Escape from (VFD) noise?

[ because it seemed to belong in a number of different forums... I dumped it here instead ]

I want to preface this by saying I'm no stranger to VFD noise - I've fought with it before, I've fought with it on this rig (this Huanyang VFD broadcasts crap like a talk radio station in campaign season), and I've seen it showing a lot of these same symptoms in the past - so I have little doubt that there's an major EMI issue at play. It's the surrounding circumstances I'm a little puzzled by.

First things first: I've got a half-a-Lead 1010 being driven by (for the moment) OpenBuilds CONTROL under linux, via an OpenBuilds BlackBox that also controls a (very much not Openbuilds) Huanyang 1.5kW VFD and (before I unplugged it out of desperation) a shopvac on an IOTRelay. I probably also have a variety of other bits of related technology that I didn't think to mention here, but probably should have. I'll apologize once someone calls me on it.

Earlier tonight, everything was working normally (ie, the machine was making massive holes in my workpiece in all the wrong places, and I was swearing profusely and throwing things). Then the toolpaths started... stopping. Not aborting, not stalling or losing steps - the machine acted like it was a feed hold, while CONTROL was acting like... well... also like it was in a feed hold, but without showing any sign of being paused. Or like it was still cutting, except without moving. Or like it got tired and decided to insert an indefinite dwell into the program so it could go wander down to the bar & grab a pint or three. Since it was behaving sorta-vaguely-paused-ish, I tried pausing & then unpausing it, which worked pretty much the same as if it had been in a feed hold. Except that it happened again a couple minutes later. And then again a little after that. After another iteration or three, it stopped unpausing.

After fiddling with various things for a while (including a lot of things that could easily be mistaken for doing the same thing over and over again while expecting different results), I started occasionally seeing some noise in CONTROL's console log - assorted fragments of grbl MPos status lines, mostly - that I'd seen coming from the a noisy VFD a few times in the past (...and resolved... and lost the notes explaining how), making it apparent that at least a major part of the problem was VFD EMI (and, yes, the log noise would go away when I shut off the spindle, and would get even worse at the lower speeds - so if not causation, definitely close correlation).

As the night went on and my troubleshooting went nowhere progressed, the noise in the log became more and more pronounced - instead of the occasional line or two in the log, I'd get several a second as soon as the M3 went through - and then CONTROL (or the BlackBox, or the GRBL grmlns) started dropping the USB connection - sometimes requiring me to un/re-plug before reconnecting.

It's the rapid escalation more than anything else that has me most worried - as if something's getting ready to die, explode, or maybe give birth to some unspeakable noisebeast - and, despite having absolutely no rational reason to think so, I can't help being paranoid that one component is getting ready to wipe out all the others in some sort of nuclear EMP murder-suicide.

My cable routing is far from ideal and altogether too much of it is unshielded or ungrounded... but I've already cut a few different projects in the week or two since it'd last been touched, so I have trouble believing it's responsible for a sudden onset problem like this (I'd also rather not get into another big rewiring project right now, only to finish & discover it was the electronics at fault).

• I've tried moving around cables, with no clear impact.
• I've tried unplugging the motors & other lines from the BlackBox on the chance that one was acting as an antenna, with no change.
• I've tried unplugging even the TOOLHEAD line from the VFD (leaving nothing but the USB & PSU attached & running the VFD from the front panel), and, while there might have been a bit less noise (might), the problem certainly didn't go away (the noise in the log, that is - by this point I'd long stopped even trying to run a toolpath).
• I tried lowering (and even raising) the VFD's carrier frequency, and all it did was take longer to accelerate past the extra-noisy slow speeds.
• Hell, I've even tried [insert other things that I was too- -impatient to take notes on and don't quite remember now].

So, long story sh

So... any thoughts on what might be causing the problem - or, more importantly, causing it to suddenly accelerate from unnoticeable to unworkable like this? Is this a known failure mode for Chinese VFDs? Is there a dying filtration circuit in the BlackBox? Are there mutant termites gnawing the shielding off my spindle power cable? Is CONTROL getting resentful about all the bug notes it sees me taking?

...or were things just starting to go too smoothly for me?


-Bats
(...did I overflow the posting character limit? Did I lose my audience a page or two back? Did... are you still reading this?)

 

Definitely possible, at least in general terms. Right now I've got shielded robotic cable on the spindle, but I've never been happy w/the ground situation on the VFD - a screw in a plastic case is a sorry excuse for a chassis ground. New routing w/a metal box for the VFD was on the eventual To Do list already - although I'd worry that, unless we're thinking of two different things, the "flex" Conduit I've worked with isn't nearly flexible enough to keep up with the spindle movement - but why would it wait weeks to manifest, then ramp up out of nowhere over a couple hours?

That really is the sort of problem I'd have, isn't it? Except that I tried both lowering and (for no good reason) raising the carrier frequency. And, again, the wire length hasn't changed recently, and certainly - barring those mutant metal-munching termites-not gradually over the course of the night.

