Back
[02:16:38] <Deejay> moin
[08:20:19] <Demure_> Hello all! I'm in way over my head I'm afraid and could use a chat to understand some things better... I was setting up my lathe with steppers and a Mesanet board, Leadshine drivers et all.. I still needed the power supply, but then I ran into someone selling a whole bunch of mechanics on a flea market, including a whole set of ballscrews (Won't be using) but more interestingly a whole set...
[08:20:20] <Demure_> ...of Faldic servo amplifiers and servos that look near new.
[08:21:18] <Demure_> Now since I already have a mesa card for step / dir, I was wondering if there's a way to use that as the input for these servo amplifiers? The language is a little bit beyond what I know. Spec sheets says it takes a pulse input string, which is configurable in a few different ways
[08:21:55] <Demure_> 0: Command pulse/command sign,
[08:21:57] <Demure_> 1: Forward/reverse rotation pulse,
[08:21:58] <Demure_> 2: Two signals with 90-degree phase
[08:22:00] <Demure_> difference
[08:22:59] <gregcnc> 0 should be step direction, do you know what the resulting step increment would be?
[08:23:04] <Demure_> I have the 7i76 card from Mesa, would there be any way of controlling these? If not, what would I need, if yes, what else do I need to control them or is it that simple?
[08:23:22] <Demure_> Step increment as in the amount of steps it takes?
[08:23:26] <Demure_> It has a 16 bit encoder, so quite a few
[08:23:53] <gregcnc> with 16 bit encoder I'd spend the money and get the right mesa card
[08:24:08] <Demure_> What would the difference be?
[08:24:14] <gregcnc> analog output
[08:24:24] <Demure_> The encoders go straight into the amplifiers, so in theory the Mesa card doesn't have to deal with those
[08:24:27] <Demure_> Ah, I see.
[08:24:37] <gregcnc> oh
[08:24:57] <gregcnc> what model amp?
[08:25:14] <Demure_> Faldic RYB 201S3 VBC
[08:25:56] <archivist> you send the encoder to the mesa card as well then you get the servos in the loop
[08:27:18] <Demure_> Wouldn't it be better to let that up to the amplifier?
[08:27:35] <gregcnc> no the system is open loop in that case
[08:27:58] <Demure_> And the amplifier would still work just fine if I wouldn't plug the encoder into it?
[08:28:19] <archivist> plug into both
[08:28:21] <gregcnc> I was going to do the same thing, with 7i76, but got a 7177 instead.
[08:28:34] <enleth> we need a more fine-grained definition of open loop/closed loop
[08:28:38] <gregcnc> damn manuals need login
[08:28:42] <Demure_> I have one here
[08:28:47] <Demure_> I can send it to you if that helps
[08:28:58] <gregcnc> i found one
[08:29:46] <gregcnc> encoders are serial
[08:30:00] <Demure_> You'll have to explain that one to me
[08:30:12] <Demure_> Does that mean it doesn't detect direction?
[08:31:11] <Demure_> The servos are GYS 201 DC1 - CA
[08:33:03] <archivist> the signal from an encoder can take many forms, quadrature, serial, absolute, differential and a mix of those
[08:33:06] <Demure_> And ultimately I'd like to know if this is worth going for over the leadshine AM822 with some 2A steppers, since I have both systems here now (First conversion).
[08:33:28] <gregcnc> still downloading, 7i77 is for quadrature encoders, but I think there are cards for conversion of encoder signal
[08:33:31] <archivist> steppers are simple but less good
[08:33:39] <Demure_> I'm only familiar with quadrature
[08:33:50] <gregcnc> what machine?
[08:33:57] <Demure_> An Emco 5 CNC
[08:34:16] <gregcnc> for that i may not bother
[08:34:29] <gregcnc> the original steppers were 72 steps?
[08:34:30] <Demure_> I see
[08:34:34] <Demure_> Yup
[08:34:48] <Demure_> But I have 400 step oriental motor steppers here atm
[08:35:01] <gregcnc> right so that's a big improvement already
[08:35:10] <Demure_> I've never used the original, however
[08:35:24] <Demure_> It came without the motors and control system when I bought it, simply the mechanics and spindle motor
[08:35:31] <Demure_> So I've ordered a KBIC controller for the latter
[08:36:00] <Demure_> So the only thing left if I went stepper is a +-80v power supply, which is harder to find than it seems
[08:36:14] <Demure_> Are the Chinese eBay ones really acceptable or too dodgy?
[08:36:34] <Demure_> (Sorry for all the questions; I try to solve 'em all myself but at some point I just get stuck. Still a lot to learn.)
[08:36:45] <archivist> dont vibrate them
[08:38:08] <archivist> I was testing a stepper loose on the bench and the psu let out the magic smoke :)
[08:38:15] <Demure_> I have the AM882H Leadshine drivers which seem like AM882's but with a heatsink + fan attached
[08:38:21] <Demure_> Ouch
[08:38:26] <Demure_> Because of the vibration?
[08:38:32] <archivist> I think so
[08:38:52] <archivist> I lolled
[08:40:03] <archivist> stepper is the chunky one in the vice
http://www.collection.archivist.info/archive/DJCPD/PD/2013/2013_12_30_JS_hobbing_machine/IMG_1731.JPG
[08:40:52] <archivist> that image after the psu went phut
[08:41:18] <Demure_> So it's not the nice controllable psu that went bust I hope?
[08:41:53] <Demure_> The drives say they can take 24-100v VDC or 18-70 VAC. I'm assuming more = better, right? I'd like to stay a little under the max (+-80vdc) but I'm lost on where to find a decent one, any tips? I need to supply power to two drives with 2A motors.
