Triggering an SCR from the output of an LM324 configured as a comparator

Thread Starter

davismltc

Joined Mar 11, 2021
15
My goal is to create a fail safe circuit (or rather...more of a battery protection circuit) for my 14.6v lifepo4 batteries that are charged by 33v PV panels in the event of a failure of the BMS to stop excess charging. My plan is to configure an LM324 as a voltage comparator with one input a 5v reference and the other a voltage divider. My issue is how to properly utilize the output of the LM324 to turn on an SCR.

Additional circuit info: Once the SCR is triggered, the PV panel will be shorted to stop any excess charging. For additional safety, I've opted to isolate the higher PV panel voltage from the battery voltage by a small relay.

I'm confused about the gate voltage and current requirements necessary to turn on the SCR. I've selected an MC106-6G SCR to trigger a small relay. My interpretation of the datasheet is a gate trigger voltage between 0.2 and 1 volt and a gate trigger current greater than 200uA. From searching other posts and reading about SCR gate triggering (I'm a novice trying to learn), it seems as though the current is the most important criteria for triggering.
Datasheet for the 4A SCR: ( taydaelectronics.com/datasheets/files/A-1497.PDF )

I've created the circuit with the help of an online simulator. If I put a current limiting resistor on the output of the LM324 (which should be 20mA or so), the output voltage is around 14v (much higher than the 0.2 - 1v called for on the datasheet). If a voltage divider is utilized, the current flow is below the triggering current threshold of 200uA. Can someone help with the triggering solution?

A screenshot of the circuit simulator is posted. FYI...The simulator doesn't offer a PV panel, so I've substituted a 33v battery and a 10.5A relay to simulate the solar panel when it is shorted out. The SCRs shorting the panel are BT151-500R rated at 12A each.
Datasheet for the 12A SCR: ( taydaelectronics.com/datasheets/files/A-1485.PDF )

Thank you in advance for any advice or help!

Prior to LM324 trigger:
LM324 Prior to trigger.jpg
.
LM324 triggered:
Triggered LM324.jpg
.
Post triggered LM324 (lower battery voltage) prior to reset:
Post triggered LM324 (lower battery voltage) prior to reset.jpg
 
Last edited:

Thread Starter

davismltc

Joined Mar 11, 2021
15
Oops... 22k current limiting resistor on 12A SCRs should be 1.5k
Prior to LM324 trigger:
LM324 Prior to trigger.jpg

LM324 triggered:
Triggered LM324.jpg

Post LM324 trigger (lower battery voltage) prior to reset
Post triggered LM324 (lower battery voltage) prior to reset.jpg
 

BobTPH

Joined Jun 5, 2013
8,807
Why dump the current into a short instead of simply disconnecting the panel? If the panel is producing 10A at 33V, the SCR would have to dissipate 330W!

Bob
 

DickCappels

Joined Aug 21, 2008
10,153
@BobTPH you have good point, but what the SCR saturates at 1 volt?

Overall it looks like a pretty good circuit and a lot of thought went into it. What is the purpose of the circuit? I don't see any sensors to make the Schmitt trigger trip.

The LM324 cannot source current to a load that is connected to its positive power supply pin. I recommend that you ground the cathode of the SCR so you forward bias the gate. The relay would be in series with the anode.

The table below from the Littlefuse data sheet gives the conditions for the gate voltage and gate current when the SCR is off.
1652190856295.png

For the lager SCRs:
1652191351574.png
I would go for the 15 milliamps on the gate.

You might want to add a little bit of positive feedback (hysteresis) to keep the circuit from oscillating.

I would look for an actual comparator rather than an op amp, and if you have to use an op amp for some reason, the LM324 may be a poor choice because the large amount of bias current supplied by its inputs can throw off your voltage divider. At the very least lower your voltage divider values by a factor of 10.
 

crutschow

Joined Mar 14, 2008
34,281
Some comments:
  • The 4A SCR is better connected to common with the relay coil connected in series with the SCR cathode to the plus voltage.
  • The SCR gate looks like a forward-biased diode, so the gate current is the critical parameter. which should be more than 500µA and less than 0.2A for the 4A SCR.
  • You can't reliably trigger the two panel SCRs by connecting their gates in parallel. Use a separate gate resistor for each SCR.
For your interest, below is the LTspice simulation an SCR trigger circuit using the TL431 programmable reference:

1652193668258.png
 

ci139

Joined Jul 11, 2016
1,898
No offence , ? using an SCR to trigger a RELAY to trigger a set of SCR-s ????
woudn't the SSR do the required at all in one
 

ElectricSpidey

Joined Dec 2, 2017
2,758
I'm confused...

