G'day Guys,
How large an electrolytic cap can be placed at the output of a 7812 regulator?
The datasheet suggets 0.1uF ceramic cap but I would have envisaged something larger: 100-470uF.
The use case is to drive some EC2-12NU, 12vdc small signal relays for a source selector.
Looks like the brand supplied at my local brick and mortar retailer is National semiconductor.
How large an electrolytic cap can be placed at the output of a 7812 regulator?
The datasheet suggets 0.1uF ceramic cap but I would have envisaged something larger: 100-470uF.
The use case is to drive some EC2-12NU, 12vdc small signal relays for a source selector.
Looks like the brand supplied at my local brick and mortar retailer is National semiconductor.
The output capacitor value placed next to the 78xx and 79xx linear fixed series voltage regulators does not really matters, performance-wise. 10uF is a typical value because anything larger requires - if you follow the datasheet recommendations - the addition of a protection diode to avoid backflow trough the regulator if the input voltage drops below the output when power to the device is removed. This may damage the regulator.
If it is really only being used to drive some relay coils, I'd just put the recommended .1uF near the regulator for stability. Relay is not going to care. I tend not to use regulated supplies for relays. Really no need. They usually accept very wide voltage ranges. Your 12V relay specs for example show operation down to 9V with a max voltage of 1.5x12 = 18V. Do make sure you put a diode on the relay coil to prevent voltage spikes on turn off if the relay does not include one internally.
If reliability is a concern, a resistor followed by a zener diode is a better solution, by far. Relay supply may be left unregulated, but the zener+resistor does add a little bit of protection to resist voltage spikes and polarity inversion.
The output capacitor (if required by the mfr) is to ensure stability of the regulator at high frequencies.
Some regulators have excess noise in the kHz range, and a capacitor up to 100uF will reduce it.
See figure 3 for a good example.
https://sound-au.com/articles/power-supplies3.htm
Some regulators have excess noise in the kHz range, and a capacitor up to 100uF will reduce it.
See figure 3 for a good example.
https://sound-au.com/articles/power-supplies3.htm
Last edited:
The backflow diode is only required for 7808 and upwards
Negative and low drop regulators are fussier about load capacitance for stability, just follow the datasheet advice
NO!
That's terrible advice. DO NOT do that; it'll ring like a bell.
This idea of 0.1uF at the output of a 3-pin reg (or any other!!) is a constant source of nonsense, regulator misbehavior and general disappointment it is not funny. It's an utter f'idiocy, that persists when - if anyone bothered to read them, for a start in so many commercial, diy and other designs - it's NOT even a good place to start. Properly-read the average 3-pin reg datasheet (LM3xx, LM7xxx), and the output cap is not required for stability;
and conversely, a small-value , low-ESR cap (like a ceramic) IS proven, the best way to provoke instability in any conventional regulator*!
Put any bypass cap - at the load; as large as you like. Here's one take, mine on the Lm317, on why:
http://www.acoustica.org.uk/t/3pin_reg_notes3.html
*low-dropout types using inverted pass device architectures are different. If you understand why, you don't need my thoughts.
That's terrible advice. DO NOT do that; it'll ring like a bell.
This idea of 0.1uF at the output of a 3-pin reg (or any other!!) is a constant source of nonsense, regulator misbehavior and general disappointment it is not funny. It's an utter f'idiocy, that persists when - if anyone bothered to read them, for a start in so many commercial, diy and other designs - it's NOT even a good place to start. Properly-read the average 3-pin reg datasheet (LM3xx, LM7xxx), and the output cap is not required for stability;
and conversely, a small-value , low-ESR cap (like a ceramic) IS proven, the best way to provoke instability in any conventional regulator*!
Put any bypass cap - at the load; as large as you like. Here's one take, mine on the Lm317, on why:
http://www.acoustica.org.uk/t/3pin_reg_notes3.html
*low-dropout types using inverted pass device architectures are different. If you understand why, you don't need my thoughts.
So for the purpose of driving a relay coil approximately 20-50mm away (depending on which relay is being driven), no output cap would be a more stable solution?
Use a dropping resistor to get about the right voltage-- NOT critical. Supply spikes won't hurt. Polarity reversal is user-screw-up... do you expect that? Is it even worth protecting against?
Relays have been used by the billions in elevators, telephony, automobiles, appliances. Often by very ignorant designers and users. Don't over-think it.
I found I needed bigger caps when used with relays as flywheel diode spike pushed voltage/current spikes back into the rail.
Same goes for H bridge chips but even more so.