Hacker News Reader

Comment by Syonyk | Hacker News Reader

Supply chain issues are killing everyone.

You want to build a product? Great, here's the list of what you can't get this week. It's different from last week, next week will be different too. Good luck!

Parts that used to cost $0.50 are now $5+ (various ADCs and DACs come to mind as one I've been dealing with). God help you if you want trailer axles. Etc.

There's only so much you can do with flexibility and "Redesign every few weeks for what's available for this batch" before your design costs and parts cost increases chew up all the available profits, even if you've designed for flexibility. "We can use any uC in this entire line!" "That line is on 54 week lead time. Yes, all the variants."

Such is life in the decline.

View on HN

I'm a co-founder of a niche multi-effect pedalboard company and we're filing for bankruptcy in Germany this very week.

It's not only the crazy price changes and random stocking issues. Small players have zero leverage with suppliers and are a much weaker link in this whole chaos. Good luck trying to find investors willing to put money to drive growth.

Hardware startups is absolutely brutal. No surprise so many fail after a "successful" Kickstarter campaign.

I'm actually surprised we're only going under nearly 30 months after the pandemic started. I would bet that millions of people will be unemployed because a gazillion of small companies will shut off in the next 18 months.

I'll never start (or work for) a hardware company for the rest of my life. Software is so much a better business to invest that makes me cry just thinking how much money and years I lost of my life.

I'm so sorry to hear about that, it seems really frustrating. I hope things work out for you.

I fell you! I am thinking about nice 3D sensor for robotic applications. I have hardware prototype and rudimentary software. However no parts are available for sale, so I am not able to advance to production ready version. Very frustrating! Happily I can fund my development from other sources and have long runway. I dodged a large caliber bullet not going all in with this single project 18 months ago.

I don't want to deposit fear into your dreams. You should follow what your heart tells you. But not going all in was probably the smartest decision you've made. Good luck!

das tut mir wirklich leid. nur das beste für die zukunft. manchmal wird mut nicht belohnt :(

Every big fall I had ended up being a huge improvement in my life in hindsight, after a year or two. You couldn't have foreseen these issues and should feel proud of the experience you had. I hope everything works out well for you.

I appreciate the sympathy, really. The thing that I probably regret the most is really the time invested. That's the only thing I can't make up for. Even if I try to look back fondly of the learning experience I'm still severely burnt by it.

What I did in fact learn was to give absolutely tons of respect for people who pull this off. There's a lot of trash talk articles about Peloton and other struggling hardware companies but they have no clue how freaking difficult it is. Now at least I can say I do.

-- Edited for clarity

The future is still bright. I was lucky enough to have two business going and the second one is a thriving consulting software company. Sorry if it looked like I was in a dire situation personally. Not really the case.

The EEs are spinning alternative designs, the SWEs are writing cart sniping bots, the Supply Chain guys are trying to figure out if anyone with chips can be compensated for their consideration, ASML is desoldering chips from washing machines so they can make machines that make more chips, Influencers are using clout to beg for chips

https://www.youtube.com/results?search_query=chip+shortage+a...

Et tu, ATMEGA328P?

It's getting very real very fast out there.

>ASML is desoldering chips from washing machines

It's not what ASML does - ASML exec reported that some other company does this

And according to the New York Times, this is the reason that Russian troops are looting major appliances from Ukraine, and trucking them back home.

There must be some interesting chips in washing machines these days.

they want a direct line to NSA

Why not just pull the circuit boards and truck those home? Do you really have to take the rest of the machine?

Maybe they are... alternatively maybe they aren't and it's simply easier and faster to tell troops to chuck washing machines onto trucks rather than spend time dismantling them and learning how to identify the relevant parts and remove them without damage, not to mention Russia has a lot of spare fuel now.

I find the connection to acquiring electronic parta dubious. No one would transport washing machines on tanks when other transportation means exist. Plundering by badly disciplined troops looks much more likely.

