Relay Question

[Buffalo Tom]

I'm using the Panasonic TQ4-L-5V, which is a single-coil latching relay. I'll be triggering it using an ATtiny to send set/reset pulses. How long should the pulse duration be?

The datasheet specifies a maximum set time of 3 ms (without contact bounce??). Does that mean I should aim for exactly 3 ms, or is something like 5 ms better to ensure reliable switching?

Also, I'm not using a flyback diode across the coil. From what I've read, it's common not to use diodes with single-coil latching relays that only receive brief pulses. Can anyone confirm if this is the correct approach?


https://api.pim.na.industrial.panasonic.com/file_stream/main/fileversion/4514

[antonis]

From what I've read, it's common not to use diodes with single-coil latching relays that only receive brief pulses. Can anyone confirm if this is the correct approach?

Not me...
(unless, of course, there is no back EMF when relay is released..)

[Buffalo Tom]

From what I've read, it's common not to use diodes with single-coil latching relays that only receive brief pulses. Can anyone confirm if this is the correct approach?

Not me...
(unless, of course, there is no back EMF when relay is released..)

OK! So you are using a diode on single coil latching pulse relays?? I dont understand how? The polarity is changing for set and reset. On all DIY kits I found (Jack Orman etc) There is no diode...

Ohh! Found this post: https://forum.pedalpcb.com/threads/flyback-diodes-and-relays.3576/#post-28437

[antonis]

Back to back Zener diodes or resistor+capacitor configuration..

[merlinb]

The datasheet specifies a maximum set time of 3 ms (without contact bounce??). Does that mean I should aim for exactly 3 ms, or is something like 5 ms better to ensure reliable switching?

Datasheet, page 17:
"Regarding the set and reset pulse time, for the purpose of reliable operation under ambient temperature fluctuations and different operating conditions, we recommend setting the coil applied set and reset pulse time to 10 ms or more at the rated coil voltage."

[Buffalo Tom]

Datasheet, page 17:
"Regarding the set and reset pulse time, for the purpose of reliable operation under ambient temperature fluctuations and different operating conditions, we recommend setting the coil applied set and reset pulse time to 10 ms or more at the rated coil voltage."

Oh! Thanks!

[Buffalo Tom]

Back to back Zener diodes or resistor+capacitor configuration..

Ok thanks! So two diodes! But I never seen it on any of those true bypass relay kits available that's using Attinys and one coil latching relays. For example here is the AMZ version.


Did you read this: https://forum.pedalpcb.com/threads/flyback-diodes-and-relays.3576/#post-28437
He came to the conclusion: So now we know that the ATTINY can drive the relay reliably and it can absorb the inductive kick-back without damage.

This makes me unsure if I should use diodes or not? The pcb is small and I have two relays there and a Attiny 84... So it will require 4 diodes

[merlinb]

For single-coil relays it depends on details of the driving circuit, whether H-bridge or half bridge, using MOSFETs (which have body diodes) or BJTs (which don't), how the logic works etc.

• If either end of the driving circuit goes completely 'open' after the pulse, then you may need flyback protection.
• But if the driving circuit goes open after the pulse but there are built-in diodes to the rail (there often are inside ICs and MOSFETs), then flyback can be suppressed through the power supply itself.

So it depends. But back-to-back Zeners can't hurt, if you want to be certain.

[Buffalo Tom]

Im trying this without diodes. If it works for Jack Orman it should work for me. To stress test the ATtiny I came up with this code. So my ATtiny84 is cranking out over 3,000 full relay flips every hour. I will let it ride and we'll check in after 24 hours for a full 72,000+ operations report.

Code Select Expand

#define RELAY_A_POS 6
#define RELAY_A_NEG 5
#define RELAY_B_POS 7
#define RELAY_B_NEG 4

const unsigned long PULSE_DURATION = 12; // Pulse width in ms

// Delay pattern in milliseconds
const unsigned long INTERVALS[] = {
  3000, 3000, 3000,   // 3 sec
  1000, 1000, 1000,   // 1 sec
  500,  500,  500,    // 0.5 sec
  250,  250,  250     // 0.25 sec
};
const uint8_t NUM_INTERVALS = sizeof(INTERVALS) / sizeof(INTERVALS[0]);

uint8_t currentIntervalIndex = 0;

void setup() {
  pinMode(RELAY_A_POS, OUTPUT);
  pinMode(RELAY_A_NEG, OUTPUT);
  pinMode(RELAY_B_POS, OUTPUT);
  pinMode(RELAY_B_NEG, OUTPUT);

  digitalWrite(RELAY_A_POS, LOW);
  digitalWrite(RELAY_A_NEG, LOW);
  digitalWrite(RELAY_B_POS, LOW);
  digitalWrite(RELAY_B_NEG, LOW);

  // Initial RESET pulse on startup
  pulseRelay(RELAY_A_NEG, RELAY_A_POS);
  pulseRelay(RELAY_B_NEG, RELAY_B_POS);
}

void loop() {
  unsigned long delayTime = INTERVALS[currentIntervalIndex];

  // SET both relays
  pulseRelay(RELAY_A_POS, RELAY_A_NEG);
  pulseRelay(RELAY_B_POS, RELAY_B_NEG);
  delay(delayTime);

  // RESET both relays
  pulseRelay(RELAY_A_NEG, RELAY_A_POS);
  pulseRelay(RELAY_B_NEG, RELAY_B_POS);
  delay(delayTime);

  // Move to the next pattern step
  currentIntervalIndex++;
  if (currentIntervalIndex >= NUM_INTERVALS) {
    currentIntervalIndex = 0; // loop back to start
  }
}

void pulseRelay(uint8_t highPin, uint8_t lowPin) {
  digitalWrite(highPin, HIGH);
  digitalWrite(lowPin, LOW);
  delay(PULSE_DURATION);
  digitalWrite(highPin, LOW);
  digitalWrite(lowPin, LOW);
}


[Buffalo Tom]

The code has now been running continuously for 20 hours, during which the ATtiny84 has performed:

Approximately 60,624 SET+RESET operations (relay flips)
Around 121,248 total pin pulses driving the relay coils
There are no signs of damage, and the relays continue to operate flawlessly.

This strongly suggests that the ATtiny84 can reliably drive the Panasonic TQ2-L-5V latching relay and handle the inductive kick-back without issue.