Hello. I have a homebrew linear 13.8V 20A power supply I have been using
for some years to power an IC-735, a 2M rig and some accessories. The PS
has these characteristics:

- well filtered (35000uF)22AMP, 21 volt DC output from diode bridge circuit
- LM-317 regulator circuit (as per LM-317 datasheets)
- MJ2955 NPN feeder transistor from regulator driving 4 - 2N3055 PNP power
transistors in parallel
- large heat sinks (I don't think heat is a problem)
- good air circulation with fan

I have the regulator set to 13.7 VDC out. Lately however, I have had some
problems with voltage drops: while drawing 20 amps the voltage drops to
about 11.3 volts and I have had reports of a rough CW signal as a result.

I have tested the reg circuit and even replaced the LM-317 to no avail. I
cannot see where the problem lies, unless perhaps the MJ2955 feeder
transistor is not working as it should? (I don't yet have a replacement
transistor to test that idea, hence this post).

Any ideas greatly appreciated.

Larry VE7EA

--
For direct email remove fake.

Larry Gagnon wrote:
Hello. I have a homebrew linear 13.8V 20A power supply I have been using
for some years to power an IC-735, a 2M rig and some accessories. The PS
has these characteristics:

- well filtered (35000uF)22AMP, 21 volt DC output from diode bridge circuit
- LM-317 regulator circuit (as per LM-317 datasheets)
- MJ2955 NPN feeder transistor from regulator driving 4 - 2N3055 PNP
power transistors in parallel
- large heat sinks (I don't think heat is a problem)
- good air circulation with fan

I have the regulator set to 13.7 VDC out. Lately however, I have had
some problems with voltage drops: while drawing 20 amps the voltage
drops to about 11.3 volts and I have had reports of a rough CW signal as
a result.

I have tested the reg circuit and even replaced the LM-317 to no avail.
I cannot see where the problem lies, unless perhaps the MJ2955 feeder
transistor is not working as it should? (I don't yet have a replacement
transistor to test that idea, hence this post).

Any ideas greatly appreciated.

Larry VE7EA

It's probably not dropping to a constant 11.3V -- it's probably dropping
out on the valleys of the rectifier's output. A steady low voltage
would either render your transceiver inoperable or lower your output
power; the "rough CW" is probably modulated at 120Hz.

Most likely that "well filtered (35000uF)" has turned to "lots of hum
(dried up electrolytics)".

I assume you don't have an oscilloscope. If you do, clip it onto the
power supply output and take a look while you run it close to it's
limit. Any filtering problems should become immediately apparent.

If you don't have fancy equipment just measure the output with your
voltmeter switched to "AC" -- most VOMs will filter out the DC and just
let you see the wiggly part. Do this under both heavy load and light --
if there's a honking big AC component as the voltage sags then there
needs to be some new capacitors in your future.

--

Tim Wescott
Wescott Design Services
http://www.wescottdesign.com

Do you need to implement control loops in software?
"Applied Control Theory for Embedded Systems" gives you just what it says.
See details at http://www.wescottdesign.com/actfes/actfes.html

"Larry Gagnon" wrote in message
newsp.uj37o6zpaiss39@localhost...
Hello. I have a homebrew linear 13.8V 20A power supply I have been using
for some years to power an IC-735, a 2M rig and some accessories. The PS
has these characteristics:

- well filtered (35000uF)22AMP, 21 volt DC output from diode bridge
circuit
- LM-317 regulator circuit (as per LM-317 datasheets)
- MJ2955 NPN feeder transistor from regulator driving 4 - 2N3055 PNP power
transistors in parallel
- large heat sinks (I don't think heat is a problem)
- good air circulation with fan

I have the regulator set to 13.7 VDC out. Lately however, I have had some
problems with voltage drops: while drawing 20 amps the voltage drops to
about 11.3 volts and I have had reports of a rough CW signal as a result.

I have tested the reg circuit and even replaced the LM-317 to no avail. I
cannot see where the problem lies, unless perhaps the MJ2955 feeder
transistor is not working as it should? (I don't yet have a replacement
transistor to test that idea, hence this post).

Any ideas greatly appreciated.


In addition to Tim's suggestion, If a rectifer diode opened, you might also
experience a similar problem. With a scope, it'll have 60 Hz ripple instead
of 120 Hz ripple when loaded.

-NM

Howdy,


A few of random suggestions...

The MJ2955 is a PNP and the 2N3055 are NPN.
The pass devices are NPN that make an emitter
follower output which provides current gain.

Load the supply until the voltage drops.
Then measure the base voltage at each 2N3055.
The bases should all be the same, a base to
emitter voltage drop higher than the output
voltage. One of the pass devices may have
failed or the drive current could be insufficient.

I'm guessing without a schematic but the MJ2955
is probably boosting the LM317's output and may
also be part of an over current foldback
protection circuit. Check the collector voltage
it should be nearly the same as the 2N3055 base
voltage. If it's much higher that points to a
problem with the pass devices. If it's not then
there's not enough loop gain.

You may find a high power low value resistor near
the MJ2955 for current limiting, check the value
of that too. If it's damaged or drifted high
you could be hitting the current fold back
protection.

The 21V bulk supply will sag if there is a bad
rectifier so a check of that under load is a good
test. If it's much below 18V that could be too
low. The LM317 needs about 3V head room and the
2N3055 will need a Vbe drop.

