Questions on op-amp based adjustable power supply

[[email protected]]

I was pointed to this link in from another post of mine regarding a
high-current adjustable power supply:
http://ludens.cl/Electron/Ps20/Ps20.html

I have all of the parts that I need and have started building a
prototype. I have a few questions before I can finish.

1) I could not find a LM336Z-5.0, but I did get a LM336Z-2.5. In the
page's notes, it says this is OK, but to change the values of R2 & R3.
For my 2.5v version, what would these two values be?

2) I am hoping to get an adjustable voltage range from 11v to 20v or
so. I would ideally like to use an 18 VAC transformer, but could not
find one locally with a high enough current rating. I figured 18 VAC
would translate into about 25 VDC after a full wave bridge and a large
filter cap. The only one I could find with high enough current is one
with a 26 VAC secondary. This makes about 40 VDC after the bridge &
filter cap. I understand it will cause the transistors to dissipate
more power, but I am mostly concerned about exceeding the operating
voltage of the 741. Is there an easy way to make this transformer work?
If I'm exceeding the supply voltage of the 741, would it work to use,
say, a 12v regulator for the 741's supply voltage? I do not have any
experience in working with op-amps.

I should expand on my application: I will be using this power supply to
charge a bank of capacitors rated at 700,000uF. This will demand A LOT
of current for a very short amount of time. I need them to charge
within a second or two, so if I use a power resistor between the supply
and the cap bank, it will have to be small. I have the transistors on
some pretty beefy heat sinks.

Regards,
Kevin Schneider

 

[colin]

I was pointed to this link in from another post of mine regarding a
high-current adjustable power supply:
http://ludens.cl/Electron/Ps20/Ps20.html

I have all of the parts that I need and have started building a
prototype. I have a few questions before I can finish.

1) I could not find a LM336Z-5.0, but I did get a LM336Z-2.5. In the
page's notes, it says this is OK, but to change the values of R2 & R3.
For my 2.5v version, what would these two values be?

Increase r3 or reduce r2/P1 etc. the exact ratios depends on what range of
output voltage you want,
there will now only be 2.5v acros R2 and P1 and output voltage less 2.5v
accross r3.
the voltage is in proportion to the resistance.

2) I am hoping to get an adjustable voltage range from 11v to 20v or
so. I would ideally like to use an 18 VAC transformer, but could not
find one locally with a high enough current rating. I figured 18 VAC
would translate into about 25 VDC after a full wave bridge and a large
filter cap. The only one I could find with high enough current is one
with a 26 VAC secondary. This makes about 40 VDC after the bridge &
filter cap. I understand it will cause the transistors to dissipate
more power, but I am mostly concerned about exceeding the operating
voltage of the 741. Is there an easy way to make this transformer work?
If I'm exceeding the supply voltage of the 741, would it work to use,
say, a 12v regulator for the 741's supply voltage? I do not have any
experience in working with op-amps.

I should expand on my application: I will be using this power supply to
charge a bank of capacitors rated at 700,000uF. This will demand A LOT
of current for a very short amount of time. I need them to charge
within a second or two, so if I use a power resistor between the supply
and the cap bank, it will have to be small. I have the transistors on
some pretty beefy heat sinks.

I wouldnt recomend altering the supply of the 741,
I would suggest a transformer of the right voltage,
it would mean a smaller transformer and less heat/smoke !

Colin =^.^=

 

[[email protected]]

Increase r3 or reduce r2/P1 etc. the exact ratios depends on what range of
output voltage you want,
there will now only be 2.5v acros R2 and P1 and output voltage less 2.5v
accross r3.
the voltage is in proportion to the resistance.



I wouldnt recomend altering the supply of the 741,
I would suggest a transformer of the right voltage,
it would mean a smaller transformer and less heat/smoke !

Unfortunately a lower-voltage power supply is not an option right now.

