We’ve been contacted by [Cedric], telling us about the smallest ARM MCU he’s ever seen – Huada HC32L110. For those of us into miniature products, this Cortex-M0+ package packs a punch (PDF datasheet), with low-power, high capabilities and rich peripherals packed into an 1.6mm x 1.4mm piece of solderable silicon.
This is matchstick head scale computing, with way more power than we previously could access at such a scale, waiting to be wrangled. Compared to an 8-bit ATTiny20 also available in WLCSP package, this is a notable increase in specs, with a way more powerful CPU, 16 times as much RAM and 8-16 times the flash! Not to mention that it’s $1 a piece in QTY1, which is about what an ATTiny20 goes for. Being a 0.35mm pitch 16-pin BGA, your typical board house might not be quite happy with you, but once you get a board fabbed and delivered from a fab worth their salt, a bit of stenciling and reflow will get you to a devboard in no time.
Drawbacks? No English datasheet or Arduino port, and the 67-page PDF we found doesn’t have some things like register mappings. LILYGO promised that they will start selling the devboards soon, but we’re sure it wouldn’t be hard for us to develop our own. From there, we’d hope for an ESP8266-like effect – missing information pieced together, translated and made accessible, bit by bit.
When it comes to soldering such small packages, we highly recommend reflow. However, if you decide to go the magnet wire route, we wouldn’t dare object – just make sure to send us pictures. After all, seems like miniature microcontrollers like ATTiny20 are attractive enough of a proposition that people will pick the craziest route possible just to play with one. They say, the madness of the brave is the wisdom of life.
We thank [Cedric] for sharing this with us!
it was so small I didnt even see it. I kept looking at the voltage regulator thinking it sure doesnt have a lot of ports. doh!
+1 :)
+1
Some Lattice FPGA dev boards have a big arrow pointing to the main chip because it is so small it would get lost among the other components :)
I actually had to follow the PCB traces from the pin holes to find the chip.
Interesting. ST makes one that’s almost that small–the STM32G031y8y6 is 1.9×2.1mm, in WLCSP-18, with a whopping 8K of RAM and 64K of flash, double what this has, but it’s $3.06 in QTY 1. I’ve got a couple on my desk, but haven’t found a fab that does microvias (although I haven’t looked hard.)
Very excited to have more wee μCs to choose from for future projects! Aside from this and the ATtiny20, there’s also the Maxim MAX32660, which is 1.6mm×1.58mm (A whopping 20% larger! What a gargantuan chip!). It’s a Cortex M4 sporting 256KB Flash and 96KB SRAM, and can be clocked up to 96 MHz. I’d love to know what other sub-2mm microcontrollers have slipped under my radar….
There’s a PSOC-4 and a Delfino C28xx in similar tiny packages too.
So it’s comparable to an atmega128 in terms of ram/flash?
Here’s a 3 x 5 mm LGA-14 I soldered to with wirewrap wire: https://i.imgur.com/4AZtEUA.jpg
The leads are 0.8 mm pitch and one layer deep. This BGA package of the component in the article is 4 x 4, so the inside 2 x 2 balls are a layer deep, and the 0.35 mm pitch is more than twice as fine.
This would be really difficult to solder up with wire.
Here’s a 0.35mm DSBGA package I did last year with a $30 eBay microscope.
It’s not easy, but doable.
https://imgur.com/gallery/PbFXaSt
There surely comes a point, where the cost of a PCB manufacture outweighs the diminishing returns of a small device? Where the reduced pin count impacts the ability to interface with real world I/O?
This is not the ideal component for a hobbyist Think of a “smartphone”, where you already have the expensive PCB and need just a little smarts somewhere in a small corner. That’s what it is aimed at.
Atmel understood the importance of making entry-level-components. So they had the DIP atmega168 for hobbyists, and hoped that those who had experience with them as a hobbyist, would buy massive quantities when they recommended the same chips to their workplace later on.
Now the TQFP can be seen as “can be used by a hobbyist” and “entry level” has increased, but this is the non-entry-level version of this chip. (They might not even HAVE an entry-level version).
The PDF says there are other packages available – QFN20, TSSOP20 and TSSOP16.
JLCPcb has those other packages in their library if you want to build a board to play around with the device.
Does anyone have an idea of how solid of a company Huada is? There is currently an insane number of startups in China introducing CM0 MCUs and BL-SOCs. Very unlikely that all of them survive.
Just to add: There is a lesson to be learned from the Paudauk “3 cent” MCU, which basically turned into unobtanium right now. The only easily accessible source LCSC does have almost zero stock, and what is there is not very competitive in pricing anymore.
The machine translation of the Chinese pdf is good enough to find out more about the part, which surprised me. The translation is in no way perfect, but it is far better than I was expecting it to be.
