Generating two complementary signals on 12 V supply on/off

[CircutScoper]

Sorry about my inputs earlier. Looks like my multimeter is reporting faulty results. When checked with scope voltages are 12.4 and 5.2 Max peaks. So I guess we are good there.

Adding 1K resistor did not help. I interchanged Q1 and Q2 just to check if the Xsistor is faulty. But both are fine.

Somehow the required biasing is not getting achieved for Q1. Can you pls relook at the ckt. At the moment I've 1k also connected from +ve of the cap to gnd.

Please connect your scope to the + end of the cap. What do you see there when you disconnect +12V?

 

[dashy1981]

On +cap, on 12v disconnection, I see a decaying signal. Its almost 3.8 sec long.
Additionally, I also checked when switching on, Q2 base receives a short pulse of 0.78V wrt gnd and it corresponds to the time period of approx 35-38ms. But that we know it works.

 

[CircutScoper]

On +cap, on 12v disconnection, I see a decaying signal. Its almost 3.8 sec long.
Additionally, I also checked when switching on, Q2 base receives a short pulse of 0.78V wrt gnd and it corresponds to the time period of approx 35-38ms. But that we know it works.

That slow decay is the problem and Analog Kid astutely warned that it might occur when he said: If the output of the 12 V supply wanders down to 0 V slowly (very possible with a light load), the slew rate might be below the 10 uF differentiator time constant.

It's too slow for the 20ms timeconstant of C1 to detect it.

One fix would be to increase C1 to 2000uF.

 

[dashy1981]

That slow decay is the problem and Analog Kid astutely warned that it might occur when he said: If the output of the 12 V supply wanders down to 0 V slowly (very possible with a light load), the slew rate might be below the 10 uF differentiator time constant.

It's too slow for the 20ms timeconstant of C1 to detect it.

One fix would be to increase C1 to 2000uF.

2000uf also did not help...

 

[CircutScoper]

2000uf also did not help...

2000uf also did not help...

Is the 1k resistor still there?

 

[CircutScoper]

2000uf also did not help...

The basic problem seems to be that +12V has hardly anything loading it to pull it to zero when you turn it off. That's why its fall is too slow for my circuit to detect.

In your intended final application, what will be connected to and powered by the +12V supply? Will it still be practically nothing?

 

[dashy1981]

The basic problem seems to be that +12V has hardly anything loading it to pull it to zero when you turn it off. That's why its fall is too slow for my circuit to detect.

In your intended final application, what will be connected to and powered by the +12V supply? Will it still be practically nothing?

No... It shall be powering two nema17 motors

 

[dashy1981]

And yes 1k is still connected

 

[CircutScoper]

No... It shall be powering two nema17 motors

Then I suggest you connect two nema17 motors (or an equivalent load = 3Amp = 4Ohms?) and see how the circuit behaves then.

 

[dashy1981]

Then I suggest you connect two nema17 motors (or an equivalent load = 3Amp = 4Ohms?) and see how the circuit behaves then.

Tried... No luck both with 2000uf and 10 uf

 

[CircutScoper]

Tried... No luck both with 2000uf and 10 uf

What does the 12V discharge curve look like now?

 

[CircutScoper]

On +cap, on 12v disconnection, I see a decaying signal. Its almost 3.8 sec long.
Additionally, I also checked when switching on, Q2 base receives a short pulse of 0.78V wrt gnd and it corresponds to the time period of approx 35-38ms. But that we know it works.

Okay. How about this change of plan? Add our old friend the 555 to generate nice snappy logic transitions when 12V crosses 10V (when turning ON) and 5V (turning OFF) no matter HOW long it takes to do it.

It adds complexity but avoids the transition time issue altogether.

 

[dashy1981]

Okay. How about this change of plan? Add our old friend the 555 to generate nice snappy logic transitions when 12V crosses 10V (when turning ON) and 5V (turning OFF) no matter HOW long it takes to do it.

It adds complexity but avoids the transition time issue altogether.

View attachment 55594

I like the idea... But Is this dependent on any way on the reliability of my cheap 12 v smps. It seems to give more than 12 v. So do you suggest any 12 v regulator ic at input ?

 

[CircutScoper]

I like the idea... But Is this dependent on any way on the reliability of my cheap 12 v smps. It seems to give more than 12 v. So do you suggest any 12 v regulator ic at input ?

I thought you said your 12V supply is giving 12.4V. That's only about 3% off. But if you think you need better, you can buy an efficient regulated switching supply for $10 or so.

 

[dashy1981]

I thought you said your 12V supply is giving 12.4V. That's only about 3% off. But if you think you need better, you can buy an efficient regulated switching supply for $10 or so.

That's right at the moment it's 12.4 but have seen cheap 12 v supplies over the usage deteriorating and giving higher voltage. So just wanted to check if the trigger for 555 can in any way made indepent from the variations.

555 I've ordered today will get in couple of days. Will update once tried

 

[CircutScoper]

That's right at the moment it's 12.4 but have seen cheap 12 v supplies over the usage deteriorating and giving higher voltage. So just wanted to check if the trigger for 555 can in any way made indepent from the variations.

555 I've ordered today will get in couple of days. Will update once tried

The 555's trigger points are proportional to the 5V supply. So long as it's "close" to 5V, "12V-ON" will occur at 10V, "12V-OFF" at 5.

 

[dashy1981]

Nope... Same results... Q2 never seems to turn on... So no pulse on 12v on.

 

[dashy1981]

Nope... Same results... Q2 never seems to turn on... So no pulse on 12v on.

12v off produces a pulse as earlier. :-(

 

[CircutScoper]

12v off produces a pulse as earlier. :-(

What do you see on your scope when connected to the common node of R4&R5 when you cycle 12V on/off?

It should be a ~0.7V positive pulse when 12V cycles off, and ~-0.7V negative pulse when 12V cycles on.

 

[CircutScoper]

What do you see on your scope when connected to the common node of R4&R5 when you cycle 12V on/off?

It should be a ~0.7V positive pulse when 12V cycles off, and ~-0.7V negative pulse when 12V cycles on.

PS: I suspect you'll see a larger amplitude negative pulse. I.e. > -2V. If you do, this will indicate Q2 is either defective or not properly connected. Most likely the latter.

It's also possible that Q2's collector connections are faulty, in which case the R4&R5 negataive pulse will be normal.