Відео

Thank you :) I want to understand the charging and charge measuring process better. A year ago I tried to just put the piezo nozzle, charge electrode and charge sensor together and tried to measure the charge, but it did not work very well and when I broke one amplifier I also saw that the result was not repeatable with other amplifiers. Now I want to try it step by step, starting with large drops at low frequency and a simple setup. My plan is to build my own amplifier next for measuring these drops and if everything works well I want to increase the drop frequency and reduce the drop diameter until ultimately ending up at 0.1mm drops at 40000 to 60000 drops per second for the CIJ printer. What I learned so far is that shielding seems to be very important to not pick up noise like the mains frequency with your charge sensor on the way to the amplifier.

@@dominik_meffert DIY MJF printing here we go!

Hi Michael, I'm trying to figure out what I have to pay attention to when charging and measuring drops. At the moment I try out a bunch of AD620 amplifier modules to make sure that the reading is repeatable, because last year I did a similar test and I later found out that I only got the reading with one of the amplifier boards (after I broke it 🤦). When I can get the same readings with all boards I want to try building a more optimized amplifier, try out different charge sensor designs and later I want to try out detecting smaller drops with higher frequency for finally getting a reliable charge sensor for my CIJ project.

Anything electrically conductive can cut, even graphite, I wish I had one

The best machine on UA-cam.

Nice :)

Maybe it could work, but you would need more powerful piezos or maybe something driven by an electromagnet, since solder paste has a much higher viscosity than water + ink.

Thank you :) Here is the project on Hackaday: hackaday.io/project/167446-diy-inkjet-printer

Hi, thank you :) I just learned about it while doing it. When I find a project I want to do, I search online for everything I can find about it and try to understand it well enough to get it done. I'm really not that smart, just motivated 😅

If you want you can use a hollow brass or copper tube for EDM drilling which has the benefit that it is possible to pump water through it for flushing out the metal particles that are created during cutting. You can also use thin sheet metal plates for cutting.

It's an early version of a DIY piezo inkjet printer. Here is the project on Hackaday: hackaday.io/project/167446-diy-inkjet-printer

Yes, I want to build something like a small hydraulic setup with this valve keeping the pressure stable while the pump can continuously run. The unused ink should get feed back into the reservoir tank as well as the return line from the printhead.

For this circuit I used an ACS712 current sensor for detecting when the electrode touches the workpiece, instead of using a voltage divider for that. I used stepper motors for the X and Y axis and another one similar to a 3D printer's direct extruder for moving the electrode. The Arduino Mega is running Marlin for driving the stepper motors and receiving GCODE and the Arduino Nano is just used for generating a square wave for the MOSFET of the EDM circuit.

Danke 😁 Ich habe mal Schreiner gelernt und arbeite momentan als Lackierer (kein Scherz) 😅

Hi, the ink stream is broken up by a piezo on the nozzle into droplets which are evenly spaced apart so that they have no electrical connection to each other. The nozzle has a ground connection so that the ink stream is at ground potential until it breaks up into droplets. At the position where it breaks up there is an electrode (brass block with a slit) located. When the droplet breaks loose from the ink stream it gets a negative electrostatic charge from a positive voltage (312V in this test). When the charged droplet now passes the angled high voltage electrode (with 4.4kV in this test) it gets attracted to it and leaves it's straight flight parth. In this test I flipped the switch that powers the 312V supply to show that the stream of charge ink droplets gets deflected. The ultimate goal would now be to deflect only single droplets with exactly the needed charge to form characters and pattern for printing, but there is more work to do to make this happen. Here in the details section is a diagram of it: hackaday.io/project/178762-cij-printer Thanks for your interest :)

@@dominik_meffertthanks for your explain

I'm just guessing, but I think he's trying to make an inkjet 3d printer... no idea really. It's an interesting idea.

Hi @OfficialSeccour, I haven't dared to make commentated videos, yet, because I have not much experience in speaking English, so it would likely sound weird. But maybe I can make a project summary with a text to speech app for the next major update. Until then I write a buildlog for every update on Hackaday: hackaday.io/project/178762-continuous-inkjet-printing/details Thank you for your interest in my project :)

Hi @Unmannedair, I'm working on a Continuous Inkjet Printer like they use in the industry for printing the expiry date on food and doing other product labeling. The goal of the project is to build a useful (fast + good quality) inkjet printer at home that needs no parts from commercial inkjet printers/printheads so that you can build your own inkjet projects like part labeling machines or binderjet/polyjet 3D printers as private or even commercial project based on it.

​@@dominik_meffert that's fair. Also don't worry about your spoken English you can only progress as you try :) My only issue with the updates is that it's a bit difficult (at least for me) to really link what I see on the video and the things you detailed in writing. Maybe inserting more pictures next to the relevant text would make things easier ? Anyway I like your project as I'm planning to either modify or current printer or build one from scratch to print on our packagings

I managed to cut a 20mm aluminum bar with a wire EDM setup, but it took around an hour. I got the best results in 1mm and 2mm aluminum sheets.

@@dominik_meffert Nice!

