3 cent MCU

[Mechatrommer]

Program Memory Size... 1KW... right.

I did this project in under 1K

the 2nd time misunderstood so let me clarify, i wasnt complaining about the size, i was complaining the SI unit used i have bunch of pic10f206 all right with "512W" i've made them some useful stuffs that i'm more than happy with enough troll now, since its persistent use in the datasheet, i know now its Kilo Word :|

back to topic, i think if this even true, it must be something with low capital investment, like low cost of 2nd hand manufacturing facilities. labor cost, we know its super cheap in that communist governed country. the RnD maybe none, they just buy some template/blueprint/lithography whatever the name is from some company 10-50 years old design. so yeah maybe its possible in ChinaLand. i will concern most on reliability (0¢ QC?) since recently i bought a clone TC1 Transistor Tester with claimed "super anti disturbance" STC15W104 mcu in it from China but it turned out i got a fried FW chip at least thats what i think it is.

[SiliconWizard]

Sadly It's OTP

It's dead cheap. You can afford to waste a few.

Of course. Those MCUs are for very high-volume production and we are talking about 3 cents retail price! There's probably no way they could embed even 1K of Flash for that cost.

The annoying point with this is not to waste parts but since there's apparently no way of running code from external memory, this should be very annoying during the development/debugging phase. Having to replace the chip for every software build? Sucks... So you should probably make some kind of dev board with an SO8 socket to be able to replace chips very quickly.

[coppice]

back to topic, i think if this even true, it must be something with low capital investment, like low cost of 2nd hand manufacturing facilities. labor cost, we know its super cheap in that communist governed country. the RnD maybe none, they just buy some template/blueprint/lithography whatever the name is from some company 10-50 years old design. so yeah maybe its possible in ChinaLand. i will concern most on reliability (0¢ QC?) since recently i bought a clone TC1 Transistor Tester with claimed "super anti disturbance" STC15W104 mcu in it from China but it turned out i got a fried FW chip at least thats what i think it is.

Modern semiconductor vendor have very little capital investment beyond their mask sets. The investment is mostly by the companies running the fabs and the assembly and test facilities. These Padauk MCUs are in a 180nm process. Until very recently nobody was implementing low end MCUs in such a fine process. There are no old designs to pick up and run with cheaply. I presume Padauk have engineered these MCUs from the same process baseline as people who were making complex 180nm chips 10 years ago. Things like the toy and commercial gift markets provide opportunities for staggering numbers of small cheap MCUs. Re-engineering old low end 500nm MCUs for a 180nm process and bring substantial cost benefits if done right, and the volumes are certainly there. It sounds like they are hitting the same noise related stability problems most vendors had when they first tried making 180nm devices for a market that uses 1 and 2 layer PCBs. Luckiy for them, they have a huge potential market to serve which doesn't care too much about stability, as they learn what everyone else in the MCU world has been learning about robust fine geometry designs over the last several years

Padauk is Taiwanese, not Chinese. Their costs, and general environment are little different from any European, American or Japanese vendor.

[coppice]

Those MCUs are for very high-volume production and we are talking about 3 cents retail price! There's probably no way they could embed even 1K of Flash for that cost.

At 180nm, 1k of flash or 1k of EPROM probably doesn't make a huge difference to the die area. What is key to keeping cost low is not putting a voltage pump on the die, so the flash can be self programming. I presume the programmer for this device applies a high voltage during the programming cycle. You might even find the device is actually flash based, and they just haven't provided the ability to erase the array.

[wraper]

The annoying point with this is not to waste parts but since there's apparently no way of running code from external memory, this should be very annoying during the development/debugging phase. Having to replace the chip for every software build? Sucks... So you should probably make some kind of dev board with an SO8 socket to be able to replace chips very quickly.

Not annoying for any serious people. Buy in circuit emulator and develop as much as you want. https://www.yoycart.com/Product/568282464626/

[nctnico]

back to topic, i think if this even true, it must be something with low capital investment, like low cost of 2nd hand manufacturing facilities. labor cost, we know its super cheap in that communist governed country. the RnD maybe none, they just buy some template/blueprint/lithography whatever the name is from some company 10-50 years old design. so yeah maybe its possible in ChinaLand. i will concern most on reliability (0¢ QC?) since recently i bought a clone TC1 Transistor Tester with claimed "super anti disturbance" STC15W104 mcu in it from China but it turned out i got a fried FW chip at least thats what i think it is.

