I keep seeing this schematic when looking for a variable supply. I was told on another thread (that i cant seem to find now) that i should look into lm723 for a variable bench supply. so as one of my wishes is to power universal heaters for a bunch of valves i plundered, i would like to give this circuit a go as it is 0-70v. But looking at the comments at the bottom of there page i wonder if the only issue is that there is an error with the pin numbering 10 and 11. Please can someone look this over and give thoughts there thoughts on this design
http://www.eleccircuit.com/0-70-volt-2-amp-dc-variable-power-supply/
http://www.eleccircuit.com/0-70-volt-2-amp-dc-variable-power-supply/
Hello jimmiegin
Funny, I am right now in the same process as You, regarding making a 0 - 30 Volt 3 Amp. powersupply, and I have looked a lot of places to find the best circuit. A lot of people, including those, that has made the application notes, has omittet the drawbacks of the old, but fine uA723.
Let us make a brainstorm on this topic!
First: 2 Amp, and a raw supply of more than 70 Volt, that gives more than 150 Watt dissipating power, if the output is short-circuitet. I would use no less than 6 pcs. 2N3055 as pass transistors.
Regards Soeren Poulsen Denmark
Funny, I am right now in the same process as You, regarding making a 0 - 30 Volt 3 Amp. powersupply, and I have looked a lot of places to find the best circuit. A lot of people, including those, that has made the application notes, has omittet the drawbacks of the old, but fine uA723.
Let us make a brainstorm on this topic!
First: 2 Amp, and a raw supply of more than 70 Volt, that gives more than 150 Watt dissipating power, if the output is short-circuitet. I would use no less than 6 pcs. 2N3055 as pass transistors.
Regards Soeren Poulsen Denmark
Now i have looked i can see there is only a couple of changes to the 70v from the 30v supply. A couple of minor resistor values and the 50v transformer =70vdc as you pointed out. Is a transistor power disipation based on using no heatsink? And are you saying you would use 6 times 2n3055?. I think the power disp is 90w for 2n3055. Is this circuit even worth looking at?
Hello jimmiegin.
I know that circuit too. it is good, but it has one minor issue: The operational amplifiers can only cope with +/- 18 Volt at the maximum, and that is 36 Volt from + to -.If the - pin is held at -6 Volt, the + pin may at max be 30 Volt.
As the pass transistors is configured as NPN darlington transistors, the output of the integrated circuit must go 2 Ube voltages higher, and that limits the output voltqge to 30 Volt - (2X0,65Volt)! If instead the constructor had used a PNP driver, the circuit would not have this limitation, but then there comes other issues! Still, let us go for the old uA723, because when we know all the limitations, we can come around them, and get a great powersupply!
The power dissipation in the pass transistor has nothing to do, with the power dissipation the datasheet says as the maximum. If You want 70 Volt out, You got to have at least 75 Volt raw supply, and shorting the output, pulls the maximum current of 2 Ampere, that gives 150 Watt.
That would cause one single power transistor to be glowing red!
An other thing, I forgot is, that the good old 2N3055 has a maximum Uce of 70 Volt, so we can not use it for that high a voltage. It calls for a 2N3773. - and more of them!
Regards Soeren
I know that circuit too. it is good, but it has one minor issue: The operational amplifiers can only cope with +/- 18 Volt at the maximum, and that is 36 Volt from + to -.If the - pin is held at -6 Volt, the + pin may at max be 30 Volt.
As the pass transistors is configured as NPN darlington transistors, the output of the integrated circuit must go 2 Ube voltages higher, and that limits the output voltqge to 30 Volt - (2X0,65Volt)! If instead the constructor had used a PNP driver, the circuit would not have this limitation, but then there comes other issues! Still, let us go for the old uA723, because when we know all the limitations, we can come around them, and get a great powersupply!
The power dissipation in the pass transistor has nothing to do, with the power dissipation the datasheet says as the maximum. If You want 70 Volt out, You got to have at least 75 Volt raw supply, and shorting the output, pulls the maximum current of 2 Ampere, that gives 150 Watt.
That would cause one single power transistor to be glowing red!
An other thing, I forgot is, that the good old 2N3055 has a maximum Uce of 70 Volt, so we can not use it for that high a voltage. It calls for a 2N3773. - and more of them!