My shop's basically a tree of power strips - so while they're not in the same immediate strip (I like to have independent control of the VFD & water pump, so they're on their own strip) they go to the same ground within a few feet. That doesn't necessarily rule out some sort of odd ground loopy thing, but, as above, I'm not sure how it would explain the sudden onset & ramp-up.


-Bats
(apparently written by elves while I was sleeping & left waiting for me to send. In case of offensive and/or incoherent content, I plead the Roseanne Barr Ambien Defense )
(annnd... apparently the grbl grmlns kept it from being sent when I finally did hit send)

 

Soooo...... Good news and bad news.

The bad news is the magic blue smoke has escaped from the VFD.

The good news is I get to start shopping for a less-noisy replacement.


-Bats
(or maybe that's backwards - I'm pretty sure dealing with shopping is the bad news)

 

Honestly, I'm still a little undecided on that bit. I think I was kinda hoping someone would tell me authoritatively "Oh, those are just noisy VFDs - replace it with a ______ and all your troubles will be over". At which point I could skip the shopping, get a new VFD, and have a spare noisy piece of sitting on the shelf for some less noise-sensitive application (I'd been considering the knee mill, even though the current motor is single-phase).

Instead, I'm left with a week of research & shopping (seeing as my cunning plan to offload all the work & fish a recommendation out of the community appears to be failing), and no spare.

Still, at least it means I'll be able to rule out the VFD as the source of my problems?


-Bats
[insert frustration and sad emoji here. and maybe some emoji showing smoke leaking out of expensive components]

 

Oh, naturally. Listening in on a detuned radio, even the BlackBox makes a racket, to say nothing of the motors. But there's "making noise" and then there's getting this in the console log:

Code:

.000|FS:0,0>
.000|FS:0,0>
.000|FS:0,0>
:1.000,0.000,0.000|FS:0,0>
:1.000,0.000,0.000|FS:0,0|Ov:100,100,100>
.000|FS:0,0>
.000|FS:0,0>
s:1.000,0.000,0.000|FS:0,0>
0.000,0.000|FS:0,0>
000>
000>

(my assumption was that the noise was mangling some of the grbl status lines such that CONTROL's filter didn't recognize & mute them)

Minus the metal box (which was the slightly longer-term plan anyhow), that sounds about like what I did. And, for quite a while, it seemed to work. Then, out of nowhere, I started getting noise spewing all over the GRBL log whenever I... well... did the stuff I prattled on about for far too long in the first post.

Unfortunately that's a 220V, which is going to raise the price substantially (especially without space in the breaker box - otherwise I'd probably have tried running a line already) and it looks like their 110V options are a choice between the same Huanyang model that just died - which I'm a little leery of, although at $130 it's beyond cheap enough - or the even cheaper (and worse-regarded) no-brand lookalike knockoff. [edit: never mind that last bit - mixed up my search windows]

Of course, just replacing it with another cheap import has an undeniable appeal - or just an identical HY unit, which would make replacement a simple drop-in job - and did I mention "cheap"? But I'm worried I'd just be setting myself up for a repeat of the problem - whether immediately, or three months after the warranty expires.


-Bats
(I used to like shopping! What happened to me?)

 

Unfortunately ferrites won't help a VFD turn on once you let all the smoke out...

I'd considered picking up (or scrounging from the junk pile) a handful of ferrites while it was still alive, but my bigger concern was why it'd suddenly started making so much noise in the first place. and while it wouldn't hurt to have a handful of ferrites kicking around when I set up the new one, I feel like picking out a replacement should be an opportunity to get something that works as well as mine did a week ago, rather than as poorly as it worked two days ago - in which case they look more like a fashion accessory than a requirement.


-Bats
(that's it! I'm ditching all this CNC nonsense and going into business making ferrite jewellery! I'll be rich! RICH!)
(spellcheck thinks I'm going to make ferret jewellery. realistically, that would probably be a better seller)
(although I'm not sure how to get a ferret to hold still long enough for ear piercing)

 

One other question, Rob (since you've got the details fresh in your mind & probably have a better grasp of the tech than I do in the first place):

What happens if you try to use an underpowered VFD (say, a 1.1kW rather than a 1.5kW) on a spindle? Do you just get slower acceleration/braking & less overall power out of it, or does the higher draw end up cooking the unit?

I ask because I don't see many 110V/1.5kW drives in the less-Chinese brands - they tend to top out at 1.5HP/1.1kW. I don't know whether this is because the Chinese 1.5kW units aren't really 1.5kW & just have ridiculously inflated ratings (like air compressors), or if it's because trying to make a 110V/1.5kW drive is just a bad idea & more reputable brands know better.