Modern semiconductor vendor have very little capital investment beyond their mask sets. The investment is mostly by the companies running the fabs and the assembly and test facilities. These Padauk MCUs are in a 180nm process. Until very recently nobody was implementing low end MCUs in such a fine process. There are no old designs to pick up and run with cheaply. I presume Padauk have engineered these MCUs from the same process baseline as people who were making complex 180nm chips 10 years ago. Things like the toy and commercial gift markets provide opportunities for staggering numbers of small cheap MCUs. Re-engineering old low end 500nm MCUs for a 180nm process and bring substantial cost benefits if done right

I think you got a zero wrong here. You probably mean 18nm but I doubt they'll use such an advanced process for a very cheap microcontroller. It is more likely these are produced in an old (written-off) factory which can do a 180nm process.

[mubes]

Sadly It's OTP

The thing is only 16 MHz and six or eight pins, so it should be easy to emulate it (including timing) on a faster CPU. I'd convert its machine code to C and compile that to ARM/MIPS/RISC-V/xtensa native, with a busy-wait for the right clock cycle after each instruction. A 64 or 80 MHz CPU might just about be workable with a lot of care. A 256 or 320 MHz one would be cake.

This.

Given that they're 8 bitters pretty much any half sane CORTEX-M should be capable of emulating one of them in realtime...and then you could add in watchpoints, breakpoints and all the other goodies you'd want for realworld development....Feels like going back 30 years but would be fun at $0.03! Personally I'd try and go the interpreter route so you can run the actual code on the debugger, rather than re-targetting.

I've got some Bluepills laying around here somewhere...hmmm...emulating a 3c micro on a $2 micro with USB...such decadence, but oddly compelling.

Anybody up for it with me?

[wraper]

This.

Given that they're 8 bitters pretty much any half sane CORTEX-M should be capable of emulating one of them in realtime...and then you could add in watchpoints, breakpoints and all the other goodies you'd want for realworld development....Feels like going back 30 years but would be fun at $0.03! Personally I'd try and go the interpreter route so you can run the actual code on the debugger, rather than re-targetting.

I've got some Bluepills laying around here somewhere...hmmm...emulating a 3c micro on a $2 micro with USB...such decadence, but oddly compelling.

Anybody up for it with me?

LOL, real time in many cases ARM is much slower than 8 bitters. If you are doing computational heavy tasks, ARM certainly is faster. But if you need fastest response with minimum of computations, in many cases 8 bitters are much faster at the same clock. For real wold development you don't bother with something this stupid and just buy ICE for $100.

[mubes]

LOL, real time in many cases ARM is much slower than 8 bitters. If you are doing computational heavy tasks, ARM certainly is faster. But if you need fastest response with minimum of computations, in many cases 8 bitters are much faster at the same clock. For real wold development you don't bother with something this stupid and just buy ICE for $100.

Do we actually know what the ICE can do other than just give you a re-writable device? I guess I've missed a piece of documentation somewhere?

I love the idea of Uber-cheap chips but development time is much more expensive than silicon nowadays (unless you're into 'proper' volumes) and I've no wish to go back to the bad old days of 8051-flashing-led debugging....I _like_ having stepping, register output, watchpoints and breakpoints and all that good stuff!  I got lazy when I got old.

For sure you're on the limit of what an M3 will do in I/O given that this chip is claiming single cycle instructions (Bluepill runs at 72MHz, and will do about 16MHz in I/O...very good by ARM standards), but the rest of it is eminently feasible. Anyway, this would be one of those for-fun projects; If I were being sensible then I'm back with the 40c LPC812s!

DAVE

[wraper]

Do we actually know what the ICE can do other than just give you a re-writable device? I guess I've missed a piece of documentation somewhere?

[SiliconWizard]

The annoying point with this is not to waste parts but since there's apparently no way of running code from external memory, this should be very annoying during the development/debugging phase. Having to replace the chip for every software build? Sucks... So you should probably make some kind of dev board with an SO8 socket to be able to replace chips very quickly.

