# Learning moment: Can this guy's relay be replaced with this circuit? Video attached

#### Tonyr1084

Joined Sep 24, 2015
5,570
This guy has issues with his PVP and inverter. When he connects a specific load it kicks the inverter out because of inrush current exceeding the capacity of the inverter. He builds a bypass circuit using a resistor to drop current below the cutoff threshold and uses a relay to bypass the resistor once the inrush is past. He's calling it a soft start though it's not really a soft start.

I got to thinking (dangerous thing for me to do) and wondered if I can do the same thing without the use of the relay but instead using an SCR. Watch the video below and then comment on my proposal of using the P channel MOSFET (part unspecified) and the SCR to replace his approach of using the relay.

This is purely a learning exercise, so don't ask me what I'm trying to do. I'm just trying to learn to apply different approaches. And likely what I have drawn, since I know so little about MOSFET's, is probably incorrect. Your correction on the drawing would be appreciated along with an explanation of what I got wrong. Thanks.

#### nsaspook

Joined Aug 27, 2009
7,898

#### AlbertHall

Joined Jun 4, 2014
10,790
I see a possible problem which depends on details we don't have.
When the circuit is switched on, the capacitor holds the MOSFET gate low, thus the MOSFET and thyristor will both be off - so far so good.
The 3R3 will be feeding the output and during the time dealy the output voltage will rise, to what level we don't know.
Now the MOSFET gate voltage is rising and the source voltage cannot be lower than the (unknown) output voltage.
With the values shown the gate voltage can rise to only 6V and to switch on the MOSFET that needs to be greater than the source voltage by enough.
The problem could be helped by eliminating the lower 100k resistor which would then allow the gate to reach the full 12V.

#### Tonyr1084

Joined Sep 24, 2015
5,570
Did you watch the video? The guy considered that but didn't like the idea of the reduced current. And the time it would take the thermistor to cool down and increase its resistance so that if there's a momentary interruption in voltage the unit won't shut down.

And he's trying to keep it simple, so I'd discount the idea of an expensive solution. Just wondering how I'd go about building such a circuit and not use a relay.

Last edited:

#### Tonyr1084

Joined Sep 24, 2015
5,570
I see a possible problem which depends on details we don't have.
When the circuit is switched on, the capacitor holds the MOSFET gate low, thus the MOSFET and thyristor will both be off - so far so good.
The 3R3 will be feeding the output and during the time dealy the output voltage will rise, to what level we don't know.
Now the MOSFET gate voltage is rising and the source voltage cannot be lower than the (unknown) output voltage.
With the values shown the gate voltage can rise to only 6V and to switch on the MOSFET that needs to be greater than the source voltage by enough.
The problem could be helped by eliminating the lower 100k resistor which would then allow the gate to reach the full 12V.
Thanks Al. This is all theory and nothing else. Just a though on how to accomplish the same thing in the video without using the relay. In the video the relay coil is a constant drain, which, if battery powered, may be an issue. I'm just trying to think outside the box. Or maybe in the same box just in a different corner of the box.

#### nsaspook

Joined Aug 27, 2009
7,898
Did you watch the video? The guy considered that but didn't like the idea of the reduced current. And the time it would take the thermistor to cool down and increase its resistance so that if there's a momentary interruption in voltage the unit won't shut down.

And he's trying to keep it simple, so I'd discount the idea of an expensive solution. Just wondering how I'd go about building such a circuit and not use a relay.
No. I have my own practical experience and knowledge on the issue. YMMV

Using properly designed thermistors eliminates power drop issues (The short cycle issues are still there with a relay/ssr based solution due to the compressor refrigerant issues). I tried several DIY relay and small soft-start devices on my 12VDC 2000W Inverter setup running a 5000 BTU aircon before sticking with thermistors for the last few years with zero inverter problems caused by switch-on current loads.

Last edited:

#### Tonyr1084

Joined Sep 24, 2015
5,570
No. I have my own practical experience and knowledge on the issue.
I respect that. However, I'm curios about the circuit I posted. Al suggested eliminating one of the two 100KΩ resistors, which makes some sense. However, then there's no method for draining the cap and gate capacitance when shut down.

I'm not building anything. Was just curios about what is called "Soft Start". Heard it discussed a few times on this site; so when I saw a YouTube video I looked at it. Then wondered if it could be done a different way. In the video the guy used a relay on a 1 second delay, channeling current through a 3.3Ω resistor. After the delay the relay jumped out the resistor. Said his only concern was the current drawn by the relay, which is small, and he could live with that. I wondered if there was a way to turn it on without using the relay. So I thought about an SCR.

All this is just an exercise in theory. Not building anything like it. Not today. Not anytime in the foreseeable future. But I do appreciate the input from everyone. Even those who say I did it wrong.

#### AlbertHall

Joined Jun 4, 2014
10,790
Al suggested eliminating one of the two 100KΩ resistors, which makes some sense. However, then there's no method for draining the cap and gate capacitance when shut down.
There is still a path through the remaining 100k, 3R3, and the load but you could add a diode from MOSFET gate to output (cathode to output) to provide a discharge path.

#### Tonyr1084

Joined Sep 24, 2015
5,570
There is still a path through the remaining 100k, 3R3, and the load but you could add a diode from MOSFET gate to output (cathode to output) to provide a discharge path.
Oh. OK.

Like I said, I'm not building anything. Was just an exercise.

Thanks all. Will modify my drawing and keep it for if ever comes a day.

#### nsaspook

Joined Aug 27, 2009
7,898
I respect that. However, I'm curios about the circuit I posted. Al suggested eliminating one of the two 100KΩ resistors, which makes some sense. However, then there's no method for draining the cap and gate capacitance when shut down.

I'm not building anything. Was just curios about what is called "Soft Start". Heard it discussed a few times on this site; so when I saw a YouTube video I looked at it. Then wondered if it could be done a different way. In the video the guy used a relay on a 1 second delay, channeling current through a 3.3Ω resistor. After the delay the relay jumped out the resistor. Said his only concern was the current drawn by the relay, which is small, and he could live with that. I wondered if there was a way to turn it on without using the relay. So I thought about an SCR.

All this is just an exercise in theory. Not building anything like it. Not today. Not anytime in the foreseeable future. But I do appreciate the input from everyone. Even those who say I did it wrong.
There are different types of slow-start loads. Highly-capacitive loads like high power LED module power supplies or high power high voltage supplies, highly-inductive loads like motors/magnets, non-ohmic loads that can interact with the ‘hiccup’ protection power supplies or over-stress devices. What works best for one type might not be the best for another. Soft-start phase control of SCRs might cause harmful harmonics in some circuits but be perfectly fine in others.