build a simple 5v power supply for digital circuit

[robb]

"Jim Thompson" <[email protected]>
wrote in message

Hint: The 7805 forces 5V across R1. Where does the current through
R1 go?

Hi Jim thanks for the reply/hint,

I did not miss your hint, it just seems to have a depth i will
need to figure out.

it seems to be a trick question and i do not want to get it wrong


thanks for help
robb

i will take a wild guess and say the current goes to BR1 (-) and
(Pin 2) ? yes ?

 

[robb]

Yeah, there's an error. The connection between the (-) output and R1
should not be there.

hi, thanks Sphero,

wait a minute are you toying with me ?

well i can not figure which connection you are talking about iis
it the same as the one i described ?

thanks for help,
robb

 

[Phil Allison]

"Fred Abse"

You really think so?

** I know so - you ASD fucked, fuckwit.

Try this:

** Drop dead.


...... Phil

 

[Spehro Pefhany]

hi, thanks Sphero,

wait a minute are you toying with me ?

well i can not figure which connection you are talking about iis
it the same as the one i described ?

thanks for help,
robb

http://server2.hostingplex.com/~zstoretr/fix.gif



Best regards,
Spehro Pefhany

 

[Jim Thompson]

"Jim Thompson" <[email protected]>
wrote in message

Hi Jim thanks for the reply/hint,

I did not miss your hint, it just seems to have a depth i will
need to figure out.

it seems to be a trick question and i do not want to get it wrong


thanks for help
robb

i will take a wild guess and say the current goes to BR1 (-) and
(Pin 2) ? yes ?

Once you apply Spehro's correction eliminating a short, the current
flows through the selected resistors below R1... R4 or R5 or R6 or R7
or R8 or R9 or pot R2

...Jim Thompson

 

[robb]

http://server2.hostingplex.com/~zstoretr/fix.gif

not the wire i was thinking, but i was close
thanks for the correction
robb

 

[robb]

Some 78xx regulators oscillate a bit especially with out
the proper load on them.
even then, you may need to place some by pass caps on it
as you did.

so the oscillation that probably contributed to frying my first
7805
and is it typical that the oscillation shows up on the bridge out
?

thanks for reply,
robb

 

[Jamie]

so the oscillation that probably contributed to frying my first
7805
and is it typical that the oscillation shows up on the bridge out
?

thanks for reply,
robb

board lay and lead run lengths plays a roll in that.

I assume you had a electrolytic cap on the bridge +/- outputs?

in any case, that does not stop the high freq you saw in there, you
don't always see this problem in all 78xx, I've seen it in some
depending on who made the component where is others, they work fine
with out any compensation. Also, if you constructed the bridge your
self instead of using a ready made unit, it's smart to put some small
bypass caps across the diodes..

Other issues is when voltages on the output side exceed the voltage
on the input side.. the internals of the regulator have an issue with
this how ever, like I said before, it depends on who made the component.
I've seen some that included the bleeder diode.
In a case like that, a bleeder diode from the output to the
input is a desirable component to use. For a + type, the anode would
be on the output, cathode on the input. Basically, that simply causes
a by pass of the regulator if the output voltage happens to be higher
than the input voltage by 0.7V or more.


Hope that did something for you.

 

[ehsjr]

(images on a.b.s.e)

i built and experimenting with the 5v DC psu

The images do not show a filter cap after the bridge and
before the 7805. Add a large cap (4700uF shown) across
the brideg output:

--------
| +|-----+------in[7805]out---
| Bridge | | + |
| | [4700uF] |
| | | |
| -|-----+-----------+

 

[T]

Yeah, there's an error. The connection between the (-) output and R1
should not be there.


Best regards,
Spehro Pefhany

I noticed that when I looked at the schematic. I'm planning on building
this one since it's a variable power supply, hell I might just build two
of them because I've got a couple products in mind that use 5V for logic
and 12V to drive things like steppers, etc.

 

[T]

hi, thanks Sphero,

wait a minute are you toying with me ?

well i can not figure which connection you are talking about iis
it the same as the one i described ?

thanks for help,
robb

If you look at the schematic the connection between the lower end of R1
and the other end of C3 should not exist.

 

[T]

http://server2.hostingplex.com/~zstoretr/fix.gif



Best regards,
Spehro Pefhany

Nice graphic! Thanks for that, I used a plain black pen to cross out the
connection on my schematic.

 

[pom]

Jim Thompson a écrit :
///Hello
nobody mentioned it but the schematic is faulty!
Negative output should not be connected to the "mass" leg of the
regulator. Else, output voltage cannot be different from 5V - other
effects neglected
pom

 

[Phil Allison]

Hello
nobody mentioned it but the schematic is faulty!
Negative output should not be connected to the "mass" leg of the
regulator. Else, output voltage cannot be different from 5V - other
effects neglected

http://www.qsl.net/lu3mgp/electronica/UniversalPowerSupply.pdf

** The design is purest crap.

Output regulation wil be poor, as will hum rejection - since Iq is a highly
variable quantity.

