At the conclusion of our webinar, Dealing With Component Shortages That Impact Battery Packs Designs, we had several questions submitted to our presenter, Randy Ibrahim, Battery Development Consultant at Epec. We have compiled these questions into a readable format on our blog.
Q&A From Our Live Battery Webinar
Quick Links:
- I noticed that you skimmed over the topic of counterfeit parts. I would like to know how we can be certain that we didn't receive counterfeit parts from a broker.
- When you are going to have a compact and high-performance design, for example, you are going to design a mosfet based board, you may select a mosfet based on Rdson, voltage, current to have a narrower track, and smaller board, and so on. So, I think when you want to have efficient and compact design, you may not be able to change the components?
- Why use highly integrated parts to start with if you only get extra bells and whistles?
- If Epec is our CM, can we assume that you have already scrubbed our BOMs for the easy solutions, such as tube vs. reel delivery?
- How do you know that our design is using a common footprint in its layout?
- Can you please go into a little more detail on what variables need to be considered when using scrap parts?
- Wouldn't firmware have already been used instead of hardware, you know, whenever it was possible to keep the costs low anyway?
- In addition to counterfeit parts, we have had a variety of issues with trying to confirm the handling history of parts from brokers. Packages that look like they have been opened without proper ESD handling, moisture considerations, etc. Any thoughts or recommendations about this topic?
Watch the Recording Below:
Question: I noticed that you skimmed over the topic of counterfeit parts. I would like to know how we can be certain that we didn't receive counterfeit parts from a broker.
Answer: This one I probably skipped over. To be honest with you is it's because I'm still suffering from PTSD on this for a really bad experience. Sometimes it's really tough to determine if you got counterfeit parts, and I remember this kind of brings back memories of quite a while ago. I was working for another company, and we had a $3 million recall, and that's a lot, even though that's close to 20 years ago. And it was due to something really kind of stupid. It was a 3 Amp counterfeit diode, if I recall, and, the counterfeiters were pretty clever. The part looked real, it had all the correct markings, right size, markings on the reel.
It was interesting. And it passed all our production tests, of course, but we started noticing some failures out in the field, and we double-checked our design to make sure that we didn't spec in the wrong part or wrong diodes and, that would be dumb that we overlooked. We kind of found that everything was looking pretty good, so we decided to send the parts back to the manufacturer to see why they're failing and see what they could do or come up with, and we were just expecting maybe a simple X-ray, what have you. But they carefully examined the part and the factory found that the date codes did not exist in their system. So, these counterfeit parts used date codes that they never produced, which was definitely a red flag. And then also the printed logo was slightly different. It was, definitely, when we started looking at all the other parts that they produced, it was a slight fade on one side; it was not a legit logo. And, then they also decided to examine the part under X-ray and what they revealed was there's a 1 Amp diode die, which is much smaller than the 3 Amp die that would be expected in the part.
So, essentially, the counterfeiters just relabeled 1 Amp die, same packages as the 3 amp, and passed it on and grabbed some money there. Looking back, probably the takeaway here is that if there's any doubt your parts are counterfeit, obviously, use a reputable broker, but as a minimum, if you suspect anything, just contact the manufacturer to confirm the dates, date codes, in our case, the logos on the part, before using them. You can send pictures and what have you. And it takes a little time, but it can definitely cost you a massive recall that I personally experienced and still suffer today, looking back at it.
Question: When you are going to have a compact and high-performance design, for example, you are going to design a MOSFET based board, you may select a MOSFET based on Rds(on), voltage, current to have a narrower track, and smaller board, and so on. So, I think when you want to have efficient and compact design, you may not be able to change the components?
Answer: Yes. We agree that this would not work for all products and industries. In a compact design, the only possibility is to have multiple layouts available for different parts.
Question: Why use highly integrated parts to start with if you only get extra bells and whistles?
Answer: That's a fair question. A couple of things that would come to my mind would be, one thing, the highly integrated parts. They do offer less design risk since they've been fully tested at the factory, at the fab house, and, they're used in thousands of projects in different industries. So, every one of these parts is used by a lot of other customers, a lot of history on the parts, and so the risk is very, very low that these parts will misbehave.
The other thing I know is, and this even goes back to college days, now, some of the stuff we studied, but when you get everything on one die, you get better accuracy, better stability. So, these parts have much higher accuracy and better temperature stability, just the fact that they're on a single die. And, generally, they require less board real estate, especially if you're looking at redesigns and you are using discrete components, that can actually increase your board size, so they definitely have some real estate. Also, they can save on pick and place, it's just one IC versus many.
The big factor here is also time to market. There's less testing involved because there's a lot of Amp nodes, you have demo boards, what have you, so you're out of the gate really quick with these highly integrated ICs. A good example would be, you know, like, a battery charger IC. It offers a very stable DC-DC power conversion over a very wide input voltage range and also output voltage ranges. If you use a lot of discreet parts to build these DC-DC converters, sometimes it can be a little difficult to get stability with a very large variance. And voltage range is also current, so highly integrated ICs kind of eliminate all that. And as long as you stay within their expected design parameters and bulk capacities on the output, you're good to go. And also, the charge algorithm, that's also built in, all the LED indicators, that's all built in, in one IC.
