Hi all,

Before wasting time and money, would you kind enough to check some schematics found on websites?

For 0-12 V Variable power supply, D3 and D4 are not the best way.
Better with a CT transformer and negative to get -1.24V


Thanks a lot, and nice day.

jmCS 9012

 

D3 and D4 are a cheap and simple way to drop the LM350’s minimum output voltage of 1.25 volts to zero on the output terminals.
It works and is failsafe. An added bonus is they prevent current back flow if the AC power is lost.

 

I like this part for a bench power supply. It does go down to 0V on the output. There is also a LT3081 . LT3080

 

For 0-12 V Variable power supply, D3 and D4 are not the best way.
Better with a CT transformer and negative to get -1.24V

You're right. Putting diodes on the output of the voltage regulator is a lame idea. You no longer have a regulated output voltage because it depends on the current through the diodes.

The second schematic doesn't have very good voltage regulation either.

 

Below is the LTspice of one way to get an LM317 (or LM350) to adjust to zero volts with good output regulation without requiring a CT transformer:
It uses a capacitor/diode voltage-doubler type circuit to generate a negative voltage, which is regulated to -1.25V by the inexpensive LM337 to offset the +1.25V minimum output of the LM317.
The LM337 can be a small package version since it only has to supply about 10mA of current.

 

Hi AAC members and Schmitt Trigger, Ron Simpson, dl324, Crutschow

Thanks a lot for your advices.

Always a pleasure to learn with your replies.

Nice day.

jm

 

Below is the LTspice of one way to get an LM317 (or LM350) to adjust to zero volts with good output regulation without requiring a CT transformer:
It uses a capacitor/diode voltage-doubler type circuit to generate a negative voltage, which is regulated to -1.25V by the inexpensive LM337 to offset the +1.25V minimum output of the LM317.

View attachment 364532

Hi AAC members and Schmitt Trigger, Ron Simpson, dl324, Crutschow

Thanks a lot for your advices.

Always a pleasure to learn with your replies.

Nice day.

jm

Hi Crutschow,

One more time, thanks alot ::)).

I need an output of 2A to be comfortable with the device.

Good choice i think.

Nice day.

jm

 

I need an output of 2A to be comfortable with the device.

Then you want to replace the LM317 in my schematic with an LM350 or similar.

 

Then you want to replace the LM317 in my schematic with an LM350 or similar.

Thanks.

 

keep in mind that LM350 must have a large heatsink. without it IC will shut down due to thermal limit.

 

keep in mind that LM350 must have a large heatsink. without it IC will shut down due to thermal limit.

Hi panic mode,

Thanks for your advice.

Now, i get strong advices to get a '' bullet-proof '', an reliable PSU.

Going to compute thermal calculations about Heatsinks; i have a lot of docs about them - and like it - .

Nice day.

jm

 

your regulator Pd is about 25-30W depending on package and manufacturer. some may claim 80+W but that will not work without cryochamber

 

your regulator Pd is about 25-30W depending on package and manufacturer. some may claim 80+W but that will not work without cryochamber

Or if you live in Pluto.

 

Or if you live in Pluto.

That's one heck of a commute.

 

Hi Schmitt Trigger and Crutschow,

Sadly, you are right ::::((((((

I made briefly some calculus with a spreasheet.











So, i think if you told me to ride to Pluto, it was a good reason!!!!

Would you kind enough to check if i made mistake(s) or ?

I put a 40°C in Ambient Temperature, and bad idea of course......

With a fan, is it possible to have an LM350 delivering 2 A, not more?

Found some heat sinks but bulky.

Another answer is to immerse the whole thing in Mediterranean Sea :::::))))))) closer than Pluto.

Thanks a lot and nice day.

jm

 

The make heatsinks with built-in fans ( as typically used for computer processor cooling) which can dissipate a significant amount of heat with a relatively small size.
You might look at using one of those.

 

The bottom line is that you cannot realistically have that extra wide input to output ratio with a linear regulator. You will have to consider an additional switchmode stage somewhere.
I would attempt a linear to drop 800 to 380V, followed by the switch mode for the 380 to 12V conversion.
Having said that, for the first stage finding a series pass transistor with that much SOA, will be a significant challenge.

 

The make heatsinks with built-in fans ( as typically used for computer processor cooling) which can dissipate a significant amount of heat with a relatively small size.
You might look at using one of those.

Hi ,

Thanks a lot.

Nice evening.

jm

 

The bottom line is that you cannot realistically have that extra wide input to output ratio with a linear regulator. You will have to consider an additional switchmode stage somewhere.
I would attempt a linear to drop 800 to 380V, followed by the switch mode for the 380 to 12V conversion.
Having said that, for the first stage finding a series pass transistor with that much SOA, will be a significant challenge.

Hi,

Thanks also for your fast reply.

Nice evening.

jm