The component distributor TTI has released its first quarter market report and the outlook is ominous: 28 passive electronic component types have increasing lead times, while 24 saw price increases. Tantalum molded chip cap lead times are now up to 32 weeks.
Lead times for most connectors remain stable, although prices are climbing. The exception is TE, whose lead times are climbing.
Memory supplies are also generally getting tighter.
With some component manufacturers now requesting 18 month forecasts, the risk for double-bookings is on the rise. Beware! Someone always gets stuck holding the bag of chips.
Most of the electronics design world is by now aware that we’re in a very serious period of components shortages. Hardest hit seem to be ceramic capacitors, but other passives as well as a variety of connectors and silicon parts are also caught up in the shortage storm. Allocation and shortages hit every few years, but this one seems to be the worst in recent memory. It could be a problem until 2020, and the supply chain and world of components manufacturers will likely be a different animal coming out of it.
So, you might ask, isn’t that just a problem for high volume producers? No, I would answer. It affects anyone regardless of volume. The exact way that it hits you and what you can do about it may differ, but it has or soon will hit all of us.
Here’s five things you can do to minimize the effects. I’m going to go backwards and starting with the most important thing for people who need low volumes manufactured:
1. Check the availability of all of your parts immediately before sending us your bill of materials.
The. very. last. thing. before sending us your BoM. It’s not uncommon for a part to be in stock one day and out the next. We’ve even seen cases where the part’s in stock in the morning and out by the afternoon. If you’re having us quote and order your parts, verify they are in stock as the last thing you do before sending your files to us.
Almost every BoM we see these days has one or more parts that are out of stock. We send you an email about the parts being out of stock. We can’t do anything else until we hear back from you. We can’t build without parts and we don’t know your design like you do, so we can’t guess at substitutions. A last-minute check can save days of delay.
2. Put one or two alternate part numbers in your BoM, especially for passives.
As I said above, we don’t know your project so we can’t pick a sub for you. Give us some alternates. Put them on the same line as the original part, to the right. And be sure to tell us in the special instructions that you’ve put alternates in the BoM.
3. Consider your parts values carefully. You may be able to pick something with better availability.
The 0.01?F capacitor is the hardest hit component. It’s the most commonly used bypass capacitor. Some designs need exactly that value, but many don’t. It may be easier to find a 0.022?F, a 0.0047?F, or something else close enough. If that’s the case, choose a close enough value that has better supply, or put one in as an alternate.
4. You might need a slight redesign to use a smaller package.
Since smaller packages can be used in more applications, many suppliers will be allocating more of their foundry capacity to smaller form factors like 0402 and 0201 sizes. Some component manufacturers have said they’ll be permanently discontinuing anything bigger than 0402 parts except when absolutely necessary.
Stick with 0402 size passives. It may be easier to find the parts you need in that package, and those size parts will be the first ones to come back in stock.
5. If we send you a message about a part we can’t find, respond as quickly as possible.
We do our best to avoid any delays in this process, but we can only do so much. Help us out by getting back to us as soon as possible, and don’t be afraid to give us more than one part number to try.
This can be a pretty annoying problem and it can cause delays and other problems. The good news is we’re having this problem because the design world is booming and technology is advancing. It will get better, and following these five tips can help prevent delays. Don’t forget to check your parts for availability right before sending your BoM in to us. I mean it!
Duane Benson Parts, parts everywhere, but not an 0805 to solder
Interesting report on counterfeit component trends, prepared by ERAI. PLICs and microprocessors are the most commonly reported counterfeited parts.
One big takeaway: “Suspect/counterfeit parts that have not been previously reported are constantly entering the electronic supply chain and the threat of encountering one of these parts remains very high.”
All that said, the number of fake parts reported is minuscule — just 774 were reported to ERAI. As epidemiologists know, the best way to reduce risk and occurrence of negative outcomes is through research and communication.
Have you purchased any electronics components lately? Have you tried and failed to do so lately? Allocation is the word of the day and substitutions are your friend.
