K12G Quiet Channel

[jollyllama]

Hi folks,

I've got a K12G that I've been running for about 2 years, and recently I noticed that one channel is about 30-50% quieter than the other. It's enough to be noticeable, but not a complete failure of the channel. I've read some other threads here about this type of issue, but I've got a question about testing:

- Safety first: How do I need to discharge the capacitors before I start poking around?

- Is there a good way to measure the signal coming out of the outputs with a multimeter without hooking up speakers? Since I'd have to have the speakers at pretty high volume to test this, and since I have have a toddler and baby at home, I'd really rather be able to do this silently (and hopefully more accurately) if possible.

- Alternatively, I've also heard of people tracing the voltages at each component (compared with the working channel) to isolate the problem - could someone elaborate on how to do this?

Thanks very much!

 

[BinaryMike]

Cap discharging is usually done with a resistor of a few hundred to a few thousand ohms and clip leads. I recommend a wirewound power resistor, because they have the best surge current survival characteristics.

You will need dummy load resistors to stand in for speakers. Get something around 8 ohms with at least 10W power rating. 8.2 and 10 ohms are commonly available. Shops use huge 8.00 ohm 1% non-inductive loads.

DC voltages are normally measured relative to ground. If you post your measurements along with a schematic, lots of folks will try to assist. Swap tubes first, though. Maybe you'll get lucky.

 

[azazello]

You can resolder all caps in cathodes of tubes.

 

[jollyllama]

Cap discharging is usually done with a resistor of a few hundred to a few thousand ohms and clip leads. I recommend a wirewound power resistor, because they have the best surge current survival characteristics.

Okay - I think I understand. After I put together a resister/clip lead set up, do I just leave that on each lead of the caps for a few seconds?

You will need dummy load resistors to stand in for speakers. Get something around 8 ohms with at least 10W power rating. 8.2 and 10 ohms are commonly available. Shops use huge 8.00 ohm 1% non-inductive loads.

Two questions: How do I connect those to the outputs, and how do I take measurements after the dummy loads are in place?

DC voltages are normally measured relative to ground. If you post your measurements along with a schematic, lots of folks will try to assist. Swap tubes first, though. Maybe you'll get lucky.

Okay, so one probe of my meter grounded, and one measuring from each point along the signal path while the amp is on, correct?

Thanks for the reply, and thanks for your patience with my very basic questions - unfortunately that's where my skill level is right now. I'm learning!

 

[jollyllama]

You can resolder all caps in cathodes of tubes.

Can you give me a little more info? I'm not sure I understand what the "caps in cathodes of tubes" refers to. Thanks!

 

[azazello]

You can find number of legs for every tube using Google. Then You can find cathode /every tube has anode, grids 1 or 2... and cathode/ to cathode are connected resistor and cap/capacitor/. At last You can solder leg and another ends of resistor and capacitor.

 

[BinaryMike]

Okay - I think I understand. After I put together a resister/clip lead set up, do I just leave that on each lead of the caps for a few seconds?

Two questions: How do I connect those to the outputs, and how do I take measurements after the dummy loads are in place?

Okay, so one probe of my meter grounded, and one measuring from each point along the signal path while the amp is on, correct?

After using the cap discharge rig, check residual charge with a voltmeter. This will give you a good feel for the necessary discharge time for subsequent use.

Dummy loads must be connected to a tube amplifier under test even if volume is set to zero, but for troubleshooting purposes you would apply an equal amplitude 1KHz sinewave to both inputs at a level that yields output power somewhat below maximum (6~7VAC in this case), then use a scope or AC voltmeter to check signal amplitude along the path from input to output.

DC voltage readings are usually taken with volume set to zero, even though power supply voltages may drop under high power output conditions. If there's a defect serious enough to produce an obvious gain difference between channels, then your DC bias voltages are likely to be far into the weeds and the cause will be obvious to many of us.