Regards Soeren
The circuit p12 of this AN is much more serious: http://www.jaycar.com.au/images_uploaded/FN817.PDF
It can be scaled to basically any current/voltage with suitable components
It can be scaled to basically any current/voltage with suitable components
Hello Elvee.
That circuit looks fine, and You are right, it can be configured at almost any high voltage, because it has the Q4 on the output of the IC. That is exactly, how I am intending to use the old uA723. NPN pass transistors, and a PNP driver for the pass transistor, then the uA723 must pull low, when more output current is needed. Then again, we must respect the maximum Uce of the output transistor in the uA723. Regards Soeren Poulsen
That circuit looks fine, and You are right, it can be configured at almost any high voltage, because it has the Q4 on the output of the IC. That is exactly, how I am intending to use the old uA723. NPN pass transistors, and a PNP driver for the pass transistor, then the uA723 must pull low, when more output current is needed. Then again, we must respect the maximum Uce of the output transistor in the uA723. Regards Soeren Poulsen
Soeren Poulsen and Elvee you input is very much appreciated. Please could you/someone tell me in dummy terms if....1. The 723 circuit could remain basic the same but with the replacement and addition of exe times 2n3773 and how this may be configured to make the thing workable. And 2. If the tl081 circuit would function as desired if higher voltage op amps where used instead?. And lastly, if I where to go for the fig 14 option in datasheet link Elvee posted, what additional transistors would be required to give 2 amperes max current please? Thank you guys.
Last edited:
I took a look at the Intersil CA3130 design for .1 to 50vdc output Elvee posted. Does anyone know what is the limiting factor for preventing the output to going to zero volts.
For 50V, 2 paralleled 2N3055 would be the bare minimum. For 70V, 2 2N3773.
Note that most of the parts of the schematic are obsolete, but can easily be replaced with more modern ones: the opamp could be a TLC271, and the CA3086 can be replaced by a 15V zener and a temperature compensated one like the 1N825.
Adding a series resistor between the base of Q4 and the opamp output would be a good idea too.
With some modifications, the current limit could go at or near zero
Hello Elvee.
Have You consideret using the little but dammed smart TLC431?
- a variable zenerdiode! - used in almost every SMPSU.
Soeren Poulsen Denmark
Have You consideret using the little but dammed smart TLC431?
- a variable zenerdiode! - used in almost every SMPSU.
Soeren Poulsen Denmark
0-70V lab supply would require huge heat sinks, consider range from 30V -70V or use a multi-tapped transformer.
Ok, so as most of this is way above my head I have started learning about transistors and so far learned a little about common emitter/base/collector config's and characteristics of each. And a little about characteristic curves and biasing. But as I search to find what I think is relevant to my progress in learning of power supply design, all I find is amp circuits. Please could someone point me in the right direction and maybe outline a few basic things I should be learning in order to gain some understanding of what is going on in these psu schematics. Thanks for your input all, maybe soon I will understand more of what you are talking about.
Hi
ESP site is a reasonable place to peruse ideas for linear PS projects, it is practical with a little theory thrown in the mix. ESP Projects Pages - DIY Audio and Electronics
always keep devices SOA under consideration esp as voltages go higher http://en.wikipedia.org/wiki/Safe_operating_area
precision lab supplies should always have a good current limit scheme, IMO bullet proof LAB PS are the most challenging PS project you can do.
ESP site is a reasonable place to peruse ideas for linear PS projects, it is practical with a little theory thrown in the mix. ESP Projects Pages - DIY Audio and Electronics
always keep devices SOA under consideration esp as voltages go higher http://en.wikipedia.org/wiki/Safe_operating_area
precision lab supplies should always have a good current limit scheme, IMO bullet proof LAB PS are the most challenging PS project you can do.
Last edited:
Hello jimmiegin.
I so respect, that You want to learn about electronics at component level. As You might have guessed, I am an old (61 year) radio-tv-tech, who has studiet a little more. A variable power-supply, is the first, and most nessary equipment, to have, after the multimeter. I have never wanted to scare You away, but, there are some electronic facts, You can not omit. Start a project file. Write on top: 0 - 70 Volt variable powersupply 0 - 2 Ampere currentlimit around a uA723.
Then I will walk You through each block of the schematic, and I so hope a lot of other are watching in, and correcting me, if I make an error.
The first thing: You need 70 Volt out, and 2 Ampere, current limited.
You must have a raw voltage supply larger than the maximum voltage, You want out of Your circuit. I am "shooting from my hip" saying, that You will need at least 75 Volt, raw voltage.
