[Mogens V.]

Too many of my (guitar) rack devices use simple 78/79 regulators.

I've seen some with a transistor supposed to shunt noise to ground, but
seems more like a phase inverter, depending on accurate resistors to
match the emitter resistance. Seemed to also depend on how many, and
size of, capacitors dropped around the board.
Another one used two high speed swithing transistors.

Would be nice merely replacing the 78/79 and lifting the output leg to
hook in a filter.

Anyone have links to simple add-on filters, so I don't have to invent a
commen wheel?

--
Kind regards,
Mogens V.

[GregS]

In article , wrote:
Too many of my (guitar) rack devices use simple 78/79 regulators.

I've seen some with a transistor supposed to shunt noise to ground, but
seems more like a phase inverter, depending on accurate resistors to
match the emitter resistance. Seemed to also depend on how many, and
size of, capacitors dropped around the board.
Another one used two high speed swithing transistors.

Would be nice merely replacing the 78/79 and lifting the output leg to
hook in a filter.

Anyone have links to simple add-on filters, so I don't have to invent a
commen wheel?

I don't know what your trying to accomplish. Those devices are pretty good,
but a LM317 can be used which is a little better, especially with the
additional capacitor. I don't know what noise spectrum your trying to get rid
of. If your not too interested in a stable DC level, capacitors are very
effective at reducing 60 hz and above noise.

greg

[Geoff]

Mogens V. wrote:
Too many of my (guitar) rack devices use simple 78/79 regulators.


What do you imagine is wrong with a 78/79xx regulator ?


Would be nice merely replacing the 78/79 and lifting the output leg to
hook in a filter.

If you feel it's necessary it's easy.

Anyone have links to simple add-on filters, so I don't have to invent
a commen wheel?

Yeah sure, a 10u electolytixc shunted by a 100n ceraminc capacitor. It
should be already there.


geoff

[isw]

In article ,
"Mogens V." wrote:

Too many of my (guitar) rack devices use simple 78/79 regulators.

I've seen some with a transistor supposed to shunt noise to ground, but
seems more like a phase inverter, depending on accurate resistors to
match the emitter resistance. Seemed to also depend on how many, and
size of, capacitors dropped around the board.
Another one used two high speed swithing transistors.

Would be nice merely replacing the 78/79 and lifting the output leg to
hook in a filter.

Anyone have links to simple add-on filters, so I don't have to invent a
commen wheel?

Those are actually pretty good regulators, and the problem may be
somewhere else. I'd suggest you go back and check out the way the power
circuitry is wired. It's very common for that to get done wrong, and
cause all sorts of hum problems.

No current flows from the transformer during most of the power cycle.
Only when the voltage there is greater than the voltage in the filter
caps will any current flow at all, and when it does flow, the capacitor
looks like a short circuit and the current can be very high. The voltage
drops from those high currents, if not isolated, can couple directly
into the output of the regulator -- and even a perfect regulator would
not help because the hum has cleverly bypassed it entirely.
Specifically, the resistance of even a short wire can be sufficient to
cause problems. Also, there can be large magnetic fields due to the
large currents, and those must be isolated too.

The wiring should be like this (in order):

From the transformer to the diodes; from the diodes to the terminals of
the filter capacitor; from the capacitor terminals to the input and
common of the regulator. Do not connect the regulator common to any
other place than the appropriate terminal of the capacitor --
specifically not to any other place on the common lead, say nearer the
diodes or transformer. To minimize the magnetic field problem, use
twisted pair -- from the transformer to the diodes, and from there to
the terminals of the capacitor.

Take the power from the output and common of the regulator to the
circuitry. Make no ground connections anywhere else. The common pin of
the regulator should be the ONLY place the input common and output
common are connected. It should be a "zero-length" lead.

Isaac

[Mogens V.]

isw wrote:
In article ,
"Mogens V." wrote:


Too many of my (guitar) rack devices use simple 78/79 regulators.

I've seen some with a transistor supposed to shunt noise to ground, but
seems more like a phase inverter, depending on accurate resistors to
match the emitter resistance. Seemed to also depend on how many, and
size of, capacitors dropped around the board.
Another one used two high speed swithing transistors.

Would be nice merely replacing the 78/79 and lifting the output leg to
hook in a filter.

Anyone have links to simple add-on filters, so I don't have to invent a
commen wheel?


Those are actually pretty good regulators, and the problem may be
somewhere else. I'd suggest you go back and check out the way the power
circuitry is wired. It's very common for that to get done wrong, and
cause all sorts of hum problems.