If a SCR is not used as a switch in series with a load, how does it not see the entire source voltage?

Someone please explain, I'm not really up to speed on PV systems.
 

Thread Starter

davismltc

Joined Mar 11, 2021
15
@DickCappels - Thank you so much!!! Moving the SCR cathode to ground explains the low trigger current I was confused about. Thanks to everyone else for the suggestions about not shorting the panels...just disconnecting them. I'm contemplating that change as well. Was busy creating a new schematic (attached)

So the 4A SCR will need 200uA to trigger, or is that the max current when triggering it in DC mode?

LM324 Prior to trigger SCR cathode groiunded.jpg
 

Thread Starter

davismltc

Joined Mar 11, 2021
15
@crutschow Thanks!!

Since the 4A SCR needs 200uA to trigger and as much as 500uA @ 40*C, I've changed the current limiting resistor to 24k. I also think I need to reconsider the PV panel side of the circuit. The way I'm showing electrically isn't correct, because the simulator doesn't have a PV panel selection. The 12A SCRs would short the panel completely and absorbing the 350w output. It would be better to disconnect them as previously suggested, rather than short them out. I'm a novice, so it will take me a while to sort that out and create a schematic. I've shown the PV panel as a 33v battery with a current limiting resistor in the following schematic. Not perfect, but a little closer to reality...Triggered LM324.jpg
 

eetech00

Joined Jun 8, 2013
3,858
@crutschow Thanks!!

Since the 4A SCR needs 200uA to trigger and as much as 500uA @ 40*C, I've changed the current limiting resistor to 24k. I also think I need to reconsider the PV panel side of the circuit. The way I'm showing electrically isn't correct, because the simulator doesn't have a PV panel selection. The 12A SCRs would short the panel completely and absorbing the 350w output. It would be better to disconnect them as previously suggested, rather than short them out. I'm a novice, so it will take me a while to sort that out and create a schematic. I've shown the PV panel as a 33v battery with a current limiting resistor in the following schematic. Not perfect, but a little closer to reality...View attachment 266867
I don't see how that circuit is failsafe. The SCR's can fail shorted, or triggered "on" by some other circuit failure.
 
Last edited:

Thread Starter

davismltc

Joined Mar 11, 2021
15
@eetech00 - If any of the SCRs fail shorted, they prevent the PV panel from overcharging the Lifepo4 battery, which is the component being protected. The batteries are the highest cost element, as well as a fire risk.

History: What has happened in the past is the PV panel's charge controller fails and outputs full PV panel voltage, 33v or higher, (depending on the number of PV panels and how they are configured) to the 14.6v lifepo4 battery. The BMS allows the charge until the battery becomes fully charged and then disconnects the charge. When it disconnects, the FETs of the BMS aren't rated for the full panel voltage and fail shorted, which reconnects the full PV panel voltage to the battery. The lifepo4 batteries are found overcharged and ruined. It would be ideal if the BMS FETs were rated at full panel voltage, but none of the suppliers provide that info and they are made competitively priced.
 
Last edited:

Thread Starter

davismltc

Joined Mar 11, 2021
15
The above scenario is why I'm reluctant to just "disconnect" the PV panel instead of shorting them. If the disconnect fails shorted, I'm back in the same boat of overcharging the batteries.

While monitoring only, the Falstad simulator shows a 2mA draw on the battery until the relay and alarms are triggered. Once triggered, the current to the battery increases to 35mA. My battery is 200Ah on my system. The low current draw gives me quite a bit of time to discover the alarms and find out what triggered them.

What would be better is to disconnect the batteries completely with a DC circuit breaker from the entire system. But the circuit I've proposed is considerably lower cost than 200+A DC circuit breakers that can be remotely tripped. I don't think it's that common for the PV panel charge controllers to fail shorted, so the budget friendly fail-safe I'm proposing is the route I'll probably take, as long as it works reliably.

Another option would be a fuse with a crow bar circuit. Being a novice with electronics, the crowbar is probably above my skill level at this point, and 200A DC fuses aren't that inexpensive either. My approach is pretty budget friendly.
 
Last edited:

Thread Starter

davismltc

Joined Mar 11, 2021
15
If I were to replace the LM324 with an LM311 comparator, can someone help with the pull up resistor value? The LM324 output has a current limiting resistor (24k) providing 550uA to the 4A SCR in my schematic. Would the 24k resistor remain on the output of the LM311 as well?
LM311 Pinouts.png

Thanks in advance!!
 
Top