Not annoying for any serious people. Buy in circuit emulator and develop as much as you want. https://www.yoycart.com/Product/568282464626/

Oh sure, OK. Didn't see that on the page mike posted.

Still, some issues may not be easy to spot (or impossible) with an ICE, such as EMC-related issues. So that still could be a little annoying during the development/testing phase.

[mubes]

Ha! That'll teach me to go posting from an airport departure lounge....now it starts to look _really_ interesting... let's see if I can get one of these ordered.

DAVE

[wraper]

I've no wish to go back to the bad old days of 8051-flashing-led debugging....I _like_ having stepping, register output, watchpoints and breakpoints and all that good stuff!  I got lazy when I got old.

With modern 8051 you usually get a very decent debugging.

[glarsson]

So that still could be a little annoying during the development/testing phase.

We survived this development process when we used those shiny new PAL thingys. And they didn't cost 3c...

Btw, you will not get annoyed when ordering one million processors at less than 3c each.

[gamalot]

Please don’t be too surprised, otherwise those Chinese electronics engineers will feel that you are ignorant. 

[coppice]

back to topic, i think if this even true, it must be something with low capital investment, like low cost of 2nd hand manufacturing facilities. labor cost, we know its super cheap in that communist governed country. the RnD maybe none, they just buy some template/blueprint/lithography whatever the name is from some company 10-50 years old design. so yeah maybe its possible in ChinaLand. i will concern most on reliability (0¢ QC?) since recently i bought a clone TC1 Transistor Tester with claimed "super anti disturbance" STC15W104 mcu in it from China but it turned out i got a fried FW chip at least thats what i think it is.

Modern semiconductor vendor have very little capital investment beyond their mask sets. The investment is mostly by the companies running the fabs and the assembly and test facilities. These Padauk MCUs are in a 180nm process. Until very recently nobody was implementing low end MCUs in such a fine process. There are no old designs to pick up and run with cheaply. I presume Padauk have engineered these MCUs from the same process baseline as people who were making complex 180nm chips 10 years ago. Things like the toy and commercial gift markets provide opportunities for staggering numbers of small cheap MCUs. Re-engineering old low end 500nm MCUs for a 180nm process and bring substantial cost benefits if done right

I think you got a zero wrong here. You probably mean 18nm but I doubt they'll use such an advanced process for a very cheap microcontroller. It is more likely these are produced in an old (written-off) factory which can do a 180nm process.

I got the zeros right. Most new ARM MCUs being made today are in 180nm or 130nm. The finest processes you see used for MCUs right now are 65nm. Super fine geormetries don't work so well for most MCUs. The costs don't hit a sweet spot, and the processes leak too much for good sleep operation. The biggest incentive for going to 65nm has more to do with integrating RF than making a good MCU, which is why most general purpose MCUs are still in 180nm or 130nm.

[Mr. Scram]

Not released yet:

I think someone on here ordered a programmer for these. There's a topic on them. I'll try to find it if I can.

[SiliconWizard]

Software and the role of an MCU definitely have an higher probability of going through debugging and adjustment phases than a simple PAL IMO, and EMC matters were also often a second concern back then compared to what we are concerned with these days.

I definitely have examples in mind where it could get annoying, such as if you're using this MCU to implement an SMPS and have to adjust some parameters during testing to pass EMC testing or such. Of course when you're using this kind of parts you're probably heading for millions of units sold and thus have ample resources for this. Either that, or you may work in a company that doesn't care about regulatory matters. Keeping my SMPS example - just tear down many cheap chinese PSUs and you'll see we're probably often in the second case.

As for me I would probably pay a few cents more a chip and select a vendor that has at least some NON-OTP versions of the chosen MCU for development. But that's just my opinion of course.

I still hear them say far in the background: "but 3 cents!!"
 

[mikeselectricstuff]

I definitely have examples in mind where it could get annoying, such as if you're using this MCU to implement an SMPS and have to adjust some parameters during testing to pass EMC testing or such.

In that case you'd make them software-tweakable for development via a serial interface (bit-bashed,obviously) etc.

