-2
\$\begingroup\$

When it comes to selecting a microcontroller, processor, or SOC, the company that produces the chip has an impact on how smoothly the project will go. Documentation, engineering support, openness & open source involvement, quality history (amount of errata/ silicon bugs), toolchain, and many other factors are involved.

Are there any good, independent, reviews or analysis for comparing processor makers on the qualities that impact how well a project goes?

I'm looking for well researched review sites, or publications, rather than the opinions of people here about any particular manufacturer.

\$\endgroup\$
5
  • \$\begingroup\$ How long they have been in the market is a big factor. \$\endgroup\$
    – Trevor_G
    Commented Oct 26, 2017 at 18:05
  • 3
    \$\begingroup\$ Basically, it's your experience or the experience of the people you work with. And how well developed is the relationship between your company and theirs. \$\endgroup\$
    – The Photon
    Commented Oct 26, 2017 at 18:11
  • 1
    \$\begingroup\$ Easy. Microchip. \$\endgroup\$
    – jonk
    Commented Oct 26, 2017 at 18:56
  • \$\begingroup\$ Whether comprehensive reviews of chip manufactures exist or not does not seem opinion based to me. If this question does not fit with electronics.stackexchange, all good. But if it does fit would one of the closers please help me edit the question? I must be missing some crucial terminology for everyone to misunderstand... \$\endgroup\$
    – hauptmech
    Commented Oct 26, 2017 at 22:57
  • \$\begingroup\$ In addition to the opinion-based aspect, there is a rule here saying that shopping recommendations are off-topic, which this would also count as. \$\endgroup\$
    – Lundin
    Commented Oct 30, 2017 at 12:55

1 Answer 1

3
\$\begingroup\$

I've voted to close this, with others, because it is admittedly "very opinion based." But I also have an opinion. ;)

I've been working with microcontrollers since the Intel 8080A with the Altair 8800 I built from a kit in early 1975 (ordered it January.) I've been working with microcontroller manufacturers as a small business measurement instrumentation manufacturer since about 1986. I started with Microchip quite early (the PIC16C54 through PIC16C57 were all that was available to people like me [less than 1 million unit orders, as I wasn't in the rice cooker business] from Microchip at the time.) I have been also buying from a variety of other suppliers, including Atmel and TI, and working through each of their FAE systems.

Microchip has always supported me quite well, regardless of my order level. They've treated me as if I were a million unit per year customer without a break. And they support their chips "forever," continuing to make them when other manufacturers had long since stopped making comparable parts. This is the same for their support tools, as well. I had a programmer tool that was 15 years past the date when they stopped making and selling it and the power switch became "flakey." I called them up in the afternoon, asking about getting a switch for it. The response was unexpected. Instead, they sent me a new (or beautifully refurbished) replacement unit that day, along with a box to return the old unit. There was no charge. And I received the replacement unit and the box the very next morning!

I have found chip bugs (in the PIC18F series, using their ICE2000 system to capture and demonstrate the bug) and had them redesign and refabricate the parts within 6 weeks, shipping me functioning units. Also, Microchip actually does fix their errata from time to time. Not like TI, who documents them but pretty much NEVER fixes them if there is any way to avoid it. (This can be a good thing or bad thing, depending on your perspective, I suppose.)

Anyway. I have not had a single instance, ever in 30 years of working with Microchip, where Microchip failed to provide beyond-expectation support. By contrast, every other company has failed me in significant ways from time to time. Where it mattered and where it cost me substantially in either time or money or both. (And Atmel more than most, by the way. Now that Microchip has bought Atmel, I'm considering using them once again. Maybe.)


There are good reasons to choose other companies. For example, you cannot find a good 16-bit ADC capable of 1 million samples per second, built into a microcontroller, other than with the SiLabs C8051F061. Not in a 32-bit CPU, not in a 16-bit CPU, just in that 8-bit thing (which includes a much-needed DMA unit.) For example, you cannot find a better clocking and timing system than the TI MSP430 B7 timer, nor a better low powered device than TI's MSP430. Microchip's so-called nanowatt devices are a pathetic far-cry, by comparison.

But when it comes to support; supporting their buyers and supporting their tools and supporting their parts.... as a matter of business relationships?? It's Microchip 100%.

\$\endgroup\$
5
  • \$\begingroup\$ +1. I've also found the Microchip support for design engineers the best I've encountered. Their FAEs are well trained, and are real company employees. They didn't farm out support to distributors, like some of their competitors did. The right people are accessible too. That, and the fact that they don't obsolete products makes them my first choice for anything that isn't a commodity item. I'm designing products now that may have over 20 years production life. Nobody else has the track record to make me feel as comfortable about that as Microchip. \$\endgroup\$ Commented Oct 26, 2017 at 20:29
  • \$\begingroup\$ Looking for comprehensive review sources rather than opinions. Not asking which chip maker is best. A source that elaborates on your comment about TI Vs microchip practices with errata, except with 8 other companies compared is the type of thing I'm looking for. \$\endgroup\$
    – hauptmech
    Commented Oct 26, 2017 at 20:32
  • \$\begingroup\$ @OlinLathrop Yes. And when other companies (Atmel) were putting smaller customers on 9 month allocation (or longer), Microchip was still treating me fine. Support matters to me more than whether or not feature A is better than feature B. There are times when I've had to select someone else -- the SiLabs ADC for example because of the competitive advantages. But if I have any reasonable option, I stay with Microchip. \$\endgroup\$
    – jonk
    Commented Oct 26, 2017 at 20:36
  • \$\begingroup\$ @hauptmech Ah. I didn't clearly get that from the question. Sorry. I'm glad I managed to include something you could use (errata fixes.) On that score, Intel also does repair errata in later revisions. TI pretty much never does. You may want to re-phrase your question better and ask it again. Perhaps it can be made to clearly avoid "opinion-based" answers. I and others here saw your question as asking for opinion. That may be more a matter of how you phrased things and something you can fix. \$\endgroup\$
    – jonk
    Commented Oct 26, 2017 at 20:38
  • 1
    \$\begingroup\$ @hauptmech Though I have to say that looking for periodicals isn't a good way. Such articles are often "bought and paid for" (I know, since I've bought and paid for them.) A researcher, an honest one, would need to carefully design the questions to be asked on the basis of research into successful and failed projects, and everything in between. From that, some self-demarking criteria might be found (how data shows natural demarcations is itself a whole study.) Those could then be used to fabricate a list. And so on. But I don't know of such. \$\endgroup\$
    – jonk
    Commented Oct 26, 2017 at 20:43

Not the answer you're looking for? Browse other questions tagged or ask your own question.