David Bourgeois wrote:

Hi,

My TDS540 doesn't start anymore. When I switch power on and press the
ON/STBY button, the fan starts and panel light up. Then within 30 secs
the display should start but it doesn't do it anymore.
I bought it second hand last year and it had the problem already, but I
let it turned on for a couple of hours and then it finally started
normally, all tests passed. After that, it could take from 10 secs to a
few minutes to start. I noticed that sometimes I couldn't turn it off
with the ON/STBY button, I had to power it off from the mains switch.

I didn't power it on for a few months and now it doesn't start at all
even after 24hours powered on.

I found the service manual and followed the diagnostic procedures (low
voltage power supply is OK), and eventually ended up at the line saying
"replace the DRAM processor/display board". The display should be OK,
when I turn it on I have a flash (line that quickly expands on the
screen). The digit LCD on the DRAM board displays hex from '3' to 'E',
not '1' to 'E' as stated in the service manual.

It seems there is something like an initial check, can it be temperature?
I've read that electrolytic capacitors on these DSO leaks and should be
replaced. I checked mines and they didn't leak. Can this still be the
cause? Should I get a cooling spray or heat some parts of the board to
check if that makes any difference?

Thanks for any help.

David Bourgeois
If that has the fluorescent type display, you may want to check the
inverter for the hight voltage to the display..
if it's a LCD., it should have lamps and that also has an inverter
that can cause the same problem..

But I would look first at the inverter circuit. Aging crimps inside
of Electrolytic caps give for poor connection and generate ESR.
Switching supplies have issues with this because you'll have
in effect a soft response on the cap and it may not start or not
generate the desired effect in over all performance of the circuit
depending of where it's located.

It's not common to have ESR meters lying around, even though the
actual construction of one is very simple and low part count.

An alternative would be if you have a square wave generator and a
second scope lying around.
Place a low value R from the output of the generator that fits with
in the specs of the unit, like a 50 ohm for example. What you want to
do is make a L network, R coming from the generator in series and the
DUT (device under test) in this case a suspected Cap, to the other end
of the R then to common. Attach a Scope across the Cap and monitor the
wave being generated. What you're looking for is the base line of the
square wave. Adjust the output frequency of the generator so that you
can see a full charge on the cap before it reaches the right side of the
ON duty cycle of the square wave.

When the duty cycle reverts to off, you should see a near perfect
0.0 volts. This is because the response time of the square wave is
fast enough to cause the cap to create near perfect 0 ohm condition
how ever, if there is internal issues with the cap, you'll see a non
base line effect or even noise generated at the initial skew point of
the wave..

years ago, I made up a test jig box that has a square wave generator
in it and employs an analog meter along with a scope signal so that I
could debug either way..
I used a bridge to detect the differential offset via 2 small caps
to take a reading only at the falling edge of the square wave. This
results would go into a sample and hold amp.

Have fun..


http://webpages.charter.net/jamie_5"