[08:41:54] <archivist> no just a cheap ebay 60/80 volt supply
[08:42:43] <archivist> a transformer, rectifier and capacitor is often used
[08:43:37] <Demure_> I've heard this before and saw some available from Antek, but I don't quite understand the difference of that and a cheap power supply
[08:43:46] <gregcnc> you only need the voltage if speeds are high enough to make use of it
[08:44:26] <archivist> you dont need stabilised for the motor power supply
[08:44:53] <Demure_> In the Emco 5 they are geared down so I figured the speed is not a bad thing in this case
[08:45:34] <gregcnc> you have to do the math and look at the stepper spec
[08:47:52] <Demure_> http://catalog.orientalmotor.com/item/stepping-motors/pk-series-stepping-motors/pk268-02a?jumpMenu=%23 These are the ones I currently have (Was mistaken, 200 steps, but since they're geared I don't think that should cause an issue, esp. with microstepping et all)
[08:54:45] <Demure_> Is the leadshine a bipolar or a unipolar drive? Or is that simply just a different way of wiring it?
[08:55:13] <archivist> will be a bipolar drive
[08:55:33] <archivist> no one sensible uses unipolar these days
[08:56:45] <Demure_> The motor has 6 leads, so that should also be able to be used as bipolar, correct?
[08:56:56] <Demure_> In which case the max voltage would be 120v, so that should not have to be an issue
[08:57:55] <Demure_> Is my idea of "I just want to be able to run it as fast as possible" a bad thought?
[08:58:07] <Demure_> I mean, I can always run it slower with the same setup, no?
[08:58:35] <archivist> the drive is the voltage limit, any motor voltage quoted is DC and the current is the important figure not the voltage
[08:59:34] <archivist> in actual use, you want the best speed not max
[09:00:07] <archivist> max is only useful for non contact moves between cuts
[09:00:15] <Demure_> That I understand, but with a higher voltage I can always set the drive to make the motors go slower and for the ideal speed, and if I ever need to go that fast, I will be able to, no?
[09:00:28] <Demure_> I'd like to do a lot of large threading so I figured the speed could be useful for that
[09:00:48] <Demure_> I see. The drive takes up to 100v, probably an overclocked version of the AM882 which takes up t0 80v, so I'd like to keep it to that (Not sure I trust the 'souped up' bits.)
[09:01:38] <archivist> I never use max, always allow some headroom
[09:01:58] <Demure_> With voltage, you mean?
[09:02:08] <archivist> any thing
[09:02:20] <Demure_> That's why I figured anywhere 60-80v would be good, since it's rated up to 100.
[09:02:27] <archivist> one needs a safety factor
[09:02:36] <Demure_> True.
[09:03:10] <Demure_> I'm going to have to lay a cable from the kitchen to even have ground... /Shitty appartment
[09:04:01] <archivist> keep lathe in the kitchen
[09:04:17] * archivist admits nothing
[09:04:31] <Demure_> It's been on my mind, but alas we do not have the space for it...
[09:05:38] <Demure_> What would be your preferred option to supply that 60-80v to the drives? I've found things like www.antekinc.com/ps-5n80-500w-80v-power-supply/ and www.ebay.com/itm/NEW-900W-80V-DC-11-25A-Output-Switching-Power-Supply-with-CE-for-CNC-/182097903903 but the differences are a little bit lost on me I'm afraid.
[09:06:34] <Demure_> And am I correct in thinking that while i.e. 11.25a is overkill, it will only draw what is required?
[09:07:02] <JT-Shop> I like Antek power supplies
[09:07:14] <archivist> it will only draw what is needed, some overhead means less stress on the supply
[09:08:35] <Demure_> Ok, good to know
[09:08:45] <Demure_> For the Antek, I would need to add a power supply cable / case possibly a fan?
[09:09:07] <archivist> probably dont need a fan
[09:09:10] <Demure_> I like that they have an option for 80v + 24v so I can run a toolchanger stepper, too.
[09:11:25] <Demure_> Can I wire it up to the same power supply plug that leads to the kbic controller or would it be better to have two plugs for that to not stress the cable too much?
[09:12:55] <Demure_> & since the price is identical for the Antek power supply variations, is there any reason to not go for the 80v + 24 + 5 rather than the 80v + 24 or 80v only model?
[09:23:37] <enleth> I'm looking for a motor power supply right now too, wondering if there's an advantage to using regulated switching supplies over the oldschool transformer/rectifier/capacitor combo
[09:25:47] <pcw_home> switching supplies typically dont deal with back EMF as well
[09:25:48] <Demure_> From what I understand there's actually an advantage to unregulated as it can deal with back flowing energy.
[09:25:59] <Demure_> But I'm still learning so don't take it from me
[09:26:37] <pcw_home> switching supplies will have very small output capacitors (due to the high switching frequency)
[09:27:57] <enleth> What I noticed with the CC drives is that the +/-10V input regulates output voltage relative to the supply voltage - which also means that, if the supply voltage dips when another motor kicks in, so does the motor voltage given the same control input, even if it wasn't at max. But I guess that's of no real concern with a position loop.
[09:28:03] <pcw_home> so any energy dumped back into the supply will raise the voltage much faster than a 50/60 hz supply with its large output capacitors
[09:28:47] <enleth> It just was noticeable when bench testing
[09:29:35] <pcw_home> normally +-10V systems use velocity feedback around a current feedback loop so supply voltage should not change much
[09:30:07] <pcw_home> but if you run the drives in direct (PWM) or voltage mode you will get that effect
[09:31:15] <pcw_home> what I mean is changes in supply voltage are compensated for in the current loop so you should not see much change unless you are running in voltage mode
[09:31:45] <pcw_home> ( voltage mode means 10V is 100% of bus voltage)
[09:33:06] <pcw_home> there are some changes with bus voltage (ideally the current loop gain should change with bus voltage)
[09:37:31] <enleth> if I understand the manual correctly, I'm pretty sure it came configured in voltage mode and I could not be bothered changing it yesterday, just wanted a quick test to see if the drives are operational
[09:39:43] <pcw_home> One disadvantage of voltage mode is that as you discovered the motor drive voltage varies with bus voltage
[09:39:45] <pcw_home> (so you cannot tune FF1 to compensate for BEMF very precisely)
[09:40:42] <enleth> I'll have more time to actually set it up for velocity feedback later this week, going step by step according to the manual
[09:41:39] <pcw_home> voltage mode has some advantages ( a fair amount of inherent high frequency damping due to the low output impedance of the drive)
[09:43:07] <pcw_home> but is harder to tune and typically needs a higher servo thread rate
[09:43:08] <pcw_home> ( since the velocity feedback part must now be handled by LinuxCNC )
[09:43:50] <enleth> makes sense
[09:43:58] <gregcnc> there is exactly one line describing voltage mode in the AMC manuals are we talking the same thing?
