power supply questions

[sparky!(Sean)]

Oh didn't see the last four posts, it didn't send me an email I think I will stick with the lm317, but I would like a 7 segment led voltmeter

 

[Ubergeek63]

For example, as the capacitor ESR goes down the diode peak current goes up, which means the power being delivered is being done over a shorter time, and with higher peak current the RMS current goes up even though the same amount of power is being delivered.

i will have to take exception here. that is only true of a capacitor of infinite value. of for practical purposes, when the voltage ripple is held to a minimum value by the capacitor.

if the ripple voltage is allowed to be 1/2 the peak then the ESR has very little effect on the peak diode current.

 

[bountyhunter]

This thread is getting very techy . . . but fundamentally it shows the difficulty of building a 10A PSU using a linear circuit.

As somebody pointed out: a smart designer needs to recognize the difference between possible and feasible.

You don't see 300W linear lab supplies made anymore: guess why.......

 

[ChrisP58]

Oh didn't see the last four posts, it didn't send me an email I think I will stick with the lm317, but I would like a 7 segment led voltmeter

There are lots of digital voltmeter modules.

Search google or ebay for "panel digital voltmeter"

 

[MrAl]

In reference to capacitor ESR and diode peak current......

i will have to take exception here. that is only true of a capacitor of infinite value. of for practical purposes, when the voltage ripple is held to a minimum value by the capacitor.

if the ripple voltage is allowed to be 1/2 the peak then the ESR has very little effect on the peak diode current.

Hi Uber,

Well im not sure why you would say that. The ESR come into play almost without regard for the capacitor value. To quote a few extreme values, with 10 ohms ESR that resistance could be almost equal to the load itself, which means the capacitor would see much less ripple current and after all the capacitor is what causes the peak current in the diode for the most part unless of course it is very small in value.
With a very large cap value and low ESR yes the diode current peak value will be limited only by the other resistances, but that doesnt mean it has zero effect for other values of capacitance. If you look at a picture of the diode current you'll see peaks that are there that are mainly caused by the capacitance even if it is only 1000uf. Increasing ESR lowers those current peaks, and decreasing ESR makes those peaks higher. With zero ESR the current peak is maximum, with infinite ESR the peak is minimum, and this is with almost any reasonable value of capacitance.


Other notes:
A more reasonable 300 watt DC power supply that needs a linear output would be built using a switcher on the input and linear on the output.

 

[Ubergeek63]

Well im not sure why you would say that. The ESR come into play almost without regard for the capacitor value. To quote a few extreme values, with 10 ohms ESR that resistance could be almost equal to the load itself, which means the capacitor would see much less ripple current and after all the capacitor is what causes the peak current in the diode for the most part unless of course it is very small in value.

I was going from knowledge...having designed and built a variety of supplies....

It is a bit application dependent and i was in error to say imply that the ESR would have NO effect.

It is a fact that reducing the ESR has much less effect at low capacitance values. In a simple bridge at 220V in and a 1A constant current load varying the capacitance and ESR we get:

ESR / 500μF / 5000μF
10 / 3.9A / 3.9A
1 / 8.1A / 8.4A
0.1 / 12.8A / 17A
0.01 / 13.2A / 29A

In these figures from LTSpice you can see that in both cases the peak current doubles for every order of magnitude drop in ESR UNTIL you get to 0.1 ohms at which point the 5000μF doubles but the 500uF REMAINS RELATIVELY CONSTANT

this is caused by the charging of the cap being spread over a longer period of the cycle.

Dan

 

[The Electrician]

You might consider just buying one of these:

https://www.mpja.com/28V108A-POWER-SUPPLY/productinfo/18927+PS/

 

[sparky!(Sean)]

You mean the lm7805 and yes it will but I am looking for variable voltage and that can't handle enough current