Many, many parts are in short supply, or unavailable with extraordinarily long lead times. Sure, that happens every now and then in this industry. It’s a periodic nuisance, but what should you do for the long term? We’re are getting some interesting stories from component suppliers that might help.
What we’re hearing is that many passive manufacturers will be trying to move their customers to smaller sizes. They want to consolidate on as few packages as is possible. That means we may be seeing the end of 1206, 0805, and maybe even 0603 form factors for many passive values.
It kind of makes sense. Right now, there might be several dozen different varieties of 0.1uF, 16V capacitor. Does the industry need that? And if there isn’t enough fab capacity to make all of the variations, why not consolidate and run more of fewer variations? It won’t surprise me if we start seeing fewer voltage ranges as well. In most cases, a 16V part will be just fine if you’re calling for a 6V or 10V part.
The chip industry has been doing this for a while. Many of the newer components just come in BGA or QFN packages. Fewer and few leading edge parts come in large through-hole or SOIC packages.
Consider using smaller components, like standardizing on 0402 parts. I know it can be a pain to use smaller parts, but any potential for future proofing your design now can prevent delays or otherwise unnecessary redesign cycles. You might just be able shrink your board size and save some money on the board fab too.
Keep approved substitutions close by, and look for newer chips that are more likely to stay in production. For microcontrollers, pick parts that have multiple memory capacity or speed range variants all in the same package.
This looks to be a pretty extreme allocation cycle, and I have a feeling that the industry will be different when we come out of it.
Duane Benson Which is worse Being the missing link or the weakest link?
There are a lot of components that require special handling. Some days, “special” requirements seem more the norm than the exception. But, every now and then, we see something that puts even those special components to shame.
Not long ago, we received a parts kit that contained a component so fragile, that most of them didn’t survive the trip with the shipper. It’s a 10 x 9mm (well, actually 9.68 +0.00/- 0.08mm x 8.64 +0.00/- 0.08mm, to be precise) sensor that’s only 0.05mm thick. That’s 1/4 as thick as the diameter of the solder balls connecting it to the PCB.
The part has solder balls on the silicon, with no other packaging. The dice has to be that thin, as the light-sensitive area is on the other side. That doesn’t make for a very robust component. It would require special handling all around. Unfortunately, no matter how careful we might be, if they’re broken when we receive them, there’s not much we can do (other than take pretty pictures).
In taking these closeups, I noticed that the registration in ball placement isn’t all that great. In the image below, take a look at the ball on the left, second from the bottom, and the ball on the far right.
The datasheets call out all non-specified tolerances as +/-0.001mm. With these being 0.2mm diameter solder balls, I’d have to say this is way outside of that tolerance. I’m sure the part would have adhered to a decent board just fine, but if the PCB were off a similar amount in the opposite direction, you may very well have a problem.
Duane Benson You could make a very tiny sundial out of this. But, could you use the shadow parallax to calculate the distance to the sun?
Surface mount components are carefully packaged up in strips, tubes or trays, because they’re machine-assembled. The assembly robots need order and organization to properly do their job.
Through-hole parts, on the other hand, are almost always manually inserted by actual human-type people. That being the case, the manufacturers and distributors are sometimes more lax with their packaging. They assume that, since a human is picking the parts, a jumble is okay. Sometimes it is, but not always.
In the case of these through-hole DB-25 connectors, the jumble was too much and lead to a number of bent pins. Slightly bent pins usually aren’t a problem, but some of these ended up a lot more than “slightly” bent.
To make matters worse, these pins are small thin-wall tubes, which are more susceptible to breakage when bent than are solid wire pins. For the connector on the bottom of the image, some of the most horribly bent pins may not be straightened without breaking. If they are, they’ll certainly be weakened.
The moral of the story is that through-hole parts need care too. We can’t toss them around just because they aren’t the latest technology.