As far as over-powering the op-amp, I am somewhat confused on how they
are rated. According to the spec sheet for the 741 that I have, the
supply voltage is rated for "+/- 18v". Since I am using it with a
single-voltage supply, would I double this rating to find out the
effective supply voltage it is rated for? (i.e. if it normally operates
from -18v to +18v, would the supply voltage on a single-voltage supply
be 18*2 = 36v?)

Thanks
Kevin

 

[Donkey D...]

Unfortunately a lower-voltage power supply is not an option right now.

As far as over-powering the op-amp, I am somewhat confused on how they
are rated. According to the spec sheet for the 741 that I have, the
supply voltage is rated for "+/- 18v". Since I am using it with a
single-voltage supply, would I double this rating to find out the
effective supply voltage it is rated for? (i.e. if it normally operates
from -18v to +18v, would the supply voltage on a single-voltage supply
be 18*2 = 36v?)

Thanks
Kevin

To answer your most recent question, a '741 can operate from a 0V/+36V
supply, (absolute maximum).

Regarding the transformer, 18VAC will convert to about 24VDC after the
bridge rectifier, if the transformer's load approximates it's current
rating. e.g. A transformer rated at 12VAC, 5A will usually be within about
10% of 12V if a current of 5A is drawn. The voltage will increase as the
load is decreased and vice versa. Using a 26VAC transformer will result in
approx. 36VDC, ((26 x 1.414)-1.2), which may be higher if the typical load
of that transformer is not met, exceeding the rating of the '741. It would
not hurt to use a regulator, (at the minimum, a Zener diode), to keep the
'741's supply voltage within specs.
.... Donkey

 

[[email protected]]

To answer your most recent question, a '741 can operate from a 0V/+36V
supply, (absolute maximum).

Regarding the transformer, 18VAC will convert to about 24VDC after the
bridge rectifier, if the transformer's load approximates it's current
rating. e.g. A transformer rated at 12VAC, 5A will usually be within about
10% of 12V if a current of 5A is drawn. The voltage will increase as the
load is decreased and vice versa. Using a 26VAC transformer will result in
approx. 36VDC, ((26 x 1.414)-1.2), which may be higher if the typical load
of that transformer is not met, exceeding the rating of the '741. It would
not hurt to use a regulator, (at the minimum, a Zener diode), to keep the
'741's supply voltage within specs.
... Donkey

Thanks for the reply. Come to think of it, 36V is even high for a
standard voltage regulator. How would I use a zener diode, and what
value would I choose? If I do lower the 741's supply voltage, do I need
to connect any of the voltage reference components (U1, P1, R2, R3 in
schematic) to the same lower supply voltage or would they remain on the
+36V rail as they are in the schematic?

http://ludens.cl/Electron/Ps20/ps20.gif

Thanks again,
Kevin

 

[Donkey D...]

Thanks for the reply. Come to think of it, 36V is even high for a
standard voltage regulator. How would I use a zener diode, and what
value would I choose? If I do lower the 741's supply voltage, do I need
to connect any of the voltage reference components (U1, P1, R2, R3 in
schematic) to the same lower supply voltage or would they remain on the
+36V rail as they are in the schematic?

http://ludens.cl/Electron/Ps20/ps20.gif

Thanks again,
Kevin

You'll need to apply the lower voltage to the op-amp, the LM336Z/R1 and the
R2/P1/R3 voltage divider. You'd need a 1W Zener, something like 33V. Having
chosen the Zener, you'd then need to recalculate the resistor values to
suit.

.... DD

 

[Donkey D...]

regarding & less 20v is bridge A transistors result

You'll need to apply the lower voltage to the op-amp, the LM336Z/R1 and the
R2/P1/R3 voltage divider. You'd need a 1W Zener, something like 33V. Having
chosen the Zener, you'd then need to recalculate the resistor values to
suit.

... DD

Correction: You don't need to apply the lower voltage to the R2/P1/R3
voltage divider, if you calculate the correct values for these.
....Donkey

 

[colin]

range

To answer your most recent question, a '741 can operate from a 0V/+36V
supply, (absolute maximum).

