I have been thinking about what to implement as a soft start for my F5T V3 build for a while now. Something simple enough and uses readily available, inexpensive parts.
The plan that I arrived at involves the use a NTC and then shunt it with a relay after a short delay - the typical approach. To set the delay time, I am going to use a NE555 timer which allows setting a wide range of delay time simply by turning a trim port on the board. The timer does need a power supply (5, 12, or 24V). I am going to buy one of those tiny SMPS (12V) to power it. In my plan, the timer will turn on (with a preset delay time) a relay (higher current and voltage rating than the one on the timer board) to which the NTC is connected to control AC flow.
Basically, when the main AC switch is on, AC will flow through the relay and the primary of the transformer with the NTC in series. The timer circuit is also activated. After the preset delay time, the relay on the timer board will be turned on, which in turn will trigger the second relay to short out the NTC (to place the NTC in parallel with the relay contact). In this arrangement, even if both relays failed and never turn on, the transformer will still be connected to the main via the NTC when the switch is turn on. I should be able to catch that since there will be no relay clicks.
One aspect that I am pondering is how the relay contact resistance compares to the NTC resistance when the steady state is reached (the NTC is in parallel with the relay contact). I am worry that if there is always current flowing through the NTC, it will never cool back to ambient temperature. As such, it may not be able to offer protection if the main power suddenly got tripped off and then on again (long enough to deplete the cap bank but too short for the NTC to cool back to ambient).
To overcome the potential problem mentioned above, I can change the relay set up to completely switch the NTC out of the circuit. However, that involves switching the relay contact when the steady state is reached. I am thinking about possible arcing when this switching occurs and shortens the life of the relay. In addition, in a set up like this, the power supply will be momentarily off when the switching takes place. The charge stored in the capacitor bank will drop. According to spec sheet of the relay that I am think of using, the maximum switching time is about one cycle ~150 ms. Based on the current drawn of the amp (being class A, the drawn is relatively constant) , my estimation indicated that the capacity of the bank will be dropped by about 20% (from the max capacity). Thus, once the power re-connected, no big current spike is expected. So I think I am good.
The parts required for this set up is a small - a 12 V smps, a simple 555 timer board, and a relay board with the NTC installed to control the AC flow. My estimate for the total cost for a soft start as such is < 10 $ and all I have to do is wire them up (no PCB and soldering required).
Any comment or suggestions is welcome.
Regards,
The plan that I arrived at involves the use a NTC and then shunt it with a relay after a short delay - the typical approach. To set the delay time, I am going to use a NE555 timer which allows setting a wide range of delay time simply by turning a trim port on the board. The timer does need a power supply (5, 12, or 24V). I am going to buy one of those tiny SMPS (12V) to power it. In my plan, the timer will turn on (with a preset delay time) a relay (higher current and voltage rating than the one on the timer board) to which the NTC is connected to control AC flow.
Basically, when the main AC switch is on, AC will flow through the relay and the primary of the transformer with the NTC in series. The timer circuit is also activated. After the preset delay time, the relay on the timer board will be turned on, which in turn will trigger the second relay to short out the NTC (to place the NTC in parallel with the relay contact). In this arrangement, even if both relays failed and never turn on, the transformer will still be connected to the main via the NTC when the switch is turn on. I should be able to catch that since there will be no relay clicks.
One aspect that I am pondering is how the relay contact resistance compares to the NTC resistance when the steady state is reached (the NTC is in parallel with the relay contact). I am worry that if there is always current flowing through the NTC, it will never cool back to ambient temperature. As such, it may not be able to offer protection if the main power suddenly got tripped off and then on again (long enough to deplete the cap bank but too short for the NTC to cool back to ambient).
To overcome the potential problem mentioned above, I can change the relay set up to completely switch the NTC out of the circuit. However, that involves switching the relay contact when the steady state is reached. I am thinking about possible arcing when this switching occurs and shortens the life of the relay. In addition, in a set up like this, the power supply will be momentarily off when the switching takes place. The charge stored in the capacitor bank will drop. According to spec sheet of the relay that I am think of using, the maximum switching time is about one cycle ~150 ms. Based on the current drawn of the amp (being class A, the drawn is relatively constant) , my estimation indicated that the capacity of the bank will be dropped by about 20% (from the max capacity). Thus, once the power re-connected, no big current spike is expected. So I think I am good.
The parts required for this set up is a small - a 12 V smps, a simple 555 timer board, and a relay board with the NTC installed to control the AC flow. My estimate for the total cost for a soft start as such is < 10 $ and all I have to do is wire them up (no PCB and soldering required).
Any comment or suggestions is welcome.
Regards,