[Ian Bell[_2_]]

I am building a two channel mic pre with a 48V phantom power supplying
each input. W|orst case load on each is a pair of paralleled 6K8
resistors to earth or 3K4 or about 14mA, so say a max of 30mA for the
pair which at 48V amounts to a mere 1.44 watts..

The regulator consists of a 7824 with a 24V zener up its ground leg
which is convenient as the regulator ground current is 5 to 8 mA which
is just right for the zener. I have a 47K across the smoothing cap and
another across the output to discharge things when power is turned off.
Total current draw worst case less than 40mA or under 2W.

Now the transformer I am using is a little 7VA toroidal with a pair of
22V windings in series i.e.peak output (ignoring diode drops) is 44 root
2 = 62V. The 7824 can have 40V input to ground which means with the 24V
zener up it ground I can have 64V at the input from the smoothing cap.
Close but there's no drops or transformer regulation allowed for so it
should be OK


Anyway, I built it and the problem is with just 10mA load, the smoothing
cap has 74V across it!

The phantom power on/off switching is arranged so when it is off another
pair of 6K8s are connected to ground so the load does not vary too much,
but even so I cannot rely on there being sufficient dc load to ensure
the 74V across the smoothing cap is brought down enough.

So I am concluding the best solution is to slug the ac secondary with
about 5VA or so. What do you think?

Cheers

Ian

[Phil Allison[_3_]]

"Ian Bell is a DOPE"


The regulator consists of a 7824 with a 24V zener up its ground leg ..


** Change it to a TL783.

Which is what nearly everyone is using for 48 volt phantom.

http://uk.farnell.com/texas-instrume...h-v/dp/1703396



..... Phil

[Ian Bell[_2_]]

flipper wrote:
On Tue, 27 Oct 2009 21:52:10 +0000, Ian Bell
wrote:

I am building a two channel mic pre with a 48V phantom power supplying
each input. W|orst case load on each is a pair of paralleled 6K8
resistors to earth or 3K4 or about 14mA, so say a max of 30mA for the
pair which at 48V amounts to a mere 1.44 watts..

The regulator consists of a 7824 with a 24V zener up its ground leg
which is convenient as the regulator ground current is 5 to 8 mA which
is just right for the zener. I have a 47K across the smoothing cap and
another across the output to discharge things when power is turned off.
Total current draw worst case less than 40mA or under 2W.

Now the transformer I am using is a little 7VA toroidal with a pair of
22V windings in series i.e.peak output (ignoring diode drops) is 44 root
2 = 62V. The 7824 can have 40V input to ground which means with the 24V
zener up it ground I can have 64V at the input from the smoothing cap.
Close but there's no drops or transformer regulation allowed for so it
should be OK


Anyway, I built it and the problem is with just 10mA load, the smoothing
cap has 74V across it!

That seems a tad high but small VA transformers often have rather
large regulation values so it's not unusual.

Out of curiosity, why did you pick such a large secondary voltage?
Dual 15V would have been more than enough.


Er, 30V * root 2 is only 42

Even 2 * 18 * root 2 is only 51


Cheers

Ian


The phantom power on/off switching is arranged so when it is off another
pair of 6K8s are connected to ground so the load does not vary too much,
but even so I cannot rely on there being sufficient dc load to ensure
the 74V across the smoothing cap is brought down enough.

So I am concluding the best solution is to slug the ac secondary with
about 5VA or so. What do you think?

Seems a shame to burn power just to make up for a hot winding but if
you're going to do that then why not a shunt regulator? That way at
least the burn off power is actually available for use.


Cheers

Ian

[Ian Iveson]

Ian Bell wrote:

I am building a two channel mic pre with a 48V phantom
power supplying each input. W|orst case load on each is a
pair of paralleled 6K8 resistors to earth or 3K4 or about
14mA, so say a max of 30mA for the pair which at 48V
amounts to a mere 1.44 watts..

The regulator consists of a 7824 with a 24V zener up its
ground leg which is convenient as the regulator ground
current is 5 to 8 mA which is just right for the zener. I
have a 47K across the smoothing cap and another across the
output to discharge things when power is turned off. Total
current draw worst case less than 40mA or under 2W.

Are we getting near the point yet?

Now the transformer I am using is a little 7VA toroidal
with a pair of 22V windings in series i.e.peak output
(ignoring diode drops) is 44 root 2 = 62V. The 7824 can
have 40V input to ground which means with the 24V zener up
it ground I can have 64V at the input from the smoothing
cap. Close but there's no drops or transformer regulation
allowed for so it should be OK

Oh dear. Why should it be OK when you haven't allowed for
the regulation? If the secondaries are specified as 22V,
then that would be at full load. Hence at less load, more
volts. Transformers are quite cheap, so the best solution is
to get one with the correct output.

Anyway, I built it and the problem is with just 10mA load,
the smoothing cap has 74V across it!

Even taking recent experiences into account, I'm surprised
you couldn't work that out before you built it.

The phantom power on/off switching is arranged so when it
is off another pair of 6K8s are connected to ground so the
load does not vary too much, but even so I cannot rely on
there being sufficient dc load to ensure the 74V across
the smoothing cap is brought down enough.

So I am concluding the best solution is to slug the ac
secondary with about 5VA or so. What do you think?

Someday when it needs fixing, some tech is going to discover
you're a hopeless bodger.