[09:44:41] <enleth> gregcnc: CC manuals are a bit more verbose on this topic
[09:44:46] <gregcnc> I was telling someone yesterday to not use voltage and use torque mode instead
[09:46:06] <enleth> apparently switching modes on a CC drive is a little bit more involved as everything in it is analog and you have to set a few trimpots to zero for voltage mode, otherwise it might go bonkers trying to do something inbetween
[09:46:14] <gregcnc> he had no tach though
[09:46:42] <pcw_home> torque mode will be less affected by supply voltage changes but does require a faster servo thread times
[09:46:44] <pcw_home> (but may be you best choice is you have no tach )
[09:46:55] <pcw_home> if you
[09:47:05] <gregcnc> I set my torque mode to 4k servo thread
[09:47:24] <enleth> in my case velocity mode is probably the go-to solution as I have built-in tachs in the servos
[09:47:46] <enleth> and that's how the machine is set up now with the original drive
[09:47:56] <pcw_home> torque mode has the advantage that you can see the actual torque used
[09:48:58] <enleth> what's that good for in practical terms?
[09:49:07] <MacGalempsy> good morning
[09:49:18] <enleth> other than the obvious, like overload sensing
[09:50:00] <pcw_home> Checking how close you are to the drive limits
[09:50:02] <pcw_home> integrating to get the sum of work on the job (for maintenance)
[09:50:26] <pcw_home> limiting torque when homing
[09:51:48] <enleth> the CC drive has a current monitoring output
[09:52:36] <archivist> measuring tool wear
[09:53:20] <enleth> I could hook that up to an analog input on 7i77 and monitor the current there
[09:55:55] <gregcnc> anyone have Soidworks dialog boxes that disappear off screen? what a pain
[09:56:35] <enleth> is there any further advantage to using torque mode when current monitoring under voltage mode is possible?
[09:56:44] <enleth> s/voltage/velocity/
[10:00:09] <gregcnc> seems CNC machines generally use velocity, at least most online discussion indicates
[10:00:36] <enleth> that was my impression as well
[10:01:13] <enleth> I guess torque mode might be useful in huge machines intended for extremely heavy hogging at slow speeds?
[10:02:08] <enleth> but as long as current monitoring is possible with velocity mode, I kind of don't see the point, at least for my mill
[10:03:42] <MacGalempsy> can anyone tell me if the 010v spindle control
http://linuxcnc.org/docs/html/examples/spindle.html code lines can be inserted anywhere in the HAL file?
[10:05:51] <archivist> almost anywhere but... the order of them matters so you need the loadrt before you can refer to its use, then you initialise, then connect
[10:07:44] <archivist> and connection order can matter, so an input uses the current cycle data rather than previous cycle
[10:10:21] <MacGalempsy> right now it the spindle goes on, but only to 100%
[10:10:45] <MacGalempsy> so I hope the 010v example will suffice. the VSD has internal PID control
[10:35:45] <gregcnc> how important is shaft ground when using VFD on a motor?
[10:39:32] <enleth> you mean electrical grounding?
[10:39:56] <archivist> motor outer body must be grounded
[10:41:07] <enleth> in any case, a VFD provides galvanic isolation
[10:44:29] <enleth> TBH I've never considered shaft grounding to be important with 3-phase motors
[10:45:32] <Demure_> What are the risks of running the entire system of drivers, steppers and spindle motor without ground? <:
[10:47:04] <enleth> that depends, but most importantly, if you don't have galvanic isolation for the whole thing, electrocution is among the risks
[10:48:57] <pcw_home> I think VFDs tend to generate more shaft --> frame currents so shaft grounding is used (mainly to protect bearings)
[10:49:25] <pcw_home> (the currents slowly EDM the bearings)
[10:50:15] <gregcnc> right i'm thinking bearings
[10:57:26] <pcw_home> http://est-aegis.com/TechPaper.pdf
[11:33:51] <Loetmichel> PCW_: is that a reasonable problem?
[11:34:07] <Loetmichel> my 3 phase china spindles have isolated shafts
[11:34:22] <Loetmichel> unless seriously high side load
[11:34:36] <Loetmichel> the bearings seem to be well greased ;)
[11:45:24] <archivist> or ceramic balls
[11:45:43] <gregcnc> $6900 for a 9" SB lathe?
http://bn.craigslist.org/tls/5562409563.html
[11:46:52] <jdh> I saw 500 for a used HF 7x
[11:46:53] <archivist> lots of money
[11:47:38] <archivist> reconditioned...threw some paint at it
[11:48:53] <gregcnc> maybe it learned something during it's time at the school
[11:49:20] <pcw_home> Loetmichel: I think most water cooled spindles have electrostatic shielding so its not as much of an issue
[11:49:31] <Erant> I just bought a very nice 618 (with a ton of extras, ER32 spindle with full set of collets and whatnot) for $600. Came with the original manual, receipt and shipping tickets.
[11:49:53] <Erant> Full set of change gears too.
[11:49:57] <pcw_home> you can put a scope on the motor shaft to find out
[11:51:04] <archivist> my southbend came worn out for £120 about 20+ years ago
[11:51:59] <Erant> Actually, I had a question. So, I want to add a variable speed motor to this guy. Everyone seems to be using the treadmill motors, but that's a lot of potential torque to push through this lathe... I'm looking at a 1/3HP or 1/2HP motor now. Am I looking at the right power range for a 618?
[11:52:16] <Erant> It has the original 1/4HP (?) motor right now.
[11:52:29] <Erant> Which runs surprisingly smooth for a 50 year old motor.
