Definition of jelly bean/jellybean/jelly-bean components?

[DWSmith]

I did a search here and perhaps I didn't go down the list far enough but I didn't find a definition.  I did find definitions on DigiKey.

How is it usually being used here?

Thanks.

[ledtester]

I gave a couple of AIs this prompt:

How would you define a jelly bean electronic component?

and I think Claude gave the best answer:

A jelly bean electronic component is a low-cost, broadly available, general-purpose component for which numerous commercially available equivalents can perform essentially the same role without requiring the design to depend on unusual or tightly controlled characteristics.

The key idea isn't simply “cheap.” It's “generic and replaceable.” A $0.10 component with very unusual characteristics isn't necessarily jelly bean, while a somewhat more expensive but ubiquitous standard component might be.

For instance, you need an NPN transistor to drive a small relay. There are numerous ones that will work - e.g. 2n3904, BC547, 2N2222, ... For a lot of applications these are interchangeable.

FWIW, this seems to be Digikey's page on jellybean components:

https://forum.digikey.com/t/what-are-jellybean-electronic-components/46180

[abeyer]

Dave's take in his series:

https://youtu.be/uq1DMWtjL2U?si=pnEQpKNH0qvCJzQ1&t=60

[Sensorcat]

It looks as if Claude took Dave's definition from his video series for its answer. Both are essentially the same, but Dave was first. AKAIK, there is no common definition, everybody can use the term as they want.

When watching Dave's videos on this topic, I was confused at first, because I wondered what problem he wants to solve with it. At that time, I wanted to order a selection of 'standard parts', such that I don't have to order components in tiny quantities (value and number) again and again. Biased with this thought in mind, I assumed this was the goal. But messages on this forum later indicated that was never the case. So if you ask what the definition of jellybean might be, the obvious next question is: "Now that I know, what do I do with it?" In my humble opinion, not much. In engineering, we usually try to minimize a cost function. Even the hobbyists do that, when they ask for advice on projects or equipment purchase. But jellybean components rarely help with this optimization: If you want to minimize ...

• development cost, using higher quality and performance components usually reduces design time, test time and all kinds of problems.
• board space, using modern, higher performance components usually results in less space and component count.
• BOM cost, you would select cheap components, but not care how long they are on the market.
• procurement problems for high volume, you would talk to suppliers directly, not check stock of distributors.
• the risk your component is discontinued when you don't want to discontinue your product, you monitor if suppliers give the respective warning, then order (low volume) or talk to the suppier (high volume).

In any case, it is of little use to have a 'generic' component for a particular function in a particular commercial product. 'Generic' counts for the lab stock I mentioned above, but for this purpose, cost is not a big issue.

Example: There are certainly jellybean OpAmps, but instrumentation amplifiers not so much. If you need to make a decent InstAmp first time, you usually find out that it's not worth (any more) to make it from three jellybean OpAmps ($0.20 in a quad package) instead of using the integrated option ($1.50 typ.).

[CatalinaWOW]

I would disagree a little that there is no application for jelly bean parts. 

1.  If you are in modest scale production you usually don't have the pull with the manufacturer to assure that your selected optimum parts will stay in production or the capital to do a lifetime buy.  Jelly beans address this issue.

2.  Again, for low a moderate production, there is less product to amortize the engineering to find the perfect part, learn its peculiarities and get it into production.  Having a narrow set of always available components very noticeably reduces the design cost.  Many applications don't require a great deal of invention, merely application to a new problem set.

3.  In products which have a long fielded life, repair remains possible.  Many of the TE forum are too aware of the inability to fix a still useful piece of gear because the IC that perfectly solved a problem went out of availability a decade or more ago.  Now if your company doesn't care, or even prefers that your product doesn't stay in service this doesn't apply.

[Sensorcat]

I would disagree a little that there is no application for jelly bean parts. 

1.  If you are in modest scale production you usually don't have the pull with the manufacturer to assure that your selected optimum parts will stay in production or the capital to do a lifetime buy.  Jelly beans address this issue.

2.  Again, for low a moderate production, there is less product to amortize the engineering to find the perfect part, learn its peculiarities and get it into production.  Having a narrow set of always available components very noticeably reduces the design cost.  Many applications don't require a great deal of invention, merely application to a new problem set.

I don't see a lot of contradiction between your post and mine. I did not write that there is no application for these parts, I wrote that they do not help much with optimization. If you select a no-time-for-optimization example, it is a different case. A valid case, I would say, but not the most common. Which is opinion, of course.

3.  In products which have a long fielded life, repair remains possible.  Many of the TE forum are too aware of the inability to fix a still useful piece of gear because the IC that perfectly solved a problem went out of availability a decade or more ago.  Now if your company doesn't care, or even prefers that your product doesn't stay in service this doesn't apply.

This is usually and unfortunately not something that makes a difference in success for a manufacturer. It would be a better world if it would, but it clearly doesn't. It even does not make a difference for how popular or desirable a piece of equipment is. What do many people here on the forum would like to have on their test bench? A 3458A and a Tek 2465B, both with unobtainium inside. These are the most admired instruments, despite the difficulties to repair them.

[MariuszD]

In engineering, we usually try to minimize a cost function. Even the hobbyists do that, when they ask for advice on projects or equipment purchase.

Equipment can be expensive, I understand that someone who uses it for hobby purposes looks for compromises. I do the same, I use some tools once a month, and in that role, top-notch equipment would go to waste. But with electronic components, it's different; they are cheap, and in projects made in a single copy, the biggest cost is labor, bom is not important. I observe how people struggle who optimize the cost of components right from the start. A hobby that was supposed to bring joy becomes work for an imaginary gready boss.
I was designing for money, I had to optimize the cost. When I design for myself, I can be free from these constraints. The goal is the design process, not time to market.

But jellybean components rarely help with this optimization: If you want to minimize ...

• development cost, using higher quality and performance components usually reduces design time, test time and all kinds of problems.
• board space, using modern, higher performance components usually results in less space and component count.
• BOM cost, you would select cheap components, but not care how long they are on the market.
• procurement problems for high volume, you would talk to suppliers directly, not check stock of distributors.
• the risk your component is discontinued when you don't want to discontinue your product, you monitor if suppliers give the respective warning, then order (low volume) or talk to the suppier (high volume).

If someone wants to optimize a hobby project, it means they have long destroyed their work-life balance and treat their hobby like a job. First, he should work on himself.
I'm writing about this because I made the same mistake myself.

The fact that generic parts minimize the cost of the bill of materials is a fact. That's why we always encounter them in cheap Chinese devices with poor performance.
I just don't understand why someone would want to make  low-quality devices with mediocre parameters as a hobby. And in the end, struggle with additional difficulties during startup and repairs.
My hobby is free from enshitification.

[Solder_Junkie]

When I worked in the electronics industry for a major European manufacturer, it was surprising how many different transistors were used for more or less the same job. We were not allowed to substitute parts and had to use the exact part number as the original unless there was a design change notice for the board. As you can imagine, it caused problems for the parts department as well as service.

IMO jelly bean parts are items such as the 2N3904/3906 transistor, or their SMD version, MMBT3904/3906. These transistors have been around for many years and are available cheaply and in huge quantities from several manufacturers.