Solenoid driver: unwanted oscillation

[crutschow]

Sorry, I measured at pin 3 of 555

The voltage at pin 3 is 0V when the MOSFET is off.

 

[earckens]

The voltage at pin 3 is 0V when the MOSFET is off.

When running with a resistive load on the power mosfet then pin 3 pulses nicely. When the solenoid is connected pin 3 is showing the behaviour described above: counter-emv pulses, and no low voltage.

 

[crutschow]

When running with a resistive load on the power mosfet then pin 3 pulses nicely. When the solenoid is connected pin 3 is showing the behaviour described above: counter-emv pulses, and no low voltage.

Then you need to determine where this counter-emv is coming from, since that's obviously not normal.
It sounds like a ground problem.
You need to think about where the high ground current flows and follow good grounding and signal flow design.
Random connection of signal paths can result in random circuit operation.

• The 555 should be decoupled as I mentioned in post #39.
• The return ground from the MOSFET source should go directly to the supply common and that should be the only return to the supply. The 555 circuit ground connects to the MOSFET source terminal also (single point ground).
• You should have a decoupling capacitor (100μf or more) directly from the V+ solenoid connection to the MOSFET source terminal.
• Is the diode connected directly across the solenoid terminals?

 

[earckens]

Then you need to determine where this counter-emv is coming from, since that's obviously not normal.
It sounds like a ground problem.
You need to think about where the high ground current flows and follow good grounding and signal flow design.
Random connection of signal paths can result in random circuit operation.

• The 555 should be decoupled as I mentioned in post #39.
• The return ground from the MOSFET source should go directly to the supply common and that should be the only return to the supply. The 555 circuit ground connects to the MOSFET source terminal also (single point ground).
• You should have a decoupling capacitor (100μf or more) directly from the V+ solenoid connection to the MOSFET source terminal.
• Is the diode connected directly across the solenoid terminals?

Rebuilt the breadboard setup, it works well now! I suspect I had not made proper ground connections and/or any of your recommendations. Thank you crutschow!
Very nice sound to hear the pulses going through the coil. Current consumption can be diminished to about 60mA (1/6 duty cycle, 16%) before the coil cannot hold anymore.
Great circuit!

Edit: yellow is pin 3, blue is mosfet drain.

 

[crutschow]

Glad you got it working.
Those waveforms look very good, and not that different from my simulation (always nice to know when the actual circuit performance is close to the simulation).

 

[earckens]

...always nice to know when the actual circuit performance is close to the simulation...

I imagined, that is why I included them, and as a thank you for your work on this subject.

 

[earckens]

Glad you got it working.
Those waveforms look very good, and not that different from my simulation (always nice to know when the actual circuit performance is close to the simulation).

Hi crutschow, I have two fabrication questions:
1. for the trimpot would you use a simple 270 degree pot, or a multiturn trimmer?
2. would you replace the 2x 1N4148 by Schottky diodes (less voltage drop)?

 

[crutschow]

1. for the trimpot would you use a simple 270 degree pot, or a multiturn trimmer?

I would use a multiturn trimmer since it gives much better resolution and you are not likely to change its setting very often.

2. would you replace the 2x 1N4148 by Schottky diodes (less voltage drop)?

The feedback timing signal at that point switches been 0V and 12V so the few tenths of a volt difference in voltage drop between the 1N4148 junction diode and a Schottky does not cause any important effect in the circuit operation.

Schottky's are great if their lower voltage drop is helpful, such as for power rectifiers or low level signal rectification, but here it has no significant advantage.
You can use them if you like, however.

 

[earckens]

I would use a multiturn trimmer since it gives much better resolution and you are not likely to change its setting very often.
The feedback timing signal at that point switches been 0V and 12V so the few tenths of a volt difference in voltage drop between the 1N4148 junction diode and a Schottky does not cause any important effect in the circuit operation.

Schottky's are great if their lower voltage drop is helpful, such as for power rectifiers or low level signal rectification, but here it has no significant advantage.
You can use them if you like, however.

Can I send you a pm?

 

[crutschow]

I prefer to reply to all questions here.

 

[earckens]

I prefer to reply to all questions here.

If you wish I can send you a preassembled pcb when ready.

 

[crutschow]

If you wish I can send you a preassembled pcb when ready.

Sorry, I don't quite follow.
Why would be the purpose of doing that?

 

[earckens]

Sorry, I don't quite follow.
Why would be the purpose of doing that?

Hi, there is no purpose. This happens sometimes; and anyone in this thread wishing the same just needs to let know.

 

[crutschow]

Here is a variation of the circuit for your interest, that uses no current when off.

 

[earckens]

Here is a variation of the circuit for your interest, that uses no current when off.

View attachment 112013

Dammit, I submitted my board for fabrication yesterday; this will become version v2