I got, like, FOUR Arduinos in a bin out in the garage somewhere. I'm gonna be rich!

This is not true.

I laugh when people claim chips are desoldered by anyone at all, even Russia.

Anyone with a minimal amount of knowledge knows that it would cost tons of time and money (and be impossible for anyone not well educated/experienced in the craft) to:

1) design a new circuit board around the chip 2) print said circuit board in sufficient quantities 3) Write software (from scratch) for the chip

ASML is most certainly not involved, regardless.

I personally know people in low volume industries where they have pulled parts from prototypes, rejects and devkits to keep production moving. It absolutely gets done when it makes sense. For example with FPGA's.

I promise you that there are plenty of reasons to get a chip professionally desoldered.

Your steps 1 thru 3 could cost $100k or more, depending on the chip.

If you're selling 1,000 to 10,000 units of something in the short term, you'd definitely pay the premium for a removal and repackaging

steps 4+) get new electronics through FCC, UL, CE, etc. certification, which is slow and expensive, strongly motivating keeping the components the same

You desolder to avoid 1, 2, and 3. That's why it makes sense. Washing machine has the stupid microcontroller you need and costs half of scalper prices? Washing machine it is.

Obviously the economics only work in low volume / high NRE sectors -- like semiconductor tooling.

Yep.

We use smart PLCs (programmable logic controllers - basically an Intel Atom powered realtime platform with modular high-speed analogue/digital I/O) for a lot of our projects. For one project, we were planning on deploying a cheap PLC plus a thermocouple module and some SPI modules. Because of availability, that turned into buying thermocouple and SPI interface chips ourselves, and designing a custom PCB to interface those to a Raspberry Pi. And now of course Raspberry Pis are impossible to get, so while we managed to get hold of enough for the first run of deployments, we don't have any more of them - we even had to ship our tester Pi! It's a real scramble, and we've definitely lost a bunch of money on that project.

Do you really need multicore linux computer with gigabyte of ram to send temperature over the network?

That depends on how fast you're sending it, how you are sending it, how precise it is, and if edge calculations have to be run, and calibration workloads.

All valid points in 1985. Nowadays smallest cheapest microcontroller has more processing power than fastest desktop from 1995.

And in the last 30+ years do you think that maybe some of what we do with temperature data has changed? Or how we interact with a network? Or the data that sensors can produce?

If you think you can answer parents problems without even knowing the problem space, you're behaving foolishly.

Hey, it's alright - rasz just asked a question, and it was a fair one :)

There's asking a question to learn and there's indicating that someone did something the wrong way without understanding the problem space.

Understand the problem, then propose solutions.

Anything else is ass backwards.

OP described remote temperature sensor for a PLC. What you might want to do with data might changed, but how we gather it only got optimized down to specialized controller chips. Amplification, compensation, linearization - all taken care of by dedicated interface chips, something like AD849x, MAX6675/MAX31855 or ADS1118. There is nothing left to do at the remote end that cant be done further down the line. Measurement is pretty much Push Button, Receive Bacon to the point you can build your own soldering station with arduino, one of those chips, mosfet and a display module in 100 lines of code.

When it comes to sensors and latency, you can use a 200mhz cortex m3 with a bespoke runtime you write in a couple weeks, or you can throw a real OS on top of multiple Ghz and cores in order to make up for the overhead of having a real OS.

SRAM running at 200mhz with 1 cycle latency can perform what looks like miracles when placed side by side with DRAM running at LOLWTF Ghz that is stuck behind 100 cycles of delay talking to the CPU that first tries to work its way through multiple steps of onboard cache and lookup buffers and all sorts of fun things.

Embedded controls can do stupid amazing things with very little resources because all of a sudden you have removed a huge constant delay from all memory accesses.

Throughput, however, they don't do so well on throughput. :-D

Yep. DRAM latency is no joke -- ask your fancy 32 core 5GHz super-duper-scaler CPU to chase a linked list and it turns back into one of those old beige boxes with the turbo button, but with the turbo button not pressed!