[11:54:56] <enleth> pcw_home: do you happen to know how are input stages built in interpolators compatible with current signal encoders? There's an abundance of information online on working with voltage signal encoders, mostly 1Vpp, but very little on the 11uApp interface.
[11:54:57] <archivist> 1/2hp 3phase and vfd
[11:55:54] <enleth> If I had to guess, I'd try a Wheatstone bridge first, scale output to 1Vpp then feed that to a voltage signal interpolator
[11:56:53] <pcw_home> we made one using the IC-HAUS interpolator, I think we have an opamp current to voltage converter with a
[11:56:54] <pcw_home> (shortable by a jumper) series resistor for 1V mode
[11:58:19] <pcw_home> so differential amp in for 1V, differential current in for 11 uA
[12:00:20] <pcw_home> you also need to offset the amp voltage up to typically 1.65V for most 3.3V micros A-D
[12:00:29] <enleth> right, thanks. do you remember if the opamp had to be something special/unusual for this?
[12:00:33] <pcw_home> (0 to 3.3V input typically)
[12:01:21] <pcw_home> No but a couple hundred KHz bandwidth at least
[12:02:41] <witnit> still working on these encoders? pcw_home you have the highest support integrity of any product developer i know
[12:04:05] <enleth> witnit: still? that's just the beginning
[12:04:59] <enleth> at least I know what I'm up against now
[12:06:11] <witnit> well about 12 hours ago I was thinking you were just learning of your encoder type
[12:07:04] <enleth> that's correct, I didn't really care to check their output before
[12:07:07] <Erant> archivist: Any particular reason that's better than a DC motor?
[12:07:45] <enleth> it appears that current signal encoders aren't terribly popular so anyone could have fallen for this
[12:07:50] <Erant> Brushes?
[12:08:23] <archivist> Erant, quieter, more reliable etc
[12:08:53] <witnit> enleth:
http://tinyurl.com/zdasg86 I bought those, 12,500 lpr they work wonderfully
[12:09:35] <witnit> I mean at that cost and that quality and plug it in and go, I couldnt pass them up, ntm they have the little spirally door stoppy flex thing
[12:12:07] <enleth> witnit: well, I guess I could replace them, but OTOH I can take a little setback in retrofitting my mill, it works as is right now with the old control, and I've been looking for an excuse to dust off my EE skills
[12:14:51] <witnit> enleth: what is the travel of each axis?
[12:15:25] <enleth> 450mm, 250mm, 125mm, as far as I remember
[12:18:16] <witnit> you could use linear encoders on the machine and leave the other encoders on the motors, then you could use both controllers
[12:19:19] <witnit> I have no clue if this would even be worth your efforts but I like having ideas
[12:19:43] <maxcnc> hi all from a darken germany
[12:19:54] <maxcnc> thunder growing outside
[12:20:25] <witnit> hello
[12:20:36] <witnit> tis sunny here
[12:21:15] <maxcnc> smile if you are off work
[12:22:29] <witnit> =D
[12:23:38] <enleth> witnit: it would be rather difficult to mount linear encoders on this mill in an elegant and safe way
[12:24:23] <enleth> and I don't really have the funds for this
[12:30:12] <enleth> witnit: for example, I'd have to install the Y encoder inside the knee, with the read head on the Y ballnut bracket. this requires a specific length of the scale that can fit inside the ballscrew compartment but still be long enough for the whole axis travel
[12:31:16] <enleth> getting a glass scale cut to length is not very high on my expenses priority list
[12:33:21] <enleth> same thing with the quill, there's only so much vertical room in the head casting
[12:35:29] <enleth> I could do that for fun sometime in the future, when the mill earns back some of the money I spend on it
[12:48:24] <witnit> I thought about buying three 6" linear gage probe type encoders with magnet bases, so I could just slap them on a x/y/z and hit play :)
[12:54:43] <zeeshan> darn it my gamble didnt work
[12:54:49] <zeeshan> found an almost identical vise to mine
[12:54:52] <zeeshan> it sits 66 thou higher :p
[12:55:30] <pcw_home> mill it down
[12:55:51] <enleth> witnit: I wouldn't particularly trust that to position a 1.5ton mill with over 1kW servos
[12:55:58] <zeeshan> pcw_home: thats the plan :D
[13:11:19] <witnit> jeesh enleth no faith, no faith at all :P
[13:12:11] <witnit> if a fly lands on the probe and it doesnt go retract back, fly buzzes off, KABLAM
[13:16:56] <maxcnc> im off GN8
[13:57:54] <witnit> pcw_home: I obviously need to do some reading but here is the idea, press a button as input via 7i37 > activate bash file, do you know an obvious way to accomplish this or know where I can find documentation for such a feat? So far im looking at activating an M command with button, but that doesnt seem like the most direct route
[14:02:01] <witnit> anyone who knows this answer would be great. I feel like im looking at it from the wrong perspective but essentially I want a one touch, automatic backup and one touch update and reload ngc from a remote location, ip, usb, or something
[14:11:51] <djdelorie> witnit: isn't there an app for that? ;-)
[14:13:47] <enleth> pcw_home: thanks again for all the info. If I understand the problem correctly, just about any standard transimpedance amplifier circuit straight from an opamp application note will do. At 3100rpm max and 125 pulses/rev, I'd be getting 38750Hz at most. With many simple application note circuits reaching 2-4MHz at higher gains than I need here, this seems to be easier than I anticipated.