No current flows from the transformer during most of the power cycle.
Only when the voltage there is greater than the voltage in the filter
caps will any current flow at all, and when it does flow, the capacitor
looks like a short circuit and the current can be very high. The voltage
drops from those high currents, if not isolated, can couple directly
into the output of the regulator -- and even a perfect regulator would
not help because the hum has cleverly bypassed it entirely.
Specifically, the resistance of even a short wire can be sufficient to
cause problems. Also, there can be large magnetic fields due to the
large currents, and those must be isolated too.

The wiring should be like this (in order):

From the transformer to the diodes; from the diodes to the terminals of
the filter capacitor; from the capacitor terminals to the input and
common of the regulator. Do not connect the regulator common to any
other place than the appropriate terminal of the capacitor --
specifically not to any other place on the common lead, say nearer the
diodes or transformer. To minimize the magnetic field problem, use
twisted pair -- from the transformer to the diodes, and from there to
the terminals of the capacitor.

Take the power from the output and common of the regulator to the
circuitry. Make no ground connections anywhere else. The common pin of
the regulator should be the ONLY place the input common and output
common are connected. It should be a "zero-length" lead.

Isaac

Of cause you and G & G are right. I'm probably overreacting a Bit.
Might've been a good ide to say a bit more about it.
I'm about repairing an LXP-1 with the two 78/7912's burned.
This and other devices are used mainly for guitars. No problem with the
regulators for this use, but as this is also part of my homestudio, I
thought the 80dB rejection in the 78/79's might be a limitation.

In most of my gear, the regulators are fairly well laid out, like in
your common practice comments above.

I'm about to start a general overhaul of things to improve S/N and tone,
replacing old technology opamps, capacitors et al.. where appropriate,
and thought I might as well take a look at the power supplies as well.

--
Kind regards,
Mogens V.

[Geoff]

Mogens V. wrote:


Of cause you and G & G are right. I'm probably overreacting a Bit.
Might've been a good ide to say a bit more about it.
I'm about repairing an LXP-1 with the two 78/7912's burned.
This and other devices are used mainly for guitars. No problem with
the regulators for this use, but as this is also part of my
homestudio, I thought the 80dB rejection in the 78/79's might be a
limitation.

It's not assuming there is also a smoothing capacitor between the bridge
rectifier and the 3T reg ?!!! One thing worth checking is that there is a
ceramic or tant cap on the 3T output to reduce chance of oscillation.

In most of my gear, the regulators are fairly well laid out, like in
your common practice comments above.

I'm about to start a general overhaul of things to improve S/N and
tone, replacing old technology opamps, capacitors et al.. where
appropriate, and thought I might as well take a look at the power
supplies as well.

If you axctually really have a problem with s/n and tone, loook at getting
good gear in the first place ! But I suspect you are possibly obsessing
unnecessarily, or there is another problem (gain structure ?) in the way
you've got things hooked up.

geoff

[Mogens V.]

Geoff wrote:
Mogens V. wrote:



Of cause you and G & G are right. I'm probably overreacting a Bit.
Might've been a good ide to say a bit more about it.
I'm about repairing an LXP-1 with the two 78/7912's burned.
This and other devices are used mainly for guitars. No problem with
the regulators for this use, but as this is also part of my
homestudio, I thought the 80dB rejection in the 78/79's might be a
limitation.


It's not assuming there is also a smoothing capacitor between the bridge
rectifier and the 3T reg ?!!! One thing worth checking is that there is a
ceramic or tant cap on the 3T output to reduce chance of oscillation.

Oh yes, in the LXP-1 case, there's a 1000µ cap there and ceramics too.
Somewhat the same with other devices.

In most of my gear, the regulators are fairly well laid out, like in
your common practice comments above.

I'm about to start a general overhaul of things to improve S/N and
tone, replacing old technology opamps, capacitors et al.. where
appropriate, and thought I might as well take a look at the power
supplies as well.


If you axctually really have a problem with s/n and tone, loook at getting
good gear in the first place ! But I suspect you are possibly obsessing
unnecessarily, or there is another problem (gain structure ?) in the way
you've got things hooked up.

Obsessiveness likelyhood As said, started looking at designs because
of repairs. Well also, I can't afford (new) outboard gear hooked up
through lightpipe or spdif, so I shop used gear that I run through
analog S/R loops. For this reason, I'm just trying to have as good a
signal chain as possible.

--
Kind regards,
Mogens V.