Maker Pro
Maker Pro

Designing A Power Supply

Javier

Mar 25, 2012
49
Joined
Mar 25, 2012
Messages
49
Hi all; I have a switching power supply which I took from the forgotten circuit box and I would like to make use of it; this switching power supply provides 42 VDC across its output but I want to make this voltage level be variable or adjustable from 1.25V to 30 VDC by mean of an integrated voltage regulator such as LM317 or LM350K. The problem is that these regulator ICs can be connected to 40 Vdc as maximum input voltage and my switching power supply delivers 42V therefore these regulators can be not used. Is there any other regulator IC making the same task as the LM317 but allowing more voltage at its input? a friend of mine told me to place several diodes in series with the output of the switching power supply until obtaining a safe value of voltage for the LM317 like 35V for example; any answer or suggestion from you would be of great help to me. thank you and God bless you
 

Harald Kapp

Moderator
Moderator
Nov 17, 2011
13,700
Joined
Nov 17, 2011
Messages
13,700
A few issues here:
  1. Regulating from 40 V (or 42 V) down to 1.25 V...30 V by a linear regulator is not the best idea. The linear regulator needs to drop the voltaeg difference and produce proportionally high waste energy in the form of heat. It can be done, but a rather big heat sink will be required.
  2. For this reason I recommend you use a switch mode regulator to convert the output of the junk PSU to the required level. These regulators can be bought off the shelf for various voltage/current levels. You may even be able to find one which accepts 42 V input, so no change to the existing PSU is required.
  3. Reducing the 42 V to 40 V by series diodes is possible, but inefficient for the reason mentioned before. Typically such a power supply has a feedback mechanism (voltage divider) by which it senses the output voltage and controls the regulation such that the output is stable at the set voltage level. By manipulating this feedback (change the resistor ratio) you should be able to reduce the 42 V to a more mangeable output, e.g. 40 V or even less to minimize waste power in the subsequent regulation.
    It is probably not a good idea to try to achieve the full range you envision by manipulating this feedback as the components within the PSU are probably not well suited to output voltages very much lower than 40 V.
 

Javier

Mar 25, 2012
49
Joined
Mar 25, 2012
Messages
49
harald, thank you for you response, I have no choice to achieve what I want but I like the second option; someone told me there is a LM317 with the final suffix HV which can be connected with 60 vdc in its inpu but I don´t know if it is sold in my country. regard to the second option, I would like to learn more about the switch mode regulator; if you have any information about it, please, show it to me; thank you harald, THANK YOU...
 

Javier

Mar 25, 2012
49
Joined
Mar 25, 2012
Messages
49
Harald, Thanks for your help; Yes, that LM317HV is a high input voltage and I think it is the easiest way to achieve what I want; of course I need a large heatsink to cool the regulator; I have many heatsinks, the problem is that maybe it is difficult to find a LM317 with this suffix HV because most of ICs sold here have the suffix T; I am gonna read the information you gave me; thank you Harald
 

Javier

Mar 25, 2012
49
Joined
Mar 25, 2012
Messages
49
it was so wonderful to be a truth; LM317HV is not sold in my country; it is not even known in the electronic parts sale stores :( ..reading a little about the switch mode, maybe this method will not work as I wish because I wanna use a potentiometer to control the output voltage for the power supply obtaining voltage levels from 1.5v to 30 volts; I want to try the LM317, I have no choice; Harald, I have many large heatsinks, I think one large can prevent the regulator from breaking down, well, I will wait for your opinion and help. God bless you.
 

Harald Kapp

Moderator
Moderator
Nov 17, 2011
13,700
Joined
Nov 17, 2011
Messages
13,700
I have many large heatsinks, I think one large can prevent the regulator from breaking down
The HV variant is specifically released for high input voltage. This is not a matter of the size of the heatsink but of the dielectric strength of the semiconductor elements within the chip. Using a non-HV type will almost surely lead to destruction.

reading a little about the switch mode, maybe this method will not work as I wish because I wanna use a potentiometer to control the output voltage
Switch mode regulators can be controlled by potentiometes, too.
See e.g. this module.
 

davenn

Moderator
Sep 5, 2009
14,254
Joined
Sep 5, 2009
Messages
14,254
The HV variant is specifically released for high input voltage. This is not a matter of the size of the heatsink but of the dielectric strength of the semiconductor elements within the chip. Using a non-HV type will almost surely lead to destruction.


