Ha! Can't afford the controller yet so it will have to!Quote:
Originally Posted by Clive S
Printable View
Ha! Can't afford the controller yet so it will have to!Quote:
Originally Posted by Clive S
Hi everyone, I have uploaded the next episode of my build, not that exciting this time just making holes in gland plates! I'm working on the ali angle that supports the cable chain next...
https://youtu.be/60zoxk6o2LI
Woohoo! Thanks Gary! My IP-M has arrived. https://uploads.tapatalk-cdn.com/201...b4b6d63984.jpg
They look good, don't they! But just wait until you start wiring the bottom row of terminals right next to the cable guides. I was really glad at that point to have used crimp-on ferrules as they gave you something to get hold of with tweezers while guiding wires into place, and without the risk of odd stray shorting strands. Using differential wiring to the stepper drivers?
Out of interest, how will you run the ethernet cable out of the box? I ended up using a bulkhead ethernet connector which effectively sits inside an XLR connector shell. Standard RJ45 inside but protection and latching of XLR. Means that I can disconnect all cables from cabinet very easily without trailing wires, but everyone has their own way of doing things.
Yes I have ferrules and they are great as you say.
I was planning on using cy cable for the connections with one side of the shield connected to the star earth.
Is this the kind of rj45 connector you mean https://www.ebay.co.uk/itm/152664475439
Sent from my iPhone using Tapatalk
I used something more like this although I can't remember now where I bought it. This is the panel-mounting socket, and you can get a matching XLR plug shell that takes a standard RJ45 cable and connector. Beware - some of these go on the cable before the RJ45 connector is fitted and some can go on to existing ready-made cables. My steppers plug into the box via 4-pin XLR connectors so these match.
In my control box, I used UTP cable between IP-M and stepper drivers to get the benefit of twisted pair cabling with differential signalling. UTP gives you 4 pairs, so I used one cable for two stepper drivers. The twisted pair is possibly more effective than CY screening for noise rejection, along with the IP-M's capability of differential output. Well, it works for me, but on this forum you only have to wait a few more minutes for an expert to come along!
I just could not resist:hysterical: But I am not that one. I use a very similar bulkhead fitting like yours and also 4 pin XLR's for the steppers. Although is is easier to wire the steppers direct to the drives.Quote:
Well, it works for me, but on this forum you only have to wait a few more minutes for an expert to come along!
Edit http://www.ebay.co.uk/itm/like/17212...6243387&crdt=0
Certainly true! My first version did exactly that. I then found that I was going to need to take the cabinet out of the machine several times during testing and development, and I ended up changing to XLR just for ease of use. I had the connectors from my previous machine anyway. My cabinet has only one cable going out through a gland and that's the mains cable, so it can completely detach from the machine itself if needed. Of course, the experts get it right first time so never need to disconnect anything thereafter!Quote:
Originally Posted by Clive S
Excellent build Joe read the whole thread from beginning to end yesterday! The PSU bit is a real god send to me right now as building my own and you really helped demystify it for me, thanks ;)Quote:
Originally Posted by JoeHarris
I have a quick question what thickness Red/black wire do I need to get to connect from the rectifier to the capacitors then to the output?
Also where did you get those terminal blocks I could only find them in China or America although I found a suitable replacement I think with barrier terminal blocks.
https://www.rapidonline.com/truconne...-block-21-3073
Thanks
Desertboy
Glad it helped - I was pretty scared of this part of the build!! Clive and Dean were super helpful as usual. I think I used 2.5mm tri-rated, I'll try and remember to check for you when I get home.Quote:
Originally Posted by Desertboy
In fact you can see in the pictures it says 13AWG which is 2.5mm. It hasn't seen any hard use yet so remains to be seen if it will work well, but it switches on without killing me or catching fire, so that's a start!Quote:
Originally Posted by JoeHarris
It was Clive that put me on to you and convinced me I could build the supply. It looks really simple but at the same time I can totally understand the fear lol I was petrified but now I think it'll be relatviely easy just trying to work out if I can cram this into an ATX power supply casing if I move the fan externally and add a 2nd fan either so it blows in one side and sucks out the other.Quote:
Originally Posted by JoeHarris
I want to try and fit the stepper drivers and self build PSU in a PC mini tower case with a whole pc as well.
