- For me pin-compatibility is not that important and 0.7" is still small enough, so I would vote for the design that is easier to DIY.
- A F3-ready design would be awesome indeed.
Community rework of standard maple and maple mini.
(181 posts) (14 voices)-
siy
Member> - A F3-ready design would be awesome indeed.
I think that F3 requires completely new design. There are sensible differences between F1 and F3 in LQFP48 package. I think it will be enough to say that STM32F103Cxx has 40 non-supply pins while STM32F37xC has 43 non-supply pins.
Posted 2 years ago # -
gbulmer
ModeratorSorry for the delay posting info about about my boards. Siy spurred me on to try harder (thank you very much Siy), and I think I did. Thanks to Siy's inspiration it is 0.6" pitch.
So there are initial pictures and PDF's of the schematic and PCB for a 'Maple-mini-like' board across at my blog
It is very similar to Maple-mini, however there are several changes. There are notes on the schematic about the changes, but I'll reproduce the list here to save folks time:
- Redesign Maple-mini’s 4-layer PCB, as simpler, double-sided PCB
- Single-sided Surface Mount Devices (SMD) (rather than Mini two sides SMD), for simpler DIY assembly
- Larger 0805 parts, replacing 0402 parts, easier (for me) to make
- Much higher capacity Voltage regulator, aim is full power from 9V input
- Polyfuse protecting Host’s USB-sockets power
- USB Electro-Static Discharge (ESD) protection for Host-USB socket
- More compact USB termination circuit (inside USB ESD device)
- Simplified USB ‘pull-up resistor’; signals ‘USB-device-type change’
- Through-hole USB socket intended to be more robust than SMD USB socketThis is pretty much the same list that we have been working on for a while. IMHO Siy's pioneering work should be noted because he made such a significant contribution with his designs in 2012.
It is not exactly Maple-mini pin compatible. One pin has changed.
The 'av+' pin, which provided 3.3V, from the on-board dedicated analogue voltage regulator, is now simply a 3.3V 'Vcc' pin supplied by the much larger capacity LM1117-3.3V. I don't think most users will notice the difference. However by removing the small SOT-23 CP17xx regulator from the pin-out, I hope the board will operate without overheating or damage at higher input voltages.
It should be Maple-mini bootloader and code compatible as all signal pins are the same.
WARNING: this is 'In-Work', and untested. So please don't assume it is 'ready'.
I'm aiming to upload to github later this week.
Posted 2 years ago # -
gbulmer
ModeratorI haven't carried across all of the Maple-mini's silk-screen onto my design.
I'm not confident that some of the PCB houses can do such small text.
I also wanted to make the text a bit bigger for me :-)I'd like to use a 'notation' to identify ADC and PWM pins.
Arduino used '~' for PWM, which seems okay, though it might be nicer for analogue signals.
I did think '[' looks a bit like some of a square wave :-)I was thinking of something for ADC.
My thoughts: use the text 'overline' which uses very little space, or maybe '%' which sort of implies more resolution than 0 or 1 :-)
I was thinking of '*' for pins with more functionality, that needs to be looked up.Or do people like all the small Maple-mini text?
Posted 2 years ago # -
gbulmer
ModeratorSiy -The STM32F302/303 LQFP48 pin-outs are different to the STM32F372/373. As you say, the STM32F37x has more I/O.
I think the LQFP48 STM32F302/303 are pin compatible with the STM32F103, and hence could operate on the same PCB. Are you saying that is incorrect, or only that the STM32F37x are not pin compatible with the ST32F103?
I am more interested in STM32F303 than the other STM32F3 parts. It has the most RAM, 4 fast ADCs, 2 DACs, and lots of interesting analogue (4 op amps and 7 analogue comparators), as well as plenty of timers and other peripherals. There is also an STM32F3Discovery board which can be used to test and cross check code, and it has the ST-LINK/V2 on board for hardware debugging.