I'm also seeing enough oddly conflicting numbers on the hardware that I don't know which to believe (or if they're somehow all correct, and I've just forgotten even more than I remembered). If you look at the pics on my other plea for help, the spindle is stamped 8A(first pic), while the VFD lists a 13A output on the nameplate (second pic), but has a Rated Motor Current setting that won't go over 7A (no pic - use your imagination. picture some 7-segment displays glowing a bright red "07").


-Bats
(Err... sorry, the setting wouldn't go over 7A. It ain't goin' anywhere now.)

 

Sounds like something to keep in mind when I'm trying to re-purpose an existing VFD, but not something to plan for when shopping for a new one.

Speaking of numbers, I got the following in a reply from ger21 (who I generally consider to have one of a lot more of a clue about CNC-related topics than I do) over on CNCZone that I found puzzling took me days altogether far too long (and only after the third time reading your post) to understand:

"If you look at the specs of something like a 2.2Kw Hitachi VFD, you'll see that it can draw 25+ amps at 240V. So a 120V 2.2KW VFD would need over 50 amps. That would translate to a 1.5KW 120V VFD needing 35+ amps."

I couldn't figure out how the hell he was getting those numbers when the Hitachi VFDs I looked up were all rated at 10-11A. It wasn't until I was combined that with a re-read of this post and then dug up a Hitachi spec sheet that it hit me that the ratings on all these VFDs referred to nothing but output figures. Somehow I'd been blithely doing the math that, ok, 1.5kW / 120V = 12.5A = Hey, no problem for a 15A circuit!

For anyone else who's as slow as me and eventually stumbles on this post, that's (from the specs for the NES1-022SB 240V 2.2kW 1PH to 3PH drive):

10A (rated output) * 1.732 (convert to 3PH) = 17.32A + 30% (conversion loss and... CYA factor?) = 22.49A (matching the spec sheet's 22.5A rated input current)

And, double(ish) for 120V, that's 45A for 2.2kW, or around 30A for a theoretical 1.5kW unit.

So, maybe not quite the same as ger's numbers (not that I'll quibble over a couple amps), but yeah... the math still adds up to ouch.

The one thing I still don't quite understand is how I was able to run the (potentially) massively amp-sucking VFD along with a "6.5HP" (yeah, whatever) Shopvac, the CNC itself (7-10A PSUs), and a couple room's worth of assorted (mostly LED) lighting, all off of a single 15A circuit, unless I'm underloading everything a whole lot more than I'd ever realized. Time for much more aggressive chipload, apparently. Or I need to find someone who'll sell me a 400W ER16 spindle.

I sorta figured - otherwise I wouldn't have linked it. I'd just act passive-aggressively bitter about the fact that you hadn't replied to it, until you asked me what was wrong and I could throw a tantrum.

I hope they've started adding hazard pay to your overtime by now.

As I tried to explain a little awkwardly in the CNCZone thread, it's not even the "Chinese" part, or even the reputation of importy quality (hell, even Mitutoyo & a number of the other classic metrology brands make some of their tools in China these days). It's more - and I quote (myself) - more to do with the problems that almost inevitably come with buying products from companies based there. Cryptic & poorly translated manuals, battling the language barrier with overseas support (not that domestic companies are always great about this either, mind you), and the dim prospects of ever trying to return a potentially defective product (warrantied or not) to China all add up to a loss of confidence in even the best product before even opening the box. After all (and I suppose I can blame American capitalism for the cynical viewpoint) why should a company worry about quality when there's so much isolating them from their customer?

After a hundred minor frustrations with the HY unit (even leaving aside the end-of-life issues), I'd have a hell of a lot more faith in ordering from a brand with a strong domestic footprint (or from somewhere at least more commercially-legally entangled with the US, like Germany or Japan) - even if it's all still Chinese-made.

Of course, the fact that the math says a real 1.5kW VFD for real-world 110V circuits is apparently impossible (or at least highly improbable) suggests I probably won't be finding a matching drive from many honest suppliers.

I'd looked at a few of their big black Not-So-Micro drives that seemed close, but always turned out to be 110V in, 240V out.

Oh, I'm sure most VFDs would've let the smoke out. As we learned from Ghostbusters, crossing the streams is Bad. The problem (that somehow, in two different threads, no one seems to have noticed or speculated on) was the suddenly-escalating noise that led to the increasingly desperate troubleshooting that eventually caused the de-smokifying.

I don't think it has any Super Top Sekrit settings, unless they have to be unlocked with the Konami Kode - all the params seem to be either documented or locked. I suspect it has more to do with the math up above. If it's anything like the Hitachi - with a 2.25x draw based on rating - then a 7A rating would mean a 15.75A draw. Allowing for looser safety margins and a little more tweaking by the marketing department, that could come out around 13A.

I haven't gotten popped it open yet, since it's still wired up (I was waiting until I could document the wiring, and then just didn't get around to extricating & disemboweling it), but it seemed like an awful lot of smoke for a MOSFET. I guess we'll see.


