Good job and novel approach. Ive done this before, very similar, but trying to go from from pwm back to analog can be difficult. What I did was same front end: dc:dc , pic, but I just sent the adc value via serial over optocoupler. i used 12 bit adc, and the way I sent it was 3 bytes: upper 4bits held a code: 0010 - lower 4 bits hold nibble 0 0011 - lower 4 bits hold nibble 1 0100 - lower 4 bits hold nibble 2 On the receiver side when I got nibble 2, then I knew I had a complete 12 bits. to go back to analog, you can use a dac, or just use the digital to create a dispay.
Thank you, I can imagine a DAC being much better for output quality (no switching ripple). And I suppose serialising the digitised input should allow for higher bandwidths too, making better use of the optocoupler's bandwidth. Also in the majority of systems the signal will be digitised anyway so in that case just placing the ADC on the isolated side and transferring the serial data makes much more sense.
@@electrarc240 but I do like your attemp, its a good exercies to know how things work, in my lab I have lots of failures :) to learn from. FYI: the reason I use a split byte for sending data is so I dont have to worry about syncing data and stuff like that and since its sending fast, a missed byte doesnt really matter much. a simple 3 case swith works: switch (inByte &0xF0) { case 0x02: tval = inByte & 0x0F; break; case 0x03: tval += (inByte
I am a retired electronics and software engineer, and I watch your channel because I love to see a young mind going through the steps of gaining experience. I hope many other students and hobbyists for that matter watch you channel. Keep up the GREAT work man! :)
Just found your Chanel where have you been 😂 very informative content loved it 👍 can i ask you to do some videos about inverters so called (Grid tied invertes) how do they sync with the grid frequency....ty
Analog Devices has one that is totally optical isolated. I've seen some that use a piezo transducer to send ultrasonic s down a glass rod. A PWM output on locally floating sense feeds optical back. Good for surgical patient contact isolated system.
I thought HV battery in EVs was on same ground as the low voltage. They do have leakage from HV to chassis detection and safety interlock that cuts the battery off when any connections are exposed.
Just did some Googling I can only find stuff that talks about full galvanic isolation of the HV. That’s how it’s done in Formula Student so I assume it would be somewhat similar in road cars. We use an insulation monitoring device to ensure there’s never a current path to chassis ground from either terminal.
Hi.❤❤your knowledg is very good.if can help to us to design and build in action to build smps sytems.ducumentary is very but are very different and unbuildable.
THe have the PERFECT KEYBOARD, ....I would never use any other, ....but apart from not being made any more, and has several semi-serious design issues (like the dreaded volume control issue), it is INSANELY PRICED!!!!!!!!!!!!!!!!!!!!!!!!
I have a pile of PIC12F683's as well. BUT I UTTERLY DESPISE the fact that the internal cpu clock is way less that the max, which would require an external xtal. This is a unique 8-PIN DEVICE (not in any larger pacakge),... that is INSANE and EVIL. ....I will NEVER EVER forgive them for this! Theirs and (ATTiny13's as well) cpus are BANNED FOREVER in my designs, ...including their recent cpu's that do not do this.
Good job and novel approach.
Ive done this before, very similar, but trying to go from from pwm back to analog can be difficult. What I did was same front end: dc:dc , pic, but I just sent the adc value via serial over optocoupler.
i used 12 bit adc, and the way I sent it was 3 bytes: upper 4bits held a code:
0010 - lower 4 bits hold nibble 0
0011 - lower 4 bits hold nibble 1
0100 - lower 4 bits hold nibble 2
On the receiver side when I got nibble 2, then I knew I had a complete 12 bits.
to go back to analog, you can use a dac, or just use the digital to create a dispay.
Thank you, I can imagine a DAC being much better for output quality (no switching ripple). And I suppose serialising the digitised input should allow for higher bandwidths too, making better use of the optocoupler's bandwidth. Also in the majority of systems the signal will be digitised anyway so in that case just placing the ADC on the isolated side and transferring the serial data makes much more sense.
@@electrarc240 but I do like your attemp, its a good exercies to know how things work, in my lab I have lots of failures :) to learn from.
FYI: the reason I use a split byte for sending data is so I dont have to worry about syncing data and stuff like that and since its sending fast, a missed byte doesnt really matter much. a simple 3 case swith works:
switch (inByte &0xF0)
{
case 0x02: tval = inByte & 0x0F; break;
case 0x03: tval += (inByte
I am a retired electronics and software engineer, and I watch your channel because I love to see a young mind going through the steps of gaining experience.
I hope many other students and hobbyists for that matter watch you channel. Keep up the GREAT work man! :)
Any chance you can include a schematic of the circuit used for each series of the test, beyond just telling us what or how each element is connected?
Such a good video! Also custom FS Inverter 👀😁
Try with IL300 or VOA300 analog isolated devices of Vishay. Nice work.
An IL300 is in the Mail right now! Video on that plus a custom board will be coming in a few weeks
This an was excellent very informative video, thank you for taking the time!
Just found your Chanel where have you been 😂 very informative content loved it 👍 can i ask you to do some videos about inverters so called (Grid tied invertes) how do they sync with the grid frequency....ty
Nice bench.
Analog Devices has one that is totally optical isolated. I've seen some that use a piezo transducer to send ultrasonic s down a glass rod. A PWM output on locally floating sense feeds optical back. Good for surgical patient contact isolated system.
👍👍
I thought HV battery in EVs was on same ground as the low voltage. They do have leakage from HV to chassis detection and safety interlock that cuts the battery off when any connections are exposed.
Just did some Googling I can only find stuff that talks about full galvanic isolation of the HV. That’s how it’s done in Formula Student so I assume it would be somewhat similar in road cars. We use an insulation monitoring device to ensure there’s never a current path to chassis ground from either terminal.
Could you give some details on the pic12 model and programming ? I am trying to replicate the circuit.
I'm going to have a look for the code over the next few days, should be fairly simple
You can use Arduino too.
@@adaminsanoff Yes and probably much easier to get going
Hi.❤❤your knowledg is very good.if can help to us to design and build in action to build smps sytems.ducumentary is very but are very different and unbuildable.
Can pls mention the name of ic used for pwm generation?
It was just a PIC12 programmed to output a duty cycle proportional to the ADC reading
hi.without schematic circuit diagram expelaination is poor. thank you
THe have the PERFECT KEYBOARD, ....I would never use any other, ....but apart from not being made any more, and has several semi-serious design issues (like the dreaded volume control issue), it is INSANELY PRICED!!!!!!!!!!!!!!!!!!!!!!!!
I have a pile of PIC12F683's as well.
BUT I UTTERLY DESPISE the fact that the internal cpu clock is way less that the max, which would require an external xtal.
This is a unique 8-PIN DEVICE (not in any larger pacakge),... that is INSANE and EVIL.
....I will NEVER EVER forgive them for this! Theirs and (ATTiny13's as well) cpus are BANNED FOREVER in my designs, ...including their recent cpu's that do not do this.
Poor Microchip 🥲. Though I do agree!