Yamaha CA-1000 runaway DC offset

Mico

Well-Known Member
I have just recapped amp modules on Yamaha CA-1000 replacing three electrolytic caps.
Everything seemed fine, easily adjusted 15V on both sides, A/B and A class idle currents are stable as well and DC offset on the right side. But left side DC is acting up, on cold startup it takes off from -150mV and than starts the countdown: in about 5min DC drops to - 65mV and continues to move in that direction.
After 15min. value is already zero , only to stabilize around 15-20mV. when full operating temperature was reached. Relay clicks in immediately even at -150mV so all this does not trigger protection. DC is the same regardless of the operating mode (A/B or class A) and amp/preamp bridge has been disconnected.
Never seen this before and not 100% sure if it was the same before recap, probably not as I would have noticed the extreme.
 
Last edited:
Never seen this before and not 100% sure if it was the same before recap, probably not as I would have noticed the extreme.
Was it doing this before you replaced the parts?
 
Did you use a bi-polar capacitor for C603 -- 47µf/50v ?
 
Did you check the fusible resistors are in spec on the power amp boards (replaced in some versions with ordinary resistors).
As a side question, did you replace the 2SC458LG on the heat sink.
 
Lets find the problem before willy nilly throwing more parts at it.
 
For drifting, slow to stabilize offset . We need to focus on TR602 and 603 long tailed pair.
 
That would cause bias issues if it was bad.
The drop from -150mV to 0mV as the amp warmed up pegged my interest. Yeah, unlikely to be cause however
the transistor does control how hard each half of the amp works, maybe missing some +mV
 
Did you use a bi-polar capacitor for C603 -- 47µf/50v ?
Yes C 603 is bipolar 47µF/50V and the rest is properly installed too. I am pretty sure DC was not that wild before.
Fusible resistors tested fine, will look into TR602 /603.
 
TR602 /603.
Look your board over real good for stray solder bridges and iffy solder connections around those transistors.
The transistors need to be pretty closely matched for Hfe.
 
It seems I measured fusible resistors twice on the same side... After repeating everything on the left channel, 680Ω fusible measured 310Ω .
Will order new one and see what happens, looks like problem was there before recap.
 
Last edited:
R622 measured 310Ω outside the circuit, new 680Ω resistor is already in. No change in DC stability, still takes 30min. to stabilize completely.
Will replace TR 602/603 with matched pair, can not see anything else being wrong.
 
Last edited:
Will replace TR 602/603 with matched pair
Try to match both hfe and Vbe.

If you are bored, could you measure the Vdc at TP3 and GND/chassis. Repeat for TP4 and GND. Which if any is bouncing
Repeat for Pin C, note if it's stable/bouncing...
 
Try to match both hfe and Vbe.

If you are bored, could you measure the Vdc at TP3 and GND/chassis. Repeat for TP4 and GND. Which if any is bouncing
Repeat for Pin C, note if it's stable/bouncing...

4am here...not the best time to fry the outputs, will do it latter. What really bugs me is that temperature shift. When I measure dead-cold amp DC starts at 150mV, if I warm up heat sink with hear dryer for 15- 20 sec.and than measure,
DC starts way lower, around 50-60-70mV.
 
What really bugs me is that temperature shift. When I measure dead-cold amp DC starts at 150mV, if I warm up heat sink with hear dryer for 15- 20 sec.and than measure,
DC starts way lower, around 50-60-70mV.
Need to sleep on that. Not sure that's real significant.
The outputs and bias transistor are on the heatsink. The heatsink is quite large, did you heat it up evenly, maybe one output hotter etc
or the bias transistor thermally out of whack with the outputs. Transistor Vbe drops as temp increases. A good observation anyway.
 
While there I will also replace 2SC458 bias transistor with KSC1845, can not hurt to try as those Hitachi devices have pretty bad reputation.
 
Did not change anything yet but did some thermal testing. While lightly cooling TR602 with compressed air, DC offset immediately goes down relative to initial value. When doing the same on TR603, DC starts moving back to its cold start position. Will see what happens when two new KSA 992 go there.
 
Back
Top Bottom