[gamalot]

Just in case if there is anyone looking for anything like this, cheap MCU with more I/O pins, Flash, ADC, DAC ...... 

https://www.electrodragon.com/w/Logicgreen

[SiliconWizard]

I definitely have examples in mind where it could get annoying, such as if you're using this MCU to implement an SMPS and have to adjust some parameters during testing to pass EMC testing or such.

In that case you'd make them software-tweakable for development via a serial interface (bit-bashed,obviously) etc.

Sure. Provided that you have at least one free IO (over only 6) and that you're confident enough that this extension of the firmware would not mess anything else during normal operation... and also that it would behave exactly the same at power-on as when you program parameters via this interface...

Anyway, not saying that those 3-cent MCUs are not useful. Just saying that they are definitely not silver bullets and that you're bound to pay for the low-cost one way or another. There's no free meal. A PIC10F for instance is more than 10 times the cost. I reckon those MCUs are mainly targeted at asian companies that HAVE to keep manufacturing costs way down and often don't allow the use of non-asian components for various policy reasons (beyond sheer cost).

[langwadt]

back to topic, i think if this even true, it must be something with low capital investment, like low cost of 2nd hand manufacturing facilities. labor cost, we know its super cheap in that communist governed country. the RnD maybe none, they just buy some template/blueprint/lithography whatever the name is from some company 10-50 years old design. so yeah maybe its possible in ChinaLand. i will concern most on reliability (0¢ QC?) since recently i bought a clone TC1 Transistor Tester with claimed "super anti disturbance" STC15W104 mcu in it from China but it turned out i got a fried FW chip at least thats what i think it is.

Modern semiconductor vendor have very little capital investment beyond their mask sets. The investment is mostly by the companies running the fabs and the assembly and test facilities. These Padauk MCUs are in a 180nm process. Until very recently nobody was implementing low end MCUs in such a fine process. There are no old designs to pick up and run with cheaply. I presume Padauk have engineered these MCUs from the same process baseline as people who were making complex 180nm chips 10 years ago. Things like the toy and commercial gift markets provide opportunities for staggering numbers of small cheap MCUs. Re-engineering old low end 500nm MCUs for a 180nm process and bring substantial cost benefits if done right

I think you got a zero wrong here. You probably mean 18nm but I doubt they'll use such an advanced process for a very cheap microcontroller. It is more likely these are produced in an old (written-off) factory which can do a 180nm process.

I got the zeros right. Most new ARM MCUs being made today are in 180nm or 130nm. The finest processes you see used for MCUs right now are 65nm. Super fine geormetries don't work so well for most MCUs. The costs don't hit a sweet spot, and the processes leak too much for good sleep operation. The biggest incentive for going to 65nm has more to do with integrating RF than making a good MCU, which is why most general purpose MCUs are still in 180nm or 130nm.

afaik most if not all Cortex-M is 90nm

[cdev]

Wow, thats very inexpensive.

I expect the biggest impact will be in developing countries where cheap chips and homebrew machinery that uses them can free up lots of workers to work on other things.

(Note: I edited this when reading back it became clear to me that I had made some fundamental mistakes on the functionality of the chip)

Here, the difference between $0.03 and $0.50 or so may not matter much at all to hobbyists much of the time, but it certainly would to any manufacturer and in some countries that amount of money represents much more cost to the average person than in the US or Australia.

To sell them here, I'd expect to see cheap basic dev boards that made use of them - a manufacturer would be smart to make dev tools for them available- and cheap, even selling them at their cost. (with emulation so the design could be hashed out before burning it to the code's home on the 3 cent chip) Price would have to be much lower than $103 for hobbyists.

That is a lot of money to spend for an emulator thats used for one chip only, that likely will only be useful for a few years as electronics changes so quickly.

It would be much more useful to hobbyists if there was a sister, code compatible version which could be written to once or reused, like the AVR chips - allowing its use for development and then the 3 cent chip would complement it, be super cheap, and be write once.

I am guessing many would certainly use a very cheap chip that could be effectively protected from code extraction.

[Siwastaja]

Blah, you wouldn't make anything super complex with these. Not something absolutely requiring debugging or emulation. Getting timings right by looking at the assembly listing would suffice as well.