I think if the bulk cap had dried up you'd hear
hum even at lower load current.

You can trouble shoot this with a voltmeter,
stage-by-stage, once you load it until it's
exhibiting the problem. Once you identify the
failing stage a careful visual inspection followed
by some resistance and junction checks should
solve the puzzle.


73,
Grumpy



"Larry Gagnon" wrote in
newsp.uj37o6zpaiss39@localhost:

Hello. I have a homebrew linear 13.8V 20A power supply I have been
using for some years to power an IC-735, a 2M rig and some
accessories. The PS has these characteristics:

- well filtered (35000uF)22AMP, 21 volt DC output from diode bridge
circuit - LM-317 regulator circuit (as per LM-317 datasheets)
- MJ2955 NPN feeder transistor from regulator driving 4 - 2N3055 PNP
power transistors in parallel
- large heat sinks (I don't think heat is a problem)
- good air circulation with fan

I have the regulator set to 13.7 VDC out. Lately however, I have had
some problems with voltage drops: while drawing 20 amps the voltage
drops to about 11.3 volts and I have had reports of a rough CW signal
as a result.

I have tested the reg circuit and even replaced the LM-317 to no
avail. I cannot see where the problem lies, unless perhaps the MJ2955
feeder transistor is not working as it should? (I don't yet have a
replacement transistor to test that idea, hence this post).

Any ideas greatly appreciated.

Larry VE7EA

Oh!

I forgot to say if the 2N3055's don't have individual
base drive resistors, if their bases are all wired
together, then you'd have to pull them and check the
base-to-emitter and base-to-collector junctions with
the VOM to know if one had failed. I've seen a smaller
supply built like that and one of two 2N3055 had failed.
It made 8A instead of 12A output before going out of
regulation.

If there are individual base resistors the 2N3055 with
the higher base voltage, because there's no current drawn
though its drive resistor, is the bad one.


73,
Grumpy


Grumpy The Mule wrote in
:

Load the supply until the voltage drops.
Then measure the base voltage at each 2N3055.
The bases should all be the same, a base to
emitter voltage drop higher than the output
voltage. One of the pass devices may have
failed or the drive current could be insufficient.

Thanks everyone above for all your input and suggestions. Very useful. I
will get into it and do a few more tests.

hey larry

I see trouble with the bypass caps on the LM317 regulator. They have
opened up and now the adjust pin, which is an op amp input, isn't
bypassed, letting in massive quantities of rf and overloading the op
amps ability to compare the reference voltage to the adjust pin
voltage.

73 OM
n8zu



On Nov 4, 5:06 pm, "Larry Gagnon" wrote:
Hello. I have a homebrew linear 13.8V 20A power supply I have been using
for some years to power an IC-735, a 2M rig and some accessories. The PS
has these characteristics:

- well filtered (35000uF)22AMP, 21 volt DC output from diode bridge circuit
- LM-317 regulator circuit (as per LM-317 datasheets)
- MJ2955 NPN feeder transistor from regulator driving 4 - 2N3055 PNP power
transistors in parallel
- large heat sinks (I don't think heat is a problem)
- good air circulation with fan

I have the regulator set to 13.7 VDC out. Lately however, I have had some
problems with voltage drops: while drawing 20 amps the voltage drops to
about 11.3 volts and I have had reports of a rough CW signal as a result.

I have tested the reg circuit and even replaced the LM-317 to no avail. I
cannot see where the problem lies, unless perhaps the MJ2955 feeder
transistor is not working as it should? (I don't yet have a replacement
transistor to test that idea, hence this post).

Any ideas greatly appreciated.

Larry VE7EA

--
For direct email remove fake.

On Nov 4, 5:06*pm, "Larry Gagnon" wrote:
Hello. I have a homebrew linear 13.8V 20A power supply I have been using *
for some years to power an IC-735, a 2M rig and some accessories. The PS *
has these characteristics:

- well filtered (35000uF)22AMP, 21 volt DC output from diode bridge circuit
- LM-317 regulator circuit (as per LM-317 datasheets)
- MJ2955 NPN feeder transistor from regulator driving 4 - 2N3055 PNP power *
transistors in parallel
- large heat sinks (I don't think heat is a problem)
- good air circulation with fan

I have the regulator set to 13.7 VDC out. Lately however, I have had some *
problems with voltage drops: while drawing 20 amps the voltage drops to *
about 11.3 volts and I have had reports of a rough CW signal as a result.

I have tested the reg circuit and even replaced the LM-317 to no avail. I *
cannot see where the problem lies, unless perhaps the MJ2955 feeder *
transistor is not working as it should? (I don't yet have a replacement *
transistor to test that idea, hence this post).

Any ideas greatly appreciated.

Larry VE7EA

--
For direct email remove fake.

Sounds like my power supply. One of the emitter resistors on a pass
transistor opened. The transistor did not fail. I replaced the
resistors and decided to upgrade the 2n3055 to some NTE181s. Also
modified the power supply for remote sense to allow for IR losses in
the power cable.

Larry Gagnon wrote:
Thanks everyone above for all your input and suggestions. Very useful. I
will get into it and do a few more tests.

Let us know how it works out.

--

Tim Wescott
Wescott Design Services
http://www.wescottdesign.com

Do you need to implement control loops in software?
"Applied Control Theory for Embedded Systems" gives you just what it says.
See details at http://www.wescottdesign.com/actfes/actfes.html