Bent and broken nozzle tips

[michael]

I've bent three nozzle tips today during:

• Printing the precircuit curve and line test
• Printing a backlash Vernier scale
• Printing my own circuit half way

I'm worried about this, because each time I put in a new nozzle tip five minutes later it's bent. I'm not touching the tips (no cleaning) and I'm not removing the cartridge except to replace bent nozzles. My configuration is default (from restore defaults) so print height is .1mm for conductive ink. I've used -2 to -8 for ink level, which is not too much.

Suspicion

I suspect that the machine is bending the nozzles during the calibration because the nozzle tip is then lightly pushed against the two pressure sensors. There's no place on the PCB that the tips are making contact with, like a 2 ounce copper pad or hole.

Question

What steps should I take to avoid more bent (and eventually broken when I try to straighten them) tips?

[michael]

Errata

The suspicion of the second pressure sensor (x and y axis detection) bending a tip may be dismissed after advice from Voltera folks. The lateral pressure from the tip striking the soft plastic is not enough to bend metal.

Cleaning

A factor worth investigating is if the first z axis spring loaded pressure sensor is dirty enough (more than the print height .1mm) to cause the probe to report a differing value relative to the conductive ink nozzle. The nozzle tip might strike parts of the substrate then.

Advice includes lightly scraping with plastic or metal if alcohol doesn't remove the residue, and it's very important to keep the z axis spring sensor clean but not important to keep the x and z sensors clean.

Exotic features

If you have exotic features like my design with traces near the PCB edge, probing and selective printing with uneven or missing brackets, tape over PTH holes, and other nonsense, they are all candidates for problems leading to bent nozzle tips.

All it takes it .1mm or less to bend a nozzle!

Work around

A good idea to test is if a completely broken off tip can be manipulated (print height and dispense) to print basic circuits without fine pitch features.

[michael]

Update

The weekend is drawing to a close and I'm still getting bent tips during conductive ink printing, however I've reduced the problem to a 5-10% bend angle by superclamping with brackets. A slightly bent tip will still produce a good print, even for fine pitch traces and pads. Aligning to predrilled holes with less than perfectly straight tips is a problem, but more manageable than broken tips.

[jazozaya]

Hi @michael,

The machine should not bend the nozzle tips in normal operation. If you find the tips are breaking, then something wonky is going on. If you could pinpoint when they break we'd be able to diagnose better.

If the tips bend because they are printing too close to the surface, then it implies the height calibration is wrong. Could be caused by:
- A dirty bottom Z axis switch.
- A dirty gold pad (X and Y switches)
- A substrate that wasn't clamped down properly.
- A defective probe (unlikely)

The switches should definitely be inspected for any dried ink and cleaned before a print. We find a Q-tip with some rubbing alcohol works well. Out of curiosity, in your machine, is the nozzle contacting the bottom Z switch in the center of the cylinder or off to a side?

Can you confirm your print height is 0.100 mm? This is the default setting.

A good idea to test is if a completely broken off tip can be manipulated (print height and dispense) to print basic circuits without fine pitch features.

A completely broken off tip can work... to a degree. Ink will still flow but your resolution will go out the window. Also the ink flow may slow down during a print and you may need to periodically click the + button. If you are only working through hole components then this might be fine.

[dcahill-dtx]

We were also regularly bending tips, as per jazozaya's comments our issue was build up of paste / ink on the z-axis switch.

[michael]

build up of paste / ink on the z-axis switch

I don't think this was the problem in my case, although I've taken the advice anyway and clean the spring loaded metal contact sensor after each use. I shine a flashlight onto the surface and compare it's appearance before and after swabbing it with a small paper towel, alcohol, and the end of a plastic tooth tool.

This method soaks less alcohol than a qtip and I can apply more pressure to the metal surface. The first time I cleaned conductive ink from it it didn't work at all with a qtip, due to the chemical binding of the ink with the sensor's metal surface.

[michael]

Thanks @jazozaya for the advice, as well as your colleagues who have helped.

---

In your machine, is the nozzle contacting the bottom Z switch in the center of the cylinder or off to a side?