-Bats
(oh, so that's what happened.... apparently I blew the smokistor!)

 

Ferrites are definitely on the shopping list (I'm also shopping for a metal enclosure for the VFD itself this time around), since it looks like I'll probably be getting another of the same drive after all (*sigh*). What still bothers me, though, is that ferrites might have helped with the baseline noise, but the ramp-up of the problem suggested something was getting worse, and getting worse fast. I don't know if putting a bandaid on it would've fixed things, or just delayed the (visible) onset of symptoms a bit longer - potentially until the underlying problem could do some real damage to the hardware.

You know, like, without waiting for me to damage it.

Funny story, that... Sadly enough, the power strips were the rewiring *wince*

My "shop" is approximately half of the furnace room - although, like a machine shop Genghis Khan, I'm slowly working on annexing the other half - and I got sick of living with one and a half outlets (the "half" being a 30ft extension cord from the next room that feeds the PC - to keep it independent of tool-induced blackouts) and always having to find an extension cord to plug anything in. Eventually, as my path of conquest advanced, I gained a second outlet (this one a prosperous 20A), but it was distant, difficult to govern, and promptly overwhelmed by a ravenous air compressor.

So, in an attempt to make life more comfortable for my loyal subje ... for my cruelly oppressed vassals... for my tools, without exposing my weakness to neighboring warlor housemates by attempting to retain the services of a foreign electrician (who might bankrupt the empir me, because there's no room left in the breaker box for the new circuit or two I so desperately need), I went out and grabbed a half dozen 4ft power strips that were on sale.

Long story and overextended metaphor short, one strip goes to the outlet, the other strips go into that one, and the tools/lights/stuff-that-needs-power-not-including-conquering-warlords/etc go into the strips (which run along the supports of the forehead-cracking storage loft that runs over the homelan the CNC machine & the original core of the work area. It doesn't give me any more power, but it does mean I (almost) always have an outlet in reach of where I need to work.

I've thought about trying to do a "proper" job of it, but haven't figured out a great alternative - the overhead loft that holds most of the strips is a sort of rickety plywood-and-2x4 construction (not my work), and never feels quite stable or permanent enough to run conduit to from the wall. Not to mention trying to star-wire some 50-60 15A outlets would take an ungodly amount of wire, and daisy chaining them basically just leaves me with a sturdier-looking version of what I have now. Of course, I always kind of liked the idea of a room with the walls/ceiling completely covered in routed wiring (maybe some sort of holdover from watching Serial Experiments Lain)... but unfortunately that much wiring turns out to be really expensive.

Funny, my psychiatrist told me the same thing. Right before telling me she couldn't deal with working there and was dumping me off on someone else.

Her replacement doesn't seem to have much of a sense of humor.

It's probably deliberate.

Yeah, yeah, everyone enjoys the first one... but somehow by the fifth one their patience starts wearing a little thin.


-Bats
(Just ask yourself "What would Genghis Bats do?" Maybe I could roll up with my army, sack the laundry room, and ride off with their 220 line...)

 

If you think that's good, I could make you fall in love with your CNC rig all over again by showing you the one I started out with.


-Bats
(the loft was even lower then, too)

 

Seriously... I had no idea until (quite literally) I was halfway through that reply.

On a long enough timeline, everything's temporary.

And I suppose the bat flu is making things more temporary than ever for a lot of people.

Troll? I bet you were the one trying to sell his little sister and the stolen Pakistani nukes, aren't you?!

Oh... right... fishing trolling.

(not to be confused with phishing trolls)

(or bridge trolls - even ones that do eat fish)

I'm pretty sure I've seen a fair number of reputable brands offering 1.1kW 120V unit. I have no idea whether they're intended for use on a scrawny little 15A circuit, but - since I wasn't drawing enough to blow a fuse before - I can't imagine a better drive with a lower rating sucking more power.

Considering they sell "3.5HP" routers designed to run on 15A circuits? Yeah. Granted, I've always also wondered about the difference between the "1500W" on my little 3PH spindle vs a pulley-driven spindle with a big honkin' capacitor-start 1500W motor on it.

Granted, I might well be able to get more power out of the 3/4HP motor mounted on my mill than a 2HP spindle on something as light as my CNC rig.

Err... if I had my mill running, that is.

Maybe I can finally get that wrapped up while I wait for a VFD to ship.

Oh, trust me, I've been all sorts of tempted... I was under the impression there was all sorts of codey/insurancey stuff about dryers needing dedicated circuits or something.

Also, I'm not sure I could close the door between rooms anymore if I ran a fat enough extension cord for a 40-50ft run of 30A+ 240V.

Granted, I'm purely guessing on both of those - I assume you've got a far better picture of the realities on this.

It definitely didn't seem normal... but at the same time it seemed to be the usual way of doing things when it came to 110V Chinese spindles.

