SystemD LiteAmp

Jump to Latest
S

sumeshak

Member
Joined 2014
Interesting.....

Looking great..!
Is this project availablefor DIY?
Can youpost some more details?
Thank you,
Sumesh

Reactance said:
Lockdown allowed for time.

I experimented with a dedicated supervisor circuit.
I have managed to complete 70% supervisor layout and 90% of the firmware is done as well.

Here is what it can do.

- x2 dc monitor voltage +/- up to 100V (programmable 1V steps).
- x1 dc offset +/- (programmable 1V steps).
- x2 +/- current rail monitor 20A (programmable)
- x3 temperature sensors each programmable.
- x1 signal shutdown command output.
- x1 switching frequency monitor.
- x1 4-wire fan output + Fan RPM monitor.
- x1 solid state relay. (will upgrade to x2 for full bridge application)
- x1 SPI control and communication. (I2C "master mode" is a dog to get right on a PIC16F18857)

The cost of doing this is no more than $20 parts, but the firmware took a while to get right. I have a breadboard of it working on the desk.
Click to expand...
 
Reactance

Reactance

Member
Joined 2005
sumeshak said:
Looking great..!
Is this project availablefor DIY?
Can youpost some more details?
Thank you,
Sumesh
Click to expand...

Hi - what is your interest in this? I think I have supplied enough details.. what other details are you interested in? most of the work DONE has been pushed into firmware, because of COVID I have not sent the R&D boards for fabrication, but I have messy breadboard version on my desk with the completed firmware. The circuit is basic analog configurations (dividers, RC filters, Integrator) feeding into ADC's for further sampling. (making ultra use Core Independent Peripherals).

I do need to take some critical measurements notably around the "shutdown command reaction speed" for OC and OV, I have no faith in the "programmable on-chip over current trips" they seem have a reaction time of ~300us and when destroyed along with the chip the output stage / amplifier is at the mercy of fuse protections or vaporized pcb tracks/silicon. Analog transistor current mirror trips are still by far the fastest sense lines, coupling that with firmware needs to be design carefully, its ISR needs priorities setup correctly for the different things, the PIC im using is a rather basic a "fun part" from that point of view no advance nested vector ISR system. I could have used an ARM chip but then I would be drowning in other complexities there good benefits and keeping the architecture straightforward. simple!

I used C with lots of MACRO functions and a limited deep stack on the PIC I have 30% flash remaining and I get an internal EEPROM. yes some say for real-time its C IS slow but this is half truth without measurement for the application in question. Some microcontrollers are on the dual core bandwagon (radio stack cores for RF are no longer a shared with the main processor) I guess these cores will find use in other areas over the years and become dirt cheap. (just not now and many years after COVID era has gone)

Problem will remain, when will innovation like this reach the "shores of DIY" others with the nohow is keeping groundbreaking research and development close to their chest while the rest play with ir20XX in the year 2021.
 
Last edited:
S

sumeshak

Member
Joined 2014
Thank you

Dear Sir,
Thank you for the reply.
I am learning PIC programming.
I found this project interesting and became curious.
Regards,
Sumesh


Reactance said:
Hi - what is your interest in this? I think I have supplied enough details.. what other details are you interested in? most of the work DONE has been pushed into firmware, because of COVID I have not sent the R&D boards for fabrication, but I have messy breadboard version on my desk with the completed firmware. The circuit is basic analog configurations (dividers, RC filters, Integrator) feeding into ADC's for further sampling. (making ultra use Core Independent Peripherals).

I do need to take some critical measurements notably around the "shutdown command reaction speed" for OC and OV, I have no faith in the "programmable on-chip over current trips" they seem have a reaction time of ~300us and when destroyed along with the chip the output stage / amplifier is at the mercy of fuse protections or vaporized pcb tracks/silicon. Analog transistor current mirror trips are still by far the fastest sense lines, coupling that with firmware needs to be design carefully, its ISR needs priorities setup correctly for the different things, the PIC im using is a rather basic a "fun part" from that point of view no advance nested vector ISR system. I could have used an ARM chip but then I would be drowning in other complexities there good benefits and keeping the architecture straightforward. simple!

I used C with lots of MACRO functions and a limited deep stack on the PIC I have 30% flash remaining and I get an internal EEPROM. yes some say for real-time its C IS slow but this is half truth without measurement for the application in question. Some microcontrollers are on the dual core bandwagon (radio stack cores for RF are no longer a shared with the main processor) I guess these cores will find use in other areas over the years and become dirt cheap. (just not now and many years after COVID era has gone)

Problem will remain, when will innovation like this reach the "shores of DIY" others with the nohow is keeping groundbreaking research and development close to their chest while the rest play with ir20XX in the year 2021.
Click to expand...
 
You must log in or register to reply here.
Top Bottom