Switch mode regulators can be controlled by potentiometes, too.
See e.g. this module.

yup, agreed
no need to look for any other solutions


Dave
 

Javier

Mar 25, 2012
49
Joined
Mar 25, 2012
Messages
49
thank you Harald and Dave for your answers; You are right Harald; I am going to read the article you posted; thank you for wasting your valious time on my requeriment; thank you.
 

73's de Edd

Aug 21, 2015
3,613
Joined
Aug 21, 2015
Messages
3,613

Seeeeeeeeeeenor Javier
. . . . . . .(and thats being heaveeee on the Seeeeeeeeeeeeee)


Try This One

Breadboard it up and . . . . . . . . check it out . . . . . . .

CAVEAT . . . . . breadboarding up this circuitry reveals a need to establish a safe low voltage reliability threshold of ~ 10 VDC . . .so use it for 10VDC, upwards. . . . of DC output voltage.
Also watch current pull . . . as dissipated wattage across the '317 increases rapidly, at ever increasing higher voltage output levels.


csbooBo.png



In more recent times, when I needed a high voltage regulator . . . BUT . . . . didn't want to take a quantum leap back in neanderthal design technology,by having to use 6L6 or 6080 kluge vacuum tubes.

I would just use this type of circuitry quirk, which permits using the common LM317 for its control element, by the control of the max voltage being available across the LM317 by the shunt zener . . . .voltage limiter.

I have made up to a 250V supply using this methodology . . . .but at those high voltages, naturally, a lower current draw is the norm.Since a mere 100 ma at 250 volts equates to 25 watts of power !

BUT . . . . one could also use the technique of a shunting power transistor across the LM317 in and out to up the devices power capability.

Consult NATIONAL data sheets for the LM317 for more info. . . . . or I think that Tejas Instrumentos. . . . . in being their new Sugar Daddy / Moneybags . . . . has carried them forward into their tech files and application notes.


Errata:
I now see that the adjustable resistor pair for voltage scaling, had the K of 2.2 K omitted, and the variable 10K might have to be scaled up or down a bit, to trim in your voltage adjustment end limits.





73's de Edd
 
Last edited:

Harald Kapp

Moderator
Moderator
Nov 17, 2011
13,700
Joined
Nov 17, 2011
Messages
13,700
@73's de Edd : Your circuit will work for out output voltages down to 61 V ( a bit more considering the accuracy of the zener).
The op required 1.25 V ... 30 V. This is not going to work as the zener will become conducting, overriding the LM317's regulation.
 

73's de Edd

Aug 21, 2015
3,613
Joined
Aug 21, 2015
Messages
3,613
Sir Harald . . . . . .et al . . . .

That circuitry is that which I used to replace an OA3 gas tube regulator.
It "woiks" magnificently . . . . .even with no omnipresent glow..
In the case of the lower thresholds of the lower voltages mentioned, the 39V Zener still protects the LM317, while the other zener transitions into being a 1N4006.

Thasssitttt . . . .

73's de Edd
 
Last edited by a moderator:

Harald Kapp

Moderator
Moderator
Nov 17, 2011
13,700
Joined
Nov 17, 2011
Messages
13,700
@73's de Edd : I still disagree: At Vin=42 V and Vout=1.25 V, the voltage across the LM317 needs to be 40.75 V. There's no other way this can be achieved with this circuit. At 40.75 V the 39 V zener diode will be conducting, however, whic will lead to a current that increases the output voltage to (42 V -39 V) = 3 V, thereby anihilating the regulation down to 1.25 V.

Or which detail am I missing?
 

