Why do touch buttons on my new iInterface go completely crazy.? Either they respond before I touch them or they overreact by activating other buttons.
Usually EMI in the environment around it. Take it away from your workshop. Make sure you are on Firmware v1.51. Power it up using a USB cable. During the Boot process, when the big logo appears on screen, press and hold the MODE button on the left side until you hear a beep. This is a device test mode where you can test it just like the factory does. NB make sure you are on Firmware v1.51 first as thats the version we added the Mode button code described above to allow customers to also run the Device Test. docs:interface:firmware-update-control [OpenBuilds Documentation] Taking it away from your workshop for this test tells you if a nearby spindle, faulty earthing, fluorescent lights, etc in the workshop is causing interference: docs:blackbox:faq-emi [OpenBuilds Documentation] If it tests out fine away from the shop, we know the hardware is good. If you fail the test, contact the store via Support and they will be happy to get a new one out to you
We'll help you work the problem if it is that. But lets try that initial tests first. That will tell us which of the two issues it is first.
I have now grounded the frame on my cnc and it helped when I go a little away from power sources. And then I ordered ferrite clamps. But when I mount it on the magnet holder, it goes crazy. Could it be because it is not the original magnet the store sells.?
I almost doubt it directly related - but what if you take the magnet and hand-hold it against the back, but move away from the machine? Does that make a difference? (May be conducting some interference from the frame to the Interface via the mount?) That said, we'd really still like to make sure the hardware is good by performing that test above ^ - before we start chasing interference issues: What if it is the hardware and we could have had the store send out a new one already. Please first try that - away from the shop - it will give us the confidence that a) the hardware is good (or bad) and then b) yes we have to try to find interference. You can repeat the QA test with the magnet on the back too - away from the shop - just to cement any observations. It guides you through testing each button in turn...
I have done the test and it works. And it works after I grounded the cnc away from the magnet. As soon as I mount the magnet away from the cnc, it goes crazy. But I will test it with the magnet in a other room like I did without. The magnet is very powerful.
Good stuff, that mode will allow you to trace things as you move forward. I will say we've tested loads of different magnets, both the "car mount" type we recommend, bare neodyniums, etc over the years - never seen a magnet itself cause issues like that. But do keep us posted
Ferrit Clamps hjalp desværre ikke.. Men jeg har opdaget at når jeg eller bordet rører ved alupladen der er monteret på bagsiden, til magneten så sker det.. Men først når den er tilsluttet Blackboxen. For når jeg laver de test du beskriver, med en magnet på i et andet rum, er der ingen problemer.
Ferrit Clamps unfortunately did not help .. But I have discovered that when I or the table touches the aluminum plate that is mounted on the back, to the magnet, it happens .. But only when it is connected to the Blackbox. Because when I do the tests you describe, with a magnet on in another room, there are no problems.
Add a grounding wire to that point? See if you can earth out any interference coming from there? Further up the stack, something is generating interference, do take a look around - flourescent lights from the 80s/90s, VFD spindles, any other electrical transformers or similar?
Only when Interface is connected to cnc there are problems. And that's only when I or something touches the plate with the red arrow, even though I'm standing as far away as possible. If I or something touches all that around with green arrow, there are no problems. So how do I get the plate grounded.? The table on which the magnet is mounted is grounded.
So, long story short, something from the machine frame is conducting some EMI to it. Mount the magnet to the table, a shelf, or something else non-conductive - not on the machine itself. What machine is it? Running a VFD? We could send a replacement, but as you confirmed the test away from that area is perfect. You will need to address whatever is the source of interference in that area. We had seen one guy who had a chaffed through electrical cable in his garage floor that conducted up his body if he stood on the floor, went away if he sat on a plastic chair. Another guy's subpanel didn't have an earth so all the electrical devices were not earthed. Then of course Chinese VFDs are a good known source of interference too. Tell us whats around
Today I have strange beviator too : All homing are complete at begining, All parts ares grounded. This stranges moves not appear before today ? I make device test mode and all are good. Firmware are 1.51 Can you help me with that ?
Usually EMI in the environment around it. Take it away from your workshop. Make sure you are on Firmware v1.51. Power it up using a USB cable. During the Boot process, when the big logo appears on screen, press and hold the MODE button on the left side until you hear a beep. This is a device test mode where you can test it just like the factory does. See docs:interface:device-test-mode [OpenBuilds Documentation] NB make sure you are on Firmware v1.51 first as thats the version we added the Mode button code described above to allow customers to also run the Device Test. docs:interface:firmware-update-control [OpenBuilds Documentation] Taking it away from your workshop for this test tells you if a nearby spindle, faulty earthing, fluorescent lights, etc in the workshop is causing interference: docs:blackbox:faq-emi [OpenBuilds Documentation] If it tests out fine away from the shop, we know the hardware is good. If you fail the test, contact the store via Support and they will be happy to get a new one out to you
Thanks for fast reply. I make all of your previous post before posting question with that ;-) I try with exact same setting and nothing today ? This is strange. Thank you for your support
So, what if we update everything, run the MODE test and get stuck at X- / X+ test. Something is wrong with the way the touch is interpreted. -Any suggestions for troubleshooting and/or fixing that?