I have a mini 80plus PC power supply I can fit somewhere in the case (It's about 1/3 size of ATX) and I think I can put the AM882 where the drive bays are. I could ditch the Hard disk and run from a fast USB flash drive or SD card it will be quick enough for linuxcnc. Or I could mount the hard disk somewhere else. I think as long as the PSU's extract their heat to the air and I add a bathroom extractor fan as well for the pc and stepper driver waste heat I'll be ok. Maybe add an active intake as well sucking air a distance away from the extraction point.
Joe A picture paints a thousand words I tend to point people to your P/S vids (if that is ok) as they are an excellent way to show how easy they are to build. There is one point and you have it correct is that you must not but an earth (ground) to the bolt on the top of the transformer that runs through it. It is ok at the bottom end ie the plate.
If you do die can you get some one to send me a PM saying get 4mm wire lol ;) I missed that but actually managed to read the part number off the connector blocks and find them on ebay lol.Quote:
Originally Posted by JoeHarris
No worries Clive - just glad they are of some use to people!Quote:
Originally Posted by Clive S
Btw I have made a playlist of the two videos here to keep them together: https://www.youtube.com/playlist?lis...zD30sZjtp_VyqY
Those connector blocks are pretty huge and a total overkill! I recon 2.5mm should be ok as it is rated for 30ampsQuote:
Originally Posted by Desertboy
Another little part made...
https://uploads.tapatalk-cdn.com/201...4508200558.jpghttps://uploads.tapatalk-cdn.com/201...8e331b9cab.jpghttps://uploads.tapatalk-cdn.com/201...5d3ffde25b.jpghttps://uploads.tapatalk-cdn.com/201...64cd94e5e2.jpghttps://uploads.tapatalk-cdn.com/201...92b9edd3c2.jpghttps://uploads.tapatalk-cdn.com/201...dcdbf8d944.jpghttps://uploads.tapatalk-cdn.com/201...ac8778f87f.jpg
Beautiful!!
Skickat från min SM-N910C via Tapatalk
Just stunning work, made me realise I forgot to buy a vice yet ;)
Or a big hammer:thumsup:Quote:
Originally Posted by Desertboy
Brian / Dean have I got this right?! https://uploads.tapatalk-cdn.com/201...9ac593bf52.jpg
That looks fine. I've just had a quick look in the IP-M manual and that is all OK for X. Are you using master/slave on one axis? As I remember, channels 0, 1, 2 are used for X Master, Y, Z, and channel 3 for X slave.
One small point to watch (and I only say this because I managed to get this wrong on my own machine) - be very careful to connect + to +, - to -, just as you have in your diagram. If you happen to swap the wires in the pair so you have + to -, - to +, nothing blows up and it all seems to work fine. However, if you have done this on the Step signal, then actually what happens is that every time you change direction on that axis, you lose a step. In my case, I managed to get the Z axis wrong. When you are just doing simple profile cuts, you don't see an issue. Then you start a complex 3D carving job and the cutter gradually goes deeper and deeper into the work. Best solution is not to get it wrong! However, if this does happen (and the IP-M manual tells you how to test this - p11/12 in my copy) you can either swap the wires so that they are correct, or change the step signal to be active low instead of active high. Not a big deal if you are aware of the possibility.
Thanks that is really useful: this diagram in the am882 manual confused the hell out of me: https://uploads.tapatalk-cdn.com/201...fbc9a0bc6f.pngQuote:
Originally Posted by Neale
The (+) and (-) on the pulse and direction terminals relating to differential signals was also a revelation!!
I'm only using three of the outputs on the ipm (0,1 and 2) as I am using one motor and a long belt for my x - axis.
Really appreciate the help.
Absolutely - the diagram is correct for most motion controllers/BOB's, and you could wire the IP-M that way, if you could interpret the over-complicated diagram. But the IP-M is much better wired the way you have it in your own sketch.
Are you also going to use the alarm outputs from the AM882s?