[14:17:43] <PCW_> Yeah its easy
[14:17:59] <PCW_> you do need a fast A-D
[14:18:51] <PCW_> the IC Haus one are like 50 MSPS but 2-5 should do
[14:20:38] <enleth> I might try to play with some DSP-enabled uCs, quite a bit of those were released in the past several years
[14:21:10] <PCW_> (when we considered doing one we were going to just do a big table lookup for the interpolation)
[14:21:12] <PCW_> 8 bits of sine and 8 bits of cosine concatentated to a pointer and 64k lookup table
[14:23:23] <enleth> was it for sine encoder support in the I/O cards? too bad it got scrapped
[14:23:50] <PCW_> yeah not really enough demand
[14:24:07] <PCW_> I woud use a STM32 now
[14:27:15] <witnit> its getting closer
http://pasteboard.co/11liCGBg.jpg
[14:28:07] <_methods> nice
[14:28:53] <witnit> I decided to put the controls right into the sliding door, I will get a new clear cover for it once im done tossing it around
[14:31:47] <witnit> The silly thing is, I already have a job for the lathe and am rushing the build so I can get the order started, its just 6 pieces but its the only machine I have with a bore big enough to handle the bar, basicly zap in the XZ and manulally actuate all the relays
[14:59:00] <alex4nder> hmm
[14:59:09] <alex4nder> anyone ever machined horn blanks?
[14:59:34] <alex4nder> like, from a buffalo
[15:19:47] <CaptHindsight> http://cdn.shopify.com/s/files/1/0190/2338/products/BuffaloHorn2_large.jpg?v=1439533162 like these?
[15:36:40] <XXCoder> alex4nder: nope, but likely you will need to probe heavily in order to contour if engraving
[15:44:25] <XXCoder> alex4nder: from material properies that I do know, you probably can treat it like wood, only bit more brittle
[15:49:20] <alex4nder> yah
[15:49:34] <alex4nder> as far as I understand it, it's basically hair,.. so it seems like you want to keep temps low
[15:49:51] <XXCoder> you probably dont want to burn it and smell it
[15:50:08] <XXCoder> got scrap horns to test on?
[15:53:02] <CaptHindsight> proteins like fingernails
[15:53:12] <CaptHindsight> organic polymers
[15:53:21] <XXCoder> Ketrin or something
[15:53:23] <XXCoder> cant spell it
[15:53:37] <CaptHindsight> with mineral deposits
[16:06:37] <alex4nder> XXCoder: I need to order some up
[16:06:52] <XXCoder> cool. whats tour plan anyway
[16:07:02] <alex4nder> make some eyeglass frames
[16:11:10] <gregcnc> google says that's actually a thing. glasses made from horn
[16:11:22] <XXCoder> yeah
[16:11:47] <XXCoder> just be sur its not horn from endangered animal
[16:15:01] <SpeedEvil> XXCoder: Not costing $100K/horn is probably a clue.
[16:15:02] <SpeedEvil> :/
[16:15:22] <SpeedEvil> (though yes, there are probably less endangered animals)
[16:15:22] <XXCoder> maybe. heh
[16:16:17] <Lowridah> ne1 has exotic animal horns /msg me $$$
[16:16:26] <SpeedEvil> I have some unicorn horn.
[16:16:34] <SpeedEvil> It's a rarity.
[16:17:20] <XXCoder> 0 is lower than any other numbers yeah
[16:38:46] <Deejay> gn8
[16:41:14] <Erant> Any thoughts on this VFD?
http://www.ebay.com/itm/ABB-ACS55-01N-02A2-1-U-1-2hp-VFD-1-120V-9-5A-In-3-230V-2-2A-0-250Hz-Out-/222067034223
[16:45:14] <XXCoder> bees.
https://i.rmbl.ws/s8/2/3/x/u/t/3xuta.caa.1.mp4
[16:46:49] <Erant> Guess I could take a gamble on this one...
[16:46:50] <Erant> http://www.ebay.com/itm/ABB-ACS55-01N-02A2-2-1-2hp-Frequency-Drive-200V-240V-/191852822354
[16:47:52] <FloppyDisk> Ha! That picture looks like a cat burglar.
[16:48:16] <FloppyDisk> Not sure what the golve is for...
[16:48:51] <XXCoder> "this item is totally hot! moving rapidly will be gone soon"
[16:49:32] <Erant> Haha.
[16:50:03] <Erant> Yeah, I'm just not sure what the whole max frequency deal is on a VFD...
[16:50:15] <Erant> And how that relates to RPMs
[16:50:39] <SpeedEvil> Max frequency * your motors frequency to RPM ratio = max RPM
[16:51:28] <SpeedEvil> For example, if your motor goes 1500RPMs at 50Hz (3000 cycles/min), then simply halve maximum frequency to get rotations per second, and multiply by 60 to get RPM
[16:51:29] <malcom2073> So if you have a three phase 1750rpm motor, at 400hz you can run it at 11krpm :-d
[16:51:35] <SpeedEvil> (I know, 1450)
[16:52:00] <malcom2073> silly 50hz'ers
[16:53:01] <malcom2073> I've heard that a lot of three phase 1750rpm motors are identical internally except for wiring to their 3500rpm variant, so can be run at 120hz
[16:53:22] <malcom2073> No clue how much truth there is to that though
[16:53:26] <XXCoder> I'm consisdering buying router speed control
[16:53:38] <malcom2073> XXCoder: Like the cheap SCRs?
[16:53:38] <XXCoder> would it weaken spindle though? (really a router)
[16:53:44] <Erant> Got it. So the fact that the max freq on the VFD is 250Hz is fine, because I probably will run it between 0-60Hz
[16:53:49] <malcom2073> Is your spindle motor a universal motor?
[16:53:50] <XXCoder> malcom2073: link to one?
[16:53:54] <malcom2073> Or three phase?
[16:54:08] <XXCoder> malcom2073: it apparently is as it sparks as it runs (brushed)
[16:54:58] <SpeedEvil> XXCoder: It can compromise cooling at lower speeds
[16:55:16] <XXCoder> malcom2073: its a router, with normal uk plug so doubt its 3 phase too
[16:55:31] <XXCoder> SpeedEvil: good to know, will make sure it says cool.
[16:55:41] <SpeedEvil> you could in principle add a fan.
[16:55:44] <malcom2073> Yeah, if it's jsut a normal router you can get a router speed control for it, which is just a SCR
[16:55:44] <SpeedEvil> Or add a thermistor
[16:56:01] <malcom2073> I actully have a 3HP one heh
[16:56:26] <XXCoder> SpeedEvil: though do it get weaker with lower power? ie easier to stop
[16:57:38] <malcom2073> XXCoder: Yes
[16:57:40] <SpeedEvil> The peak torque of a parallel wound motor will drop I think as voltage drops.