Regarding the transformer, 18VAC will convert to about 24VDC after the
bridge rectifier, if the transformer's load approximates it's current
rating. e.g. A transformer rated at 12VAC, 5A will usually be within about
10% of 12V if a current of 5A is drawn. The voltage will increase as the
load is decreased and vice versa. Using a 26VAC transformer will result in
approx. 36VDC, ((26 x 1.414)-1.2), which may be higher if the typical load
of that transformer is not met, exceeding the rating of the '741. It would
not hurt to use a regulator, (at the minimum, a Zener diode), to keep the
'741's supply voltage within specs.
... Donkey

unfortunatly you cant reduce the positive or negative supply to the op amp,
if you reduce the negative supply,
the output transistors will not be able to turn off,
and the positive supply needs to be higher than the input to the op amp
(obviously)
however there are op amps with a higher voltage rating wich might be ok,
the other alternative is to reduce the inputs to the op amp so that the
positive supply can then be regulated,
so that its used like a differential amplifier,
or a string of big diodes folowing the rectifier ... you only need to lose 4
volts to keep the 741 happy

Colin =^.^=

 

[colin]

unfortunatly you cant reduce the positive or negative supply to the op amp,
if you reduce the negative supply,
the output transistors will not be able to turn off,
and the positive supply needs to be higher than the input to the op amp
(obviously)
however there are op amps with a higher voltage rating wich might be ok,
the other alternative is to reduce the inputs to the op amp so that the
positive supply can then be regulated,
so that its used like a differential amplifier,
or a string of big diodes folowing the rectifier ... you only need to lose 4
volts to keep the 741 happy

Actually there is another way, that is to simply put a zener in series with
the negative supply of the 741 AND its output,
you only need to lose 4 volts, I didnt notice r4/r5 forms a voltage divider
before so you can almost get away with it just in the negative supply (maybe
just reduce R5),
adding a 100R resistor from the base of the output transistors down to the
negative rail will ensure the output transistors are off with subsequent
higher op amp op voltage.

Colin =^.^=

 

[Donkey D...]

lose

Why can't a Zener be used to limit the (positive) supply to the op-amp, the
reference AND the feedback voltage divider? (The R4/R5 divider would also
need attention.)

.... Donkey

 

[colin]

Why can't a Zener be used to limit the (positive) supply to the op-amp, the
reference AND the feedback voltage divider? (The R4/R5 divider would also
need attention.)

zeners on both the inputs would make the regulation very poor.
resistor dividers making a diff amp would be more accurate.
its simpler to put zeners on the neg rail and op amp output where their poor
regulation doesnt affect the regulation.

Colin =^.^=

 

[Donkey D...]

zeners on both the inputs would make the regulation very poor.
resistor dividers making a diff amp would be more accurate.
its simpler to put zeners on the neg rail and op amp output where their poor
regulation doesnt affect the regulation.

Colin =^.^=

Of course - I'm a bit slow sometimes - must be old age.

.... Donkey

 

[James Waldby]

I am hoping to get an adjustable voltage range from 11v to 20v or so ....
I should expand on my application: I will be using this power supply to
charge a bank of capacitors rated at 700,000uF. This will demand A LOT
of current for a very short amount of time. I need them to charge
within a second or two, so if I use a power resistor between the supply
and the cap bank, it will have to be small. I have the transistors on
some pretty beefy heat sinks.

High-current voltage regulation probably is a dumb idea for
your application. Here are two possibly-better approaches:

1. Attach a full-wave-bridge to the secondary of your 26VAC 15A
transformer, and a variac on the primary. Connect the bridge DC
output to your capacitor bank through a sub-ohm resistor. Set
variac to get proper output voltage.

2. Wire one LM338 in constant-current mode* to deliver 5A to
your capacitors until they get to target voltage, then open a
relay or whatever. 1.4 seconds for 10V charge, 2.8 for 20V.
*See p. 12 of http://www.national.com/ds/LM/LM138.pdf which
shows R1 = .24 ohm, 2W

-jiw