Yes using the alarm outputs on the AM882s which I have wired + to +24v and - to pins 8, 9 and 10 of the ip-m's digital inputs. Have I got that right? https://uploads.tapatalk-cdn.com/201...f2fd0bde1f.jpghttps://uploads.tapatalk-cdn.com/201...e2bc4865c2.jpghttps://uploads.tapatalk-cdn.com/201...63222f627d.jpghttps://uploads.tapatalk-cdn.com/201...43c0865092.jpghttps://uploads.tapatalk-cdn.com/201...66f5b4d456.jpgQuote:
Originally Posted by Neale
Not sure how that is going to work, but I have to say that on my own machine, the fault/alarm signals go to a safety relay and not directly to the CSMIO. However, you don't need to use three inputs - you can wire the alarm signals in series, and take them to a single input. The AM882 can be configured to be active high or low on a fault; again, I'm not using exactly the same hardware as I have EM806 which are, in effect, a slightly later version of the AM882, so I'm not sure exactly how they are configured fresh from the factory. Looking at the extract of the manual above, you just need to take +24V to the + alarm connection on the first driver, take its - connection to the + on the next, and so on. When all are connected in series, take a wire from the - alarm connection on the last one to one input on the CSMIO. Then, in Mach3, you can use this as the "alarm" signal (can't remember what it's actually called, but it's something like that). In my case, because I want a stall alarm signal to act as an e-stop, I also connect my three e-stop switches in series with the alarm signals as well; a "fault" from any of them will then activate the e-stop actions.
From talking to Dean, I think this sketch shows the correct wiring: https://uploads.tapatalk-cdn.com/201...43f51281c7.jpg. He also mentioned linking this into the estop and safety relay circuit but I haven't figured that out yet. Suffice to say I had it wrong before!
Joe - that doesn't look right. I'll try to sketch out a quick drawing tomorrow when I'm not quite so tired! Are you also using a safety relay? Do you want the driver alarms to trip the safety relay or just the CSMIO?
That diagram is correct. Think of the Alarm output as a basic single throw switch, that is either open, or closed.
The opto input needs to see 24V. So 24V from the PSU goes to the opto positive, then the opto negative goes to the Alarm positive, before finally completing the circuit by connecting the Alarm negative to the PSU negative.
There a couple different ways you can connect to an e-stop/safety relay, however it depends on what you're using.
Thanks guys - as always appreciate the help! As far as my specific estop set up I have explained this on my video log. Link below if you can be bothered to watch it!! But is essentially a pilz with two estop buttons wired in series and a contactor to switch the mains on / off to the drives' psu. Could you put all the drive alarms in series between the pins shown on my sketch above and then in series with my estop buttons?
ESTOP Video link: https://youtu.be/OUgjcH8JGMs
You can wire in them series with the e-stops, however you have to give some thought as to how you get the drives back out of alarm, if your e-stop system is also killing power to the drives.
It all gets just a bit more complicated than this. One issue is that you have your stepper drivers powered on by the safety relay. That means that the fault outputs on the drivers are the equivalent of open circuit, as they are unpowered until you have energised the safety relay. However, if you wire the fault outputs in series with the e-stop switches (p9 in the CSMIO manual shows this) then you can't activate the safety relay as it thinks that an e-stop condition exists (via the driver fault outputs) but you can't turn those on until you have put power on the drivers. Catch-22.
I have virtually the same setup as yours - Pilz safety relay (different model, but that's not important), CSMIO, digital drivers (mine are EM806 not AM882 but again no practical difference for this purpose) and power to stepper driver PSU controlled by relay controlled by safety relay. What I have done is:
safety relay uses three contacts - two of the N/O and one N/C. One N/O contact switches 24V to pin 1 on CSMIO, configured as "e-stop". This puts the CSMIO into e-stop mode when the safety relay is off. One N/O contact switches 24V to the stepper driver PSU relay (like you have done). So, no power to motors unless safety relay is on. The N/C contact switches 5V to the enable inputs on the drivers. With the relay off, this disables the drivers. Once the relay is energised, 5V is removed and the drives are enabled. So, when you hit e-stop, you remove power from the steppers, tell the CSMIO to go into e-stop, and disable the drivers.
To get round the problems with the fault signal from the drivers, I take these in parallel directly to the CSMIO, pin 2, configured as "Drive Fault". The CSMIO monitors this input specifically for this kind of situation and processes the signal without needing to talk to Mach3. I do this by taking 24V to the + side of the fault connections (all in parallel) and the - sides to the CSMIO, so when a fault occurs the internal "switch" closes, puts 24V on the CSMIO pin. The corresponding CSMIO pin 15 is taken to ground.
