I have looked at the lm723 datasheet and would like to know how in this schematic they manage current limiting.
They are setting two current levels in this example.
I was wondering if someone might tell me how they arrived at Rsc (sense resistor in datasheet) as it applies in this example in a LM723 so as to calculate maximum current...
eg say I only wanted a maximum current of say 5amps how would I calculated that?
I would like to know this as I am using digital panel meters and will not need to adjust for the ammeter full scale deflection.. I will be putting the amp meter shunt resister in the main line so need to know how to calculate for the correct resistance between pins 2 and 3...
That's what I remember as well: the voltage across the resistor is compared to a VBE so the voltage is about 0.7V, it also is very inaccurate since it has the known tempco of -2mV/C. FYI, the 723 is about 40 years old and designs using it are pretty crude as far as performance goes.
Do you rally meed 30 volts 10 amp? There are a lot of parts in this supply. Can you live with 1A? 2A? You don't need a 300 watt supply to power a PIC and LED.
There are some simple designs that will be much faster to get working.
Working with testing a high power Cockcroft Walton Multiplier which requires very low source impedance... using this to initially test the this circuit....
Look at the schematic of IC in the data sheet. There is an NPN transistor with its base-emitter brought out (pins 2 & 3 for dual inline package). When the total resistance in series with the output times the current flowing through it turns on the current limiting internal NPN transistor (Vbe about 0.68 vdc at 25 deg C) it steals drive current to series pass power transistor shutting it down.
There is a graph in the spec showing the variance in the current limiter transistor Vbe versus temp. This is because the Vbe of the current limiter transistor has -2 mV/deg C temp coefficient.
I have seen add on circuitry with a 2.5v reference shunt regulator resistor divided down to provide a reliable fixed offset voltage to an extra op amp that senses the voltage drop across the power resistor in series with output to provide a temp independent current limiter control. The output of the extra op amp feeds into the current limiter transistor on 723. Lots of extra circuitry and might as well use a more modern regulator IC that does this internally.
Thanks for the reply. Just wondering if you might detail a few "more modern regulator IC"s. Could you list some new better versions of LM723 with current limiting etc..
Do you rally meed 30 volts 10 amp? There are a lot of parts in this supply. Can you live with 1A? 2A? You don't need a 300 watt supply to power a PIC and LED..
The way I read it, it's not 300W. I think it is either 6V/10A or 30V/1A based on the schematic title. A 300W linear would need a heatsink the size of Texas.
The title is not clear to me.
"0-6/1-30V & 0-1/1-10A"
I assumed that one switch is for 0-6 or 1-30V. The other switch is for 0-1 or 1-10A. Now that I look back that does not work. I have no idea what the title said. I don't see 6V10A or 30V1A.
I was looking at a powersupply by Tony Van Roon at this **broken link removed**.
He used a lm723 and it seemed like a good step up from my last project.... Have done some more research I am now aware that this is an old chip and was wondering if there might be newer better designed equivalents out there?
That's the point: 40 years ago, there were no decent switcher controller ICs and the plentiful crop of cheap switching FETs we now enjoy had not been invented yet. High power linears are just no longer feasible, switchers are the way to go.
This design came out just about when Washington was crossing the Delaware.
Yeah this circuit came out even before electricity was discovered
The goal of the preregulator is to keep a differential voltage across the linear series pass device so as to keep dissipation to a minimum so as to get the regulation specs of a linear regulator but the power dissipation of a switching regulator. This is typical in linear regulators with a switching preregulator.
This design is old though so you'll note the 1mH inductor, which is required because the switching frequency is so low. This could be quite expensive for a high current rated device. More modern design switching regulators use lower values which are cheaper and smaller physically.
The transformer T1 appears to be a regular power transformer, which can put out the required voltage at the required current using full wave rectification. This should not be hard to replace.
Transformer T2 appears to be a pulse transformer, which only has to pass a short pulse long enough to turn either SCR on when required. You may be able to replace that transformer with a set of fast opto couplers (and required bias supply circuit) or maybe a transistor network (also with bias), or roll your own pulse transformer using a toroid core and some turns of wire for primary and secondary. Some experimentation would be required and probably a scope to check that it is operating correctly.
To keep the heat down, I found a transformer with many taps. Using a large switch I can choose the 10V, 20V or 30V taps. This is a poor mans pre-regulator.