Also, the value of R1 will determine (possibly) the discharge rate of the MOSFET gate charge, depending on how the micro +5V pin operates. So, keep the value a bit on the smaller side, like in the 1k ohms range.
If using discrete through the hole MOSFETs, look into ones like the IRL540 which are common and relatively cheap. They are rated 100V and 28A at 25C with a very small RDSon of 0.077 ohms. With 4V on the gate, RDSon rises a bit to 0.11 ohms, but still very low for triggering relay coils

 

Do you realize that having a diode across the coil will change the response of the solenoid action quite a bit?

Generally it's only the turn-on time of a pin-ball flipper that's important, not it's release time.

 

Generally it's only the turn-on time of a pin-ball flipper that's important, not it's release time.

Great, to me the release time is important is why I keep tilting the darn machines.

 

Great, to me the release time is important is why I keep tilting the darn machines.

Sorry, but I don't see what the release time of a pin-ball flipper has to do with tilting the machines(?).

 

Sorry, but I don't see what the release time of a pin-ball flipper has to do with tilting the machines(?).

If the flipper release time is too slow, the flipper cannot recover and provide maximum power to a second shot, particularly when the first shot hits a nearby post, sending the ball right back to the flipper.

You can slow down the ball by shifting the table. But too hard and you’ll tilt the machine.

Have you ever played pinball? Not being snarky; just asking.

 

If using discrete through the hole MOSFETs, look into ones like the IRL540 which are common and relatively cheap. They are rated 100V and 28A at 25C with a very small RDSon of 0.077 ohms. With 4V on the gate, RDSon rises a bit to 0.11 ohms, but still very low for triggering relay coils

I have IRF540 on hand but I think I'll go with IRLZ44ZPBF for my next go-to MOSFET. Rated 55V, 51A and Rds of 0.023Ω. Just 99¢.

 

Great, to me the release time is important is why I keep tilting the darn machines.

About what times roughly are we talking here?

And is this release time purely introduced by the pulldown resistor for the MOSFET? Because like I said before, the rest is standard pinball machine mechanics, so then these release times should be in every machine? In every newer pinball machine, the flipper buttons are running on a 5V circuit, switching the 48V circuit using a transitor, so isn't that the same?

 

About what times roughly are we talking here?

And is this release time purely introduced by the pulldown resistor for the MOSFET? Because like I said before, the rest is standard pinball machine mechanics, so then these release times should be in every machine? In every newer pinball machine, the flipper buttons are running on a 5V circuit, switching the 48V circuit using a transitor, so isn't that the same?

A high-current path to ground for the MOSFET gate would ensure a fast turn-off there. Your Arduino may not do that and the pull-down resistor is a compromise. BUT, I'm not so sure it's the MOSFET that is the issue. The mechanical spring-back might be the limiting factor in response time. So your design may be fine.

 

In addition to the mosfet switch-off time, the diode on a coil keeps the current flowing until all the energy is dissipated in the loop resistance. THAT is the mechanism by which it reduces the turn-off voltage spike. And it is very real, I have used that to quiet noisy lighting contactors, with 2 adequate diodes in series back to back, and the contactor coil across one of the diodes. A cheap and simple way to end a noise complaint in a hurry.

 

A high-current path to ground for the MOSFET gate would ensure a fast turn-off there. Your Arduino may not do that and the pull-down resistor is a compromise. BUT, I'm not so sure it's the MOSFET that is the issue. The mechanical spring-back might be the limiting factor in response time. So your design may be fine.

The Arduino current sinking is limited to 40mA (or less). A gate resistor is recommended to protect the MCU. Also, a pull down resistor on the gate is recommended, even more so than the current limiting resistor.

 

The Arduino current sinking is limited to 40mA (or less). A gate resistor is recommended to protect the MCU. Also, a pull down resistor on the gate is recommended, even more so than the current limiting resistor.