All I can say is, it all works! On a stall signal, the CSMIO stops the machine immediately via its own firmware. However, the safety relay is still energised. To get the driver(s) out of the fault condition, I hit e-stop which removes power, then hit reset to re-energise the relay and hence everything else. After a stall, you are going to need to rehome anyway, which is what Mach3 makes you do after an e-stop.
I did consider reprogramming the EM806 (and you can do the same with the AM882, I believe) to change the sense of the fault output, so in effect the "switches" on the fault output were closed for normal operation instead of open. However, you still have the problem that you have to keep the drivers powered because otherwise the fault outputs look like open switches. I took the easy route of keeping them programmed as the default, so they "close" on a fault. I have them wired in parallel so any fault output takes the CSMIO pin to 24V. At the cost of more complication, I could run the fault signals into the safety relay and in principle this would be slightly more fault-tolerant, but I accept the compromise.
My main power switch operates on all power coming in to the box. I turn that on at the start of a session and don't touch it again. I have reset and e-stop switches on the box, plus two e-stops around the machine. The safety relay does the job of resetting the drivers via the driver PSU, as explained above. I don't have a separate power switch for the driver PSU.
Thanks for this explanation - I am going to try and draw it out so I can understand then I'll come back as it will take a bit of digestion (I don't have a background in electronics or computers so the learning curve is quite steep!!) ps. is it necessary to disable the drivers if you are also cutting power to them? Belt and braces?!Quote:
Originally Posted by Neale
Happy to answer questions on it, but drawing it out and thinking through the logic will help. Sorry I don't have any wiring diagrams to post but I did it all on the fly, with just a few notes on which CSMIO pin does what. Not a recommended approach :redface:
Ok so I'm a little bit confused because apart from integrating the estop pins on the CSMIO [+24v, to Pilz NO, to CSMIO pin 1, then from CSMIO pin 14, to 0v on 24v PSU] this looks to be pretty well the setup I had originally?? Albeit using three fault pins on the CSMIO rather than one [+24v, to AM882 drive ALM+, then from ALM-, to CSMIO digital input pins 8,9,10, then from CSMIO pins 21,22,23, to 0v on 24v PSU] Have I got this right?!https://uploads.tapatalk-cdn.com/201...d665f410da.jpgQuote:
Originally Posted by Neale
Just drawn it out as I understand it... have I got this right??Quote:
Originally Posted by JoeHarris
https://uploads.tapatalk-cdn.com/201...2f8ef3f082.jpg
Bugger, just noticed I forgot to show pin 14 connected to 0V
Basically, that looks fine (once pin 14 is connected!) Main problem, which is the thing I noticed when I looked at your first diagram, is the use of three inputs on the CSMIO for fault inputs from the drivers. I'm not sure that you can configure the CSMIO to do that. The CSMIO can be configured to have one "drive fault" input; it monitors this pin and will stop the machine if it sees a fault. So you need to take all the fault inputs into the same pin. Because of the way the fault signals work, wiring them in parallel like this is fine. In my case, all the driver fault "-" pins go to CSMIO pin 2; the driver fault "+" pins go to +24V (as you have them) and obviously you need to take CSMIO pin 15 to ground.
I'm sorry if I confused things a bit; I had in mind that you could wire the fault signals in series with the e-stop switches (as shown in the CSMIO manual). However, I had forgotten that when I looked at doing this, it wasn't going to work easily for various reasons so I went back to the scheme broadly that you have, with the modification I have explained above - which actually simplifies wiring a bit as you can take each fault connection on the drivers to the adjacent driver and then one wire to the CSMIO.
One thing to bear in mind is that you have to be able to clear the driver fault. Easiest way is to remove and re-apply power and you can do this on your machine by hitting e-stop or by switching off power to the driver psu as you have a switch to do this. I think that the CSMIO tells Mach3 that an e-stop has happened on a driver fault anyway, though.
Thanks I think I'm clear, so did you just take three wires, one from each driver ALM+ and twist them together in one ferrule at pin 2 on the CSMIO?
Sent from my iPhone using Tapatalk
No, I took a wire from the first driver to the second, wire from second to third, etc, then to the CSMIO. Where two wires went to a single pin, I fitted them to one ferrule. Easy job.Quote:
Originally Posted by JoeHarris
Like this? https://uploads.tapatalk-cdn.com/201...c75a66be09.jpgQuote:
Originally Posted by Neale