The nozzle contacts off to a side. In fact during the probe, ink, and solder steps, I think the nozzle contacts the bottom Z switch in slightly different locations. I'm not certain about this because I haven't compared photos, but I wonder if it's intentional that the tip contacts in semi random places in order to avoid build up of residue?

Can you confirm your print height is 0.100 mm?

Yes, I confirm it is .1 mm. I've done almost nothing with the configuration to try to solve the bent nozzle tip problem. Once I experimented with the height for a unrelated reason (to learn) but I've since clicked 'restore defaults' at least on two occasions to be sure.

I changed the backlash value out of desperation but have since reset backlash to zero because any other value caused poor alignment.

---

It seems the list of bent tip causes is quite long:

• A dirty bottom Z axis switch.
• A dirty gold pad (X and Y switches)
• A substrate that wasn't clamped down properly
• Removing and replacing the cartridge lid cap
• Particle in the ink or paste blocking the nozzle
• Prefabricated PCBs with varying PTH and pad heights
• Working too fast to excercise extreme caution

...and my gut feeling is that there might be a few other causes. I've implemented about half of a dozen ideas which your colleague offered by phone. It seems that some (or one?) of them helped a little. Probably slowing my work (like making dozens of photos at several steps) has helped too.

It just takes 1/10 millimeter of error for the tip to strike substrate or metal, which is a key thing to keep in mind.

[michael]

Clogged particles

I just had another nozzle tip bend, this time due to the particle in the ink or paste blocking the nozzle cause. I'm not sure why a clogged nozzle would bend the tip, but I confirmed this is true during the 'prime solder nozzle' step of the workflow in any case.

1. I took a nearly new nozzle mounted cartridge to the light and confirmed that the tip was straight (after being very careful removing the cartrdige cap)
2. I rotated the gear counter clockwise to force paste out the tip
3. The gear met with resistance, telling me there was a problem
4. Instead of continuing, I reversed the direction 1/4 turn clockwise
5. I put the nozzle against the light again and saw that the tip was bent

Question

...so I guess I didn't do anything wrong this time, but just got unlucky that a irregular (too large) particle was in the solder paste today. Does that sound like what probably happened?

---

Afterword

By the way, I've since spent a couple hours with the tip in nail polish remover and stirring the paste with a sewing needle but nothing has helped and the particle has not come dislodged. No paste is flowing through this nozzle.

I have managed to correct the bend however, something that rarely works but this time it did.

[michael]

Using broken nozzles

While developing a strategy for this problem, I started using a nozzle with completely broken tip in solder application of 1206 sized parts. This came from good advice from a friendly Volteran, and it seems that

broken nozzles work just fine on larger components and maybe even SOIC with a bit of trickery.

I'm not testing this method yet with conductive ink, which might be a much more difficult strategy to develop.

Sadly now that I'm comfortable with a broken tip, my nozzle has clogged and is no longer usable even in it's broken state.

[mewertow]

Hi @michael,

It's very unlikely that the nozzle could bend because of a particle size problem. You would have ink breach the piston and spill out before you could apply enough pressure to damage a nozzle - you may have missed it on the first check, or perhaps accidentally bent the nozzle against something while rotating the gear.

If you're not getting any flow through the nozzle, it could either be a particle issue (which can happen with old solder paste - this is the main reason why the paste has an expiration date), or that the nozzle was pinched when it bent.

• Matt

[michael]

@mewertow I'm equally worried about the blocked nozzle, and your theory could be right. In the case of my broken nozzle tip, it is completely gone but may be bent at the base or something.

It was flowing fine until I clamped a second PCB. It also seems to not be completely blocked, and I've been able to complete the half finished board by waiting ten minutes and getting just a few pads done after that until the nozzle blocks.

Good to know that the tips are not so fragile to bend from paste pressure or particle blockage, thanks for the reassurance.

[michael]

Might be a crazy idea, but new nozzles are being offered on the store. They are finer and finer (150 to 100 micron) but what about the value of a thicker nozzle?

I’m getting the impression that sloppy paste application doesn’t ruin circuits like I thought, because the solder mask repels overflow. Maybe a certain degree of sloppiness is even okay for QFN and fine pitch, can somebody confirm?

If this impression is true, then we could use a fat nozzle tip for solder which never or almost never blocks.