Of course, it seems like there are only two or three companies (or labels, at least) who make most all of them in these size ranges, which would let them pretty much define their own "normal".

The 110->110V VFDs don't seem very well supported either. Huanyang lists five models at the front of their manual, but everything else in it acts as if everything's either 220V or 380V (which also results in a number of inaccurate, or impossible to set default settings being listed).

A guy over on CNCZone said that was possible with one of these WEG micro drives - apparently it's a 120->230V drive that could optionally be set to 120V output.

Unfortunately it looks like it's yet another 1.5HP/1.1kW drive, which gets into that whole "underpowered drive" question. The guy who recommended it also said there was absolutely nothing wrong with using a small VFD with a larger spindle. Of course, he was promptly jumped on for it - just not before getting my hopes up.

See, that's exactly what's been bothering me.

Well, that and the number of people who completely ignore the bizarre nonsensicality and keep telling me all my problems would go away if I just shielded my VFD cable.

Either that, or the problem wasn't entirely the VFD's fault. By the end I was looking into environmental & AC line noise sources like the nearby furnace (timing didn't correlate) or the geothermal compressor (still happened after I yanked the cutout), or the solar inverters (feed straight to the utility), or my compressor (wasn't even turned on).

I also wondered if it wasn't the VFD at all (or at least the VFD wasn't getting any worse) and if there could be an issue with the BlackBox somehow becoming more sensitive/less tolerant to interference... but I couldn't figure out any way to isolate the BlackBox from itself to test the theory blind- ing-guess.

No idea - I have a very vague grasp of ATX PSU circuit layout, and absolutely no clue about VFD design. I'll see if I can crack it open & check what the damage looks like later tonight, but here are a few glamourshots I took when I voided the warranty last year:



(just think... if I hadn't cracked it open, then right now I could be furious about the fact that it died two months after the warranty ran out)

Definitely not very helpful... And yet it's looking more and more like I'll not only be buying a new Huanyang 120-120V 1.5kW VFD, but, ironically enough, probably buying a brand new spindle along with it.

Not because there's anything wrong with my current spindle (or at least the windings test out as they should). And not because I really have any immediate use for a new stupidly oversized 80mm spindle with a dinky little ER11 collet either. But it seems like my options for a spindleless VFD involve either waiting a month or two for shipping (thank you bat flu and slow Ebay seller communication), or paying very nearly the same price as the package deal.


-Bats
(by the time I get this machine working reliably I'll have everything I need to build a second machine... except the space)

 

Exactly - so I have to assume the 1.1kW drives were intended for 20A circuits. Or were by companies who only wanted to be a little less misleading than the Chinese.

Of course, this is all in theory. And yet I've been running a theoretical 30A VFD along with a 6HP* vacuum and an assortment of other crap off that same 15A circuit.
* Not real HP.

Also, before I completely call the Chinese manufacturers liars (or, well, continue my reckless spree of implying the Chinese manufacturers are liars), it occurred to me there's one other possibility (tipped off by mactec's rather indignant statement that "only the Chinese can call it 110V"). China doesn't appear to use 110V - at least according to Wikipedia they're a sensible 220V country - but Taiwan does (although maybe I'm not supposed to make that distinction). What I haven't been able to find out (my google-fu ain't that strong (and actually it's duck-fu)) is what the average breakers size looks like there. Maybe their code is written by madmen who see no problem with piping 30A+ through NEMA 5-15 outlets.

Yikes! No - Especially with the "Accept 1/4" Shank Router Bits ONLY!" warning - and I don't think I'd trust it for even that, as it looks like the lone reviewer found out the hard way. I also like how they picture it with a big ol' lock miter bit that probably shouldn't be on a 1/4" shank in the first place - those things are scary even without an extension (especially if you're dumb enough to take one up to ~25k rpm in a cheap little Harbor Freight trim router).

Yeah... that surface grinder... I wonder if I could load it single-handed into the back of a Scion tC. And maybe get them to toss in a VFD as a bonus. Or maybe two VFDs - I didn't notice whether it was 1PH or 3PH.... *sigh*

Of course, if I'm going to that much trouble, I imagine I I could rig up a dremel toolpost grinder on the lathe and make one...

...which would probably still have all the dangerous flaws of the original, plus whatever new disasters were added by my amateurish attempts.

Yeah, I don't really think I'd want to try using a 3/4" end mill on steel in the (suddenly dainty-looking) little 80mm spindle.

Yeah, that thing.

Unfortunately I remembered that the reason it's still on hold is because I was planning to CNC a wood or MDF form for epoxifying the gap between the grease zerk and the spindle liner (since slopping it up with epoxy & then sliding the liner back in just makes a mess).

Although I suppose I could probably make it just as accurately on the South Bend (my talents on the wood lathe don't include that sort of precision - or straight lines in general), if I don't mind cleaning sawdust out of the gears.

