David Janszen just emailed me again and asked that I post some revised notes with more WARNINGS than in his prior message. He also included a pdf with a reprint of the rebuilding procedure from a past magazine article (email if you want a copy). Here are his notes:
******************
The final electrical repair is pretty easy, but the prep is not, and the
work is hazardous.
*** WARNINGS ***
WORK ON KLH NINE SUPPLIES IS FOR ELECTRONIC REPAIR PROFESSIONALS, NOT
DIY'ERS. IF YOU ARE NOT SUCH A PROFESSIONAL, DO NOT ATTEMPT THE REPAIR; IT
IS TOO HAZARDOUS. The following section is to remind professionals of the
hazards.
The stepped up AC line voltage is 1000 VAC RMS and has plenty of current for
causing pain, burns and electrocution. The stepped up audio signal is even
more dangerous, and can reach 7000 VAC RMS at full power. THIS GOES RIGHT
THROUGH THE INSULATION ON CLIP LEADS.
RULE #1: MAKE SURE YOUR BODY IS WELL INSULATED FROM EARTH GROUND.
RULE #2: NEVER USE TWO HANDS ON AN ACTIVE SUPPLY.
NEVER TOUCH ANY TWO PARTS, INCLUDING THE CASE, WITH TWO
HANDS.
NEVER TOUCH ANY ITEMS THAT ARE CONNECTED TO ANY TWO PARTS
WITH TWO HANDS.
If your body is well insulated from ground, this should keep current from
flowing through your heart and stopping it.
RULE #3: NEVER WORK ALONE ON AN ACTIVE SUPPLY. If you make a mistake,
your muscles will probably cramp due to the AC current. This makes your grip
on the electrified items tighten, so you can not let go. In this case, you
will need someone to unplug the supply or knock you down and away from the
wires before you die, perform CPR if they were too slow, and in either
event, get you to the ER for medical attention afterward.
RULE #4: NEVER FEEL COMFORTABLE WORKING WITH THESE VOLTAGES.
RULE #5: IF THERE IS ANY RULE THAT YOU CAN NOT FOLLOW, OR IF YOU ARE
THE LEAST BIT ACCIDENT PRONE, REFER THE WORK TO SOMEONE ELSE.
RULE #6: IF YOU ARE MARRIED, SHOW THIS TO YOUR SPOUSE BEFORE
CONTINUING.
WARNING: Many materials that you may think are insulators are not effective
at these voltages, including wood, rubber, vinyl, paper, most kinds of tape,
etc.
Also, don't let clip leads, no matter how well they may seem to be
insulated, touch each other (or you, of course) during testing.
Lastly, the capacitors in the voltage multipler ladder stay charged with DC
for hours. If the supply has been powered up recently, short the string out
with a clip lead before working on it. The residual charge is not very
dangerous, but discharging it through your personal body tissues hurts, is
very annoying, and can have side effects, such as a strong arm recoil.
During such recoil, a flying elbow can be injured, or can injure a second
person who is nearby.
***
Everything is held in place by the wax potting compound.
To start, make yourself a fixture from banana plugs and some phenolic or
FR-4 that you can use to position the connector strip back where it belongs
after you've made the repair and are ready to pour the wax back in.
Drill one or two 1/4" holes in one corner of the supply for draining the
wax, just through the sheet metal and no deeper. Preheat your oven to 300°F,
place the supply in the oven over a container to catch the wax, and wait.
Use a large metal can, like a coffee can, or a large glass container. Tip
the supply so the wax runs straight out and not along its lower face. Scoop
some out when it first starts to melt, so its expansion does not cause it to
overflow the supply can.
Once the wax is out, you can replace the diodes. If your speaker has been at
least partially working, but not well, everything else will probably be
okay.* Wire three 1N4007 in series as a replacement for each diode. Match
the polarity of every diode to the original.
Check the voltages. The woofer bias should be about -5700VDC and the tweeter
bias -1100VDC relative to common. You need an electrostatic meter for this;
any other type will load down the supply and you will have no idea by what
amount. These will probably be about 10% to 15% higher with the silicon
diodes than they were with the selenium diodes. This is too high and will
cause problems. Experiment with adding series resistance to one leg of the
line AC between the AC step up transformer from the AC input jack. I think a
2W resistor of about 5kΩ to 7KΩ is right. If you don't have a meter, just
add a resistor and test using the panel itself, connected with clip leads,
increasing resistance until the frying sound goes away, then increasing it
by another 10% or so.
Melt the original wax. Drive a 5/16" x 1/2" self-tapping screw
(http://www.mcmaster.com/#cad-2d/=6udcvq) into the hole(s) you had drilled,
perhaps first adding a dab of silicone to the area where the screw head
seats. Pour back the wax in two or three stages/layers, with the thinnest
one last, allowing it to cool fully each time, so it all fits and is smooth
when it cools. After cooling, the screws can be removed or left in place, to
taste.
* If the speaker has been producing no sound at all, and the fuses were
okay, then it is worth checking the transformers. Failure of either
transformer under normal operating conditions is very rare, but if it occurs
will lead to a quest for parts that are probably unobtainable, or a redesign
that uses easily obtainable parts. I know of one person who has replaced the
entire supply with a new circuit that he says works fine. I would say that
this could be an improvemenmt if done right, because it could be made to
present a low impedance load to solid state amplifiers. The original signal
circuit presents a 16 Ohm load at its amplifier terminals, which has solid
state amps operating far below their power capabilities.
To connect your meter to any lugs in the supply, you will have to either get
through the wax with the jaws of a clip lead, use a very sharp probe, or
clean the wax off. A Q-Tip dampened with solvent is useful for cleaning.
Xylene works fast on wax. Kerosene (a.k.a., lighter fluid, charcoal starter,
or mineral spirits) also works well, and evaporates more slowly.
The AC line step up transformer (the small one) has a 1000V output, and it
can supply plenty of current for measuring this voltage using a regular
voltmeter. It also has plenty of current for causing pain, burns and
electrocution. If this transformer is okay, check the signal step up
transformer windings for continuity.
Note: If the wax is uneven with areas that are bubbled up or has sinking
areas or is missing part of its wax, this means it has been abused, and the
likelihood of transformer failure is pretty high, as is woofer membrane
damage. This abuse is generally accomplished by putting foil over the signal
fuses and using a huge amplifier. Repair is quite a project in that case.
*********************
I found the article I mentioned, converted it to PDF format, and attached
it. There are a few things that are incorrect or that I do not agree with. I
will eventually post this to the JansZen web site with the following
corrections:
A 1/4" drain hole is big enough and easier to plug than a 1/2" hole.
Don't bother carefully cleaning off the multiplier board. Hit it with a hair
dryer and forget about whatever wax does not drip off. Any wax remaining on
the lugs will melt away when desoldering and soldering. Also, being bee's
wax, it smells nice when it gets hot.
In each diode replacement ("daisy chain") only needs three, not five diodes
in series.
Omitting a stage from the multiplier is a bad idea, because it will not
reduce the tweeter bias voltage. Series resistance on the AC line will
reduce both woofer and tweeter voltages proportionately.
The bias voltage is negative, not positive.
Alcohol does not dissolve wax very well.
The warnings about the voltage hazards are inadequate.
David
