Hoffman Amplifiers Tube Amplifier Forum
Amp Stuff => Tube Amp Building - Tweaks - Repairs => Topic started by: fdesalvo on June 03, 2012, 11:20:03 pm
-
**lesson learned - do not wire an amp in one huge all nighter session.
*Edit- I forgot to connect the diode to the bias supply cap. I'll wire it up and check it later. That should explai the redplating rectifier..the power tubes were pulling too much current through it. It should also explain the lack of -v at the test jacks. No more late night wiring parties. I'll update you guys once I have more info.
Hey, guys.
I am almost finished building a single channel ab763 type and decided to test the output section as I wait for parts to arrive (screen grid resistors and input jack). I used this PT - http://taweber.powweb.com/store/022798sch.jpg (http://taweber.powweb.com/store/022798sch.jpg)
Please have a look at these pics (yes, I'm buying longer machine screws for the standoffs hehe) - I have added bias test points, and am not seeing any reading across them. I'm sure I misfired something - I lost track of time yesterday around 3PM - I sat down and started drilling the chassis and the net thing you know, it's 3AM and I'm wiring. Sweet mercy.
This is a slightly modified hoffman layout allowing me to keep all caps on the board; I originally intended on doing this in a smaller chassis, but ended up using a larger one. Here are some general shots beginning with the bias circuit. My NOS 6L6WGBs and 5U4GB all red plated within seconds..I triple checked the caps - all oriented correctly, I believe.
(http://i443.photobucket.com/albums/qq151/frank_desalvo/17b5cbb7.jpg)
The bias circuit is grounded at the negative end of the filter cap you see here (black wire).
(http://i443.photobucket.com/albums/qq151/frank_desalvo/39e89b41.jpg)
(http://i443.photobucket.com/albums/qq151/frank_desalvo/be807455.jpg)
(http://i443.photobucket.com/albums/qq151/frank_desalvo/1787d08a.jpg)
:dontknow:
-
My NOS 6L6WGBs and 5U4GB all red plated within seconds..I triple checked the caps - all oriented correctly, I believe.
You don't have ANYTHING connected to pin 5 of either 6L6.
You absolutely must have proper negative bias voltage on pin 5 before applying B+ to the 6L6s in a fixed bias amp.
This rule cannot be emphasized enough.
-
Can we see a pic of the circled area in red of the bias grid coupling caps?
-
You absolutely must have proper negative bias voltage on pin 5 before applying B+ to the 6L6s in a fixed bias amp.
This rule cannot be emphasized enough.
what he said.
:-)
--DL
-
My NOS 6L6WGBs and 5U4GB all red plated within seconds..I triple checked the caps - all oriented correctly, I believe.
You don't have ANYTHING connected to pin 5 of either 6L6.
You absolutely must have proper negative bias voltage on pin 5 before applying B+ to the 6L6s in a fixed bias amp.
This rule cannot be emphasized enough.
Thanks for pointing that out - I read that the purpose of grid stoppers was to prevent oscillation - I thought they were "optional"! :BangHead:
-
Can we see a pic of the circled area in red of the bias grid coupling caps?
I just have that area in an unmounted board shot - the pic you referencd above was taken before the wire from the bias pot was added to the junction of the 220Ks. When I get home, I'll snap a pic for you.
(http://i443.photobucket.com/albums/qq151/frank_desalvo/2375c1b3.jpg)
-
I read that the purpose of grid stoppers was to prevent oscillation - I thought they were "optional"!
They are. But negative bias isn't. :wink:
-
I read that the purpose of grid stoppers was to prevent oscillation - I thought they were "optional"! :BangHead:
They can be, but you still have to feed the output tubes a signal from the PI, pin 5 is the control grid on 6V6's, 6L6's, 5881's.
It is also where you send the -bias voltage through the 2 grid return 220K's that are before the grid stop R's.