Ahhh, right. You're down where they've got Florida Man working as the code inspector, aren't you?

Or when you're trying to convince the guy who owns the house that, really, it's not a terrible idea...

How about the cord itself (as I seem to remember you working in the electrical section)? What would you use for a (say) 40-50ft run of 30A 220V?

Hey, at least they're nicer than Practical Machinist

Bending the machine and knotting together a couple circuits to drive it.

As for parameters, at least on the Huanyang there are definitely high/low voltage & rpm params, as well as rated motor voltage & current, and a wide variety of acceleration & braking options... but I'm not sure the Rated Motor Current setting will do quite what we need:

The manual describes Rated Motor Current as "It is set according to the rated value of the nameplate of the motor. This parameter can be used to restrict the output current of the inverter to prevent over-current and protect the motor. If the current of the motor has exceeded this value the inverter of AC motor will turn to self-protection"

...which sounds more like a failsafe cutout than a current limiter.

I might try to toss them an email, describe the situation, and see what they have to say for future reference, but at this point I'm inclined to just cut short the waffling, order another damned HY, and be done with it. Try to enclose & shield it better this time around, hope for the best, and if that doesn't silence the broadcasts then I'll go shopping for something else & shove the new one on the shelf for some future application (I may not have room for a surface grinder, but I'm sure I can find storage space to cram another VFD... probably)

Well, there's "a little more susceptible" and there's "a whole bat-damned -of-a-lot more sensitive" - I was thinking more of the former. Spending a couple days waving around a de-tuned radio made it pretty clear that the VFD was, by a large margin, the noisiest thing in the environment, with the steppers coming in second, followed by the BlackBox itself (although it was a little hard to find a clear division between the two). Barring the possibility of noise in the lines (not to be confused with line noise, which makes me want to find something to call with a 2400 baud modem), nothing else came close (although the furnace kicking on kicked out so much audible noise it was a little hard to listen for the electrical sort). The BB obviously isn't worried about its own noise, and it's playing a fair part in creating the noise in the steppers, so I don't know if that's going to be a major contributor, but with the VFD sitting there screeching away louder than anything, it might (which is, granted, a mighty large "might") not take a huge margin for the BlackBox to start getting annoyed by it - whereas it'd have to start actively listening to pick up the mains in the ceiling.

(interesting enough, for all of mactec's probably-correct harping about how I grounded it wrong, the spindle and its power cable were dead silent - not a trace of noise, even with the antenna run parallel next to the wires. the 120V line into the VFD definitely couldn't say the same)

There's also one other faintly possible (if probably unlikely) element in the BB scenario that might (even mightier might here) be worth thinking about - I'll send you a PM about that one later.

That's possible too, as I was running the vac along with the spindle when I started seeing the problem (although I unplugged it shortly after, in the interest of troubleshooting - as soon as it became apparently that the stops weren't a fluke or me leaning on the keyboard).

Well, I suppose running a whole mess of new isolated circuits would be one way to justify putting in a whole new breaker box to make room for a spare 220V line...

More realistically, I was adding one of these EMI filters to the shopping list, at mactec's suggestion. I'm still undecided on the direct-wiring-with-cable-glands approach to the spindle, though - both because it makes things harder to move or swap out, and because the spindle & cable were so quiet - but they certainly sound like a good idea at the VFD end.

...of course, that means a metal enclosure for the VFD that they can be screwed into, which is something I haven't picked out yet. I've seen mid-tower PC cases used before, but I'm hoping for something on the smaller side - maybe something more like this. Of course, I know the Leeson I used on my South Bend called for a bunch of support hardware, like a cutout & a pair of surprisingly expensive fuses, so I haven't even looked into what else I might need to fit in that box yet.

Well, I'm sure mactec would be offended by my saying so, but it is a Chinese design.

Based on what I could tell from aliexpress prices, it looks like almost half the spindle's cost is the shipping... if the same's true for the VFD, maybe they can get massive savings by mailing the two together.

It looks like I may not be going for another combo after all, though (and I'll admit, I was kinda looking forward to the prospect of having a spare spindle lying around), having found a couple more economical options and that one "bundle" wasn't offering quite what they made it look like. As it turns out, if I was willing to wait until early May for delivery (and I'm not sure if I am), I could get the VFD alone for $111. Or I can blow $170 for one that'll arrive around the end of next week. Or I could throw a fairly irresponsible $235 at a package deal that would get here a day later. +$63 for the extra spindle sounds like a bargain, based on the quicker one... but +$124 doesn't sound like such a deal if I compare it to the one on the slow boat.

Those are all identical Huanyang units, too - although I could get one of the knockoffs for as little as $80. There's also the $135 option of having one sent direct from Huanyang themselves, but with a June 17th delivery date? Err... Yeah. No. To both.


