Bipolar PI design

[pechkata]

Bad news; I'll have to delay porting the code and build the original first.
I'm about to order pcbs from JLPcb.

Well done - right choice.

[ivconic]

They went crazy with postage and shipment!
5 pcbs are $9.90 and $16.46 are shipping!
It wasn't like that a few years ago.
The overall world political situation is the cause of this.
It's high time things to calm down and we get back to normal life.

[ivconic]

Done!

[ivconic]

I have good news and bad news.
It's good that the finally 2 PICs arrived from a friend from a neighboring country!
Incredibly "fast" and "expeditious" postal service!
Between my place and my friend's place there are even incredible 160 km !!!
It's a long way!
When there are no clouds, my friend and I can literally correspond with smoke signals!
In the days of ancient Rome, I remember, mail traveled much faster!
These 2 PICs traveled longer than if my cousin from Mars sent them to me!
Bad news; I'll have to delay porting the code and build the original first.
I'm about to order pcbs from JLPcb.





This is an excellent illustrative example of why I am so "successful" in this business and why I am literally sitting on "millions" earned from this business!
(Things would be much worse if I didn't have good friends to help.)

[Carl-NC]

See Eduardo's schematic, I think his threshold method will work.

[algan]

Hello, Carl, I will also try correcting the circuit with the CD4066, after which I copied the TL062 integral from the TDI circuit, and continued with your circuit, and from there I will continue with the TDI sound. There is also a potentiometer for thrash, what is your advice? Thank you.

[algan]

My suggestion is to use the const keyword to define constants that don't change during program execution

---

const uint8_t TXP_PIN = 9;
const uint8_t TXN_PIN = 10;
const uint8_t ADC_POT = A1;
const uint8_t TCLKP_PIN = 2;
const uint8_t GCLKP_PIN = 3;
const uint8_t TCLKN_PIN = 4;
const uint8_t GCLKN_PIN = 5;
const uint8_t BCLK_PIN = 6;
const uint8_t LED_PIN = 13;

and of time constants

---

const unsigned long TX_PERIOD = 200;
const unsigned long TX_WIDTH = 50;
const unsigned long MAIN_DELAY = 15;
const unsigned long MAIN_WIDTH = 10;
const unsigned long GND_DELAY = 5;
const unsigned long GND_WIDTH = 40;

[Carl-NC]

between TXP turning off and TXN turning on.
The MOSFETs (IRF9640 and IRF740) specification shows that the P-channel MOSFET turns off a little slower (~100ns), while the N-channel turns off faster (~80ns).
Suggestion: We can increase the dead time to 1-2µs to make sure there is no overlap.
TXP and TXN switching order
We need to make sure that first TXP (the P-channel MOSFET) turns off, wait for the dead time, then TXN (the N-channel MOSFET) turns on.
The same applies when switching from TXN back to TXP.

---

If we're talking about the circuit in Fig 22-18 (post 169) then the NMOS switches are self-driven. TXP & TXN refer to the positive and negative TX pulses, not the PMOS and NMOS switches.

[Op04]

Hi,
gpz7k run tx with irfi 4227, there's a reason for this right?

[algan]

between TXP turning off and TXN turning on.
The MOSFETs (IRF9640 and IRF740) specification shows that the P-channel MOSFET turns off a little slower (~100ns), while the N-channel turns off faster (~80ns).
Suggestion: We can increase the dead time to 1-2µs to make sure there is no overlap.
TXP and TXN switching order
We need to make sure that first TXP (the P-channel MOSFET) turns off, wait for the dead time, then TXN (the N-channel MOSFET) turns on.
The same applies when switching from TXN back to TXP.

---

void generatePulse() {
PORTB |= (1 << PB1); // TXP HIGH (P-channel MOSFET on)
PORTB &= ~(1 << PB2); // TXN LOW (N-channel MOSFET off)
delayMicroseconds(TX_WIDTH); // TX pulse width

PORTB &= ~(1 << PB1); // TXP LOW (turn off P-channel first)
delayMicroseconds(2); // DEAD TIME (~2 µs) - increased for safety

PORTB |= (1 << PB2); // TXN HIGH (turn on N-channel MOSFET)
delayMicroseconds(TX_WIDTH);

PORTB &= ~(1 << PB2); // TXN LOW (turn off N-channel MOSFET)
delayMicroseconds(2); // DEAD TIME (~2 µs) before next cycle
}

[ivconic]

Algan no, it is for this currrent Bipolar PI, the topic is about.
...
Pechkata,
...potentiometer, but in this present code is not used yet.
Code is not finished, it is under development.
You can only test it with Atmga328P WITHOUT mosfets.
I thought it would be a good idea to make a code suggestion, so other members could contribute.

[pechkata]

Exactly.
I didn't even use pot at this stage.
So I reserved the IRQ and put in a trivial LED for now.
There is no inconsistency, but the abbreviations for the pulses do not match on the graph, in the text and in the schematic.
Although I understand English, sometimes it takes me time to decipher some abbreviations and understand them properly.
...
I made a mistake in math (I was never good at it) and instead of 5kHz I'm measuring 1/2.


Hmmm... we need Joop here!
Porting is not easy for me...

Hello ivconic
Which key will you use to control the samplers?

[algan]

Hello Ivkonich, is the code for this generator? It is interesting..

[ivconic]

Exactly.
I didn't even use pot at this stage.
So I reserved the IRQ and put in a trivial LED for now.
There is no inconsistency, but the abbreviations for the pulses do not match on the graph, in the text and in the schematic.
Although I understand English, sometimes it takes me time to decipher some abbreviations and understand them properly.
...
I made a mistake in math (I was never good at it) and instead of 5kHz I'm measuring 1/2.


Hmmm... we need Joop here!
Porting is not easy for me...

[Carl-NC]

If you see inconsistencies, then point them out. I'll fix them. Meanwhile, somewhere you will need to use interrupts or you won't be able to read the pots with the ADC.