[16:57:47] <XXCoder> ok
[16:58:14] <XXCoder> it spins so fast I can feel gyro action as I try to twist position by hand heh
[16:58:24] <XXCoder> need to slow it down
[16:58:54] <SpeedEvil> I would add something to monitor the motor temperature.
[16:59:16] <SpeedEvil> This could be if you're cheap, a bimetal switch.
[16:59:34] <XXCoder> unfortunately plastic coating means its hard to get real temp
[16:59:53] <SpeedEvil> I was meaning inside.
[16:59:55] <XXCoder> thats why i want real spindle. I designed dual adoptors to fit 52mm spindle
[17:00:14] <SpeedEvil> Or go the RC spindle-motor route.
[17:00:58] <XXCoder> http://www.aliexpress.com/item/500W-Air-Cooled-Miling-motor-AC-220V-Spindle-Motor-with-6mm-collet/1903872955.html
[17:01:05] <XXCoder> SpeedEvil: how do that work?
[17:01:40] <SpeedEvil> That's just a universal motor die grinder
[17:01:45] <XXCoder> indeed
[17:01:45] <SpeedEvil> exactly like your router
[17:01:54] <XXCoder> that IS one I bought
[17:01:59] <SpeedEvil> ah
[17:07:12] <XXCoder> anyway how do rc spindle-mnotor work anyway
[17:07:54] <joem_> neat, somebody blogged about my build
[17:07:55] <joem_> http://blog.cnccookbook.com/2016/05/18/bootstrap-diy-cnc-mill-conversion-3d-printer/
[17:09:11] <Erant> XXCoder: Going to assume it's using an RC brushless motor.
[17:09:37] <Erant> Which... They're really meant for short bursts.
[17:10:24] <XXCoder> unless modificion also removes that limit
[17:11:34] <XXCoder> joem_: nice
[17:26:31] <gregcnc> I built my spindle with RC motor guts albeit very good. BLDC is the same if RC or not, but there are many ways of doing it as usual
[17:28:29] <gregcnc> I like this
http://blog.cnccookbook.com/2016/04/05/lazy-cncers-way-run-bandsaw/
[17:28:42] <SpeedEvil> You obviously can't use them at their max 'peak' power.
[17:29:03] <gregcnc> with liquid cooling you may be able to
[17:29:06] <SpeedEvil> But RC motors can be very efficient at 75% or so of that
[17:29:21] <gregcnc> I used a Neu
[17:30:31] <gregcnc> I still have to get away from the RC motor control
[17:31:49] <SpeedEvil> There are some quite flexible and programmable 'RC' controls
[17:32:40] <gregcnc> I used stuff I have. Currently a relatively ancient Aveox.
[17:49:32] <alex4nder> hmm, what kind of step frequency top-end can I expect from a mesa board?
[17:52:23] <JT-Shop> faster than your step drive can go...
[17:53:06] <alex4nder> yah, I'm not trying to run a stepper
[17:53:17] <alex4nder> I'm looking at Teknic's clearpath motors
[17:56:34] <alex4nder> looks like the clearpath step controls are limited to 500 kHz
[18:02:33] <jdh> tehn you can use mach
[18:08:31] <jdh> sorry, that was uncalled for.
[18:11:35] <FloppyDisk> jdh - You certainly should be assessed some penalty. :-)
[18:11:48] <FloppyDisk> XXCoder For that 27krpm spindle, do you have a speed controller for it to change speed?
[18:12:18] <FloppyDisk> Like a minarik or Dart type DC controller? I don't see too many people doing that?
[18:12:28] <FloppyDisk> Kinda wondered why.
[18:12:29] <jdh> I have an unexplained bias against step/dir servos
[18:12:40] <FloppyDisk> jdh - AGREED.
[18:12:43] <jdh> or any drives that don't do closed loop via the control
[18:13:00] <JT-Mobile> An oxymoron
[18:13:09] <FloppyDisk> Oddly, Japanese Omron servo's are ONLY step and direction. Why? So dumb.. IMO.
[18:13:32] <FloppyDisk> That was a great little servo system, good price, but only step and direction... ugh.
[18:13:45] <DaViruz> there are analog drivers too
[18:13:49] <DaViruz> but they are kind of rare
[18:14:14] <FloppyDisk> I'm sure, but if you're closing the loop in the drive, why not allow +-10Vdc input and allow access to the control loops?
[18:14:16] <DaViruz> R88D vs R88A for example
[18:14:31] <JT-Mobile> Step direction is good for automation applications
[18:14:32] <DaViruz> yeah i dunno, i don't like it either :)
[18:14:55] <FloppyDisk> You already have it in there... DaViruz - you know your Omron...
[18:15:02] <jdh> you need feedback to the control
[18:15:10] <DaViruz> i have a couple omron servos
[18:15:34] <FloppyDisk> jdh - I think the feedback on a step and direction is the fault signal...
[18:16:04] <DaViruz> well they still have quadrature output, so feedback isn't lost
[18:16:23] <FloppyDisk> True.
[18:16:47] <FloppyDisk> I think part of it, in Omron's defense is they want to use the servo w/ their PLC and many of them output step and direction.
[18:17:11] <DaViruz> i don't think i've ever seen an omron servo in a cnc application
[18:17:11] <FloppyDisk> So, it's a natural fit for them. But, I think they limit the usage by not having the analog in.. imo..
[18:17:40] <FloppyDisk> Yeah - you're not writing a cnc application w/ a cp1h brick PLC - Ughhhh...
[18:17:44] <jdh> normal PLC IO is too slow for useful step/dir
[18:18:03] <FloppyDisk> You need the high-speed outputs that are hardware specific...
[18:18:07] <jdh> so if you have a separate motion control module, you coudl do either
[18:18:15] <FloppyDisk> Read one of the 1000 page manuals for that...