Javier

Mar 25, 2012
49
Joined
Mar 25, 2012
Messages
49
Harald and 73's , thank you for your help... I am very happy about men like you who loves helping others..
well, 73s has recommended me his circuit based on a 39v zener diode and Harald did not agree with it.. also I am thinking of modifying the feedback circuitry to obtain at least 35 VDC as a input voltage for the LM317; Haradld, I know your suggestion of switch mode is the best option but in a country like mine is very difficult to find that kind of ICs; if the feedback circuitry can be modified respect to its resistance divisor, I think it would be a good option, of course I just give a poor opinion, master.. I am gonna draw the shcematic of this switching power supply and I will upload to the thread.. God bless you, and thank you
 

Alec_t

Jul 7, 2015
3,587
Joined
Jul 7, 2015
Messages
3,587
Perhaps use one LM317 to drop from 42V to 33V, then go from there with a second LM317?
 

73's de Edd

Aug 21, 2015
3,613
Joined
Aug 21, 2015
Messages
3,613
Well I just saw that Sr Javier made it back to the scene and needs a more positive answer.

SO . . . . I pulled a radar IF amplifier strip assembly off my $100 an hour spot on the bench and cobbled up that regulator circuit to evaluate its operation at the very low voltage levels.
Indeed it does start getting dodgy at about 7 V, where its stops having a linear change and starts having a cramped /compressed variance until it stops.
That must relate back to the reason for the 4.7 V zener shown in the initial schematic . . . .to make an abrupt limiting knee on adjustment in the low voltage end of the spectrum.

This action / evaluation was being ascertained by using " Breadus Boardamus ".

My circuit designs should be regarded as experimental. They are unbuilt but have been tested by Spice simulation.

Back in MY initial knowledge base spectrum spread . . . . . initially in the 1946'ish era . . . . . Spice was a condiment !

Now in your situation,with a parts availability / obtainability aspect . . . . . . . . is it easy to get a LM317 there ?
It would also be able to use a pass power transistor as a preregulator between your present switch mode power supply to INITIALLY get down below that 40VDC LM 317 voltage input threshold .
Am i correct that a max of 1 amp is the power supply's current supply specification ?

Not that It will help you but I have 3 LM783 high voltage regulators on hand in my stock that work with up to 125 VDC input and will regulate down to that desired low adjustable voltage that you want.
I find that mailing one to you from Texas to Columbia will cost 20 times the initial price of one.
The Lm 783 uses a higher voltage Mosfet instead of a conventional bipolar transistor for its HV pass transistor

http://www.ti.com/lit/ds/symlink/tl783.pdf

(Refer to Page 12 Fig 20)

Or if you can supply a model number AND brand of that switch mode power supply of yours, we might research its design to see about modifying its output to get it down to just below that 40 V specification.

Am I correct, that a max of 1 amp is the power supply's current supply specification ?

By inspecting your main power supplies schematic/ design, the voltage limit may be as easy as changing the voltage division ratio of two resistors, or their associated zener diode, associated with the units optical isolators feedback circuitry.




73's de Edd
 
Last edited by a moderator:

Javier

Mar 25, 2012
49
Joined
Mar 25, 2012
Messages
49
thank you 73's and alec for your answer; I could not understand the method given by alec but it is valid to analize it; 73's: I did not know there was a linear regulator with so high input voltage such as LM738 but tomorrow I will find it out if it's sold in my country (not even imagine it in my true country Venezuela). the other option is modifying the feedback circuitry and I will draw the schematic and upload it to this thread to help me to change the resistor values necessary to obtain an output voltage lower than the original voltage; God bless you
 

Javier

Mar 25, 2012
49
Joined
Mar 25, 2012
Messages
49
I got this power supply from a Epson printer but I didn't write the model;;; maybe It can deliver approx. 2 amps
 

AnalogKid

Jun 10, 2015
2,884
Joined
Jun 10, 2015
Messages
2,884
If you are limited to the non-HV version of the LM317, there are two options to get the output voltage range you want. Actually, three variations of the same option, a pre-regulator.

Note that the input voltage range of the standard LM317 regulator is almost unlimited. It is the voltage differential, the difference between the input voltage and the output voltage, that is limited to 37 V. So if you have two LM317 circuits in series, with the first one set to deliver a fixed 34 V output and the second one adjustable from Vref to 30 V, you are there. The second variation is to make the pre-regulator with a simple power transistor and zener diode rather than an LM317, but the 317 way gives much more stable results. The third variation is to make both LM317 circuits variable using a dual-gang pot. This is a bit more complex, but it shares the total heat load evenly between the two devices for better long-term reliability.

ak
 
Top