Rant - Why do commercial kit vendors design with obsolete parts?

Discussion in 'Homebrew and Kit Projects' started by K4DJM, Apr 27, 2018.

ad: L-HROutlet
ad: l-rl
ad: Left-2
ad: Subscribe
ad: L-MFJ
ad: Left-3
ad: MessiPaoloni-1
ad: l-assoc
  1. K4DJM

    K4DJM XML Subscriber QRZ Page

    I've seen this from multiple vendors. They introduce a new kit that they designed using parts from their junk-box. They've designed a nice PCB, made the kit available on their website, and then end up discontinuing the product in a short time after their junk-box stock is gone.

    There is one vendor that has a current kit for a 5W amplifier that has a date of "2016" on the PCB silkscreen. The card uses a TO-92 packaged PIN diode that went end-of-life in 2006. It was replaced by the vendor with an SOT-23 part that went EOL in 2011. The current PCB artwork has the TO-92 footprint.

    The workflow is not complicated. OK, so you've built a neat device out of your junk-box. If you think it has commercial promise, investigate the lifecycle status of the parts and where you find parts that are NRND or EOL -- make the substitutions and re-test. Part lifecycle status changes that occur after you introduce your product are headache enough. It is crazy to introduce your product with known obsolete components.

    I've got similar complaints about the handbook. For example, would you really use the 40673 FET these days, or would you use the BF992?

    /End Rant

    73,
    Don
     
  2. KV6O

    KV6O Ham Member QRZ Page

    I hear ya. Sometimes, it's not that easy....

    I built (and sell the PCB for) an Arduino shield that emulates the functionality of the Kenwood DG5 - a digital frequency readout for the old Kenwood TS-520S hybrid. It uses the Arduino Duemilanove, which was replaced by the Uno some years back. The problem is, the Duemilanove uses a proper 16Mhz crystal for it's clock, the Uno uses a ceramic resonator. If you're measuring frequency, you want your clock to be reasonably stable - not worse than what your trying to measure. The Duemilanove boards can still be found, but it's getting harder and harder.

    Since the replacement doesn't meet the needs, what do I do? I have been contemplating just making a single board version with the Atmel 328 on board with a proper crystal, but that's going to take a bit of work.

    Sometimes, it is what it is.
     
  3. K6BSU

    K6BSU Ham Member QRZ Page

    Yes, indeed! Think of the many thousands of transistor part numbers. If not wildly popular, they soon go obsolete. Nowadays, the trick is locating some equivalent. That's not so tough to do, once the key parameters are known.

    Manufacturers often buy some particular part by the tens of thousands. They negotiate their own specs, usually to eliminate some parameter which drives up the mfg and testing costs. Since the part may not meet all specs not needed, it gets a "house" part number. Older radios, no longer supported by the mfg, have these parts which cannot be identified, except by examining the function and choosing an equal part.
     
  4. K7TRF

    K7TRF Premium Subscriber QRZ Page

    All very good points.

    FWIW, this also happens in industry where some companies don't put enough emphasis on supply chain considerations and component engineering. My first engineering job out of college was managing a component engineering group at a national research lab. Every day we dealt with engineers specifying obsolete or end of life cycle components but getting them to change their designs was like pulling teeth. Folks often design with parts they know well and have used in previous designs (or copied from other's designs) and it can be really hard to get them to look at newer parts that they don't understand as well.

    Frustrating for sure, but really common even outside the kit world.
     
    KB4QAA likes this.
  5. KL7KN

    KL7KN Ham Member QRZ Page

    It's not just kits. There are many "mainline" radios that are EOL due to parts - or lack thereof. The SGC SG-2020 was just recently such a victim. As technology advances, many get left in the dust...

    There is also the problem of MCU (computer chips) that are a specialty item with very low production rate. Once gone, they will never been seen again. Some vendors will not sell replacement MCUs for fear of piracy.

    So - what should the 'lifespan' or a radio really be?

    I recently purchased a DOA HW-7 and go it running again, using cross-reference transistors. This came out in 1972. Is this enough of a lifespan?
     
  6. K4DJM

    K4DJM XML Subscriber QRZ Page

  7. K4DJM

    K4DJM XML Subscriber QRZ Page

    Yes, part lifecycle is always an issue -- but when those mainline radios were released, the parts were not end-of-life ALREADY, were they? This is my sore point...

    Would you design a production PCB in 2016 that uses a part that went EOL in 2006??? Home project PCB with a quantity of 1 to 10 maybe, but to sell as a commercial product? Madness...

    73,
    Don
     
  8. K6CLS

    K6CLS Ham Member QRZ Page

    The latest QEX magazine has an article that uses a 2N4124 and a 1N645. Just for a test jig, but really?!!??
     
  9. KV6O

    KV6O Ham Member QRZ Page

    Unfortunately, the crystal is used for the Atmega 16U2 chip which provides the USB connectivity - not for the 328. The Crystal is needed due to the critical USB timing requirements.
     
  10. WA6MHZ

    WA6MHZ Subscriber QRZ Page

    That problem is not only in the Amateur radio industry, but also Commercial electronics. The Company I worked for the last 14 years or so made video equipment. Many products were discontinued due to the parts going obsolete and unobtanium! In fact, the one KILLER was the loss of access to a certain Microprocessor which is now obsolete. the Supposed replacement IC didn't work like the old one did. So that product is now KAPUT!!! It was also the last hope for the company, so now the COMPANY IS KAPUT!!!
     
    KC8YLT likes this.

Share This Page