If You short-circuit the output, with the currentlimiter set to 2 Ampere, You will dissipate 150 Watt of power, in Your power-transistors. That is a large amount of heat. Rod Elliot at Elliot Sound Products has written how to calculate heatsinks in relationship to the permissible temperature raise of the power-transistors. If You can get a large heatzink from old thrown-out equipment, You can never get it large enough. If You are buying a heat zink at the nessary size, I think it will be at least 2 cubic decimeter large, and that is expensive. You wrote, that the datasheet stated, that the 2N3055 could cope with 90 Watt, Yes, if You can keep the chip inside at 25 degree centigrade! OK. Lets start this project. Soeren Poulsen
I so respect, that You want to learn about electronics at component level. As You might have guessed, I am an old (61 year) radio-tv-tech, who has studiet a little more. A variable power-supply, is the first, and most nessary equipment, to have, after the multimeter. I have never wanted to scare You away, but, there are some electronic facts, You can not omit. Start a project file. Write on top: 0 - 70 Volt variable powersupply 0 - 2 Ampere currentlimit around a uA723.
Then I will walk You through each block of the schematic, and I so hope a lot of other are watching in, and correcting me, if I make an error.
The first thing: You need 70 Volt out, and 2 Ampere, current limited.
You must have a raw voltage supply larger than the maximum voltage, You want out of Your circuit. I am "shooting from my hip" saying, that You will need at least 75 Volt, raw voltage.
If You short-circuit the output, with the currentlimiter set to 2 Ampere, You will dissipate 150 Watt of power, in Your power-transistors. That is a large amount of heat. Rod Elliot at Elliot Sound Products has written how to calculate heatsinks in relationship to the permissible temperature raise of the power-transistors. If You can get a large heatzink from old thrown-out equipment, You can never get it large enough. If You are buying a heat zink at the nessary size, I think it will be at least 2 cubic decimeter large, and that is expensive. You wrote, that the datasheet stated, that the 2N3055 could cope with 90 Watt, Yes, if You can keep the chip inside at 25 degree centigrade! OK. Lets start this project. Soeren Poulsen
consider "fold back" current limit and reduced output range to get a reasonable heat sink. I think even the uA723 provides that feature. do some basic heat sink calculations 1st
Hello Infinia.
I agree on Your suggestion of using a foldback. Here I must admit, that I have never build a foldback circuit in a power supply. I think an easier way for a starter, is to use a multi-tap transformer, and make a relay circuit, stepping up or down, in relationship to the output voltage. If one has a transformer with 15, 30, 45 and 60 Volt tap, the maximum dissipation would be 30 Watt at shortcircuited output. By the way, I really like the 0 - 30 Volt, 0,002 - 3 Amp circuit in Electronic lab, but it has its problems too. 1'st problem mentionet is that 1 power transistor is too little, it runs very hot. regards Soeren Poulsen
I agree on Your suggestion of using a foldback. Here I must admit, that I have never build a foldback circuit in a power supply. I think an easier way for a starter, is to use a multi-tap transformer, and make a relay circuit, stepping up or down, in relationship to the output voltage. If one has a transformer with 15, 30, 45 and 60 Volt tap, the maximum dissipation would be 30 Watt at shortcircuited output. By the way, I really like the 0 - 30 Volt, 0,002 - 3 Amp circuit in Electronic lab, but it has its problems too. 1'st problem mentionet is that 1 power transistor is too little, it runs very hot. regards Soeren Poulsen
yes you at least need one or 2 modern 250W pass devices
2X26V secondary windings is easy to find.
make it bullet proof is the hard part.
2X26V secondary windings is easy to find.
make it bullet proof is the hard part.
Last edited:
Hello Infinia.
You are right about using one or two 250 Watt pass elements, but they are so expensive.
Finding 4 or 6 old transistors like 2N3773 or equalent in a dumpster and an appropiate heatzink will be mouch cheaper! and jimmiegin could just buy a readymade powersupply, but then he had learned nothing! All best wishes for us all! Soeren Poulsen
You are right about using one or two 250 Watt pass elements, but they are so expensive.
Finding 4 or 6 old transistors like 2N3773 or equalent in a dumpster and an appropiate heatzink will be mouch cheaper! and jimmiegin could just buy a readymade powersupply, but then he had learned nothing! All best wishes for us all! Soeren Poulsen
- Status
- Not open for further replies.