Taking a look at STM32G0 series

Anything not related to STM32
ag123
Posts: 1656
Joined: Thu Dec 19, 2019 5:30 am
Answers: 24

Re: Taking a look at STM32G0 series

Post by ag123 »

thanks
74hc4060
https://assets.nexperia.com/documents/d ... CT4060.pdf
looks like a pretty cool chip, in that the 'extra' nand gate at the front can be used for for the inverting crystal osc.
and the counters / dividers may have quite interesting uses as additional clock sources divided down from the source.
dannyf
Posts: 447
Joined: Sat Jul 04, 2020 7:46 pm

Re: Taking a look at STM32G0 series

Post by dannyf »

there are also clock generator chips for this sort of things. TI, OnSemi, and SI for example.

you may also look into PIC32 chips that has very sophisticated REFO modules.
ag123
Posts: 1656
Joined: Thu Dec 19, 2019 5:30 am
Answers: 24

Re: Taking a look at STM32G0 series

Post by ag123 »

thanks, being curious, I did some search and took a look
https://www.onsemi.com/products/timing- ... generators
https://www.onsemi.com/download/data-sh ... 5150-d.pdf
https://www.onsemi.com/download/data-sh ... 5573-d.pdf
https://www.onsemi.com/download/data-sh ... 5573-d.pdf

https://www.ti.com/clocks-timing/genera ... ducts.html
https://www.ti.com/lit/ds/symlink/pll1707-q1.pdf
https://www.skyworksinc.com/en/Products ... 5351a-b-gt
https://www.skyworksinc.com/-/media/Sky ... 5351-B.pdf

many of them are 'elaborate' clock generators with PLL, and it seemed the last one Si5351 is rather 'popular' on AliX
those are useful in situations, e.g. where there is quite a few chips requiring different frequencies

for the time being, as I've 8mhz passive crystals, and I chanced upon the Schmitt-Trigger inverter 74LVC1G14
https://www.ti.com/product/SN74LVC1G14
which apparently seemed priced somewhat lower on AliX (perhaps that it is after all more popular / common)
https://www.aliexpress.com/item/33045911365.html
so i got a small bundle of it. I'm thinking I'd take those sop8 / 'sot23' 'adapter boards'
https://www.aliexpress.com/w/wholesale- ... apter.html
so the 74lvc14 would take 6 pins and the crystal take 2 pins on the so8 'adapter board', the connections and other components e.g. caps would need to be 'soldered around' it, probably looking messy, but chances are that with that I'd get a 'quick and dirty' 8mhz oscillator
dannyf
Posts: 447
Joined: Sat Jul 04, 2020 7:46 pm

Re: Taking a look at STM32G0 series

Post by dannyf »

what about 3n502?

I have used a part that's similar to cdce937 but with only 2 pll outputs. fairly easy to program (i2c).
ag123
Posts: 1656
Joined: Thu Dec 19, 2019 5:30 am
Answers: 24

Re: Taking a look at STM32G0 series

Post by ag123 »

aw, didn't stumble into this earlier, nice chip
https://www.onsemi.com/download/data-sh ... n502-d.pdf
dannyf
Posts: 447
Joined: Sat Jul 04, 2020 7:46 pm

Re: Taking a look at STM32G0 series

Post by dannyf »

if you want to use HSE, a few options.

1. find a chip that supports HSE. F0 or PY32 for example.
2. if space is a constraint, use an external oscillator (crystal or mems - less accurate), or small foot-print clock generator.
3. if space isn't a constraint, something like HC4060, or even a STM32F103 (with its MCO) will work. the PIC32 has an amazing REFO module - fractional divider, down to 1/512 resolution.

I took the first approach :)
dannyf
Posts: 447
Joined: Sat Jul 04, 2020 7:46 pm

Re: Taking a look at STM32G0 series

Post by dannyf »

with that said, modern HSIs are fairly accurate / very accurate. I am running UARTs off HSI at 115K2bps all day long. I think between accurate HSI, factory calibration, and user-trimming, HSIs can cover most of the use cases.

