Project update 8 of 12
Hey everyone, it has been a long time since the last update. Please accept our apologies for this. The main reason for this delay was the fact that not much was progressing. We have been waiting for the last parts to arrive, and there wasn’t much to write about since we had no significant news to report.
This changes today - We finally have some good news! There is still a lot work to do before we can begin shipping, however, now that we have all the Glasgow parts on hand, we are getting ready for production! Read further for more details about all the different parts of the project and what lies ahead.
As you may recall, the main part that was causing us trouble was the "very special" TI voltage regulator that we use on Glasgow to drive the two VIO banks.
By sheer coincidence, a friend of the project (huge thanks tnt!!!) found a very similar part from ST. We jumped on it and ordered a bunch. The new part is not exactly the same, it has a slightly different reference voltage. Fortunately, we think that all that needs to be changed is one resistor. Still, we’re planning to test the part and make sure that it is the savior we were hoping for.
But, before we got to testing of the replacement part, we suddenly got a message from Mouser that the precious TI part shipped! And we got them delivered this week! To everyone’s huge surprise, we now have the original correct part on hand! Don’t worry, we will still use the ST part on future batches of Glasgow after we finish validation, they will not go to waste. ;)
As for the FPGAs, Lattice shipped all the remaining parts to us, and we are also good on that front.
The only part that we don’t have enough of is the TI TLV73312PQDRVRQ1 1.2V linear voltage regulator. It currently is advertised to have a delivery date of 7/11/2023 (yeah, that is end of next year). But it is not the end of the world. We do have almost 2k of the part already on hand, so we can start putting together Glasgows. In the meantime, we will make an alternative board design that can accept a replacement part. Although the TLV73312PQDRVRQ1 part uses an uncommon footprint, we should be able to replace it relatively easily with a something that will fit in the space we have allocated for it.
In short, we have all the chips, so we can start building the Glasgows you all are waiting for! :)
This is where we have some slightly less good news to report. Due to the part shortage and all the small orders we had to make, we ended up with all the parts sitting in our US office.
The original plan for this project was to manufacture Glasgows in China. This would have been affordable and relatively quick. To make that possible, we were planning to order the parts and ship them directly to the contract manufacturer in China.
The issue is that sending all the parts to China, assembling them, and shipping them back to the US will result in additional taxes being applied, both when exporting to China and when importing back to the US. With our already razor-thin margin due to parts costs increases, this is not something we can afford, and was not factored into the campaign budget.
We contacted a few assembly houses here in the US, in hopes that we can find someone who will give us a good deal and assemble the boards for us. Unfortunately, even with the best prices of friendly CMs, the prices are still so high that it would be cheaper to pay the aforementioned taxes than assemble here in the US. :(
So we are back to assembling the Glasgows ourselves. Glasgow is sufficiently complicated, in that it is on the edge of what we can assemble with our two person team and a vintage Quad 4000C PnP machine. We can do it, but it will not be a quick process. Our best guess is that after we ramp up production we will be assembling, testing, and packing on the order of 100 Glasgows a week. This means we can slowly start fulfilling your pledges. But it will still require patience as we work through fulfilling all the pledges of the campaign.
Now that we know that we don’t have to make any major PCB redesigns, we can check off the last few minor PCB to-do items and place an order for the Glasgow revC3 boards. We will send out another update when we receive the first PCBs.
Timon has used the extra time to improve and polish the enclosure. We have tested the previous design and found a few things that needed improvement.
Here is a short change log for the new version of the enclosure:
Here are some renderings of the new cases.
We had to make sure that the PCB design will not have to change before placing the next order for the cases. So, this week, we sent the design files to our enclosure manufacturer to get our product validation samples made. We will let you know when they arrive.
The USB-C cables and GPIO wire harnesses are here. We’ve placed the order for the Sync cables too, and they arrived last weekend.
We are finally moving forward with the project!
Thank you everyone for your patience, understanding, and support. But don’t expect to receive your Glasgow tomorrow. That said, at this point, if everything goes well and things move forward with the speed that we expect, we should be able to fulfill all the campaign rewards before the end of the year. This is much better than some of the delivery predictions we got for some of the parts only a month ago.
The early bird backers will be getting their hardware first, and should look forward to getting updates once the first batch is assembled. We are sorry for the protracted nature of this project, that was never part of the plan. The part shortage has been exceedingly frustrating to navigate, but we are looking forward to getting things assembled and shipped.
If you have any more questions feel free to join our Discord server or the #glasgow IRC channel. I promise we don’t bite. :)
Piotr & the whole Glasgow team