Ok so I too have a shiny new interface.. Buttons behave exactly as in above video. When connected to USB only the buttons seem to behave.. If connected to eithe only blackbox or blackbox and USB, the buttons are unusable. I have tried running the diagnostic, both in USB mode and also when connected to blackbox data cable.. It passes the test in both cases.. I have put a whole pack of ferrite beads on the cable, no change.. There is a vfd in the same case as the blackbox, but a) it's in a heavily shielded compartment b) the spindle is not powered up and c) I just tried with vfd disconnected entirely from power, and I still have same issue. Apart from the vfd, I'm not sure what interference there could be? I'm in my living room, testing electronics setup before adding to machine.. No fluorescent lights.. Literally nothing around that I can forsee causing issues. Unless the switch mode power supply sold to me with the blackbox is the issue? Was sold by openbuilds italia. I appreciate that a solution might be "send it back for a replacement" but I got it delivered from US when openbuilds had their "free worldwide shipping" offer. I'm in Italy. I suspect it would cost me more in shipping and probable customs fees than a new device. Interface seems crazily sensitive? I have no issues with laptop running Control connected via USB..
Its ordinarily not too sensitive, so it would be worthwhile to locate possible cause. You mention PSU (good gut feeling there) : we do sometimes see PSUs with bad Earthing or noise cause that (our Meanwells are bulletproof, but we can make a working lab machine here do the instability thing by swopping in a $20 power brick in place of the Meanwell) What PSU do they supply? A replacement unit would only apply if it was flaky on USB power / test mode. This one seems fine, but affected by something connected
I don't remember the brand, if there even was one. Certainly not a meanwell.. In fact when I received it there was an inductor disconnected inside (bad soldering and hot glue done about 2cm from the correct location). Which I repaired myself rather than deal with the replacement process. I can Say it is well earthed insofar as the work I have done, (Including from the pcb to the case of the psu) I raised concerns about the quality of the product, but didn't get a reply beyond "we will replace it if you can't fix it"
Ok so I'm narrowing it down.. First I modified the provided lead, adding a massive braided shield around it.. No change. I then disconnected blackbox from psu and ran it from by bench top supply.. The interface started working.. So, that narrows it down to either 1. The psu 2. One of the 3 little dc-dc variable voltage regulators I run from the psu to provide lower voltages for the fans, water pump and power leds... 3. Or the pump (pc water cooling) or 4. The fans.. I think we can discount the power led.. Alternativeloi guess it cod be an earthing issue? But I think I have followed "best practise" wherever possible.. Psu is bolted to chassis, there's a star earth block also bolted to chassis next to it, with all earths and cable screens terminating there.. If I do need a new psu.. Is this the right one? Does the price seem reasonable? I've seen them for 3x this price, but also a couple of places for like 20 euros.. https://www.ledkia.com/it/acquistar...o/731-alimentatore-mean-well-nes-350-24v.html
One question, I followed guide to modify blackbox too work with interface.. Do I need to undo changes to test via USB from laptop? (put back USB jumper, remove internal dongle that came with interface?) or will it still work via USB as-is? Don't want to fry anything..
No need to move anything, can just plug in the USB cable and go Looks a bit overpriced compared to https://www.digikey.it/en/products/detail/mean-well-usa-inc/LRS-350-24/7705034 (but no stock, just checking typical price) - I know its a long shot - but the Lab PSU test was promising enough that I think its worth a shot (after you eliminate the DC-DC Converters as possible noise sources)
Thanks so much for the great support. This afternoon I'll try running everything, including the dc/dc converters from the bench supply and see if it works. One thing I noticed, unless I connected the bench top supply chassis to the star ground of my setup, the interface freaked out as before, if I touched the servo motors while using it. Atraching the chassis of the supply to my star ground fixed this. Its all very complicated
Sometimes too much grounding can create Ground loops too - not sure how different your machine is from our stock setups, but our standard kits does not Ground the chassis etc for example - not real need for that.
I'm using an old aluminium pc case as my "control box" its divided into 3 shielded "zones" using folded aluminium sheet for vfd and cabling, watercooling for spindle, and blackbox, psu, circuit breakers etc.. I've tried to follow grounding "best practice" as far as I understand it (not hugely!). So everything is shielded including all cables, and, as far as possible, I've tried to avoid anything "loopy" and bringing everything to a single point. I appreciate the vfd in same enclosure is a no-no, but I wanted to see if it can be made to work before trying alternatives.. I like the idea of one box..