Here it is: hackaday.io/project/181551-sinker-edm-machine

Hi, here is the project on Hackaday: hackaday.io/project/167446-diy-inkjet-printer

Thank you :) I tried capturing a feedback signal from the stream of ink droplets. The piezo breaks the stream of ink into equal sized droplets and a test signal is applied to the charge electrode which these droplets pass after breakup. The test signal are 16 pulses with each having a different phase shift to the piezo drive signal. The purpose of this signal is to find out which of the 16 phase shift settings leads to the highest charge on the droplet. When a charged droplet passes the phase detection electrode there is a pulse visible on the oscilloscope. It's alway 16 readings/droplets and a pause. The yellow signal is only for triggering. In this video I tried changing the output voltage of the piezo driver with a potentiometer and saw that it also changes the feedback signal. So, it changed which phase setting gives the best droplet charge. I still have to work a lot on this... Normally the feedback signal should only be positive and also more stable 😅

Hi, no I'm still working at it. Last week I replaced all lines, valves and fittings with either stainless steel, rubber or plastic to prevent corrosion which was a big problem with brass and copper. I'm currently testing ink made out of Ethanol + PVB which has a higher viscosity than pure ethanol (like a clear varnish). The higher viscosity ink required some modifications, but now it's working and the ink can be used to actually print things that will not evaporate (like pure ethanol). Next, I want to work on the viscosimeter for keeping the viscosity constant at all times. Then on a corrosion resistant printhead design, the piezo driver, the charging + phase detection electronics, the code for phase detection and the code for actual printing. I guess it will still take a long time until this project is finished.

@@dominik_meffert perfect. I will wait until your printer finished

It's going to be a CIJ printer when finished. The tubes are lines for ink, vacuum and pressurized air. The ink flows in a cycle across the system and only some droplets will be deflected out of the ink stream for printing. The ink stream is ejected from the nozzle and enters a vacuum tube to be recycled over and over again.

@@dominik_meffert ah one of those! They use electric charge and magnets to deflect them right, almost like a CRT Display but ink, right? You have a conveyor belt of things being made for it to label too? Hyped to see how this goes!

Yes, thats right. The machine I'm building uses a piezo at the nozzle to split the ink stream into equal sized droplets which then get a charge to deterimine their deflection angle. After charging they pass a high voltage electrode that attracts them out of the straight flight path of the ink stream. When everything works I can try out labeling with it first and if that works I want to use it as a printhead for 3D printing on powder.

Thank you :) Yes. I'm currently working on an improved version that uses a pump for lifting the ball instead of the stepper motor.

Thank you :)

Thank you very much :) Nice idea. At the time I built this project I did know almost nothing about electronics and I think with a better built power supply + the stirring tool it could work much better.

Hm, do you mean so that it appears to windows like a real desktop printer or just a software that operates the machine like a printer without the need of writing a windows compatible printer driver?

@@dominik_meffert both ideas are great. Mainly my idea was, to improve efficiency by forwarding paper only in one direction. Easyest would be, I guess, to sort the point locations in x, y axis and let it run then in g code. But a windows pri ter compatibility wouöd be great. Next step would be to reduce the droplets and gwt an open source printer going. Everyvody is building 3d printer and cnc machines. Nobody is getting rid of those drm infested printers.

I'm currently working on a CIJ Printer prototype. If I can get it to work it will get some software for printing all sort of things with. Unfortunately, the machine will be far more complicated and expensive than the piezo pump printer, but in return it should be possible to make it more precise, reliable and faster.

Yes it is

It's deionized water for flushing

I hope it's not a motor, because I'm trying to measure the ink stream and the yellow line is only there with an active ink stream and piezo. The only motor on this machine is the vacuum pump which is always running.

@@dominik_meffert its a piezo? the sensor? maybe wind or realy the ink is being measured !!

​@@daviperes Here is another test: ua-cam.com/video/OI_NzipObTI/v-deo.html The ink droplets are measured (as far as I know) by capacitive coupling when they pass the sensor.

I tried to split a stream of ink into droplets, then charge them and try to measure the charge they receive when they pass a probe.

@@dominik_meffert very hard to think on. the droplet is flying? or is that 0.1mm on wire? if it is on the wire, you measure the whole wire, in the air maybe only can be measured if they do not stick on wall?. sry

I'm currently working on a Continuous Inkjet Printer like the ones who print the expiration date on food. The ink stream flies from the nozzle into a vacuum return line for recirculation if it is not used for printing, so there is an inkstream present at any time. The 0.1mm stream is broken up into equal sized droplets which get charged by a testing signal (blue waveform) to find the perfect "phase" for charging. The yellow waveform is the signal from the charged droplets. I just took another video where the signal is much cleaner: ua-cam.com/video/BN4DvKIaSE4/v-deo.html

It's a FNIRSI-1013D

The fluid is set to a pressure of 40 psi before the nozzle and is pressurized by pressurized air.

Maybe they will sell it or the successor of it in europe in the future. The not shorted output voltage is around 72V.

I'm currently working on another DIY printer project that aims to get a higher resolution in the end: hackaday.io/project/178762-cij-printer Unfortunately it will be a lot more complex than the single piezo pump printhead. Maybe moving the piezo pumps directly to the printhead and then using a smaller nozzle and a well adjusted pulse could make a higher resolution possible.

​@@dominik_meffert how hard would it be to just gut a working printer and control with something like a raspberry pi?

Thank you :) In this video I tried to charge the ink stream by turning on the 312V, so that it gets attracted by the high voltage electrode for altering it's flight parth.