Over voltage protection (LM2956HV short failure) using NJM7808FA

Thread Starter

gkeep

Joined Oct 21, 2017
48
Hi,

I am new to electronics. I am attempting to power an expensive DSLR camera with an 750 W.hr eBike battery to get a 24 hour time-lapse (hence the large capacity). Its a 48V ebike lipo battery, and I'll be using a LM2956HV buck converter module to drop the voltage and power the camera (and microcontroller and stepper motors and a few other things). I have had a LM2956HV fail running some tests and it failed in a way, that it short circuited the in and out positive terminals, essentially putting 48 V on the output instead of 8 V. My question is, would it be a good idea to use a NJM7808FA linear voltage regulator in this configuration to prevent over-voltage damage to my camera?

See attached simplified (no microcontroller etc.) schematic.

Any help would be appreciated greatly.voltage regulator.jpg
 

bertus

Joined Apr 5, 2008
21,312
Hello,

The circuit will not work correctly.
The NJM7808 will have a dropout voltage of 2 volts at 1.5 A.
With an input voltage of 8 Volts, there will be a to low output voltage.
You will need an input voltage for the NJM7808 of 10 Volts at least.

Bertus
 

Attachments

LesJones

Joined Jan 8, 2017
3,127
I would use crowbar protection using an SCR triggered by an LM431. This has to be installed on the output side of the fuse.

Les.
 

AlbertHall

Joined Jun 4, 2014
11,211
The maximum input voltage of the NJM7808 is 35V so you are likely to get a cascade failure with both chips failing.
I would go the crowbar route as post #3.
 

Thread Starter

gkeep

Joined Oct 21, 2017
48
Hey guys. Thanks heaps for the input. I've
Sorry I've taken s while to reply. So I did some research on crowbar circuits. I don't understand how to make 8V the threshold for current to flow through the scr. Is it solely based on the breakdown voltage of the zener diode? I don't know if I'll go with that because I'm not that confident with scrs.

I do like this circuit though, as there isn't always 2 volts dissipated like in the linear regulated circuit.

Are there better linear regulators with lower dropout voltages?

Sorry to all soo many questions
 

Thread Starter

gkeep

Joined Oct 21, 2017
48
Also, I've been buying most of my components on ebay. The cheap stuff from China. I'm guessing triac scr is for ac? But when I look in the data sheets I don't know what to look for in a DC scr, not to mention a zener diode. Like, some people say to use a zener diode with a breakdown voltage afew volts above the nominal voltage. I don't know why that would be. Does that seem right?
 

AlbertHall

Joined Jun 4, 2014
11,211
The trigger voltage is the zener voltage plus the SCR gate voltage at whatever current will trip the SCR. This is not a very precise number so generally a decent allowance is made.

More accurate circuits can be made though without dramatically increasing the complexity. The one in the wiki article uses an LM431 as the voltage sensor - pretty accurate. Note that this circuit uses a triac even though it is dc, so the trigger can be negative going as available from the LM431.

https://en.wikipedia.org/wiki/Crowbar_(circuit)
 

crutschow

Joined Mar 14, 2008
26,963
Below is the LTspice simulation of a crowbar circuit using the TL431 voltage reference as a comparator to give an accurate trip point.
I didn't have a TRIAC model, so I used an SCR with an added transistor to invert the trigger to the needed positive polarity.
The trip point occurs when the Ref voltage reaches 2.5V, as determined by the voltage divider R2 and R4.
The values shown give a trip point of about 8.54V.

(R3 is just for the simulation to limit the current since I didn't have a fuse model).

upload_2019-2-9_20-2-1.png
 
Last edited:

Thread Starter

gkeep

Joined Oct 21, 2017
48
Hi. Thanks for that example. I'll definitely investigate that some more. Although, I'm unsure of how this works. Wouldn't the adjustable zener diode maintain the same voltage at the gate of the triac? Why, in the explanation does the gate get pulled low? I thought a zener diode maintains its given zener voltage value even if the supply voltage goes over.

I'm sure I'm missing something here.
 

crutschow

Joined Mar 14, 2008
26,963
I'm sure I'm missing something here.
First, the TL431 adjustable reference is being used as a comparator switch, not as a voltage regulator or Zener.
When the Ref voltage reaches 2.5V, the TL431 cathode starts conducting, which turns on Q1 (if this is not clear, look at the TL431 data sheet),
This generates current into the SCR gate causing it to latch on.

Second, you perhaps are missing that when an SCR is turned on, it latches and stays on with a low voltage drop, independent of what the gate does after that.
 
Last edited:

Thread Starter

gkeep

Joined Oct 21, 2017
48
I've been at the wikipedia diagram for hours.

https://en.m.wikipedia.org/wiki/Crowbar_(circuit)

I can't imagine how this circuit could possibly work. The TL431 is an adjustable zener diode and I don't see how "current" through a zener's cathode would switch on a triac's gate. Zener diodes don't pull anything low. Can anyone explain this to me? Sorry for my ignorance if I'm missing something obvious.
 

crutschow

Joined Mar 14, 2008
26,963
I don't see how "current" through a zener's cathode would switch on a triac's gate.
A TRIAC will trigger when the gate to cathode current is positive or negative (current flows out of the gate), so when the TLR431 "Zener" starts to conduct during an overvoltage, current starts flowing out of the gate through the Zener and the TRIAC latches on. (See circuit below).
upload_2019-2-11_1-38-32.png
 
Last edited:

Thread Starter

gkeep

Joined Oct 21, 2017
48
Aaah! It's like you latched a triac in my brain! Thanks so much for that mate. You've done me a great service.
 

Thread Starter

gkeep

Joined Oct 21, 2017
48
Sorry. Just one more question. Will a BT139 600C triac do the job. I'm unsure of how to interpret the data sheet. The gate trigger voltage is 0.7V at 12Vd (I imagine is the voltage between the two terminals). If so, does that mean the triac won't conduct until the voltage it 0.7V above Vref?
 

LesJones

Joined Jan 8, 2017
3,127
The gate voltage will rise rapidly as soon as the input voltage exceeds the set voltage limit. Think of the TL431as a comparator IC with a reference voltage (V ref) connected to it's non inverting input and the inverting input connected the junction of R1 and R2. As soon as the input voltage exceeds V ref x (R1+r2)/R2 its output will go low which will trigger the triac.

Les.
 
Top