So, I still use a lot of them, but the key takeaway here is if you do have to use them because you're stuck with real estate, size of product, anything, make sure you stay with a TI or an Analog/Linear Tech components because historically I've never had problems getting those parts. ST Thompson, that I used in the early '90s, I had a big problem getting parts. And they shut down a project of mine with a 52-week lead time, but TI analog, they've never done that to us, so. I'm sure once this problem that we have right now in the world is rectified, they'll get back on track and be reliable suppliers, but I still am a firm believer in highly integrated ICs where it makes sense.
Question: If Epec is our CM, can we assume that you have already scrubbed our BOMs for the easy solutions, such as tube vs. reel delivery?
Answer: Correct. We work with our assembly house CMs. They will contact us when parts are showing long lead times and we assist them in finding parts in all package types.
Question: How do you know that our design is using a common footprint in its layout?
Answer: Good question. I use Digi-Key, so I kind of cheat here; it’s easy to use. I look at the website and, a lot of times, if I recall correctly, was it package or case, something like that, versus there's a column that might say something in the word supplier in it, so it might be supplier device package or what have you. But stay away from anything that says supplier package and stay with the more generic note package case, the more industry standard. So, when you call out a package case, make sure you use the generic.
Another little thing you can do is you can click on it and there's a little search, cool little thing you can do. You can click on just that package and you say find similar parts. If you do the customer one, you'll find only that manufacturer. If you do the more generic one, you'll find a lot of different manufacturers that make that same part, so, that's also a good indicator on how you can see if you're using the right pads on your board layout.
Question: Can you please go into a little more detail on what variables need to be considered when using scrap parts?
Answer: There are a number of factors that need to be considered. Has the part actually been used or is it just a part of the manufacturing fallout for reasons other than the part itself? If the part hasn't been used and the part is kind of large with minimal solder connections to the board, the risk is very, very low in reusing that part. It's easy to desolder then resolder minimal whispering on the part, so you'd have a low risk there. But if you'd look at maybe the other extreme would be, if the part came from a return scrap and it's heavily used, let's say it's in a high-temperature power supply portion of the circuit and the part is very complex with dozens of fine pitch solder connections, or maybe a leadless package might come to mind here, there the risk could be a little higher for either a part failure because it's worn out or perhaps assembly failure because there's so many connections or buried connections. And, you don't have those really nice little perfect solder balls on the bottom of the part; those are gone. You might have some risks there with being able to regulate how much solder is on your board, so those are things that you really want to look at. But all these variables, of course, need to be factored in, you know, before using the part. You’ve got to work closely with your manufacturing team, and they'll let you know if it'll be reliable to reuse.
Question: Wouldn't firmware have already been used instead of hardware, you know, whenever it was possible to keep the costs low anyway?
Answer: I totally agree. It should be. A lot of the customers we work with on new projects and, we work with a lot of customers that are retrofitting new chemistries into old products, so we see some of this. And some of the products that have been around for a while have not taken advantage of the ever-increasing functionality of the newer controllers. There could be some arcade parts in there that just weren't available back in those days.
A good example of this, I remember back in '96 on batteries, back with some of the very first battery chargers we did, off-the-line battery chargers, the PIC controllers. I've always been a big fan of PICs, but back in those days PICs, I know the young engineers won't believe me in a million years, but, the old PICs only had 8-bit A to D, that was the technology, that was state-of-the-art, 8-bits. Now there's 16 and higher. So, and, I had to use an external TI, I2C 10-bit A to D and bring that in and write code to our PIC controller, just to get the resolution or the battery voltage we needed to properly regulate this battery charger. It was just kind of crazy. What's really kind of funny is, I was talking to a customer the other day, and they're actually still producing this product today, and how long has it been? I'm surprised they can still get parts. Maybe they can get parts easily since they're so old, right? But you know what, if that product were to be redesigned today, obviously that TI I2C part would just be removed and thrown away and a newer PIC would be selected, and all that functionality would be embedded in code.
So, there are opportunities there to eliminate parts and take away, I imagine, some of the old designs, to the engineers' defense of in the old days they didn't have those options, nowadays, there's a lot more. But definitely look back at your designs, whatever age they are. There are all those opportunities; it's something not to dismiss.
Question: In addition to counterfeit parts, we have had a variety of issues with trying to confirm the handling history of parts from brokers. Packages that look like they have been opened without proper ESD handling, moisture considerations, etc. Any thoughts or recommendations about this topic?
Answer: We have had similar experiences. These issues are a red flag, and we will reject the parts. If the customer is desperate and requires us to use the parts, then we void the warranty on that particular date code range. This is only done to help the customer.