[18:18:55] <FloppyDisk> We did 4 axes of simple motion off a CP1H, never again, that was a mistake. Took 2 days to get homing to work right, with the Omron tech support.
[18:18:56] <jdh> so anyway, the ClearPath(tm) motors look pretty nifty for a lot of applications
[18:19:15] <FloppyDisk> But, it was a lowcost solution at the end of the day :-)
[18:19:40] <jdh> we have some steppers that look and act just like them. not really any cheaper though
[18:20:34] <jdh> except you have to register to see the data sheets
[18:21:12] <FloppyDisk> Hmmm, like animatics.
[18:22:00] <gregcnc> Isn't that route basically the same as gecko servo drives, but in a convenient package?
[18:22:13] <FloppyDisk> gregcnc - I'd say no...
[18:22:25] <jdh> I wouldn't use gecko servo drives eithere
[18:22:32] <FloppyDisk> because you have the controller in the motor package
[18:22:52] <FloppyDisk> Which means that if you want to use Lcnc, you need to 'program' the servo to be in 'dumb' mode
[18:23:03] <FloppyDisk> probably a better name for that, but you're not using half of what they intended.
[18:23:21] <FloppyDisk> There is a place for the all-in-one, but not so much when you need coordinated motion w/ LCNC.
[18:24:21] <FloppyDisk> I was bent when an engineer bought an animatics that he boasted about it being sooo small of a package and cheap too.
[18:25:01] <FloppyDisk> But, when we added the power supply for 120vac to 48vdc, it was larger than a gecko (by about 8x). And, the speical cables cost an arm and leg.
[18:25:13] <FloppyDisk> Just my opinion... They do have their place.
[18:25:18] <gregcnc> I've been running a gecko servo for 10 years. Seems OK. I may upgrade that machine to mesa though
[18:25:37] <FloppyDisk> gecko what? you should be able to use mesa w/ gecko, just fine...
[18:25:52] <gregcnc> 320 servo drive
[18:26:57] <FloppyDisk> http://www.geckodrive.com/g320.html
[18:27:30] <FloppyDisk> hmmm, looks like step and direction, that would work w/ a mesa 7i76 or 7i76e or picosystems, I think.
[18:27:41] <gregcnc> I have 2um resolution now. I'd increase it to 1 or 0.5um but that would be slow on the parport, so I'd use anolog drives and 7i77.
[18:28:14] <FloppyDisk> I cold see that. Sounds like a plan.
[18:28:18] <FloppyDisk> What kind of mill do you have?
[18:29:58] <gregcnc> frankenmill ™
[18:30:52] <gregcnc> http://www.dieselrc.com/x/temp/other.html
[18:31:41] <FloppyDisk> Oh yeah, I see the gecko's down the page. NICE - aerotech... that will work.
[18:32:18] <gregcnc> those motors have tachs too
[18:33:04] <FloppyDisk> I'm jealous of your emco6, looks nice.
[18:33:11] <Duc_mobile> I think I found what would be considered my crack supplier. Guy in town has 5 semi trailers full of electronics. plcs, motors, drives etc
[18:33:16] <gregcnc> yeah, I need to finish
[18:33:23] <FloppyDisk> don't we all...
[18:33:45] <FloppyDisk> I'm officially in 'don't add anything to the list' mode...
[18:34:49] <gregcnc> me too, it's just too easy. if I do this and add that.......
[18:35:34] <gregcnc> two of those geckos died this year, the 5V supply failed
[18:36:15] <Duc_mobile> have you added a encoder for the spindle yet for rigid taping
[18:37:22] <gregcnc> Maybe after I add a real motor control. But I have another mill now that taps, so maybe never
[18:37:51] <FloppyDisk> I started to, but realized you really need the encoder on the spindle and not on the motor behind the gearing, the Z-pulse is key...
[18:38:44] <Duc_mobile> cant you software gear it
[18:39:15] <FloppyDisk> I have thought about different options for that and havne't come up w/ a solution.
[18:39:27] <FloppyDisk> It's been about 4 months since I looked at it, so I'm foggy.
[18:39:29] <gregcnc> you would probably OK with gear lash if you had some compliance in the tap holder
[18:39:49] <malcom2073> need tension compression tap holders
[18:40:21] <FloppyDisk> But, the issue for me is that I hae a mesa 5i25/7i77 and the encoder inputs are handled through the hostmot2 driver, I think (not totally clear)
[18:40:25] <Duc_mobile> that should work perfect
[18:40:41] <Duc_mobile> why does that matter?
[18:40:47] <FloppyDisk> And, I would want the A and B pulse for speed & direction, but need the Z for the index and it's in the 'wrong' spot.
[18:40:56] <zeeshan> yesssssss internet is back
[18:41:02] <zeeshan> http://i.imgur.com/banY5V9.jpg
[18:41:04] <zeeshan> fixture success!
[18:41:05] <FloppyDisk> why does it matter - again, it's not as straight fwd as I would have thought.
[18:41:12] <zeeshan> http://i.imgur.com/URyTEWD.jpg
[18:41:13] <malcom2073> Nice zeeshan
[18:41:16] <gregcnc> I set up a spindle encoder on the Emco 6 with 5i25 / 7i77
[18:41:53] <FloppyDisk> gregcnc - but the encoder is 'on the spindle' and not through any gearing?
[18:42:18] <FloppyDisk> Or is the encoder a motor pulley where the gearing might change?
[18:42:23] <FloppyDisk> or motor.
[18:43:08] <gregcnc> on the spindle, right if gearing can change that's a problem unless you can tell it what gear it's in?
[18:44:01] <FloppyDisk> So, I can tell what gear it's in, but the Z-pulse is the issue, per my investigation... I thought about wiring in the z-pulse into a different I/O that would be fast enough
[18:44:08] <FloppyDisk> But, haven't gotten back to it...
[18:44:54] <gregcnc> wouldn't you put Z on the encoder input like usual?
[18:45:11] <FloppyDisk> yes.
[18:45:14] <gregcnc> got to run
[18:45:21] <JT-Shop> anyone have an absolute encoder?
