Symptoms:
trace become instable: looks like some noise creates bright sequences crossing left to right the entire screen.
The symptom persists even with CH1 and CH2 set to GND.
Triggering to LINE the bright sequences stop getting through, becoming stable as they were in sync with the sweep; indded it is triggered..

Diagnosis: it is a 50Hz line noise injected in somewhere. I immediately suspected one of those big electrolytic CAPS I've seen inside the scope sometime ago.

Resolution:
Checked the TEK 475 service manual :
- found the Low Voltage power supply output specifications
- found test points for various voltages.
All voltages were within specs but the +15V had an enormous ripple level. 
Based on the ripple shape, my basic electronic instinct said to me it was an open diode in the rectifier. In fact three out of five diode bridges, including the +15Volt one, were definitively too hot. Furthermore the symptom, absent during cold operations, suddenly appears after warm up - a diode opens due too heat. 
Checked with my friend IK1ODO who confirmed my hypothesis, suggesting me to replace all rectifiers bridges and all electrolytic CAPS.

The electric diagram was almost unreadable because of poor scanning. IK1ODO provided me one perfect photocopy.
Replaced all electrolytics and rectifiers with modern units.

Following the service  manual instructions, I verified that the load on all the LVPS lines was not exceeding specs. To do that I measured voltages between base and emitter of the current limiting transistors.
Here are the results:
Line DROPRES TRANSISTOR

My TEK475 is now again operational with a stable trace.

Back to TEK475

Back to Emanuele  Girlando's Home page