Roland KC500 bias current

[Delta362]

I'm finishing up repairing a Roland KC500 that came in with blown outputs. I've replaced Q212 (2SC4029) with 2SC5200, Q213 (2SA1553) with 2SA1943.

Q210 (2SC4793) was blown. According to the Toshiba datasheet for it, the replacement is TTC011B. A word of caution is that the pinout is opposite so installing the transistor backwards lines up the ECB pins correctly.
Q211 (2SA1837) was subbed with the current production replacement, TTA006B. Same thing with installing it opposite to align the ECB pins correctly.

Q208 (2SC1815) was shorted and replaced.
Q201 (2SA1015) was replaced since Q208 was bad.

R215 (3.3ohm fuse) and R217 (150ohm fuse) were both blown and replaced. R216 (also 3.3ohm fuse) was replaced for good measure.

R218 (emitter resistor for Q209) was reading over 18k and replaced with a 68 ohm resistor. Q209 (2SC2229) seemed to read ok so I did not replace it.

I am now able to power on the unit and do not have any DC offset on the outputs. However, I put my one meter across R232 and a second meter across R233, both 0.33ohm 5W. I watched the mV for the two rise fairly quickly at about the same pace and value. After about 5 seconds they were reading 70mV and still climbing. I shut it off at that point.

If my math is correct, that was already over 200mA (0.07V/0.33ohm). That seems like these are biased way to hot, or somehow I have a runaway condition that I need to find.

Any suggestions? Thanks!

Edit to add: I have not connected a load on the output for any of the above.

[Jazz P Bass]

What voltage do you measure at the output bases?

[Delta362]

What voltage do you measure at the output bases?

After about 10 seconds:
Q212 base was +0.670 and slowly continuing to rise
Q213 base was -0.700 and slowly getting more negative

[Delta362]

The base of Q210 is at 1.255, whereas the base of Q211 is at -1.274.

[g1]

R218 (emitter resistor for Q209) was reading over 18k and replaced with a 68 ohm resistor. Q209 (2SC2229) seemed to read ok so I did not replace it.

I think you should replace it. It's pretty much controlling the bias.

[Delta362]

I think you should replace it. It's pretty much controlling the bias.

As the 2SC2229 is obsolete, got a suggestion for a suitable replacement?

[The Dude]

https://www.audiolabga.com/mal_cart/...ch_button=find

[Delta362]

Nice find! Shipping is brutal for them, but at least they have two.

Edit: Not quite as brutal for the shipping costs as I thought. $10 and the two they had in stock are ordered. Thank you for the link! I've used them before, but don't think of them often enough.

[g1]

I'm surprised there is no idle current spec. given. I see there are 2 parallel (3K9) resistors from base to emitter of Q207. Sometimes they do that so if idle current is too high, you clip one out and it makes the bias colder.
If Q209 replacement doesn't solve it, you might have to play around with that.

[Delta362]

I'm surprised there is no idle current spec. given. I see there are 2 parallel (3K9) resistors from base to emitter of Q207. Sometimes they do that so if idle current is too high, you clip one out and it makes the bias colder.
If Q209 replacement doesn't solve it, you might have to play around with that.

Thanks, g1! I was looking at that as well thinking that's where I'd play with the bias, but I didn't want to jump the gun and do that without consulting with those who are far better versed in these types of amps than I am.

As an FYI, I did see that the KSC3503 was also a recommended replacement for the 2SC2229. I have the 2SC2229 on order from Audio Labs now though so I'll try that first, then the 3k9 resistors. I was expecting the idle bias to be significantly lower than 200mA. Aren't these typically around 20mA or less typically?

[g1]

I would think at least under 50mA, probably more like 20 as you say.
Just didn't want to start with that, rather than replacing Q209 first, as it may have been stressed by the over-current that burnt it's emitter resistor.

[Delta362]

The replacement 2SC2229 Y did not solve it as the current responded the same. I then played with R236 value and ended up with a 4k7. Doesn't seem much different than a 3k9, but it made a world of difference in the bias current. Now it's down under 20mA and working good. Thanks!