Currently the Bus Pirate 5 RP2040 chip runs at 125MHz. When updating the SDK I sew that it is now certified to 200MHz with a core of 1.15v.
Is it time to bump the speed?
Also, should we switch to the core SMPS on the rp2350 boards so that if a similar thing happens we can adjust the core voltage?
I haven’t overclocked mine yet, but now that I see it’s ’certified’ I probably will for some of the things I do.
Since 200 MHz is approved, might as well do it. All of the work of making sure the code doesn’t directly reference 125 MHz is already done, so it should be fine.
If you do this, let me know—I might need to adjust the code for Flashrom and AVRDUDE.
To make it future proof you can use the system timer function to find the clock and set your speed and delays that way. It’s used in a lot of the pio program configurations.
Yep, once I finish everything related to USB and PS/2, I’ll improve all that, since right now it’s a bit of a dirty hacky mess. 
There might be issues with third party patches or similar, like dreg is mentioning. So maybe make this a compile-time option, but default to 200MHz?
I’m not fully convinced of the SMPS design of the RP2350. Like all minor and major SMPS IC designers are capable of designing a SMPS that reliably works with a wide variety of inductors of varying orientations and values, but RasPi isn’t. While other brands have some recommended and verified passives, only RasPi strongly recommends only using their one magic inductor. To me this hints at a rushed design where they didn’t want to iterate the silicon due to mask cost reasons and now the user has to live with one strict design recipe they were able to get stable with the silicon they have.
I’m not sure if there aren’t additonal issues lurking in the shadows, like in conjunction with fluctuating input voltage due to strange loads on VOUT. RasPi now take quite some time to update their datasheet with new errata. Like the security fuse issues from their contest were publicly known at least from the CCC congress at the end of last year, but it still took them till 20th feb until they updated the datasheet. So they could even know or suspect some new issues now and don’t communicate this to the customers.
The RP2350 is only now becoming available as single ICs. So only now are people able to make their own designs with it and experiment. I guess this causes any existing issues to be found and becoming known over the next year or so.
How about using a LDO with a variable output voltage and voltage divider for the next BP7 prototypes? That would make it easier to play with different voltages and speeds than the fixed-value LDO we have now. Maybe even leave open an unpopulated resistor footprint next to the voltage divider, with values precalculated to make changing it to 1.15V easy by adding just one resistor of a common value.
You have a good point on the SMPS, and I tend to agree.
One thing stood out to me: the PICO2 board doesn’t use their super custom inductor or layout (as far as I can tell, maybe they did a custom run of a much smaller package). The spark fun dev board that came today definitely had the required at launch parts. … So maybe they already know and had issues internally and the custom part is a hack…
I like the LDO for board space, routing and simplicity. Maybe an adjustable LDO with extra spots to update. Maybe we just run the core at 1.2v and be done with it 
The abs. max. rating for the core is 1.21V. Even the better LDOs don’t offer better than 0.5% base voltage accuracy, plus temperature drift. I think you’ll end up with quite a few devices running permanently above the abs. max. rating. I don’t know if this is a good idea.
The LDO accuracy + resistor accuracy can also be an issue with the adjustable LDO of course. We shouldn’t use the cheapest 2.5% LDO for this.
Sounds perfect for me OC!
I’m sure ian is already hard at work figuring out logistics options for shipping each BusPirate with a flask of liquid nitrogen. for proper OC credibility.
You see shipping liquids and gasses is difficult. However we could ship out some kind of on site liquid nitrogen generator.
Obviously a squishy startup would want to hook you on a costly monthly subscription liquid nitrogen delivery service. You’ll find no such thing here.