New Build - left channel mia

[Bryon]

Just finished building the Crack - resistances and voltages good but when I fired her up, all that I heard on the left channel was a faint buzz.  I went back in and resoldered everything remotely related to the output and checked continuity but all appears good and when I hooked it back up there was no change.

I did put in better RCA's, the Dayton 100uf caps and a 100K DACT attenuator so it isn't exactly stock.  I checked the resistances through the DACT and the continuity to and from it and all seems good to that point.  I also of course checked to see if the problem was upstream by switching the input interconnects without effect.

Have I missed the obvious?  Can anybody suggest where to check next?

thanks for your time and expertise

Bryon

[Jim R.]

Bad tube somewhere?

-- Jim

[Bryon]

Thanks for the suggestion, Jim, but I forgot to mention that I replaced the 12au7 with a tube I know is good and put a  6AS7G in place of the 6080 without solving the problem.

thanks for the thought

Bryon

[grufti]

Headphone cable?

[Grainger49]

Just to be sure you can trace the lost signal using a meter set on AC.  Start at the Left input RCA jack and then the 12AU7 grid.  From there the output of the 12AU7, the plate (read on the other side of the coupling cap) and then the input and output of the 6080 (after its coupling cap).

Just a thought.

[Bryon]

Headphone cable?

Good idea but I've tried two different headphones and had the same problem.  Something interesting that I found while trying to get the headphones to work in stereo is that if I put the plug in half way with the cardas cable on my Senn 650's the signal is weak blurred and out of focus, but if I do the same thing with my old Sony CD350's, I get glorious mono sound from both speakers.

thanks for your suggestion

Bryon

[Bryon]

Just to be sure you can trace the lost signal using a meter set on AC.  Start at the Left input RCA jack and then the 12AU7 grid.  From there the output of the 12AU7, the plate (read on the other side of the coupling cap) and then the input and output of the 6080 (after its coupling cap).

Just a thought.

Sounds like a great idea - Could you explain it to me in the colouring book fashion -eg pin numbers on the 9 pin socket and octal socket - I might [emphasis on the word might] be able to work it out but I don't want to make a mistake.


thanks Grainger

Bryon

[JC]

It is very unusual indeed for an amp circuit to have the correct DC Voltages while it is not working, so tracing the signal path is a logical approach.

A lot of DVMs will handle at least a portion of the audio bandwidth on the "AC Volts" scale, and at the very least we know that they work in the region of 60 Hertz or they wouldn't read line Voltage.  So, play a signal through the amp and start looking for it.  If your DVM is auto ranging, just set it to "AC Volts."  If it doesn't auto-range, set it to "AC Volts and work from high to low on the Voltage scale.

I recommend you start with the working channel to get a feel for how your meter reacts, then check the path on the non-working side.  Hopefully when you get to the node where you can't find the signal, you will have found the neighborhood you should look in for the problem.

[Grainger49]

I would give step by step but I don't have a Crack or the manual and don't know the 6080.  But I can help a little. 

All measurements are with one meter lead clipped to the outer conductor of either input RCA Jack.  It doesn't matter which jack you attach the "ground" lead to. 

With something feeding the inputs of your crack you will get a varying AC voltage at both of the RCA jacks.  That is the unamplified line level input.   You should be able to see the same thing at the volume control input and output.  The manual might give you some indication of which right and left are there. 

From the volume control it will go to the driver tube, 12AU7, I believe.  The input of the driver is either pin 3 or pin 7.  I am not sure which is right and which is left.  You should see the same relative voltage there as coming from the volume control.

The output of the driver tube is on pin 1 or pin 8.  Follow those two pins and you will probably find a capacitor.  You want to measure on the other side of this capacitor and the voltage should be higher than coming from the volume control.

That should be wired to the input, grid, of the 6080.  I don't know the tube pin here.  And finally the 6080 outputs through the 100uF 160V capacitor.  You want to go to the far side of that capacitor to measure again.

Somewhere along this path I'm willing to bet there is a loose connection.

