Reverse polarity protection

Started by DeusM, May 30, 2021, 05:03:20 PM

Previous topic - Next topic

ElectricDruid

Quote from: GGBB on May 30, 2021, 07:52:31 PM
Quote from: ElectricDruid on May 30, 2021, 05:58:52 PM
Separate issue, but what's the point of R1 and R2 in parallel? Sometimes that would be done to get a very specific value (although in series is much easier to work out!), but that seems a bit unlikely when there's a 2K variable pot underneath them. You could just tweak that instead, right?

The reference implementation in the datasheet calls for 240 ohms - which is 270 || 2200 - more readily available values - in order to achieve the required voltage drop of 1.25V between the output and adjust terminals. See page 11 figure 9 https://www.ti.com/lit/ds/symlink/lm317.pdf. The posted circuit is a straight copy of that with increased output capacitance.

Thanks for the explanation.

240R is an E24 value - hardly difficult to find at 1% accuracy. Trying to make an exact value by sticking two 10% tolerance resistors in parallel in this day and age is a bit retro even for TI?!? ;)

I suppose it made sense once, and no-one has updated the datasheet since then...



DeusM

Quote from: duck_arse on May 31, 2021, 11:24:16 AM
Quote from: DeusM on May 30, 2021, 05:03:20 PM
Hi guys. I have this circuit which is power supply. I though that it was protected against reverse polarity but two exploded caps tell me otherwise.


Quote from: imJonWain on May 30, 2021, 05:43:42 PM
Is that reference actually -12V or is it GND?

we need an answer to this, still. are you putting rectified DC in, or AC from a transformer?


It's Ok. I used a 1n5822 and tried it with reverse polarity and it works fine.

The -12V is ground. I use It as a reference for the PCB. It's the ground of the DC Input Jack.
It's not the amps that kills you. It's the "mojo"

DeusM

Quote from: ElectricDruid on May 31, 2021, 06:11:11 PM
Quote from: GGBB on May 30, 2021, 07:52:31 PM
Quote from: ElectricDruid on May 30, 2021, 05:58:52 PM
Separate issue, but what's the point of R1 and R2 in parallel? Sometimes that would be done to get a very specific value (although in series is much easier to work out!), but that seems a bit unlikely when there's a 2K variable pot underneath them. You could just tweak that instead, right?

The reference implementation in the datasheet calls for 240 ohms - which is 270 || 2200 - more readily available values - in order to achieve the required voltage drop of 1.25V between the output and adjust terminals. See page 11 figure 9 https://www.ti.com/lit/ds/symlink/lm317.pdf. The posted circuit is a straight copy of that with increased output capacitance.

Thanks for the explanation.

240R is an E24 value - hardly difficult to find at 1% accuracy. Trying to make an exact value by sticking two 10% tolerance resistors in parallel in this day and age is a bit retro even for TI?!? ;)

I suppose it made sense once, and no-one has updated the datasheet since then...
}

Good point. I removed one of the resistors and added it again thinking that the person who made the circuit has more knowledge than me and decided to trust them.
It's not the amps that kills you. It's the "mojo"

GGBB

#23
Quote from: ElectricDruid on May 31, 2021, 06:11:11 PM
Thanks for the explanation.

240R is an E24 value - hardly difficult to find at 1% accuracy. Trying to make an exact value by sticking two 10% tolerance resistors in parallel in this day and age is a bit retro even for TI?!? ;)

I suppose it made sense once, and no-one has updated the datasheet since then...

I take it you didn't look at the data sheet. And I didn't explain clearly. ;) TI calls for 240. Whoever did the posted circuit decided on 270||2200 as their implementation of 240 - not TI.

Most hobbyists probably would not have 240 on hand (I don't) but 270 and 2k2 - yes (and I do). But I agree with your point - and 120+120 would have been more precise than 270||2k2 and just as likely to be found on hand.
  • SUPPORTER

PRR

Quote from: ElectricDruid on May 31, 2021, 06:11:11 PM.....I suppose it made sense once, and no-one has updated the datasheet since then...

It makes sense today. And there is no reason to hit the exact number.

Two constraints:

The /ratio/ of the two resistors gives the desired output.

The chip needs enough bleeder current to swamp the internal current flowing at the ADJ pin. I(adj) may be as much as 0.1mA. The sheet suggests you swamp this 52 times with 5.2mA in the divider. So the maximum "good" value of the upper resistor is 240 Ohms. A perfectly standard value, especially when these chips cost $3 each.

No 240 in your drawers? Use 220 and fudge the bottom resistor by 22/24.  Or use 100 Ohms and 0.4166 the bottom resistor, if you can afford the 13mA idle current.
  • SUPPORTER

Rob Strand

QuoteThe /ratio/ of the two resistors gives the desired output.

The chip needs enough bleeder current to swamp the internal current flowing at the ADJ pin. I(adj) may be as much as 0.1mA. The sheet suggests you swamp this 52 times with 5.2mA in the divider. So the maximum "good" value of the upper resistor is 240 Ohms. A perfectly standard value, especially when these chips cost $3 each.

No 240 in your drawers? Use 220 and fudge the bottom resistor by 22/24.  Or use 100 Ohms and 0.4166 the bottom resistor, if you can afford the 13mA idle current.
Agreed, unless you a designing the regulator to meet a spec you can use lower or higher resistor values no problem.  (Of course to  have to adjust the resistor to ground to match so the output voltage is what you want.)
Send:     . .- .-. - .... / - --- / --. --- .-. -
According to the water analogy of electricity, transistor leakage is caused by holes.

antonis

Never had E24 series resistor values other than 5.1 and 7.5, so I used many times 220R for Vref current without absolutely on issue..
(difference is less than 500μA..)
"I'm getting older while being taught all the time" Solon the Athenian..
"I don't mind  being taught all the time but I do mind a lot getting old" Antonis the Thessalonian..