24VAC - PhotoMOS voltage spike protection

[Sikanda]

Hello, I want to drive different solenoid valves with a PhotoMOS. I have selected the AQV252G from Panasonic for this purpose. The data sheet states that I should use an RC scrubber circuit or a varistor for AC voltages to protect the PhotoMOS.

How do I design the varistor if I don't know the inductive valve to be used beforehand? The design for protection against coils is always based on a known inductance. I have also noticed that the clamping voltage is often higher than the 60V that the AQV252G can tolerate. Would a TVS diode be a better alternative here? But here too I don't know which stored energy I should calculate.

The solenoid valves that will be connected have a power of approx. 8 to 16 W

[Zero999]

Use a bidirectional TVS diode, such as the P6KE43CA.

[Kleinstein]

For low voltages a TVS diode is often the better choice than a varistor.  With solonoid valves there can be additional energy from the moving core part. So just the inductance may not be enough to get the storred energy.  One can get an upper limit for the stored energy from power times the valve response time. Usually the energy is still quite a bit lower as much of the power loss is from the resistance.

[Sikanda]

One can get an upper limit for the stored energy from power times the valve response time.

By the reaction time of the valve, do you mean the time until the valve opens or closes? These values are very specific to the used valve.

[floobydust]

You can try measure/estimate the energy storage in the solenoid but it depends a bit on if the armature or pintle is pulled in or not.
Or simply oversize/use a large TVS as already recommended. It depends if this is cost or size sensitive design.
 
I would be more concerned about the inrush current. It can be quite high for the first few AC cycles (residual magnetism as well) and the PhotoMOS relay may not take it.

Solenoids can short-circuit with age so appropriate fusing is needed as well.

[Sikanda]

I have made a rough calculation based on a TI's description. According to TI, however, this only works if the curve is the same as in the data sheet. I have assumed a maximum current of 1.5A.  Is the value for C_CLAMP correct or am I doing something wrong here?
According to this, P6KE47CA would fit better as V_RWM is too close to the operating voltage V_Max.

https://www.ti.com/lit/an/slvae37/slvae37.pdf

V_Max = 24VAC * sqrt(2) * 110% = 37,37 V

P6KE43CA
V_RWM = 36,8V
V_R = 40,9 V_min / 45,2 V_max
Ipp = 10,3A -> 10,3 * 80% @ 85 °C = 8,24A
V_CLAMP = 59,3 V

R_DYN = (V_CLAMP - V_BR) / I_Surge = (59,3V - 45,2V)/10,3A = 1,37 Ohm

V_CLAMP = V_BR + I_Surge * R_DYN = 45,2V + 1,5 A * 1,37 Ohm = 47,26 V


P6KE47CA
V_RWM = 40,2V
V_R = 44,7 V_min / 49,40 V_max
Ipp = 9,4A -> 9,4 * 80% @ 85 °C = 7,52A
V_CLAMP = 64,8 V

R_DYN = (V_CLAMP - V_BR) / I_Surge = (64,8V - 49,4V)/9,4A = 1,64 Ohm

V_CLAMP = V_BR + I_Surge * R_DYN = 49,4V + 1,5 A * 1,64 Ohm = 51,86 V

[floobydust]

Reverse Stand-off Voltage V_WM is only at 1uA so I would not be concerned about that.
Don't forget to include the tempco of the TVS if this is used across wide temperature ranges (cold).
Littelfuse datasheets omit this but Vishay does not. Spec of 0.101%/°C gives several volts difference cold to hot, something I include in my math.

[Zero999]

I have made a rough calculation based on a TI's description. According to TI, however, this only works if the curve is the same as in the data sheet. I have assumed a maximum current of 1.5A.  Is the value for C_CLAMP correct or am I doing something wrong here?
According to this, P6KE47CA would fit better as V_RWM is too close to the operating voltage V_Max.

https://www.ti.com/lit/an/slvae37/slvae37.pdf

V_Max = 24VAC * sqrt(2) * 110% = 37,37 V

P6KE43CA
V_RWM = 36,8V
V_R = 40,9 V_min / 45,2 V_max
Ipp = 10,3A -> 10,3 * 80% @ 85 °C = 8,24A
V_CLAMP = 59,3 V

R_DYN = (V_CLAMP - V_BR) / I_Surge = (59,3V - 45,2V)/10,3A = 1,37 Ohm

V_CLAMP = V_BR + I_Surge * R_DYN = 45,2V + 1,5 A * 1,37 Ohm = 47,26 V


P6KE47CA
V_RWM = 40,2V
V_R = 44,7 V_min / 49,40 V_max
Ipp = 9,4A -> 9,4 * 80% @ 85 °C = 7,52A
V_CLAMP = 64,8 V

R_DYN = (V_CLAMP - V_BR) / I_Surge = (64,8V - 49,4V)/9,4A = 1,64 Ohm

V_CLAMP = V_BR + I_Surge * R_DYN = 49,4V + 1,5 A * 1,64 Ohm = 51,86 V

Either will be suitable. The current will still be tiny. I didn't perform any calculations, just looked at the maximum figures. The lower current will mean the clamping voltage is lower, as your calculations show, so the P6KE47CA will also be fine.