So a Mini with an STM32F303 might be straightforward.
Posted 2 years ago # -
siy
Member> Are you saying that is incorrect, or only that the STM32F37x are not pin compatible with the ST32F103?
I have (untested) design for F37xCx, so I've already checked information mentioned above. One of the supply pin is located at position 17, no supply pins at positions 35-36 - this is enough to make these parts pin-incompatible with F103.
Posted 2 years ago # -
gbulmer
ModeratorSiy - Okay. I knew STM32F37xxx are different pinouts from STM32F303 and STM32F103.
I wondered if you had been through STM32F302x/303x and identified an incompatibility between STM32F103 and STM32F303.I agree, STM32F37xxx have awkward pin changes with respect to STM32F103; compatible PCBs, especially space constrained like a mini, would be difficult to design and use.
As you wrote, pins PF6 and PF7, pins 35 and 36, of the STM32F37xxx has signal pins replace power pins. So these would need to appear on a header. (These signals also appear on STM32F051x.)
Then STM32F37xxx replaces signal pins with power and voltage reference pins to support the 'Sigma Delta' ADCs; signal pin 25 becomes VREFSD+ and signal pin 17 becomes VDD_2 supply. So pin 17, PA7, would disappear from a header. A messy board.Siy - "I think it will be enough to say that STM32F103Cxx has 40 non-supply pins while STM32F37xC has 43 non-supply pins."
Would you please explain this?I haven't examined the STM32F37xxx parts in detail (my heart belongs to STM32F303 :-)
However, I had thought the net effect of changes to LQFP48 STM32F37xxx, 2 power to signals and 2 signals to power had resulted in the same number of signal pins as LQFP48 STM32F103.Posted 2 years ago # -
gbulmer
Moderatorfeurig - " There are at least 3 good designs which fit the to fix list."
What are the three?
Edit: Are you looking for Arduino-compatible headers (like a RET6, which is where you started the thread) or bread-board friendly dual-in-line? For easy DIY?I'm very interested in understanding both priorities on fixes, and constraints.
For example people have said bigger than 0.6" pitch is okay.
Changing the Maple-mini pin out is okay, especially if it gives more I/O.
Some Tekwizz colleagues suggested using larger, easier to handle 1206 components might make it more attractive for DIY.
On my LQFP64 design, I don't bring the two USB signals out to headers (I've been in a workshop where two folks damaged the USB sockets on their laptops)Do you have a prioritised list you would share, or was the list at the start of this thread 'it', including priority?
Posted 2 years ago # -
siy
Member> Siy - "I think it will be enough to say that STM32F103Cxx has 40 non-supply pins while
> STM32F37xC has 43 non-supply pins."
> Would you please explain this?Well, if we'll look at pinout of any chip, all pins can be separated in two large groups - ones which provide various kinds of supply voltage to MCU and signal ones. So, for regular STM32F10xCx MCU's there are 8 supply pins - 3 VDD, 3 VSS, 1 VDDA and 1 VSSA. There is also VBAT pin which is, strictly speaking, supply pin as well. Everything else is counted as "non-supply" and there are 39 such pins in these MCU's (my calculation above is incorrect, I've missed VBAT). For STM32F37xCx situation looks so: 2 VDD, 1 VSS, 1 VSSA/VREF-, 1 VDDA/VREF+, 1 VSSSD/VREFSD-, 1 VDDSD, 1 VREFSD+. With VBAT this is also 9 pins (so, strictly speaking, my calculations above for F37x are incorrect as well). But unlike VDDA/VSSA it's not so obvious what to do with VREFSD+/- and VREF+/-. After all chip is targeted to analog signals processing, tying these pins to +3.3 and GND does not look like right choice because contradicts main purpose of F3. So, from my point of view, it worth to bring out VREFSD+/- and nice to have VREF+ brought out. In other words, from the layout perspective these pins look like regular (non-supply) pins. Same is true for VBAT pin for both MCU's. So, it results to 40 "non-supply" pins for F103 and 43 "non-supply" pins for F37x.
