Need help finding a suitable voltage regulator IC.

Thread Starter

rock94

Joined Apr 12, 2013
7
I'm planning to build a dual DC power supply unit over the weekend.The transformer I'm using is a 12-0-12 with a 3 amp output.There seem to be very few voltage regulators with a high output current.I do have the lm338k, but it doesn't seem to have a complementary negative voltage regulator.The only pair I've found online is the lm350/lm333.Unfortunately the lm333 is not available at any store though the lm350 is.Could anyone suggest me a pair with a current output >=3A or any other way around this.Also if I go ahead with the lm338 and build a 1.2-24v variable supply what are the drawbacks when compared to a Dual +12/0/-12 variable supply?Any thoughts or suggestions are welcome.
 

t06afre

Joined May 11, 2009
5,934
You do know. That in a setting like yours. The max recommended DC current. Is 1.8 times less the transformer RMS current rating. In your case probably (3/1.8) Ampere
 

ScottWang

Joined Aug 23, 2012
7,397
Expanding the current is another way.

As this one you could find the positive voltage power, and you can adding LM337 and 2N2955 to get the negtive voltage power.

The negtive voltage power that you can refer to linked page from Dodgydave.
 

Ron H

Joined Apr 14, 2005
7,063
If you had a transformer with two isolated secondary windings, you could make two separate power supplies, and either one could then be a positive or a negative supply. Each one could use an LM338k as the regulator.
 

Ron H

Joined Apr 14, 2005
7,063
Expanding the current is another way.

As this one you could find the positive voltage power, and you can adding LM337 and 2N2955 to get the negtive voltage power.
That circuit will have inferior temperature stability and regulation, compared to an IC regulator or a feedback regulator.
 

Thread Starter

rock94

Joined Apr 12, 2013
7
If you had a transformer with two isolated secondary windings, you could make two separate power supplies, and either one could then be a positive or a negative supply. Each one could use an LM338k as the regulator.
No.its got only one secondary winding and both the positive and negative regulators will be connected to the center-tap(ground). Will it work if i use multiple setups and connect the outputs in parallel(+ve to +ve and -ve to -ve) to get a higher current output?
 

ScottWang

Joined Aug 23, 2012
7,397
That circuit will have inferior temperature stability and regulation, compared to an IC regulator or a feedback regulator.
Thanks.
Could you explain what's your viewpoint that maybe I didn't attend to?
Because that i aslo used some other power supplies using LM723 and 2N3055, I bought those kits long time ago and they still working.
 

tubeguy

Joined Nov 3, 2012
1,157
No.its got only one secondary winding and both the positive and negative regulators will be connected to the center-tap(ground). Will it work if i use multiple setups and connect the outputs in parallel(+ve to +ve and -ve to -ve) to get a higher current output?
You shouldn't connect 2 regulators in parallel without some circuitry to properly share the load between them. Google parallel regulators.

For simplicity you might want to start with a single supply using your lm338 and existing transformer.
Then build a second identical one as Ron H suggested. Then you have more versatility.
 
Last edited:

Ron H

Joined Apr 14, 2005
7,063
Thanks.
Could you explain what's your viewpoint that maybe I didn't attend to?
Because that i aslo used some other power supplies using LM723 and 2N3055, I bought those kits long time ago and they still working.
I didn't intend to get into a detailed critique of your circuit, and I'm not saying it won't work. I'm just saying that power supplies, for temperature stability and lowest output impedance (i.e, voltage regulation), should have all temperature-sensitive components and driver transistors inside a feedback loop. I admit that it's difficult to make a well-regulated, high current linear supply. It's also questionable as to why one would want to do this, when switching regulators are so much more efficient.

Your circuit, in simulation, has an output voltage change of about 1.7V when the load changes from 100mA to 8A. That may be OK in many applications.

Thermally, the heating of the transistors and the diodes will cause thermal transients as the junctions heat and cool. Again, that may be OK for many applications.

As an aside, R9 is only good for about 100mA at 0.5W. You might want to change it to 2W.
 

ScottWang

Joined Aug 23, 2012
7,397
I didn't intend to get into a detailed critique of your circuit, and I'm not saying it won't work. I'm just saying that power supplies, for temperature stability and lowest output impedance (i.e, voltage regulation), should have all temperature-sensitive components and driver transistors inside a feedback loop. I admit that it's difficult to make a well-regulated, high current linear supply. It's also questionable as to why one would want to do this, when switching regulators are so much more efficient.

Your circuit, in simulation, has an output voltage change of about 1.7V when the load changes from 100mA to 8A. That may be OK in many applications.

Thermally, the heating of the transistors and the diodes will cause thermal transients as the junctions heat and cool. Again, that may be OK for many applications.

As an aside, R9 is only good for about 100mA at 0.5W. You might want to change it to 2W.
Thank you Ron H.

I knew that kinds of power are ok for many applications, I just want to know something deeper.

What are you saying about the temperature stability, it that using as Thermistor sticking together with the Power transistor to sense the temperature and using the circuit to adjust the current?

I want to build a new liner power supply and using the transformer, I will post it as a new thread.
 

t06afre

Joined May 11, 2009
5,934
Thank you Ron H.

I knew that kinds of power are ok for many applications, I just want to know something deeper.

What are you saying about the temperature stability, it that using as Thermistor sticking together with the Power transistor to sense the temperature and using the circuit to adjust the current?

I want to build a new liner power supply and using the transformer, I will post it as a new thread.
I have about all elektor magazines from 1975 to present day. What kind of power supply do you want to build. I am sure I can find a project for your needs
 

Ron H

Joined Apr 14, 2005
7,063
Thank you Ron H.What are you saying about the temperature stability, it that using as Thermistor sticking together with the Power transistor to sense the temperature and using the circuit to adjust the current?
Temperature stability is the ability of the circuit to maintain a constant output voltage, even if the ambient temperature and/or the temperature of individual devices within the circuit changes.
Your circuit has 4 diode junctions (3 diodes and a BE junction) between a stable voltage (LM317) and the output. The temperature coefficient of a diode junction is about -2mV/°C. The temperature of your power transistors will vary a lot between no load and full load. The 1N4001s will also heat up to some extent at full load.
 

ScottWang

Joined Aug 23, 2012
7,397
Temperature stability is the ability of the circuit to maintain a constant output voltage, even if the ambient temperature and/or the temperature of individual devices within the circuit changes.
Your circuit has 4 diode junctions (3 diodes and a BE junction) between a stable voltage (LM317) and the output. The temperature coefficient of a diode junction is about -2mV/°C. The temperature of your power transistors will vary a lot between no load and full load. The 1N4001s will also heat up to some extent at full load.
Thanks Ron H.
The circuit that I linked is not my design, but because I saw many circuits that they using the similar way, so I will also using it to many applications, normally is ok, but I found out the voltage is astable when I mearsuring as 2.00x V, but the x mV can't stay at a stable status, when the voltage is using upper than 1V is ok, but if using it under 100mV that it could be a problem.

You talking about the reason why the Temperature stability is so important, but how to solve the problem of temperature coefficient of a diode junction and the Vbe of transistor, any good solutions?
 
Top