loss of photos on this thread and my new nickname. I left myself logged in on my computer and a "friend" decided it would be amusing to delete all the images in my Photobucket account and change the password and email address for a bunch of forums that I'm on, including this one.
Yeah, funny...
So I've re-uploaded the photos that I still had (sorry for the missing ones) and they should show up now. You might have to clear your cache if you visited recently looking for them.
I haven't figured out a way to get my old nickname back, thus this one.
I would have fixed it up sooner but as some of you know I suffer from Lupus and I've been particularly sick for the last month or so.
At any rate, here's what's new:
Dead Spindle Drive
(NOTE FOR THOSE READERS UNFAMILIAR WITH THESE MILLS THAT THIS IS AN ISSUE WITH THE NM-200 SERIES 2, NOT THE ORIGINALS)
Let's get this out of the way at the top. Like many others, my spindle drive has died. Like everyone else, it went when I had what felt like a million jobs I needed to finish.
I mentioned elsewhere that a while back I found loose connection for one of the signal wires from the motor. Here's a pic of the repair:
One end of the wire (drive-side) had not been properly crimped and the wire fell out. When it did the spindle died. But repairing the wire fixed the problem, and it continued to work for a few more operations over the next week or two.
Then, trouble...
I was just getting started on the first one of those million jobs. It was really a simple, single operation - drilling out a pre-piloted .500" hole in 7075. I'm typically over-cautious when I'm running a one-off piece or something that I haven't ran before. In this case I similarly took it easy with the cut. It was a relatively low-speed (around 1500RPM IIRC - the low-end is apparently the trouble spot on this driver) and very, very light feedrate (2 IPM if I remember correctly, with the recommended being triple or more) with pecks at .050". It got about 1/8" in and died. In other words, it shouldn't have failed. But it did.
So I tried power-cycling but nothing was happening. There was no output on U, V, W to the motor. So I started checking connections. I found a few issues at the drive. The grey wire second from the right in this pic was almost out of the connector. The pin hadn't been properly seated in the plastic:
That didn't fix it.
The largish blue wire going to the capacitor board in this pic was loose enough to rattle around:
That didn't fix it.
I ended up taking the drive out and checking everything. In case anyone in the future needs these pics I'll post some pics of the drive below:
Anyway, to make a long story slightly shorter, the drive is likely dead, and presumably the motor is OK. I haven't dug into the controller beyond this point as it (supposedly) is going to be replaced by Novakon.
So I contacted Novakon about a replacement. At this point the response has left me in the unfortunate situation to have to say:
I would no longer recommend Novakon products
That's a strong statement, and I usually prefer not to take sides, but as I've learned recently that some people have been taking my earlier comments as a recommendation for Novakon (when they were not intended to be recommendation at all, positive or negative) I feel somewhat obligated to state that directly. And here, briefly, is why:
1. When I initially contacted Khai about this (which he was, and always is, amazingly good to get in contact with - emailing me a reply shortly after having a new baby) he mentioned that he'd have to check to see if I was still under warranty. Which I am.
But my issue is that this is a well-known issue with the design of these drives. Because they are faulty they should be replaced IRREGARDLESS of whether it is under warranty or not. In fact, I would go so far as to say that Novakon should have proactively contacted buyers who have the new spindle motor informing them that they will be receiving a fix. Instead, it seems to be a case of ignoring a problem until it has to be dealt with. I'm guessing it's because they can't afford to eat the cost of the repairs/replacements and they haven't been able to force Sangmutan to eat the costs. Which may be a valid justification, but it doesn't change the fact that as an owner you're left SOL.
2. The response (meaning an actual solution/delivery of product, not to emails or calls) is SLOOOOOOOOOW. I've had my issues with my limited contact with Tormach and some of their customers (ugh...) but one thing that I can say is that they seem to have decent stock levels of product as well as replacement parts. If you've read through this thread you know that it took a very long time for me to receive my 4th axis, and when it was received it wasn't usable due to a wrong cable (which Novakon, to their credit, fixed promptly - they had product on hand) and the slots being machined incorrectly.
