Re: Dell 2408WFPb / 2408WFP - 7 seconds to black
Folks, I am happy to report that my 2408WFPb is once again alive and well.
Turned out it had two problems at the same time.
1. I've built a makeshift inverter tester out of a small 5w fluorescent tube and
a couple of wires with alligator clips. All 7 inverters had fired up this tube
successfully but in all 7 cases I would still get the dreaded 2 seconds to
black. However while performing this test I noticed a faint arcing under one of
the output transformers. Per Davi.p's advice above I measured the resistances of
the secondary windings of all 14 transformers. Thirteen of them showed ~1.650k
but one transformer showed 440k. I desoldered it from the board and looked at it
under the microscope. A small portion (<1mm) of the solid Cu wire that makes up
the secondary winding had burnt off the pin it was wrapped around and soldered
to pretty much the way a glass fuse blows leaving a tiny droplet of metal on the
wire end. I can only explain the 440k (i.e. the non-infinite resistance) of this
broken circuit by the presence of the carbon deposit left by the evaporated Cu
wire which provided at least some path for the multimeter current. Luckily I had
a couple of brand new transformers on hand that I had bought from China ahead of
time so I replaced this transformer right away.
2. Back in June 2013 I had mentioned that one of the CCFL pins had dislodged
itself from the socket and that I had revived the monitor by reinserting it
back. Well, sort of. This monitor is designed in such a way that in order to
remove the inverter board one has to first remove the upper metal chassis that
holds the rest of the boards. This is not readily obvious because you can pivot
the inverter board slightly upward and wiggle it out of the 14 nests holding it
without removing the top portion of the chassis. These nests are essentially the
opposite ends of the CCFL sockets. However if you decide to take this shortcut
you will be putting enough prying pressure on the CCFL pins so that some of them
might jump out of their sockets and this is exactly what had happened with my
monitor. I have a suspicion that its previous owner did try to repair this
monitor at some point because when I opened it up for the first time one of
CCFLs was already disconnected from its socket.
Back then I thought I had managed to reinsert the pin back into its socket.
Turns out I had not. Now I know that these sockets have stiff spring loaded
contacts and that the CCFL ends require quite a bit of pressure to click their
pins home. Since the CCFL pin was just sitting on top of the clip the connection
was intermittent. This explains why for a while I was able to get the monitor to
turn on by squeezing and twisting its plastic body. Finally the arcing in the
poor connection had caused the socket to became charred to a point that the
inverter lost connection to this CCFL altogether.
I took the CCFL completely out of the housing and cleaned up the charred pin
with a blade. I also removed its socket, took it apart and gently sanded the
metal clip with some 400 grit sandpaper. Upon reassembly the monitor fired right
up and has been working fine ever since.
I have documented the above in the attached .pdf file.
One problem still remains a mystery. The big capacitor on the PSU board (C605)
used to show 168v at standby and 292v while working. Back in June I desoldered
it and measured its capacitance. It was 124µf, no drift.
Now C605 shows 78v at standby and 239v when working. It's capacitance is still
124µf, no drift.
I touched up each and every soldered joint with my Weller pencil thinking that
perhaps there was a cold joint somewhere. Now C605 shows 78v at standby and 241v
when working. I have never dealt with power stages before. I wonder if there is
a voltage divider with a drifting resistor or something like that.
Folks, I am happy to report that my 2408WFPb is once again alive and well.
Turned out it had two problems at the same time.
1. I've built a makeshift inverter tester out of a small 5w fluorescent tube and
a couple of wires with alligator clips. All 7 inverters had fired up this tube
successfully but in all 7 cases I would still get the dreaded 2 seconds to
black. However while performing this test I noticed a faint arcing under one of
the output transformers. Per Davi.p's advice above I measured the resistances of
the secondary windings of all 14 transformers. Thirteen of them showed ~1.650k
but one transformer showed 440k. I desoldered it from the board and looked at it
under the microscope. A small portion (<1mm) of the solid Cu wire that makes up
the secondary winding had burnt off the pin it was wrapped around and soldered
to pretty much the way a glass fuse blows leaving a tiny droplet of metal on the
wire end. I can only explain the 440k (i.e. the non-infinite resistance) of this
broken circuit by the presence of the carbon deposit left by the evaporated Cu
wire which provided at least some path for the multimeter current. Luckily I had
a couple of brand new transformers on hand that I had bought from China ahead of
time so I replaced this transformer right away.
2. Back in June 2013 I had mentioned that one of the CCFL pins had dislodged
itself from the socket and that I had revived the monitor by reinserting it
back. Well, sort of. This monitor is designed in such a way that in order to
remove the inverter board one has to first remove the upper metal chassis that
holds the rest of the boards. This is not readily obvious because you can pivot
the inverter board slightly upward and wiggle it out of the 14 nests holding it
without removing the top portion of the chassis. These nests are essentially the
opposite ends of the CCFL sockets. However if you decide to take this shortcut
you will be putting enough prying pressure on the CCFL pins so that some of them
might jump out of their sockets and this is exactly what had happened with my
monitor. I have a suspicion that its previous owner did try to repair this
monitor at some point because when I opened it up for the first time one of
CCFLs was already disconnected from its socket.
Back then I thought I had managed to reinsert the pin back into its socket.
Turns out I had not. Now I know that these sockets have stiff spring loaded
contacts and that the CCFL ends require quite a bit of pressure to click their
pins home. Since the CCFL pin was just sitting on top of the clip the connection
was intermittent. This explains why for a while I was able to get the monitor to
turn on by squeezing and twisting its plastic body. Finally the arcing in the
poor connection had caused the socket to became charred to a point that the
inverter lost connection to this CCFL altogether.
I took the CCFL completely out of the housing and cleaned up the charred pin
with a blade. I also removed its socket, took it apart and gently sanded the
metal clip with some 400 grit sandpaper. Upon reassembly the monitor fired right
up and has been working fine ever since.
I have documented the above in the attached .pdf file.
One problem still remains a mystery. The big capacitor on the PSU board (C605)
used to show 168v at standby and 292v while working. Back in June I desoldered
it and measured its capacitance. It was 124µf, no drift.
Now C605 shows 78v at standby and 239v when working. It's capacitance is still
124µf, no drift.
I touched up each and every soldered joint with my Weller pencil thinking that
perhaps there was a cold joint somewhere. Now C605 shows 78v at standby and 241v
when working. I have never dealt with power stages before. I wonder if there is
a voltage divider with a drifting resistor or something like that.
Comment