Tales From The Factories 👻

Happy Halloween everyone!

In line with last week’s Halloween celebration, we thought we’d share with you Manel’s nightmare tales during the mass production process while he has been in China. (We’ll be sharing more of his stories in the next update as well!)

The last few days have been a struggle for us, especially for Manel.

Manel has been in China since the 10th of October. This week was supposed to be the week when we finally ship the keyboards. But alas! Here we are again. Another delay. Don’t worry. Nothing too grave has happened but there were a few quality problems that just piled up and caused a 2-week delay.

The good news is that we’ve already finished packing some of the keyboards, around 350 units, into the travel case. There are still many left to be fully assembled and packed, but Manel and the team in China are pushing through making sure they ship all the units in mint condition as soon as possible.

Speaking to you all the way from China, here are Manel's TALES FROM THE FACTORIES.

TALE 1. The Metal Pin Nightmare

It was my second day China and we were scheduled to mount 4 Dygma Raise keyboards with all its manufacturing pieces. We needed this done in order to check the process properly before doing the pre-production assembly.

After assembling them, I felt that the joint between the two halves was so loose.

I got extremely worried because all the metal pins were already done. I started trying all the different combinations with the metal pins. I even ordered more high precision measurement tools to make sure all metal pins had the same dimensions. I checked the new metal pins and compared them with the approved samples we had so I could find the problem.

But the pins looked exactly the same as the approved samples! So why didn’t it feel the same? Why was the joint loose?

Finally, I found the problem. The electric screwdriver we used to fix the screws was too weak. I had to screw it by hand. Then it worked!

The metal pins were fine.

LEARNINGS:

• Don't lose your mind. Just keep looking for the answers until you find them. 
• The tightness of the screw should be stronger compared to how it was in the past now all the components are more precise. 
• Our metal pin supplier is reliable. I didn't trust him at first but he did a good job with this difficult piece.

TALE 2: The Pre-production Delusion

October 17 was the day scheduled for the pre-production assembly. That was my birthday. No one knew that. So I was expecting a nice present: 150 units perfectly assembled and ready to be shipped.

But nope! That didn’t happen. Instead, we faced a problem with the joint covers*, so we had to stop the assembly and postpone it.

*We’ll tell you all about the joint cover story in our next update!

I just felt depressed. It was so anticlimactic. I really thought this was finally it, we were finally assembling the keyboards!

We even had a third party QC come that day to be trained to do the quality control process. We trained him anyway but it wasn’t the same to train someone with an unfinished product.

We ended the day with a small celebration in a restaurant. I was still sad that the pre-production didn’t push through but I appreciated the effort of the team here.

TALE 3. The LED Driver Obsession

We spoke about this in our last update.

There was a big mistake in the supply chain. The PCBA factory told us that they had enough stock of LED drivers for our order. But they didn’t.

We needed 4,000 LED drivers (2,000 for the left PCBs and 2,000 for the right PCBs). But they only provided 2,000 LED drivers for the left and 200 for the right. So we only had 200 complete keyboards and 1,800 half keyboards.

I expected all PCBs to be done by the time I arrived to China, but no. That wasn’t the case at all. They told me they didn’t have more stock of LED drivers and didn’t know where to get more.

We went to the Shenzhen market, spoke to different suppliers and looked for the component but nobody had it. We were told that it wasn’t a common component. I felt like a wreck.

The next day, I contacted the original supplier from Taiwan and explained to them the situation. They told me that they could ship the microchips directly from their factory to us.

We received them 3 days later and mounted one PCB set for testing. It worked fine. The assembly factory started making all the right side PCBs straight away. They finished them all in 2 days.

But then, we encountered another issue. Continue to TALE 4: The Testing Jig Disaster

LEARNINGS:

• It’s important to understand the supply chain and to know the different suppliers of all components. This is basic to keep a reliable manufacturing schedule. 

TALE 4. The Testing Jig Disaster

Testing the PCBs is a slow process. It takes one day to flash 300-400 PCB sets.

As we were late because of the delay of the right side LED driver, we couldn’t flash them together with the left side.

One day, they called me saying that the testing jig was broken. I panicked.

We didn’t have a spare test jig and making one takes a lot of time. So we drove to the SMT factory. The workers were already making the PCBAs, but they were blocked and couldn’t test and flash them.

We got the testing jig and opened it. I called Matt but at that time it was 4:00am in Spain, so he replied 3 hours later.

We tried all possible solutions. We tried using the testing jig of the Neuron, we replaced parts so we could find what was broken.

Finally, we found out that it was the SD card of the Raspberry Pi that was broken. I talked with Matt and he helped us connect the testing jig remotely. He updated the program using the testing jig of the Neuron. As it wasn’t a hardware problem, it could easily be fixed.

We realised the fragility of our testing process. It was clear that this delay affected the whole production.

LEARNINGS:

• We need to create more testing jigs for the future.

TALE 5. The PET Case Dilemma

We use PET cases for the extra switches and keycaps.

In the DVT (Design Validation Test), our PET cases had a clipping mechanism to keep the lid intact. But we decided to remove it for the switch cases because it felt too tight and difficult to open.

When I arrived, I found around 1,000 switch cases with the lids not fitting properly. They were bent.

I told them this wasn’t acceptable. We got into an argument and they suggested that we simply use tape to keep the lid intact. But I was persistent in telling them that this case is supposed to be reusable so our customers could store their other switches in it. They had to redo the entire order.

Now the PET cases are better, the fit is better.

LEARNINGS:

• Always be exigent with quality mistakes. Don't accept compromises with quality.

Despite the issues with the metal pins, the joint, the LED drivers, the testing jigs, and the PET case, our smiles remain strong.

We are doing everything in our power to push through. Manel has postponed his flight back to Spain so he can continue working with the factory. It has all been really tough but we’re determined to solve everything and start shipment in the next 2-3 weeks.

As always, thank you for staying fully updated with our developments! We will have more to share next week. 

Till then, thanks for reading!

- Dygma Team