-purp
(hrm... cost vs delivery delays? that sounds almost like... paying for express shipping!)

 

Nope. Although I haven't smelled any smoke from the walls yet. And it's not completely dead - I did manage to blow it out of the blue a couple weeks back.

Heh. Surfacing mills are the main reason I almost made the mistake of getting a 2.2kW, just for the ER20. But an edge finder? Sure - if it's actually straight & concentric.

Maybe something like the little 400W units - the long 65 mm ones would need a pretty massive mount.

What, melt them all down in a furnace & pour in the molten metal to fill the gap? Probably, but it's a little more hardcore than I was really planning on. And that's a very specific list of scrap metal.

Oh, that part's easy - the spindle mounts quite nicely (and even rotates). The problem is that the grease zerk - which is my only way to lubricate the top bearing w/o tearing the whole thing apart - has gaps between the spindle casting and the (decidedly aftermarket) inner sleeve that the (also aftermarket, as I eventually determined) spindle runs in. Pump it full of grease, and it all flows out the back & into the ram casting before it's pressurized enough to flow into the bearing.

So the plan was to fill the gap between the two near the zerk with epoxy paste. The spindle liner is a tight fit and it shoves the epoxy out of position when I slide it in, so the plan was to make a cylindrical form of the same radius out of [MDF, wood, belly button lint], cut it in half (for easy insertion/removal), wax it (to keep the epoxy from adhering) and use it instead.

It's always the simple things...

There's a fair bit of that around here, too. Both sides used to end up buying each other rounds at the local dive (RIP)... although they'd both throw a tantrum if anyone breathed a hint of an idea that might not be up to [a drunken interpretation of] code.

Probably the same, maybe half to two thirds the distance, depending whether it follows the walls or goes overhead.

Sounds reasonable, but that's Some hefty wire to run under the door. Probably have to do some thing a little semi-permanent and go through the wall.

And yet the manual is "fine".

I'll have to see what I can find out. I was actually considering a 2.2kw VFD on sale from the same vendor I got my first one at, but I think. I've found better (ie. cheaper, faster) options since.

The only thing that always keeps me from trying all sorts of crazy
( not to be confused with bat crazy) grindy stuff on the South Bend is the fear of getting abrasive dust in everything.

Maybe I should try some live tool playing on the wood lathe, where it doesn't matter (although then I'd be stuck building my own slide, or going freehand.

Based on the way it used to bluescreen a nearby PC on an independent circuit, you're preaching to the perv converted.

for the limits, though, have you tried powered switches w/ onboard filtration like the fancier Openbuilds ones? Or is the NO thing too much of a turnoff?
(I threw some on mine during the recent rewiring, but haven't really been running enough to form any solid opinions - and the troubles I had w/my old switches weren't noise related)

You turn it back on w/o noticing the cowl of the plug had gotten cross-threaded & was on just a little crooked.

Yeah.

Ow. Yeah. I don't think I want to go there if I can help it.

It does make me wish (yet again. and again, even) for a sheet brake, though.

Barring that, I guess it's probably some thing like McMaster... although at $41 for an awfully cramped 8x8x8" or $61 for a 12×12×12, it's not exactly cheap (and digikey is even more), so I can definitely see the appeal of the hammer.

That $235 includes an ER11 spindle.

For the VFD alone, it looks like I can get it for $132 by the 14th [edit: or $172 by the end of the week]. The quick turnaround on the WEG is awful tempting, but right now I have no assurance it'll work, plus having to deal with figuring out the setup & interface from scratch (think I can talk Peter into adding RS-485 support to CONTROL? ).



The Huanyang may not be great (time, may not be good), but it's a drop-in replacement - and has a lot of support in the hobby CNC community, which is helpful.

That's the next stop today.

Bought anything there lately? Prime ain't so prime, what with the ravages of bat flu and a sudden nation of shut-ins. Shipping's still free, but it's coming from China, not the Bezopolis warehouse.

Unless I hear something exciting back about the WEG, I'm probably going to order another HY tonight and spend the week or two's wait actually working on something, rather than spend another week or two on research to find something quickier, which I'll then spend another week getting running. While my time may not be much money, it's gotta be worth something.


- Bats
(but not much, apparently, if I can afford to write these posts)

 

Ok, now that's certainly a big selling point.

Of course, if I let the CNC have 220V, then all the other tooIs are going to want it too.

It would certainly offer more flexibility in tooling. Do you think it could be done as a drop-in for a QCTP (one of the reasons I always thought about doing it with a flex-shaft pendant motor), or are you thinking of one that replaces the tool post completely?

Sounds tricky. How do I still end up with a flush fit between the epoxy fill & the liner?
Well, without some fancy grinder, at least.

