It also assumes that work is the point.

It shouldn't be.

We should be trying to find ways for people to work less (and maintain or preferably improve their standard of living).

#Elon should think over his
#Work-#Life-@Balance.
It's definetely out of #Tune.
#Grimes would agree.

#elonmusk

🕯🖤🕯🖤🕯
#keinvergessen
#KämpfenBeginntMitErinnern
#niewiederfaschismus

#niewiederistjetzt ‼️

#AllezusammengegendenFaschismus
#afdverbotsverfahrenjetzt
#stopptDieBrandstifter
#WirSindDieBrandmauer

Chanita Szklarczyk, we remember you.

#Shoah #ChanitaSzklarczyk

I have enough (even more expired) epoxy from another batch to do one more embed but after that I'm thinking of switching to something else that has better shelf life.

Ordered some of tedpella.com/Material-Sciences… and we'll see if I get decent results with it. 5-8 minute cure time, semi transparent, and specifically lists PCB and component cross sections as typical uses. Seems like a good option.

Sample was still a little soft so I did a 5 minute post cure at 100C (the datasheet recommended cure).

It seems like it might have gone a little exothermic lol. Not sure what happened, it should have been almost fully reacted.

Despite the discoloration the resin is rock hard now so I'm good to go.

The back side isn't fully covering the chip so I'm gonna quickly hand sand it flat. Shoulda used the last 2.5g of resin oh well.

Here's after a few minutes of P180 and P400 abrasive.

I'm up to the die attach paddle and some bond wires are visible, but not at the silicon yet. Debating if I should switch to P800 now or go a bit further on the 400.

Whatever, I'm just messing around. Might as well go slow and jump to 800. But if I wreck the sample its not a big deal.

Went to P800 but it's taking a long time. I don't know how big the die is compared to the paddle.

I'm not in a rush, gonna go find some breakfast and then run another few minutes on the P800 and see how it goes. Still getting a feel for the tool and nowhere near ready to actually have an optimized process or anything.

Ok, yeah the die is a lot smaller than the paddle. Got pulled away to do other stuff but I ran a good 10 minutes at P400 and I'm still not to the die. Gonna stay on the 400 until I start seeing the fillet of die attach adhesive and then start going smaller.

Also going to try measuring my rate of removal so I can better gauge progress. As of now the entire sample is 7.08mm thick on my caliper (I need something with more sig figs...) so let's see where it is after another round of polishing.

And I measure 7.11 mm after two more minutes of polishing. This caliper is clearly not repeatable enough for my needs... I should really just bite the bullet and buy a nice Mitutoyo micrometer or something.

But look at that! I was actually closer than I thought. We're starting to see some die attach now.

Gonna call this it for the P400 then do a minute or two at P800 and see how close I get to the edge of the die. Ultimately the goal is to get a mirror finish, stepping through all of the grit sizes, and end up right down the center of the ball bonds.

A minute at P800 and the die attach is much more visible and the big scratches are gone. Can't see silicon yet, but it can't be far now. Probably tens of microns left to go.

Let's keep doing this a minute at a time and see how it goes.

This is slow going because I'm doing short cycles followed by cleaning and imaging as I get the feel for the tool, plus I keep running around doing other chores etc.

I also have no intuition for process parameters or speeds yet. There's a little skidding or chattering at times which I suspect is fixable with a feed rate or down force change, but the orbital motion is different from the smooth rotary behavior of the MultiPrep I'm used to. I'm also using SiC paper rather than diamond lapping film so even for the same pressure and feed rate the material removal rate will be different.

Seems I had worn out the abrasive. Put on a new P800 pad and a two minute cycle put me into the die already.

Time to step down a lot. Skipping P1200 and going to P2500 for a short (maybe 15 sec) cycle and we'll see what that does.

Switched to the Labsmore scope since I'm hitting the limits of the Leica.

Here's Mitutoyo plan apo 5x and 20x views. Same sample as last post, no additional polishing.

Turns out no, I jumped down in size too soon. The 2500 was taking forever to get me to the target plane.

Went back to P1200 for a bit until I was touching the edges of the ball bonds and only then to 2500 to polish it up a bit more.

Now I'm exactly at the plane I want to be and just need to get all the surface defects out.

Running a cycle with 3 μm diamond to see how that goes.