Duane Benson Spider-Pin, Spider-Pin, Does whatever a Spider-Pin does
Manufacturing, especially small volume one-time-only builds (like a prototype) is hard. It’s not wise for most people to actively seek out chaos, but that’s what we do, and we do it wisely. That’s what we’ve been doing since 2003.
We do it because it’s hard and because it’s necessary.
A big part of quality manufacturing involves risk reduction. Prototyping and quick-turns inherently add in a lot of risk. While we’ve designed our processes and systems around turning that risk into a quality product, there are a few things that you, the customer, can do to help reduce risk even further.
One of the best things you can do to reduce risk is to prepare a well organized kit, as shown in this video:
You can send us your parts in short, cut strips, like you get from Digikey or Mouser, long continuous strips, full or partial reels, tubes or trays. We machine place from all of those types of packages. What’s important is clear labeling and organization.
Individual, or mixed/loose components are not good, though. Pins get bent, leads get contaminated, values get mixed… Leave them in the strip, even if it’s short. If you’ve got multiple short strips of the same part, we can still machine place. Don’t tape them together. We can deal with them as is.
Duane Benson Peter Piper Picked a Peck of Pickeled Manufacturing
Small component packages seem to be a recurring theme with me. It’s understandable, I guess. Super tiny packages are becoming more and more common and we build a lot of product with them.
The smallest we’ve built is 0.3mm pitch. Those aren’t common enough to be considered standard — they’re still an experimental assembly — but not many chips use them yet. 0.4mm, on the other hand, is something we see on a pretty regular basis.
What’s so tough about that? The biggest challenge with these form-factors seems to be footprint design and escape routing. I can see why. There really isn’t room to follow any of the standard BGA practices. You can’t fit escape vias between the pads and you can’t put vias in the pads, unless they are filled and plated over at the board house. Filled and plated vias are the easiest way to go, but it can make for an expensive board fab.
KL03 WLCSP20 on a US Lincoln penny. One of my side projects involves trying to make the smallest possible motor driver. For this project, I’ve chosen the Allegro A3903 driver. It’s a 3 x 3mm DFN (dual flatpack no leads) with 0.5mm pitch pads and a thermal pad in the middle. The microcontroller will be the new Freescale KL03 32-bit ARM in a 1.6 x 2.0mm WLCSP (wafer level chip scale) package. It also comes in a 3 x 3 x 0.5mm pitch 16 pin QFN. Without an expensive PCB, that may be my only option.
Pick your CAD package. I’m using the newest version (5.1) of Sunstone Circuit’s CAD package, PCB123, but the principles here will apply to any CAD software. If you don’t already have a copy, download PCB123 V5.1 here.
If you’ve got fast Internet, you’re done now, so go ahead and install it. You’ll need the manual too, which you can get here.
I need to eat now, so stay tuned for Part 2.
Duane Benson Nerfvana – It’s like Nerdvana, but with more foam darts.
Excitement is building here. In a little over two weeks from today, The Hobbit movie will be released to theaters. I’m sure everyone reading here knows the story, but in case you don’t I’ll spoil it for you.
It’s a story about Biblio who is, according to Spock, the bravest little hobbit of them all (google that if you don’t get the reference. You’ll be glad you did). Biblio is minding his own open source robotics business when the Wizard of Menlo Park (in CA, not NJ) invites 12 MCU designers over for a meal and discussion about the merits of hardware peripherals vs. bit-banged peripherals. The MCU guys convince Biblio to go with them to The Lonely Mountain Chip Fab and help them kill a terrible ESD Spike problem. Actually, it’s the Wizard that convinces the MCU guys that Biblio could help. The next day the MCU folks left early and Biblio ran out to catch up with them without even an ESD smock.
The ESD problem came from the North because it’s more humid up North and that tends to dissipate ESD. Our Terrible Spike didn’t like the idea of being dissipated without having first destroyed a few gold interconnect wires. The MCU guys need those gold interconnects to remain intact, so they brought a secret encryption key and enlisted help from the technician Biblio.