Some IO expanding style things, some IO timing... Power up / reset sequencing control. Adding some simple states to otherwise analog discrete DC/DC converters. Possibly some kind of zero-crossing detection using the comparator in a polling loop, giving out a pulse, or a synchronized PWM. Adding some LED magic on an already existing bus (bit-banged). Things like that. Without all the modern genericity and library crap, something around 200-300 lines of C, possibly, typically. Probably not something with 10 parameters to adjust. Maybe one or two, with a few possible values. Something you have already tested before with reprogrammable devices. The small package is quick to desolder and replace a few times with hot air.

Given a somewhat experienced developer, you can prototype your idea on any microcontroller first, using the most basic peripherals available, then port/rewrite your code at once, giving some thought to it (i.e., being careful and using your brain instead of single stepping random crap code; even though isn't trendy in programming right now), and fix all the issues in a few iterations.

Actually, the longest MCU program I have written "at once", and which worked on the first try, without fixing anything, on the first flashing on an AVR, was around 200 lines of code and IIRC it did read a potentiometer through ADC for a "speed" setpoint and did some kind of 6-step commutation for a rotorless magnetic stirrer.

Granted, your first project ever on this part could be quite annoying. Writing code for any microcontroller the first time tends to involve "reverse-engineering" some peripheral usage due to poor documentation or just your own inattention in reading the documentation properly. There are always at least some small traps involved, requiring a few rounds of trial and error. But given the simplicity (i.e., lack of features) and small number of peripherals available, you would have learned everything in a few days. After that initial learning, given a simple project with finite specification that can be represented on a single sheet of state diagram (actually formally, or in your head), it's not going to require debugging or dozens of chips.

I have seriously considered parts like these. Will probably jump in one day (when I have the excuse, i.e., something that really needs using one of these). IMO, makes most sense when you use them more widely across different projects, because the most difficult part will be the first few days of learning. Then, you have unlocked another area of expertise you can utilize later! ... and put in your resume:
"Can design ultra low cost, super mass produced, almost Chinese toys".

[coppice]

back to topic, i think if this even true, it must be something with low capital investment, like low cost of 2nd hand manufacturing facilities. labor cost, we know its super cheap in that communist governed country. the RnD maybe none, they just buy some template/blueprint/lithography whatever the name is from some company 10-50 years old design. so yeah maybe its possible in ChinaLand. i will concern most on reliability (0¢ QC?) since recently i bought a clone TC1 Transistor Tester with claimed "super anti disturbance" STC15W104 mcu in it from China but it turned out i got a fried FW chip at least thats what i think it is.

Modern semiconductor vendor have very little capital investment beyond their mask sets. The investment is mostly by the companies running the fabs and the assembly and test facilities. These Padauk MCUs are in a 180nm process. Until very recently nobody was implementing low end MCUs in such a fine process. There are no old designs to pick up and run with cheaply. I presume Padauk have engineered these MCUs from the same process baseline as people who were making complex 180nm chips 10 years ago. Things like the toy and commercial gift markets provide opportunities for staggering numbers of small cheap MCUs. Re-engineering old low end 500nm MCUs for a 180nm process and bring substantial cost benefits if done right

I think you got a zero wrong here. You probably mean 18nm but I doubt they'll use such an advanced process for a very cheap microcontroller. It is more likely these are produced in an old (written-off) factory which can do a 180nm process.

I got the zeros right. Most new ARM MCUs being made today are in 180nm or 130nm. The finest processes you see used for MCUs right now are 65nm. Super fine geormetries don't work so well for most MCUs. The costs don't hit a sweet spot, and the processes leak too much for good sleep operation. The biggest incentive for going to 65nm has more to do with integrating RF than making a good MCU, which is why most general purpose MCUs are still in 180nm or 130nm.

afaik most if not all Cortex-M is 90nm

New families of Cortex M parts are being started in 90nm, 65nm and even 45nm, but the established families are mostly 180nm and 130nm. There is a lot of cost in getting the first part out in a new geometry. The rest of the family follows much more cheaply. The means new parts expanding existing series of MCUs are usually built in the same geometry as all the existing parts.