I'm a pretty patient person but that was a bit annoying. But at least the machine could be used. Now with this spindle issue the machine is just taking up space. The problem is that they don't seem to have any stock on hand of the old drives, and Sangmutan is supposedly giving them a bit of a run around regarding a fix. Last week Khai mentioned that they received their shipment of drives from Sangmutan but that they were "recalled". So there's still no word on a fix. According to some posts by Sangmutan reps on the zone they have fixed the issue, with new motors being 8 pole, but also supposedly with a fix for the existing 4-pole motors. But it seems that this fix may just be vapor.
So to sum up #2, there's no fix at the moment, and no details about when a fix will be available. And then we have to get it from China. The rotab I ordered took the scenic route on the slow boat from China, taking over 2 months (from around Nov 8 shipping date to Jan 19 when it was received by Novakon, after a long customs hold-over - Note that this doesn't include the time it took for me waiting for it to be built, or the time for inspection & shipping from Toronto to here) so who knows how long these will take? There was a mention of paying for air shipment, which I'll have to do to get it in any meaningful amount of time, but then I've got to eat the costs for that.
Now, it's one thing for this to happen to us. The other owner of the mill (my semi-retired father) is teaching in the middle east and, obviously, isn't here to use it, and my health has limited how much I can do with it. Besides which, we are really only using it for personal projects, nothing time-critical or really business related. But I can't imagine being a business owner and having to rely on this. Perhaps when they get a better spindle drive system and it's well tested (ie. after A YEAR OR TWO of sales & testing - remember, most of the problems with this Sangmutan drive & motor have only relatively recently started coming to light) they could be recommended. The guys involved do seem genuinely nice and interested in helping. But for now, I'd stay away.
As it is, I'm thinking about retrofitting a 3 phase motor & vectorless VFD, a la Tormach and the original NM-200. This is after getting nowhere looking for an alternative BLDC drive. I tried contacting another company that makes BLDC motors & drives that should fit (Adlee, recommended elsewhere on here) but they weren't too enthusiastic about the idea, which gives me some pause for thought as to whether Sangmutan will be able to get a reliable fix. I also thought about retrofitting a Fanuc spindle motor but I think it's probably not going to be worth the time & money. Anyone have any other suggestions?
-----
Anyway, what else is new?
Coolant System Upgrade
I did 2 "upgrades" to the mill before it died on me. I'll detail them here a bit. The first was modifying the coolant system.
As most know, the stock coolant system on these mills is pitiful. The pump is rated at 200GPH, which is way too low for this mill, and the coolant tank is undersized and goes dry quickly.
My solution, which really isn't mine per se, was built off the ideas posted in this forum by other members. First, for the coolant tank I used a big Rubbermaid garbage can. I think it's the 121L. At any rate, it was just tall enough to slide under the right "wing" of the mill with about an inch clearance under the drain pipe's spout. It could have been a bit higher/larger, but I've found this capacity to be OK.
I removed the strainer that's on the top of the coolant drain and chucked it. Then I used a reciprocating saw to cut the bottom completely off of the drain. So now you have about a 3" hole straight down. I took a 3" ABS coupling and enlarged the opening on one end slightly on the lathe so that it was a slip fit over the drain. Then I drilled 2 holes through the coupling and steel drain pipe from the outlet and used some stainless nuts & bolts to hold it in place. I added a ~2' tail on the fitting with 3" ABS on the other end of the coupling. Slide the garbage can under the drain, slip the ABS onto the drain, and hold it with a couple of bolts. Now the drain empties straight into the garbage bucket, about 1' off the bottom (to account for swarf buildup).
Next, I bought a dirty water pump that pumps over 2000GPH. The good thing with a dirty water pump is that if it happens to get some swarf it'll push it through, though it likely will cause a clog up-stream. But at least the pump is safe! But to avoid the clogging altogether I bought some plastic screen used for screen doors & window screens and made, in effect, a big can out of it, sewing it together with stainless steel wire. The sides of the can reach up to 3/4 of the height of the garbage can and the motor sits inside. So as long as the swarf doesn't get that high we should be good to go.
As well, I added a 100W submersible aquarium heater in case the shop heat gets shut off.
I cut a hole for the various power cords, the drain line, and the supply line in the lid of the bucket and stuck it back on to help with spray and evaporation. The supply line is 1 1/4" nylon hose. I added a valve at the top near the coolant manifold (visible in the picture below) to allow throttling the pump. By closing the valve you increase back pressure on the pump, which is effectively the same as having a higher head, thus the flow rate drops. You ABSOLUTELY need to throttle down this pump. If I had my time back I would have bought a 1000GPH pump as even that would probably be overkill, but this thing on full output is crazy.
