 Notice. New forum software under development. It's going to miss a few functions and look a bit ugly for a while, but I'm working on it full time now as the old forum was too unstable. Couple days, all good. If you notice any issues, please contact me. |
Forum Index : Microcontroller and PC projects : BASIC Interpreter for the Raspi Pico A to D question.
| Author | Message | ||||
| chippy Newbie  Joined: 30/07/2022 Location: United KingdomPosts: 3 |
Hi, I have a couple of Raspi Pico's and have successfully installed MMBasic on one of them. Everything works fine and I would now like to experiment further with the analog inputs. The commands and instructions are clear enough in V5.07.04 of the User Manual but after reading the manual I am still unsure how many bits MMBasic uses for the A-to-D conversion. Looking at documentation on the Raspberry Pi Pico web-pages the A-to-D hardware used is 12 bit. In micro-python this is upscaled to 16 bits. Does anyone know if the same upscaling is used in MMBasic when running on the pico? i.e. Can anyone confirm if MMBasic works with 12 bits (4096) or does it upscale the same as micro-python? Thanks. chippy. |
||||
Mixtel90 Guru  Joined: 05/10/2019 Location: United KingdomPosts: 7938 |
As far as I'm aware, it's 12-bit on the PicoMite. The PIN() function returns the actual voltage on the pin though, not the binary value. Mick Zilog Inside! nascom.info for Nascom & Gemini Preliminary MMBasic docs & my PCB designs |
||||
hitsware2 Guru Joined: 03/08/2019 Location: United StatesPosts: 719 |
Accepting that the Pico has 12 bit resolution .... When MicroPython ( for whatever reason ? ) reads 16 bit values , it is moot . The accuracy is the same . Same with MMBasic .... the numbers are different , but reflect the same values . my site |
||||
| JohnS Guru Joined: 18/11/2011 Location: United KingdomPosts: 4044 |
By upscaled do you mean scaled? If not, what do you mean? I can't see any way to add more than the hardware provides (BTW it's buggy). John |
||||
| hitsware2 Guru Joined: 03/08/2019 Location: United StatesPosts: 719 |
Multiply x 8 (MicroPython) Divide / 1241 (MMBasic) my site |
||||
TassyJim Guru Joined: 07/08/2011 Location: AustraliaPosts: 6283 |
Thee picomite internally takes 10 readings, discards the highest and lowest then averages the remaining 8 'middle' readings. It then scales the value. That gives an effective improvement to the resolution but I leave it to others to make any claims as to the resultant bit resolution. I suggest you do a test by reading a slowly ramping voltage and graph the results. Repeat for micro-python. Jim Edited 2022-08-01 07:52 by TassyJim VK7JH MMedit |
||||
| JohnS Guru Joined: 18/11/2011 Location: United KingdomPosts: 4044 |
That's scaling. I don't even see it as a good idea as it can easiy mislead. John |
||||
Sasquatch Guru Joined: 08/05/2020 Location: United StatesPosts: 377 |
As explained in this Application note, oversampling is a way to increase A/D resolution and/or reduce noise: AVR121: Enhancing ADC resolution by oversampling As noted in this thread, there are several known issues with the ADC on the Pi Pico. The first is noise. The switching power converter on the Pi Pico PCB generates small fluctuations in the 3.3V power supply. This varies depending on the switching mode (PFM or PWM) and the load etc. There are many articles online explaining this: ADC noise Power Supply Noise and Spectrum Testing The ADC converters built in to Micro-controllers are typically noisy anyways due to the fact that they are on the same die (chip) and share power supply and ground with other MCU functions. These noise issues can be reduced by bypassing the Pi Pico SMPS (Hold the 3V3EN pin low) and providing a clean source of 3.3V to the Pico (i.e. linear voltage regulator) Providing a stable analog reference (ADC VREF) may also help. The second is that the Pi Pico ADC has a well documented bug in it's linearity: ADC has very high DNL spikes Raspberry Pi Confirms It Is Investigating a Flaw in the Raspberry Pi Pico This is a very basic flaw in the design of the ADC. There may be workarounds for this depending on the application. Bottom line is that while the ADC resolution and noise may be improved for some applications, If you want "better" ADC (higher resolution, lower noise) you will need to add a dedicated ADC chip to your application. Regards, -Carl |
||||
| hitsware2 Guru Joined: 03/08/2019 Location: United StatesPosts: 719 |
For each PIN() ? But not into arrays method , right ? my site |
||||
| TassyJim Guru Joined: 07/08/2011 Location: AustraliaPosts: 6283 |
The ADC command uses DMA and I am not sure what bit resolution is used. I doubt if there is any oversampling. As usual, users should do their own tests to see if it suits their application. Jim VK7JH MMedit |
||||
| zeitfest Guru Joined: 31/07/2019 Location: AustraliaPosts: 582 |
Typically for micros the hardware digitisation precision is cited as (eg) 12 bit, but then the (digital) output data has to be handled in a variable of some sort, so the "result" usually has to be stored in a 16-bit word. I don't know if MMBasic has 16 bit words - I think it uses 32/64 bits for integers? As noted above, sometimes oversampling is used which increases the precision to say 16 bit at the expense of speed. Most micro A-D's and the associated designs have issues and getting accurate precise and clean even for 12-bit measurements is a lot harder than it first appears. In general I regard the ability to measure an accurate quantitative value to be a fundament of technology so I persisted at it !. Finally I was able to get good results on a picMZ2048 (edited typo) with a careful design for the analog inputs etc, but most other micros are simply not up to it. The best way is to have a separate A-D board and converter but that is more complex and expensive of course. Edited 2022-08-01 10:53 by zeitfest |
||||
| scruss Regular Member  Joined: 20/09/2021 Location: CanadaPosts: 91 |
For compatibility between boards. MicroPython used to have a mish-mash of routines to return ADC results, and they all used different precisions depending on hardware. ADC.read_u16() will always return a reading shifted to a 16-bit range on any hardware. |
||||
| matherp Guru Joined: 11/12/2012 Location: United KingdomPosts: 10315 |
As stated above the PicoMite PIN function internally takes 10 readings, discards the highest and lowest then averages the remaining 8 'middle' readings. The ADC command uses a single sample. However, as also stated above the Pi Pico ADC has a well documented bug in it's linearity As a result the Pico should be considered an 8-bit ADC If you need more you need an external ADC and then issues of power supply also become important |
||||
| chippy Newbie Joined: 30/07/2022 Location: United KingdomPosts: 3 |
Hi again, Many thanks to everyone who responded to my question about the a-to-d conversion on the pico when running MMBasic. I now have a much better understanding of the a-to-d operation on microcontrollers in general thanks to the feedback. From comments received it would seem that for best results its better to use a seperate a-to-d converter chip with a well de-coupled/bypassed reference and clean power supply to ensure minimum noise. John wrote: "By upscaled do you mean scaled? If not, what do you mean?" *** End quote *** I mean "scaled" of course, sorry for the confussion  Des (chippy) |
||||
| JohnS Guru Joined: 18/11/2011 Location: United KingdomPosts: 4044 |
No worries! John |
||||
 Print this page |
 |
The Back Shed's forum code is written, and hosted, in Australia. | © JAQ Software 2025 |
