Hewlett-Packard 141T Display Section, serial number 1615A148XX, 8552B
Spectrum Analyser IF Section, serial number 1505A095XX and 8556A
Spectrum Analyser LF Section, serial number 1907A046XX. Report of
repairs.
This spectrum analyser system failed at turn on, after the unit had
been left unused for a few months. Investigation revealed that on the
8552B the two fuses on the back panel, F1 and F2, where blown. I
checked to see that there where no short circuit and installed new
fuses. The unit ran for 15 to 20 seconds and then blew both fuses
again. Looking at the schematic, the only way for both fuses to blow
at the same time is if the overvoltage crowbar circuit is activated.
To verify this I removed CR10, the crowbar thyristor, from the A5
assembly. Turned on the unit and it worked OK for a minute or so,
then some other type of symptoms appeared. The trace is way off at
the top, I can see some light but not the beam it self.
Checking the power supply I found all the supplies high by about 30 %.
The +100 V was 129 V, the -100 V was -128 V, the 248 V was 320 V,
-12.6 V was -16.3 V. I turned off the unit after about 20 seconds. I
removed the plug-ins and checked the supply, all four where within
limits. Put back the plug-ins, the same problem is now back. I
disconnected the 100 V going to the plug ins, all the supply voltages
where OK. Checked the +100 V regulator, as it is the master and the
three others take their reference from it. With 100 V output, the
base to emitter voltage is 0.53 V on Q2, the pass transistor. With
129 V output, the pass transistor base to emitter voltage was -2.9 V.
So the regulator circuit was sending the right signal to lower the
output. This led me to suspect the plugins, maybe a current path from
the 248 V supply to the 100 V supply. Further inspection revealed
nothing wrong with the plugins.
Back to the 141T supply I measured the current into Q2, the pass
transistor with 128 V at the emitter and -3 V between the base and
emitter, it should be cutoff, not so, the current was 407 mA!
Replaced Q2, a 1854-0294, listed as Q, Si, npn Motorola SJ-1318 with a
On Semiconductor MJ15003, silicone, TO-3, NPN, 140 V 20 A 250 W,
overkill but I had it on hand. Any 140 V 1 A 75 W TO-3 NPN device
would do, maybe a MOSFET also. Everything was back to normal.
Reinstalled A5CR10 in the 8552B, all OK. System fully functional.
I noticed some PWB discolouration under A5R3, a 4.7 k 5 % 2 W
resistor, that dissipates about 1.3 W. I changed it to a 5W resistor
and spaced it as far away from the PWB, so the increased surface area
and longer leads will give a lower PWB temperature and longer unit
life.
Daniel from Ville Saint-Laurent, QC, Canada.