First though, they had to get past the TO-92 packaged parts that wanted to squash them into jelly or tacky flux. Fortunately, despite the bumbling of technician Biblio, the Wizard bought solder with no-clean flux which made the TO-92 parts stop moving once applied. After the TO-92 parts stopped working in daylight, they made a brief stop to inspect the last Homely Chip Fab in the Silicon Valley and see where the light sensitivity came from.
Passing over the Siskiyou Mountains on the way to Oregon and The Lonely Mountain Chip Fab, it started raining so they went into a cave and ate porc for dinner. Biblio ate so much that he fell asleep in the corner behind a chair where no one could see him and his buttons popped off. The missing buttons didn’t bother him too much because those ones had a de-bounce problem anyway. Luckily, the weren’t Grayhill switches or they would have hates Biblio forever, even if he used an achient gold Tolkien-ring network to bypass more porcs.
Biblio wasn’t the most skilled technician and he caught his pine cones on fire while trying to solder new switches into place, but the wizard was able to re-layout the board using Eagle CAD and an FPGA that could take many forms and would satisfactorily control the machinery and bears at the local honey production facility. But the FPGA brought them all into the murky world of Verilog and VHDL. That would have been fine except that the search engine spiders hadn’t crawled the eleven Wikipedia pages they needed to properly map out the clock routing.
The MCU guys got hungry and wouldn’t wait for Biblio to come back with pi so they rushed in causing so much in rush current that the lights went out with a snap. After eleven clock cycles in his new hall-effect switch, Biblio knew that the de-bounce problem would be gone except when he plugged the barrel jack into his Apple computer. But with no static guards to wine too, he had no choice but to use the Apple barrel jacks to get power to MCUs and switch open the flip flop made from a streaming-transistor logic gate.
Annoyingly, they split the story in two and the movie will end at this point. We’ll have to wait another year to see if Silicon Oakensubstrate is robust enough to kill the terrible ESD spike and pass final QC.
Not long ago, I wrote about a 0.3mm pitch wafer-scale BGA we received and were asked to place. The gist of that article was that those parts are very small and we d0n’t yet have a process that we feel will give the quality, reliability and consistency that we want to deliver. That means officially, we don’t, at the moment, support that form factor.
However, as it turned out, we went ahead and built it and the x-rays all said it looked good. Whew! We still don’t officially support it, but we’re working on it. If you have one of these things, you can always give us a call and see if it’s something our manufacturing engineers are comfortable with. If they say “sure, send it in,” it will be a non-standard, essentially, experimental, operation so our normal guarantees won’t apply. It will be “we’ll do our best.”
But that’s not the point. The point is that there are still a number of unanswered questions with 0.4mm pitch, and now we have a smaller one??!!
I’ve only seen 0.3mm pitch in two places: some data from Amkor, and the datasheet for this part.The part in questions is a Maxim MAX98304 Mono 3.2 Watt Class D amplifier. The entire package is just 1 x 1mm.
There is still a lot of difference of opinion on solder mask defined (SMD) vs. non solder mask defined (NSMD) at super small pitch like this. For BGAs 0.5mm and lager, the general consensus and IPC recommendation is NSMD. At 0.4mm, the Beabgleboard folks at Ti recommend SMD to reduce bridging. But I’ve had other folks say they get good results with NSMD. For 0.4mm, we’ve had best results with SMD. It’s more than just that though, you also need to religiously follow the manufacturer’s recommended pad sizes and such.
For this part, the datasheet shows the pad size (0.18mm), but doesn’t cover the SMD vs. NSMD question. Instead, it refers to a Maxim app note (#1891) for that bit of information.
Of course, this is where it gets sticky. That app note, as of this writing, shows 0.5mm and 0.4mm, but no 0.3mm. It does reference IPC-7351, which is a very good thing, but I don’t think IPC-7351 has 0.3mm pitch covered yet. Ugh. The 0.3mm part we placed used SMD pads.
Duane Benson It’s not just Facebook where you can designate something: “It’s complicated.”