All true, and all working against a fast turn-off. I'm just not convinced the MOSFET turn-off time is the important factor in flipper responsiveness.

 

Have you ever played pinball? Not being snarky; just asking.

Yes.
But I haven't encountered the particular scenario you are referring to.

If a faster turn-off is really needed than a small resistor in series with the diode (say equal to the solenoid resistance) will significantly reduce the turn-off time
The trade-off is that it will increase the turn-off voltage across the MOSFET to twice the supply voltage.

 

I'm following this dicussion with great interest, and I thought to share some of the coil charactaristics that might influence the discussion direction:

The coil resistance I1 = 3.7 Ω (powerful "kick") and I2 = 133 Ω (weak "hold"). The plunger (pulled into the coil by magnetism) is spring loaded, so will be pulled back out of the coil if the magnetism is fading away. The coil will be in the low resistance for a very short moment of time until the plunger is pulled into the coil. From that moment the coil operates on high resistance until the power is switched off.

 

I have IRF540 on hand but I think I'll go with IRLZ44ZPBF for my next go-to MOSFET. Rated 55V, 51A and Rds of 0.023Ω. Just 99¢.

Be careful with a 55V MOSFET when you are running/switching a 48V supply. Any spike on the voltage supply itself could damage the MOSFET.
IRF540 is not good for 5V gate voltages, it should be the IRL540 (or equivalent substitute) when using a 5V device to a MOSFET gate

 

Be careful with a 55V MOSFET when you are running/switching a 48V supply. Any spike on the voltage supply itself could damage the MOSFET.
IRF540 is not good for 5V gate voltages, it should be the IRL540 (or equivalent substitute) when using a 5V device to a MOSFET gate

Agreed on both counts.

 

All true, and all working against a fast turn-off. I'm just not convinced the MOSFET turn-off time is the important factor in flipper responsiveness.

You didn’t answer my question. If you were a pinball wizard, I could take your comments more seriously. Pinball requires several basic skills and flipper control is an important one. Each machine has a different “feel” to its flippers and recovery time is an important, if not most important, attribute. To illustrate, tilt threshold and bumper angles are other attributes that contribute to successful play of at able. A wizard knows by muscle memory or feel, what that parameter is on a given table.

 

I am now suggesting a fast transistor driver circuit between the arduino output and the mosfet. A totem-pole style driver, since it is a true switching circuit. That would provide turn on and turn off levels to the gate much faster. That would solve the response time problem. And the last time I played pinball was in 1968. Not a wizard then, either. So I will take his word about needed flipper speed.

 

I am now suggesting a fast transistor driver circuit between the arduino output and the mosfet. A totem-pole style driver, since it is a true switching circuit. That would provide turn on and turn off levels to the gate much faster. That would solve the response time problem. And the last time I played pinball was in 1968. Not a wizard then, either. So I will take his word about needed flipper speed.

What does that mean for the schematic and components?

 

I am now suggesting a fast transistor driver circuit between the arduino output and the mosfet. A totem-pole style driver, since it is a true switching circuit. That would provide turn on and turn off levels to the gate much faster. That would solve the response time problem.

Even with a slow MOSFET driver the MOSFET switching is still likely in the tens of microseconds region, which is negligible compared to the tens of ms response of the typical solenoid.

 

You didn’t answer my question. If you were a pinball wizard, I could take your comments more seriously. Pinball requires several basic skills and flipper control is an important one. Each machine has a different “feel” to its flippers and recovery time is an important, if not most important, attribute. To illustrate, tilt threshold and bumper angles are other attributes that contribute to successful play of at able. A wizard knows by muscle memory or feel, what that parameter is on a given table.

I just asked a friend that has a similar setup, except he's not using a microcontroller to switch the flippers, but just a 5V circuit along the buttons and a MOSFET to switch the 48V circuit. In the video he showed me, the response times seem to be fine at first glance, but I indeed would not know how it "feels".

Would this kind of setup not suffer the slower release, or just less, or the same?