I can't imagine that such a demonstration of incompetence
would impress a safety inspector. Have you got CE
certification for the stuff you sell?

Perhaps a course in electronics would be the best way
forward.

I sincerely hope I haven't been of any help. I guess Phil
just couldn't stop himself.

Ian

[Ian Bell[_2_]]

Phil Allison wrote:
"Ian Bell is a DOPE"

The regulator consists of a 7824 with a 24V zener up its ground leg ..


** Change it to a TL783.

Which is what nearly everyone is using for 48 volt phantom.

http://uk.farnell.com/texas-instrume...h-v/dp/1703396



.... Phil




Thanks Phil. Good call.

Cheers

Ian

[Phil Allison[_3_]]

"Ian Iveson"
"Bell the Bodger"

So I am concluding the best solution is to slug the ac secondary with
about 5VA or so. What do you think?

Someday when it needs fixing, some tech is going to discover you're a
hopeless bodger.

I can't imagine that such a demonstration of incompetence would impress a
safety inspector. Have you got CE certification for the stuff you sell?

Perhaps a course in electronics would be the best way forward.

I sincerely hope I haven't been of any help. I guess Phil just couldn't
stop himself.


** I'm just to darn forgiving for my own good.

Did you see all Mr Bell's awesome qualifications on his web page?

Must have had a massive dose of amnesia since then.



..... Phil

[Alex]

"Ian Bell" wrote in message
...
I am building a two channel mic pre with a 48V phantom power supplying
each input. W|orst case load on each is a pair of paralleled 6K8
resistors to earth or 3K4 or about 14mA, so say a max of 30mA for the
pair which at 48V amounts to a mere 1.44 watts..

The regulator consists of a 7824 with a 24V zener up its ground leg
which is convenient as the regulator ground current is 5 to 8 mA which
is just right for the zener. I have a 47K across the smoothing cap and
another across the output to discharge things when power is turned off.
Total current draw worst case less than 40mA or under 2W.

Now the transformer I am using is a little 7VA toroidal with a pair of
22V windings in series i.e.peak output (ignoring diode drops) is 44 root
2 = 62V. The 7824 can have 40V input to ground which means with the 24V
zener up it ground I can have 64V at the input from the smoothing cap.
Close but there's no drops or transformer regulation allowed for so it
should be OK


Anyway, I built it and the problem is with just 10mA load, the smoothing
cap has 74V across it!

The phantom power on/off switching is arranged so when it is off another
pair of 6K8s are connected to ground so the load does not vary too much,
but even so I cannot rely on there being sufficient dc load to ensure
the 74V across the smoothing cap is brought down enough.

So I am concluding the best solution is to slug the ac secondary with
about 5VA or so. What do you think?

Cheers

Ian

You can drop excessive voltage by say a 10...12V Zener (appropriately power
rated) in *series* with the 7824's input. It is better than a dropping
resistor, because the voltage drop will be load current independent.

Regards,
Alex

[Phil Allison[_3_]]

"Alex"

You can drop excessive voltage by say a 10...12V Zener (appropriately
power
rated) in *series* with the 7824's input. It is better than a dropping
resistor, because the voltage drop will be load current independent.


** Bell the Bodger did not suggest a series resistor - but rather a
parallel one to load the unregulated DC output voltage down to the value he
considered safe for the LM7824 plus 24 volt zener.

FYI - possibly the best way to use an 78xx ( or 79xx) series reg IC is to
have a pre-regulator in from of one. The pre-reg may be a TO3 transistor
( ie a 2N3055) with a simple zener supply feeding the base - the output
voltage of which need only be 3 volts above the IC regulator's voltage.

Benefits include getting the best possible output voltage stability from the
IC, zero AC hum and minimum heat dissipation ensuring that full rated
current is available, long term.

The pre-reg transistor dissipates most of the heat and is also protected by
the current limiting built into the IC regulator.


..... Phil

[Alex]

"Phil Allison" wrote in message
...

"Alex"

You can drop excessive voltage by say a 10...12V Zener (appropriately
power
rated) in *series* with the 7824's input. It is better than a dropping
resistor, because the voltage drop will be load current independent.


** Bell the Bodger did not suggest a series resistor - but rather a
parallel one to load the unregulated DC output voltage down to the value
he
considered safe for the LM7824 plus 24 volt zener.

Alex:
I understand that he suggested a parallel resistor. This would be the most
ridiculous solution -- to waste power, generate heat, increase ripple, etc.
That is why I suggested a series dropping means, but a Zener, not a reistor.


FYI - possibly the best way to use an 78xx ( or 79xx) series reg IC is to
have a pre-regulator in from of one. The pre-reg may be a TO3 transistor
( ie a 2N3055) with a simple zener supply feeding the base - the output
voltage of which need only be 3 volts above the IC regulator's voltage.
Benefits include getting the best possible output voltage stability from
the
IC, zero AC hum and minimum heat dissipation ensuring that full rated
current is available, long term.

The pre-reg transistor dissipates most of the heat and is also protected
by
the current limiting built into the IC regulator.

Alex:
Yes, this topology with a preregulator is the best, but complex -- needs
another transistor, maybe heathink... Using a suitable Darlington instead of
2N3055 would be better -- a very small current through 51...68V Zener will
be sufficient.


.... Phil