(ref: https://translate.google.com click on “Documents” and upload the Chinese PDF and then Download the translated document with a new filename)
Don’t see the value added for DIY, this is well above what is achievable and having such tiny mcu on a big devboard is pointless (except for reverse p contest, watch your feet!).
On the other hand NXP did release LPC800 in DIP8 format, but it was discontinuated quite fast.
There’s also this guy from Nexperia in a 20-WLCSP, similar specs, maybe better documented?
There’s also this one from Nexperia in a 20-WLCSP, similar specs, maybe better documented
https://www.digikey.com/en/products/detail/nxp-usa-inc/MKL02Z32CAF4R/4156753
Sorry, forgot the link before
(It’s actually NXP, not Nexperia). This type of MCU is used in e.g. Arduino Teensy to interface between the main MCU and the USB and control the MCU’s modes so that booting and flashing are v. transparent to the user.
If it’s not STM32 then I don’t care
What IS this? A MCU for ANTS?!?
I don’t wanna hear your excuses! The chip has to be at least… three times bigger than this!
Smaller and smaller is very obviously the trend and this scale is only going to get more common.
HaD: We’ve had the “This is my worst soldering job” article recently – now how about a “This is my most impressive soldering job”? I would submit it to the tip line, only I don’t have anything to submit, I am not worthy due to age, diminishing eyesight and lack of a decent microscope!
Is there a point where there will cease to be a market for smaller?
Look at that thing compared to the USB-C connector. What sort of applications or mischief could be accomplished by embedding something with this much computing power in something like an innocent looking cable?
Looked for this part on LCSC not there. But saw the contest for CH32V307. For $3-$4.00 I would use the RSIC-V CH32V307 over the HC32L110. Unless all you want to do is collect data from some I2C devices and pass the data to the USB. This would make I2C devices look huge though. But for a slave MCU it may have a purpose.
@Chris said: “Looked for this part on LCSC not there.”
It’s there, but not in English. LCSC CN search for HC32L110B6YA all variants:
https://so.szlcsc.com/global.html?k=HC32L110B6YA
I’ll just stay with the stm32 line
I like to work with DIP’s, but when is the last time you’ve seen a DIP in a commercial product? Even if they were cheaper, which they are not, they aren’t worth the board space anymore.
The chip packaging is probably a very significant component of the cost of making chips now, possibly even the major component. Incredible.
This reminds me of of this article.
https://hackaday.com/2018/10/04/malicious-component-found-on-server-motherboards-supplied-to-numerous-companies/
While such a thing is certainly possible, those original claims didn’t survive deeper investigation, and such an attack is actually either very difficult to pull off or else it’s trivial via other means. If you’re monitoring something, you have to exfiltrate the data, and most datacenters will notice the unexpected traffic. If you’re manipulating something, it’s generally easier to taint the IPMI and do it that way. It’s an intersting attack in theory, but hard to find a practical use where it’s the best solution to some problem.
@Arsenijs Picugins said: “Being a 0.35mm pitch 16-pin BGA, your typical board house might not be quite happy with… Not to mention that it’s $1 a piece in QTY1.”
The HC32L110B6YA-CSP16TR product web page [1] says: “Package: QFN20, TSSOP20, TSSOP16, CSP16”. So while it looks like there are more prototype-friendly packages available they seem to be unobtanium, at least from LCSC that is.[2]
Nit-pick: At my post time the HC32L110B6YA-CSP16TR Qty.-1 price from LCSC is 5.68 CNY or $0.90 USD each, not $1.00.[3]
@LCSC:
Jiangsu warehouse inventory: 127
Guangdong warehouse inventory: 4835
Recent orders: 16
1. HC32L110B6YA-CSP16TR product web page:
https://www.hdsc.com.cn/Category82-1393
2. LCSC CN search for HC32L110B6YA all variants:
https://so.szlcsc.com/global.html?k=HC32L110B6YA
3. @04-March-2022, 06:30 AM UTC, 5.68 CNY = $0.90 USD:
https://duckduckgo.com/?t=ffab&q=5.68+CNY+in+USD&ia=currency
NO Lock Bit /security bit mechanism….
Big No.
It’s Chinese. I don’t use Chinese processors because you never know it is counterfeit material.
I have programmed a cortex-m0 similar to this one. It was the power sequencing part on an FPGA mezzanine card, the peripheral set was small, but included the GPIOs and ADC plus timers required to monitor the power rails. The MCU package was dwarfed by the POLRs it controlled. When board space is at a premium, these things work nicely, and you can afford to have more than one dedicated to different functions. You don’t develop on the final board, the chips are programmed before the board is assembled.
The HW engineer used the MCU because it was cheaper than a CPLD an earlier revision of the mezzanine card had for the same function.