I'm using the 7805 as a variable regulator in a power supply, as I dont need voltage below 5v. The input power of the supply is +12v -12v DC (24v total). I have other regulators as well if they'd be better suited... lm317, 7912, a coupole LDO 5v and 12v, and lastly a variable 5-40v reg.

MY QUESTION IS... What would be the best way to drop the 7805's supplied input voltage from 24v down to ~15v? Diodes, power resistor, potential divider, DC-DC buck, cascade another higher rated v regulator?? I have a large heatsink and fan connected to the 7805 so maybe i'll be fine not needing to drop the input voltage? I just have no use for the higher volts and would like as much adjustment as possible with 15 turn pot being used. I had the 7805 connected across +5v -12v lines and didn't have any overheating issues, but the 5v line is now being used for a separate circuit. My goal is to have the voltage supplied to the load to be smooth and consistent without fluctuation due to heat.

Any input would be appreciated, thanks


also the current drawn is about 500-800ma, if that info is needed for anybody

 

My first instinct is a resistor, but that will limit your highest voltage.
4C/W in a TO-220 package says 15.2 watts will go to 60.8C above 25C ambient with an infinite heat sink. That leaves 39.2C you can work with. 2.57C/W heat sink needed.

 

The input of 7805 could be like this:
Vin=15V+3V=18V

Calculating the current limiting resistor -
The rest input voltage V_rest=24V-18V=6V, R-limit=6V/1A=6Ω
W=V*I=6V*1A=6W, choosing the value of watt should be equal or great than three times of calculation values, W = 6W x 3 = 18W, choosing 20W or some more up to 30W, the values of watt will be adjust according to your application Imax current.

So you need a current limiting resistor in series with Vin as 6Ω/20W or 6Ω/30W.

Whatever you want to adding the current limiting resistor or not, it's up to you.

 

Why not use a 15V regulator?

 

The fact that you are talking about +12V and -12V implies that you have a 0V somewhere. Why not just power the regulator between +12V and 0V?

What range do you want from the output (relative to 0V)?

 

Why not use a 15V regulator?

I think using 7815 and 7805 they all have the same problem, that is the Vin greater than what the input of regulator needs about 6V.

 

The fact that you are talking about +12V and -12V implies that you have a 0V somewhere. Why not just power the regulator between +12V and 0V?

What range do you want from the output (relative to 0V)?

If the op want o using the range of output voltage are 5~15V then 0,12V is not enough.

 

If the op want o using the range of output voltage are 5~15V then 0,12V is not enough.

But if his supplies are +12V and -12V, how is he getting +15V now? Unless he is talking about +15V relative to the -12V rail.

He needs to pick his voltage reference and then talk about voltages relative to that reference -- are at least make it darn clear that he isn't.

 

I think using 7815 and 7805 they all have the same problem, that is the Vin greater than what the input of regulator needs about 6V.

If the op want o using the range of output voltage are 5~15V then 0,12V is not enough.

These two responses seem at odds with each other.

 

These two responses seem at odds with each other.

I just thought the op was using +12V and -12V to be a +24V, so the -12V was the GND.

 

I'm using the 7805 as a variable regulator in a power supply, as I dont need voltage below 5v. The input power of the supply is +12v -12v DC (24v total). I have other regulators as well if they'd be better suited... lm317, 7912, a coupole LDO 5v and 12v, and lastly a variable 5-40v reg.

MY QUESTION IS... What would be the best way to drop the 7805's supplied input voltage from 24v down to ~15v? Diodes, power resistor, potential divider, DC-DC buck, cascade another higher rated v regulator?? I have a large heatsink and fan connected to the 7805 so maybe i'll be fine not needing to drop the input voltage? I just have no use for the higher volts and would like as much adjustment as possible with 15 turn pot being used. I had the 7805 connected across +5v -12v lines and didn't have any overheating issues, but the 5v line is now being used for a separate circuit. My goal is to have the voltage supplied to the load to be smooth and consistent without fluctuation due to heat.

Any input would be appreciated, thanks


also the current drawn is about 500-800ma, if that info is needed for anybody

There's a trick described in various application notes of putting a resistor in series with the 3 terminal regulator input pin, you arrange that resistor to drop 0.7V at the current you want the regulator to "hand over" . An external PNP power transistor has its B/E across the series resistor and the collector is joined to the 3-TR as an alternate current path - this is usually used to get more than the standard 1A current, but you can also reduce regulator dissipation by reducing its current load.

But any way you look at it - if you're dropping 24V down to 5V with a linear regulator, there's going to be a lot of heat that's got to go somewhere.

With that sort of dissipation burden, its well worth considering switch-mode - the MC34063 is very popular and easy to use - there's several online calculators that you just type in your parameters and it gives you what size inductor to use.

 

With that sort of dissipation burden, its well worth considering switch-mode - the MC34063 is very popular and easy to use - there's several online calculators that you just type in your parameters and it gives you what size inductor to use.

Though keep in mind that, if "keeping things smooth and consistent" is really important, that switch-mode supplies inject quite a bit of power supply noise. The question becomes how much is too much?

 

I just thought the op was using +12V and -12V to be a +24V, so the -12V was the GND.

Yes... -12 is being used as ground... the desired Vin to the IC relative to 0v is aprox 15v. I'm looking for an adjustable 5-12v to the load so to achieve my Vmax needed the 7812 (as well as using the +12v and 0v rails) is cutting it just a bit too close as there is an LED just before the load and, from my understanding, with either the 7812 or 7815 I would need a potential divider (or other method) to reduce the regulated 12v or 15v down to the 5-12v that I'm wanting. Also, I understand that a potential divider to reduce voltage isn't the most consistent method due to the fluctuation of the resistors temperature from passing current...? I'm still pretty new to EE and I'm sure there's another simpler/better method than how I'm going about this. I'm also just trying to use parts that I have. I can draw up a scheme of the circuit so far if it would help.

Thank you for all the input thus far

 

With that sort of dissipation burden, its well worth considering switch-mode - the MC34063 is very popular and easy to use - there's several online calculators that you just type in your parameters and it gives you what size inductor to use.

This is where I'm headed next if the linear supply doesn't work out... just figured I'd work with the simpler, less time consuming method first... which I suppose could have now backfired as far as time goes lol.

 

Though keep in mind that, if "keeping things smooth and consistent" is really important, that switch-mode supplies inject quite a bit of power supply noise. The question becomes how much is too much?[/QUOTE

There are two different devices that will be powered through this supply most of the time, those being a small dc rotary motor and a pulsating (on/off about 80-110 cycles/s) electromagnet. The most important thing for me is the volts remaining consistent with no more than about +/-.5v fluctuation when powered and running.

 

If you can tolerate +/-0.5V, then you shouldn't have any problems with the noise from a SMPS. You can slave both it and the adjustment for your final output to the same pot such that as you change the desired output voltage the about voltage of the SMPS adjusts to keep the voltage drop across the 7805 pretty much constant.

 

This is where I'm headed next if the linear supply doesn't work out... just figured I'd work with the simpler, less time consuming method first... which I suppose could have now backfired as far as time goes lol.

Actually - there's SMPSU replacements for the 3 terminal regulators, and elector also did an article for those wanting to build their own.

Unfortunately, I've not heard of a variable version - maybe that's changed since last time I looked.

 

If your input side is the same with my circuit, then the circuit could auto detecting from the output voltage around 9.1V and auto selecting the input voltages from 12V and 24V, it will reducing the dissipation.