Posted 2 years ago # -
siy
Member@feurig, beside published ones I have several other versions of design for different packages and MCU series. Let me know if you're interested, I'll send them to you. Perhaps you'll find something interesting/useful there.
P.S. all designs are under CC BY-SA license just like original Maple designs.
Posted 2 years ago # -
gbulmer
ModeratorSiy - "So, for regular STM32F10xCx MCU's there are 8 supply pins ... also VBAT pin... Everything else is counted as "non-supply" and there are 39 such pins in these MCU's"
"For STM32F37xCx ... With VBAT this is also 9 pins"
Okay, we are on the same page about that; you agree with me :-)"But unlike VDDA/VSSA it's not so obvious what to do with VREFSD+/- and VREF+/-."
Yes, agreed."So, from my point of view, it worth to bring out VREFSD+/- and nice to have VREF+ brought out. In other words, from the layout perspective these pins look like regular (non-supply) pins."
Okay, got it. Agree.
Those voltage reference pins could:
a) be on headers (which is the minimum work and most flexible), or
b) have their own on-board reference (easiest for a simple user), or
c) have both.Posted 2 years ago # -
siy
Member@gbulmer, I'm going to add on-board precise reference for VREFSD (most likely LM4040 for 2.048V) and dedicated pin 2x2 header so user can either plug jumpers and use on-board reference or open them and use pins to provide external reference. After some thinking I got conclusion that there is no need to have both VREFSD and VREF exposed, so VREFSD- and VREF- can be tied together and form analog ground connected to main ground via small inductor (and exposed at mentioned above 2x2 header). VREF+ then can be connected to dedicated 3.3V on-board regulator. So, overall pin configuration will be similar to F103 version.
Posted 2 years ago # -
gbulmer
ModeratorI've published a draft-quality-CAD version of my second rework to improve Maple-mini, Orone-wide-mini.
Colleagues at Tekwizz suggested I try to make the board easier to assemble for relative beginners. So the Surface Mount Devices (SMD) are larger where practical. They have changed from Maple-mini's 0402 to 1206 size components (nine times the area). The board has increased in size by 0.2" in width and length, and the 0.1 inch, breadboard friendly, header pins are 0.8" apart.
The changes created some extra 'top' board area. So I've added a 2-pin Molex socket to make external power easier to provide, and a 'Power On' LED. The pins for this socket are on the same 0.1" grid as the header pins, hence if anyone wants to use these with their own stripboard or veroboard circuits, it should fit.
I've increased the width of tracks and pads around through-hole parts with the aim of making the PCB more robust for less-experienced solderers.
The only outstanding Maple-mini defect for me is not having the LED-controlling signal on a header pin. However, all ways to put that onto the header pins would make this pin incompatible with Maple-mini, and the GPIO is already driving one LED, so I am unlikely to do that.
All feedback gratefully received.
Posted 2 years ago # -
I have a few queries regarding the schematic. Most of them are possibly from my lack of knowledge and experience.
Why keep the MCP1703? How much Vripple is on the LM1117-3.3? Does the difference in Vripple make a huge amount of difference? Could you ditch the MCP1703 for just a LC filter and use that for AV+ instead of just Vcc directly? Lastly regarding analog, why is L1 even there?
Why is D2 there? I don't think it would be of any use and actually an annoyance like the original Maple Mini r2 where you're wanting to get Vusb to power an LCD, but can't. At a later stage, would it be possible to revamp the board to use 0603 components to obtain a board similar size to the current mini?
Either way, liking the look of the board :D Great work everyone. Sorry I haven't been responding much, but I've been keeping up to date with the thread. I would like to see STM32F303 components in the future, but for now, getting something working the same would be good enough.
Posted 2 years ago #
Reply »
You must log in to post.