Linked lists and SRAM are so much fun.... All sorts of data structures that are cache and dram unfriendly are a-ok in embedded land.

If I remember correctly, pressing the turbo button actually slowed down the CPU clock.

Linked lists are fine. Trees with tagged pointers less so.

Yep, that's a fair point: both the Raspberry Pis and the smart PLCs were indeed absolutely overkill for the project. However, they're what we know best, and minimising development time for a small prototype multi-site deployment was more of a concern than using expensive-ish hardware. For production, obviously we'd move to something less ridiculous.

No and you don't want it either. In lieu of reliable hardware (PLCs), it is what is needed for virtual PLCs.

I designed a small circuit back in January. I ended up going with a PIC microcontroller because it was the only thing I could find that supported USB and was in-stock. I ordered enough parts to build four, from Digikey.

Now, 6 months later, I want to build some more, and about half of the BOM is out of stock on Digikey. I was able to source most stuff on Mouser instead, but the switches I used weren't available anywhere and I had to go to a different manufacturer, driving the cost from $12 of switches per build to $18.

Huge pain in the ass, but at least I was able to get it done. I feel for the people who designed around particular parts and are now starting at 2023 "in stock" dates.

bonus fun if you're in a regulated industry and any BoM change is going to be, at the minimum, an external test-lab redoing all your certifications, and in the worst case, a brand new product you need to submit to your regulating authority.

No wonder there are some $5 MCUs going for $500+ each (and apparently selling well at those prices, even)

That’s not a bonus, that’s the entire point. Regulations hamstring companies and then deny them the ability to even comply with the regulations because other regulations have hamstrung the supply chain after regulations forced everyone onto vulnerable supply chains in the first place.

So instead of people getting rich by innovating and making other people’s lives better, the people getting rich are the ones taking bribes (“campaign contributions” they get to keep forever) for reducing the impact of those regulations.

The thing is, those regulations were put in place to solve a problem. They might not have been the best solution, but there is no way to know without going back and researching exactly what the problem was they were trying to solve.

That’s on top of the fact that we are always dealing with incomplete information.

Umm, I'm pretty OK with medical suppliers not being able to sub in no-name or grey-market substitutions that may have unknown silicon bugs. Regulation isn't always bad, and being in a regulated industry had some of these problems even before the component shortage.

Sure, subbing a MCU for an unknown supplier in a kids toy isn't likely to cause any serious problems, but in anything that can hurt or kill people it's objectively a good thing that there are regulations in place to prevent safety issues.

Ive heard the same complaints about food regulations, but it's objectively a good thing that a consumer can purchase anything (that isn't expired) from a grocery store with near certainty that it won't make them I'll. That same surety can't be had in countries with more lax food standards.

Look at one of the other child responses to the parent - someone who used to work on medical devices.

That this is heavily regulated is a good thing, because in the not so distant past we had people die due to the Therac-25.

I get wanting fewer regulations, really, I do. But just because YOU don't see the point to regulation a, b, c or d doesn't mean they're not necessary, just that you might not realise WHY they are necessary.

Problem is most of the world was in "just in time manufacturing" and having no inventory.

Then there were supply issues, and everybody is not only trying to get their hand on the parts they need, but they also try to stockpile some components just in case.

Essentially marked changed completely, and is throwing off all kinds of estimates and people are still trying to figure out what is what. Buyers are trying to stockpile as much as they can get, I know at least one company that bought run of 50k stm micros, not for any particular project, just in case they will need them. Hardware companies are claiming they are increasing capacity, but are doing it reluctantly, because they think it will be blown over, and they will have a lot of extra capacity, that they wont be able to sell.

I used to work in medical device supplier quality and incoming inspection and I am so happy I don't work there right now, it must be an absolute shit show with different/new part numbers, probably tons of first article inspections, and spec changes galore.

Reply on news.ycombinator.com