I had to cut an opening in a panel at the rear of the mill to fit the hose in. I connected it up top to a coolant manifold that I made out of 3/4"grey iron. I would have preferred PVC but I couldn't get any of the right sized fittings. The manifold is held on with sheet metal screws and iron pipe mounting clamps (not sure the proper term, they're visible in front in the pictures). This works really well and it's as solid as you could want. Coming out of the 2 Ts and the one 90 is a 3/4" - 1/2" reducer bushing, and then a 1/2" locline valve with NPT threaded end, followed by about 6" of 1/2" locline and a 1/4" nozzle. I have some other nozzles but the 1/4" works really well. It puts a MASSIVE blast of coolant out even when everything's throttle back, so much so that I've had to add a top to the enclosure shown below. Otherwise, just the deflected spray will go anywhere and everywhere.
The 3 separate valves, of course, gives you a lot more control over the flow, which is handy.
Here's a pic:
Really, I mean REALLY, cheap enclosure
Ok, so this is a really, really blatant rip-off of an enclosure written up in this forum a few months back by iGak. I mentioned in that thread that I really liked it, and it was eating away at me so much that I finally decided to build it.
The only problem was that the only supplier with stock of acrylic or polycarbonate within 400km wanted $500 for the 1/4" pieces and AN EXTRA $500 FOR THE ONE 1/2" PIECE!!!!!That wasn't happening.
So I figured I'd make a temporary setup to use until I can get to the nearest major centre and get some real glass. But then I tried it out and this actually worked better than expected, so now I don't know if I'll replace it or not. I guess it'll depend on how things go once I get the mill running again.
Details: I used 2 layers of 10 mil vapour barrier wrapped around 1x4 frames that were screwed to the posts. That's basically it. The frame in front of the spindle just lifts out like the other design. I also added vapour barrier across the tops of both sides to avoid spray (see the comments above about the coolant system) Total cost: $50
Beat that for cheap!
The fact that it was a know entity ( the spindle drive failure) is also my main gripe about the whole situation. Like I stated in an earlier post, I understand problems occur / exist, but common courtesy isn't too much to expect ( which in this situation to me would have been a simple notification of the issue BEFORE the failure).
Khai contacted me today and said they'll be shipping out an old-stock (ie. not updated) drive to me until the new one gets here. So at least I'll be running again. Also, from what I understand Sangmutan still hasn't completed a fix. So the wait could be a while...
Looking at the drive today I wonder if adding a fan to the heatsink would have helped? That heat sink seems really oversized for cooling a power module and a few diodes. And not in a good way - with a tiny localized heat source you need a lot of surface area per volume of metal in the heat sink to properly dissipate the heat. And at a glance that heatsink doesn't seem suitable, though admittedly I haven't ran any calculations. Adding air, though, should help somewhat.
And I wonder what temp the power module is rated for? Presumably that's what's failed...
JoeBean, have you gotten your new (old) VFD yet? Is your spindle working now?
Yes and yes, although the cause was not what they claimed in their explanation thread - the controller "unable to reset" - the power module on the drive was blown. Which was what I thought, and backs up my previous assumption that the author of the explanation didn't know what they're talking about.Originally Posted by adam1071
It may be as simple as putting a suitable heatsink on the module, as the one they're using is ridiculous and inefficient in this application. In any case I currently have the entire stock of the power modules available in North America (a whopping 3) on the way for backups and experimenting. Mouser's backordered on them for eternity and a half.
I'll be doing a writeup on my website (I don't want it disappearing) once I get them in. I also have a Pickit 3 on the way to look into reprogramming the uPIC on the controller. There's a header right on the board.
Off topic, but I couldn't help but notice the blatant use of a Harley logo on the Sanmutang board (Joe Beans 5th picture down).
Yes, that was brought up in the product announcement thread by Sangmutan on here. When it was pointed out I believe the Sangmutan response was essentially "Wow! What a coincidence!"
FWIW the newer boards from Novakon have Novakon's markings, not Sangmutan's.
Posting Permissions
| | |
LinkBack URL
About LinkBacks