Bound to be HF unstable at some settings and loads too.

That ain't no way to treat a 7805.

Just sub in a LM317 and nearly all the troubles go.



.......... Phil

 

[Spehro Pefhany]

http://www.qsl.net/lu3mgp/electronica/UniversalPowerSupply.pdf

** The design is purest crap.

Output regulation wil be poor, as will hum rejection - since Iq is a highly
variable quantity.

It's actually fairly stable with output current, less so with input
voltage and temperature. Probably better than most switchable AC
adapters wot consist primarily of a badly made multi-tap transformer
and a crappy slide switch with PCB contacts. But it could be <1%
rather than <10%, as you point out.

Bound to be HF unstable at some settings and loads too.

I have yet to see a 7805 oscillate under any reasonable conditions.

That ain't no way to treat a 7805.

It should probably have a larger output cap and a diode eg. 1N5403
across the 7805 and probably across the output too since we don't know
what will get hooked up to it.

Just sub in a LM317 and nearly all the troubles go.



......... Phil

Yup. Of course you have to recalculate the resistor values, taking
into account that you should draw a minimum of 5mA from the '317 to
keep it in regulation. Eg. 237R and 715R for 5V.

Another potential problem with the original circuit is that the output
voltage will spike as high as Vin - 1.xV during the switch transitions
unless it's a shorting type rotary switch.

That's typically considered an undesirable characteristic in a power
supply.



Best regards,
Spehro Pefhany

 

[Phil Allison]

"Spehro Pefhany"
"Phil Allison"

It's actually fairly stable with output current, less so with input
voltage and temperature.

** Natsemi data shows Iq may change by up to 0.5 mA in the range of normal
output currents up to 1 amp,
by up to 1mA over the nominal input voltage range & by 1 mA over the
allowed temp range.

Adds up to 2.5 mA.

2.5 mA flowing in 1.5 kohms = 3.75 volts.

That IS craaaaap !!

I have yet to see a 7805 oscillate under any reasonable conditions.

** A comment about you and other circuits.

Waste of type.

It should probably have a larger output cap....

** There is effectively no output stability cap.

That 0.01uF hanging off the + output finds it way back to the 7805's ground
terminal via a crippling amount of series resistance.

Yup.

** Just follow the data sheet for the LM317 for the needed R values.


The only way that horribly crude idea of converting a LM7805 into a variable
reg can be even half satisfactory is * IF * the input voltage is itself
regulated and the load current fixed.



....... Phil

 

[Spehro Pefhany]

The only way that horribly crude idea of converting a LM7805 into a variable
reg can be even half satisfactory is * IF * the input voltage is itself
regulated and the load current fixed.

It's more than okay to put a zener or LED on the GND lead.

It's also okay if you don't mind blowing all kinds of current in the
resistor network (the Fairchild data sheet recommends > 25mA rather
than the 5mA the Poptronics circuit has), which would reduce the
effect of Iq variations by about 3:1.

They also show an op-amp, but that has potential stability issues that
might actually exist in the real world.


Best regards,
Spehro Pefhany

 

[Phil Allison]

"Spehro Pefhany"
"Phil Allison"

It's more than okay to put a zener or LED on the GND lead.

** Completely different matter.

Remaining within context is a real trial for you.

It's also okay if you don't mind blowing all kinds of current in the
resistor network (the Fairchild data sheet recommends > 25mA rather
than the 5mA the Poptronics circuit has),

** The Poptronics abortion runs at circa 11mA.

5 / 470 = ???



.......... Phil

 

[Fred Bloggs]

Yeah, there's an error. The connection between the (-) output and R1
should not be there.


Best regards,
Spehro Pefhany

And the circuit description is demented, the output topology is not a
'voltage divider', he's putting some ridiculous amount of current
through that LED, the whole thing is bad. No wonder that generation of
kids just threw their hands up and quit electronics.

 

[Fred Bloggs]

It's more than okay to put a zener or LED on the GND lead.

It's also okay if you don't mind blowing all kinds of current in the
resistor network (the Fairchild data sheet recommends > 25mA rather
than the 5mA the Poptronics circuit has), which would reduce the
effect of Iq variations by about 3:1.

They also show an op-amp, but that has potential stability issues that
might actually exist in the real world.


Best regards,
Spehro Pefhany

The pos is intractable as shown. Note that S2 is more likely a
break-before-make, opening the 7805 return for even a few hundred usecs
is not a good idea, output heads for the upper limits, then filtering
the output with a reasonable value of capacitance is also problematic
with that dammed switch too.