Sounds like a great idea - Could you explain it to me in the colouring book fashion -eg pin numbers on the 9 pin socket and octal socket - I might [emphasis on the word might] be able to work it out but I don't want to make a mistake.

thanks Grainger

Bryon

JC, good idea about starting with the working channel.  That sets a base line.

It is very unusual indeed for an amp circuit to have the correct DC Voltages while it is not working, so tracing the signal path is a logical approach.

A lot of DVMs will handle at least a portion of the audio bandwidth on the "AC Volts" scale, and at the very least we know that they work in the region of 60 Hertz or they wouldn't read line Voltage.  So, play a signal through the amp and start looking for it.  If your DVM is auto ranging, just set it to "AC Volts."  If it doesn't auto-range, set it to "AC Volts and work from high to low on the Voltage scale.

I recommend you start with the working channel to get a feel for how your meter reacts, then check the path on the non-working side.  Hopefully when you get to the node where you can't find the signal, you will have found the neighborhood you should look in for the problem.

[Jim R.]

The amp is direct coupled so there won't be a cap between the voltage amp (12au7) and the 6080.

-- Jim

[Doc B.]

If the voltages are all correct you know that the problem is not in the power supply or in any part of the circuit that feeds power in and out of the tubes. So that leaves the parts of the circuit that are only in the signal path. Start at the headphone jack and work your way back past the output load resistors and through the 100 uF caps to the 6080 socket, reheating all of the connections. Then start at the input RCAs and follow the wiring to and beyond the volume pot to the 12AU7 socket, reheating all of those connections.

[JC]

"That's good knowledge to know!"

OK, so the DVM readings may be pretty funky in some spots, depending on how well the meter handles the AC signal riding on top of DC, but hopefully it will do well enough to indicate a "Signal-No Signal" condition.  And, that is why trying this on the working channel first may prove helpful. 

If the DVM is not giving you a good indication of AC signal on the working channel on those nodes where the AC is riding on top of the DC, you may need to clip in a little coupling cap before the meter probe.  It won't have to have too big of a value, but the working Voltage of the cap should exceed what you would normally expect to encounter in the circuit.  Something higher than the highest B+, in other words.

[ironbut]

One thing to double check is that the ground on both channels of the headphone output jack is properly soldered. Remember that the ground wire goes through one channels solder post and attaches to the far one. IIRC, I soldered the far end but forgot to solder the other until I noticed it was loose. Once all the other wiring is soldered in place, it's tough to see those.

[Bryon]

Thanks for all the suggestions and ideas.  This forum is amazing - the expertise and willingness to help is inspirational.

I followed through everyone's suggestions but found a short on the DACT between the ground and left channel out when I was following Doc's instructions to resolder systematically [Thanks again Doc - you da man!]  -

This wasn't a problem I would have had, had I used the stock part as the prongs on the DACT are close together and the connections on the left channel are well hidden by the ones for the right channel.

At any rate, the Crack is now making wonderful sounds on both channels.

If it sounds this good now, I can hardly wait to hear it after I install the Speedball tomorrow.

thanks again to everyone for all their help.

Bryon 

[Doc B.]

Glad to be of help. After helping out a few hundred (or maybe a few thousand?) bottleheads over the years one gets a certain sense for various troubleshooting issues. When I started at this oh, so many years ago by playing around with antique tube radios I used to pick up those old 50's and 60's TV troubleshooting books to try to learn how to go about sorting out a dead piece of gear. Those guys seemed to know so much that I would never get a handle on! But after you do this for a while you begin to figure out what is most likely to be wrong from just a few clues (you just have to break a lot of s**t along the way to get there). And the solution most usually starts with "reflow all the solder joints, adding just a little more solder as you go".

After a while you will become fearless. I've watched PB get there over the years and Shawn is getting there too. Yesterday I took apart a 75 ohm 44 position Daven T attenuator, and thought to myself "it wouldn't be so hard to turn this puppy into a 44 position 50K shunt attenuator". Well, at least it won't be so hard to ask Shawn to build it for me...;^)>