HSEs still hold an advantage in terms of timing stability, however.

so if your goal is to try out the chips, I wouldn't too worried about the HSE at this point.
ag123
Posts: 1656
Joined: Thu Dec 19, 2019 5:30 am
Answers: 24

Re: Taking a look at STM32G0 series

Post by ag123 »

well, I'd try py32 next ;)
for now i'd still want to play with my stm32g0 which I've been so lazy I've not soldered them on boards yet :lol:
but in real terms g0 without HSE OSC OUT may make some 'real' customers skip that series or at least those form factor affected by it.
but of course, these days there are single chip oscillators (with crystal) after all and that the prices are actually quite competitive vs 'passive' crystals, but they do cost a little more each, percentage wise it seemed to be about double, but that is 'flea market' (e.g. AliX) prices.

I think HSE is more an issue if one needs it not to drift, for RC HSI, it is likely it'd 'just work', but that it may drift with temperature etc.
crystals has very high q factors, which kind of force it to 'lock' to a particular frequency. Though they may still be affected by temperature.

i'd try to make a 'kitchen sink' crystal osc mainly for my 'lab' experiments with g0, at least to assure myself that my patchwork 'kitchen sink' not gate osc works :lol:
dannyf
Posts: 447
Joined: Sat Jul 04, 2020 7:46 pm

Re: Taking a look at STM32G0 series

Post by dannyf »

it is fun to rig up anything.

microchip has a good an on crystal oscillators: https://ww1.microchip.com/downloads/en/ ... 00826a.pdf

you may also check out the si part I mentioned earlier - si5351. it is widely available with ready-made pcbs. you will also find arduino libraries made for it. a good alternative.

PS: the 5351A part is easy to find. the B part (which has an voltage input to fine-tune the output frequency) is harder to find.
ag123
Posts: 1656
Joined: Thu Dec 19, 2019 5:30 am
Answers: 24

Re: Taking a look at STM32G0 series

Post by ag123 »

as stm32g030 don't have HSE crystal input i.e. OSC IN and OSC OUT
I tried to make one using the Pierce oscillator circuit
https://en.wikipedia.org/wiki/Pierce_oscillator
that not gate is simply a 74lvc1g04 single not gate I bought a handful from the 'flea markets'
https://www.ti.com/lit/ds/symlink/sn74lvc1g04.pdf
i got the schmitt-trigger versions, hopefully, they'd work as well
https://www.ti.com/lit/ds/symlink/sn74lvc1g14.pdf

It turns out the full circuit is quite a few pieces of components. The bulk is due to the use of through hole components rather than smd parts.
the parts are laid out on one of those 'proto boards', 'perf boards'
https://www.aliexpress.com/w/wholesale-paper-pcb.html
a common size it seemed is 5cm x 7cm, the parts fit in about 10 holes by 8 holes slightly wider than about 1 inch (2.54cm) in length and narrower in the width. so presumbly about 2-3 modules would fit in one of thos 5x7 pcbs

that sot23-adapter is simply one of those 'sop8, sop10' style adapter pcbs in the 'flea markets', so that the 74lvc1g04 smd part is soldered on there and the module is soldered onto the perfboard
https://www.aliexpress.com/w/wholesale- ... apter.html

initially i'm half way hoping i'd fit those components on a sop8 or sop10 style 'adapter' board, but it'd seemed very difficult to achieve it given the handful of components. hence, patching onto a perfboard is probably a next 'better' way.

I'm not sure if this'd work, but that these days there are 'single chip' oscillators that does better far smaller and compact and does away with all these.
i made do with these for now as i've got a handful of through hole parts on hand which I'd guess i may be able to put one together.
Attachments
Pierce oscillator PCB
Pierce oscillator PCB
pcb-min.png (28.24 KiB) Viewed 9073 times
Pierce oscillator circuit
Pierce oscillator circuit
sch-min.png (10.54 KiB) Viewed 9073 times
Post Reply

Return to “Off topic”