feature article
Subscribe Now

Didja Hear the One About the 3¢ MCU?

Really Small Microcontroller is Cheaper Than its Packaging

“In my day we only had 1’s and 0’s. And sometimes we ran out of 0’s!” – old programmer

Cognitive dissonance means entertaining two conflicting ideas at the same time. Your head hurts. It doesn’t make sense. Some things just should not be. 

Take the PMS150C, a microcontroller chip from Padauk, which operates around the corner from a Starbucks and the Go Eat Tapas Bar & Restaurant on the busy Gongdaowu Road in downtown Hsinchu City, Taiwan. It doesn’t seem like it should be there. 

I don’t mean the company. The company is real enough. They’re probably nice people. But their products? They make my brain hurt. 

The aforementioned PMS150C is an 8-bit microcontroller that sells for $0.03. That’s three cents. As in, pennies. That’s not even the mega-volume price you get when you order a million units. That’s for one single chip. Presumably, they get even cheaper in volume. Which is hard to imagine. 

Postage is more expensive than that. A first-class stamp in the United States is currently $0.55. You could buy 17 microcontrollers and still have change left over. Cheap penny candy costs more than this MCU. It’s literally cheaper than dirt

And the weird part is, it’s not a terrible MCU. Sure, it’s on the rudimentary side, but it’s also useful, like a hammer. It runs at 16 MHz (sort of), it has internal RAM and ROM, and you can program it in C. It plays near the same ballpark as a low-end Microchip PIC or AVR, but weirder and cheaper. Those chips are priced in double digits of pennies! Scandalous. 

According to the unevenly translated website, the PMS150C is but one of a family of MCUs from Padauk. Some have more internal ROM than others. Some have more RAM. Some have A/D converters, extra I/O pins, or a hardware multiplier. Oh… you thought hardware multiplication was standard? No, no, that’s an optional feature. Without it, you get to do integer multiplication the old-fashioned way: by shifting and rotating bits. 

Prices range from a low of $0.0318 for the PMS150C up to a whopping $0.1621 for the apparently gold-plated PMS232. I love prices with four significant digits after the decimal point. 

Coincidentally, the PMS150C compares favorably with the Apollo 11 Guidance Computer we reviewed last week. It’s got a similar instruction set and programmer’s model, but with a much faster clock. Apart from the speed difference, their capabilities are remarkably similar. 

The PMS150C and its siblings have no onboard peripherals, unlike most MCUs. It’s just the processor, RAM, ROM, a couple timers, and the I/O pins. That’s it. There is no UART, no SPI, no fancy LCD controller. If you want the pins to wiggle, you’ve got to wiggle them yourself. The PMS150C is for real programmers who hand-craft their own 1’s and 0’s. Get ready to roll up your sleeves. 

In fact, the instruction set seems like it was designed specifically for bit-banging the I/O pins. There’s the usual assortment of bitwise logic instructions to toggle bits, but also a few interesting delay and timer instructions that ought to make it easier to create software-only timing loops. All instructions execute in one cycle. It’s also easy to put the MCU to sleep and have it wake on any I/O pin activity. In a sense, all the pins are interrupt pins. You could pretty easily create your own UART or PWM or other serial interface just through software. 

There’s a set of shadow registers that allow you to preserve machine state before an interrupt or subroutine call. There’s a stack pointer, so you can find your way back. The machine treats memory and I/O as separate address spaces, so it’s just like an x86! Registers can be used as indirect address pointers, so it can do C-style pointer dereferencing. The company even provides its own IDE, complete with C compiler. The whole download takes about two seconds. 

The onboard ROM is one-time programmable (OTP), so be careful what you choose to burn into it; there’s no going back. (Some of its siblings have flash.) But, honestly, a 3-cent microcontroller is cheaper than any flash chip, so you’re better off replacing the whole MCU when it’s time for an update. Kind of like buying a new car when the ashtray gets full. 

The PMS150C is just about the cheapest surface-mount component I’ve ever seen, regardless of functionality. There are resistors more expensive than that. I have trouble understanding how Padauk is able to design, manufacture, package, and sell its MCUs so cheaply, but I guess that’s not my problem. At that price, you could replace almost any component, active or passive, with a programmable MCU. 

An MCU this cheap turns the normal economics of R&D on its head. In production, the unit cost of the component is negligible. It doesn’t really matter if you’re buying 100 MCUs or 100,000 of them. What does cost money are the development tools and the engineering time. That, more than any hardware cost, will determine the ultimate profitability of the project. In effect, all the cost is up front, in development. After that, it’s nearly free. 

As long as you don’t try to mail it anywhere.

6 thoughts on “Didja Hear the One About the 3¢ MCU?”

    1. @celsokenm:
      Creepy ? No – it’s great. I was waiting for somethingn like this for a long time.

      Cores ( at least 8-bit ones AFAICT) are trimed-down cross between PIC and 6502.
      It makes perfect sense to have many of them, most of them running small loops – tasklets.

      This makes it possible to do many synchronized things at the same time, without fiddling with interrupts, registers & PC save/reload etc etc.

      Great stuff.

  1. BTW, they have interesti8ng small FLASH-based 4-“core” parts in the catalogue, but not yet on the webpage.

    It seems they are to be presented soon…

  2. BTW2: A cole cents for micro is not unheard of.

    I think some big players have itnrduced a few names in that price bracket, but that was for high-volume orders.

    Only substatnial change here is that this one serves hign-volumes in CHina, but would like to expand into the world, hence the offer – to get known through hobbyists an small/meium shops into circulation.

Leave a Reply

featured blogs
Apr 19, 2024
In today's rapidly evolving digital landscape, staying at the cutting edge is crucial to success. For MaxLinear, bridging the gap between firmware and hardware development has been pivotal. All of the company's products solve critical communication and high-frequency analysis...
Apr 18, 2024
Are you ready for a revolution in robotic technology (as opposed to a robotic revolution, of course)?...
Apr 18, 2024
See how Cisco accelerates library characterization and chip design with our cloud EDA tools, scaling access to SoC validation solutions and compute services.The post Cisco Accelerates Project Schedule by 66% Using Synopsys Cloud appeared first on Chip Design....

featured video

How MediaTek Optimizes SI Design with Cadence Optimality Explorer and Clarity 3D Solver

Sponsored by Cadence Design Systems

In the era of 5G/6G communication, signal integrity (SI) design considerations are important in high-speed interface design. MediaTek’s design process usually relies on human intuition, but with Cadence’s Optimality Intelligent System Explorer and Clarity 3D Solver, they’ve increased design productivity by 75X. The Optimality Explorer’s AI technology not only improves productivity, but also provides helpful insights and answers.

Learn how MediaTek uses Cadence tools in SI design

featured chalk talk

FleClear: TDK’s Transparent Conductive Ag Film
Sponsored by Mouser Electronics and TDK
In this episode of Chalk Talk, Amelia Dalton and Chris Burket from TDK investigate the what, where, and how of TDK’s transparent conductive Ag film called FleClear. They examine the benefits that FleClear brings to the table when it comes to transparency, surface resistance and haze. They also chat about how FleClear compares to other similar solutions on the market today and how you can utilize FleClear in your next design.
Feb 7, 2024
10,199 views