[18:45:34] <FloppyDisk> but, for rigid tapping, it needs to see a pulse once every spindle revolution
[18:46:13] <FloppyDisk> So, if you're behind gearing, your z-pulse will be off...
[18:46:38] <zeeshan> FloppyDisk: why one pulse every rotation
[18:46:41] <zeeshan> thats pretty coarse!@
[18:46:43] <FloppyDisk> It probably would work, but you wouldn't get flat-head screws like andy pugh's grandpa used to get for locomotives:-) Probably...
[18:47:04] <JT-Shop> index pulse
[18:47:05] <FloppyDisk> zeeshan - that's the rigid tapping algorithm, I believe...
[18:47:18] <FloppyDisk> Yeah - I should use index and not z-pulse...
[18:47:40] <FloppyDisk> JT has a nice picture of his encoder on the top of his bridgeport.
[18:48:16] <FloppyDisk> I can't do that, i have a kurt power drawbar.
[18:48:16] <JT-Shop> he has a quick switch spindle on the BP knee mill
[18:48:25] <FloppyDisk> :-)
[18:48:54] <JT-Shop> Sam uses a gear and a hole
[18:49:22] <FloppyDisk> Gear - as in at the lower end of the spindle?
[18:49:28] <FloppyDisk> where the tool goes?
[18:50:51] <Duc_mobile> I saw pictures where someone place a hall sensor in the head for the index pulse
[18:51:15] <FloppyDisk> Duc_mobile - that's jon elson.
[18:51:23] <FloppyDisk> I should do that...
[18:51:34] <malcom2073> buddy of mine uses a reflective sticker and IR heh
[18:52:07] <FloppyDisk> But, I'm a wuss and getting the head apart on the supermax isn't fun... But, I should check that out (again), anyway...
[18:52:21] <Duc_mobile> LOL
[18:52:52] <FloppyDisk> It's not really that hard, but the 3 HP motor is heavy and in an awkward position.
[18:53:49] <FloppyDisk> the ir and relective tape would work as well... I have a keyence capacitive sensor (I think) that would be fast enough and thought about that for a patch. I should just do it...
[18:54:14] <FloppyDisk> Anywya, I need to run. take care.
[18:58:41] <FAalbers_> CamBam question. I want to create a 3.3 mm hole with a 3.175 mm tool (1/8 inch) using a profile cut. But it seems to only work with a max tool size of 2.8 mm . Any idea why ?
[19:02:36] <FAalbers_> Well, a pocket seems to work
[19:17:47] <djdelorie> do you have a maximum overlap set too high?
[19:21:56] <BeachBumPete> did you have it setup with a finish cut at the same time? if the finish cut does not have enough material left it will not work
[20:00:52] <Polymorphism> https://www.youtube.com/watch?v=ehovEUsEce0&t=3m0s
[20:00:58] <Polymorphism> thats some shit results
[20:03:44] <Duc_mobile> have you bought one yet
[20:06:59] <malcom2073> Wait, you're shocked that a 3d printer makes a shitty router?
[20:07:00] <malcom2073> Really? :-P
[20:09:14] <Duc_mobile> somewhat like a stripper would probably make a bad wife
[20:09:33] <SpeedEvil> yeah - you _can_ make a nice CNC machine with that geometry
[20:09:56] <SpeedEvil> It has bearings and structural members which are considerably more expensive
[20:11:47] <skunkworks> https://www.youtube.com/watch?v=hPE3Qr-ECtQ
[20:17:12] <Duc_mobile> If you guys ever need a stator laced on a motor I would not recommmend me doing it
[20:19:20] <Duc_mobile> this is a job better left to robots
[20:24:42] <djdelorie> Polymorphism: he has one more cnc router than you do...
[20:25:29] <djdelorie> Duc_mobile: likewise me and transformers:
http://www.delorie.com/electronics/bldc/transformer/
[20:28:16] <Duc_mobile> djdelorie: I should stop being a cheap ass and just pay people sometimes
[20:28:56] <djdelorie> I've learned that, as I get older, I'm more likely to solve problems by throwing money at them. Still rare, but more likely...
[20:30:19] <Duc_mobile> this was a hail mary to see if I could get it working before buying a new motor
[20:30:57] <Duc_mobile> new motor is 850 from factory
[20:30:58] <djdelorie> so, given a choice between "I have to buy a motor because the old one is broken" and "I have to buy a motor because the old one is more broken"... ?
[20:31:11] <Duc_mobile> yep
[20:32:12] <djdelorie> There are times when you ask yourself, "What's the worst that can happen?" and still get surprised...
[20:33:46] <Duc_mobile> yea I dont ask that question anymore
[20:34:32] <djdelorie> that and "is that possible?" ;-)
[20:35:26] <Duc_mobile> lol
[20:35:52] <Duc_mobile> I did also get a linear encoder from the guy for 20 bucks for the knee on the bridgeport boss
[20:36:11] <djdelorie> sweet. I have X and Y but no Z on my DRO
[20:36:54] <Duc_mobile> I will tie it into my mesa 7i77 card to act as W
[20:58:34] <Duc_mobile> 5 mins till i rehook up motor
[21:06:15] <jdh> and
[21:06:32] <Duc_mobile> fucking A, it freaking runs
[21:10:40] <Duc_mobile> and nothing blew up
[21:16:14] <enleth> Duc_mobile: what happened to the motor?
[21:25:13] <Duc_mobile> enleth: the cold saw was in a shop fire and the wiring was burned back into the motor
[21:26:14] <Duc_mobile> enleth: and the electrical was melted
[21:28:44] <enleth> ah, that motor
[21:30:01] <enleth> but you didn't have to do a complete rewinding?
[21:30:28] <Duc_mobile> nope just new leads to outside the motor and pretend like i know how to tie a stator up
[21:30:49] <enleth> well, if it works, apparently it's not rocket science
[21:32:54] <Duc_mobile> nope but I may not have it 100% correct but it work and two different speeds depending on the leads wired up