Now there's an idea... Could just plug the hole in the liner & re drill the Zerk hole through the epoxy. The grease flows through the liner - no need for any dremel work. The pumping would be easy, too - I picked up a 10 gal Binks pot (much bigger than I wanted, but much cheaper than anything smaller) to use for casting resin (which, between car & CNC headaches, I still haven't touched it), and the guy made me take an equally- too-big pneumatic epoxy gun & a few big boxes of mixing tips. I've had no idea what to do with the damned thing.

Of course, I don't have any cartridges to put in it, and that seems like an awful lot of goop, but it might make everything easier.

Assuming, of course, that I get everything right on the first try. Otherwise it'll make everything a whole lot messier.

I'll probably leave it as plan B, just because of .the fact that it uses an untested tool & has the potential for a massive mess if it goes wrong.

I wouIdn't think that a cylinder of x.xxx +/- .00x diameter should be an especially taxing project for a lathe. That's kinda its thing. Then slice it in half on the bandsaw, finish, and wax it.
Not so sure about turning MDF, though. Might have to use some cheap pine or poplar. Seems like the worst part should be cleaning up the sawdust from all the lathe's oily crap.

Or am I missing something?

That's something I should really look into. It would also save me from digging small parts out of the oily chip tray.

Some sort of rail along the back, like a taper attachment uses? Just doesn't help if you've got a face shield riding on the back of the apron like mine.

For stress relieving? I'm guessing I could probably get by with the old vacuum furnace... well, as long as the part's not more than a few inches in any dimension.

Looks like they've got LEDs, but I don't see much filtration. Unless that's what that little SMD cap is for?

Well, harder to faiI Closed then fail open, right? Or is that the other way around?

I alway worry about repeatability on those little switches, though. The fancy Openbuilds ones I've got on now are miles better than what I had been using, but I still keep toying with the idea of trying something optical.

What, no English wheel?

Is this one of those zen things, like "if a CNC machine crashes and there's no one around to hear, does the operator still swear?"?

I figure if it eats storage space, it probably exists.

Also, I figured if it made sense in the price anyhow (it seemed to before, it doesn't now), I could set it aside for some future bad idea. I've got a couple big (NEMA32?) ballscrew actuators & a pile of larger extrusion that's just waiting to be turned in to some disastrous timesink.

My thoughts exactly. Although I think I can be patient enough to wait until the14th for a $135ish offer. If the guy would just get back to me on shipping.

Yikes - I hadn't heard that yet.
So then UPS & Fed Ex get to raise their rates (more), and everything else moves to Amazon Logistics? Or are they talking about pulling the plug on all (consumer) shipping?
(or just the one they've been trying to de-fund or privatize for years anyhow?)

One way or another, sounds like I've got a lot of shopping to do. The local selection ain't what you'd call great. Or good. Or adequate. Pretty much all my hobbies are dependant on mail order. Although there is one local metal supplier I was pointed at by the guy running that blacksmithing class that I really need to check out. If it's actually still open.


-Bats
(Have I mentioned that we're ing ed we are yet? Because we're ing ed)

 

Holy bats, Batsman! Someone actually read to the end of this thing?? Quick, Rob! Hide the drugs and hookers!

*glances around*

err... Hookahs! Hide the drugs and hookahs!

err... Drunken hookahs. They keep stumbling in from the street!

That's ok, neither am I.

Quick! Stamp it out before it breeds! Do you have any idea how expensive those things are?

That's... oddly brilliant. And I'm suddenly in love with the idea of a mailbox-shaped VFD, with a display in the door and a little red flag that goes up or down with an RC servo when the spindle's running.

It looks like HD's got extra-large ones for about $27 that, at 11.7 x 15 x 25" would almost be large enough to fit it with the recommended surrounding airspace. Off the top of my head, I only see two possible issues.

First, I don't know how tightly they seal. That may or may not be a real problem, seeing as, inversely, I also don't know how long the offending wavelengths are.

Second, is that - despite what you might assume from the dainty sounding "extra-large" moniker - those things are big. And, while that's no doubt an advantage when receiving extra large parcels of top quality Columbian blow, my shop (ok, fine, my "shop") is... that other thing. The "not-big" thing. The height's not very important & the width probably isn't negotiable - I'd likely need the same with any enclosure - but with that 24" depth it's got a huge (and mostly empty) footprint.

Hey, Rob, wanna weld the door hinge (anybody need a rhyme for "orange"?) back on the surviving rear third of a hacked up mailbox for me?

In all seriousness, I absolutely love the idea - I was actually just in the other room trying to see if I had somewhere to mount it on a 4x4 post to get the whole look down. I suspect I'm going to end up going the standard IEEE (inordinately expensive electrical enclosure) path, though, just in the interest of keeping an already annoyingly complicate project simple(r). Which is a shame, because if it'd come before I'd gotten quite so frustrated with the process, it'd be an enormously entertaining mod to work on.


-Bats
(and now I can't stop thinking about how maybe it could also become extra storage space, if I stick the panel on the door & shove the rest of the unit to the back...)