Brad :icon_biggrin:
-
here's a good read on the purpose of "grid stoppers"
why use grid stoppers? (http://www.aikenamps.com/InputRes.htm)
-
Lol, I'm such a rookie. I have the mechanical skills to build anything I can dream up, but the understanding of what each piece is doing still lacks. I have a long ways to go haha.
Ok, so my tubes redplated for about 5 seconds or less total. Do you think they are ruined now? The 6L6s have some decent flashing on the inside of the tube - hopefully this was enough to mitigate any nastiness from contamination.
here's a good read on the purpose of "grid stoppers"
Thank you
-
You're probably OK. Much better than accidentally connecting 450V to pin 5. Just be aware that they have been stressed and try to be cautious and nervous when you fire it up. :icon_biggrin:
-
Just be aware that they have been stressed and try to be cautious and nervous when you fire it up. :icon_biggrin:
I'VE been stressed! :l2:
-
Haha my pleasure! I love getting the electron's eye view of the circuit!
-
Yeah! with the size and details on those pixs you can almost imagine yourself an electron following the path of least resistance! Aw that feels good!
-
Yeah! with the size and details on those pixs you can almost imagine yourself an electron following the path of least resistance! Aw that feels good!
A sad, sad metaphor for what was my educational life.
-
^ nice pull. Not may know her by her native, birth-given nomenclature. ;)
-
I have a BAssman 100 on my bench that had a red plating power tube. The customer was using Tung-Sol eissue 5881's in it, and one of them lost connection on pin 5. He only had it on until he saw that power tube red plating and shut it off. That was enough to kill the tube as it tests just about dead on my B & K 747B tube tester, and it still red plates in the amp even with a good pin 5 connection. Hopefully your tubes are still good, but be suspicious of them. Watch for red plating when you play it even if it doesn't red plate when it is idling.
greg
-
Yea I've heard that once a tube redplates it will be prone to do so again.
-
Yea I've heard that once a tube redplates it will be prone to do so again.
I would look to see if they looked like they were doing something bad, but don't worry too much about it.
Without getting into it too much, what you're describing would be true if the tube became gassy as a result of the original redplating. That said, I have a Westinghouse 6V6 with a hole melted in the plate (!) which works just fine in a 5E3 Deluxe copy and without redplating.
No, I don't know what caused the hole burned in the plate; the tube was like that when I got it used in a batch of tubes.
-
^Cool, they are not red plating now.
Update -
I added the grid resistors to the output tube sockets and still have no measurable bias at my test points. In either extreme setting, the tubes do not red plate, but I can't understand why I have no reading from these test points when the both resistors are connected end to end and to ground, and the other sides are connected to pins 1/8 of each socket. Can anyone take a stab?
Also - not sure if this is because of the dirty electricity in my apt, but there is 60hz hum in the background and I can feel the power transformer vibrating.
Otherwise, it's producing a pretty nice tone thus far with my humbucker equipped guitar, though there is no sparkle to be had. I can tweak that another time.
(http://i443.photobucket.com/albums/qq151/frank_desalvo/Scratch%20built%20amp/193b529d.jpg)
-
How much voltage do you have on pin 5 of each output tube? You do have your meter set to read DC MILLIVOLTS when connected to the bias test points?
BTW, that 1.5K resistor between pin 5 and 6 only needs to be 1/2 watt.
EDIT... I highly recommend grounding the black test point at the same place you have the PT HT center tap grounded.
-
How much voltage do you have on pin 5 of each output tube? You do have your meter set to read DC MILLIVOLTS when connected to the bias test points?
BTW, that 1.5K resistor between pin 5 and 6 only needs to be 1/2 watt.
EDIT... I highly recommend grounding the black test point at the same place you have the PT HT center tap grounded.
ok -
-44.5v on pin 5 of each
ok - my apologies - I have no mV setting, but I can put it on 2V and read .019 at one extreme and .015 at the other. That doesn't seem right.
What's the logic regarding the ground? Just curious.
-
I have no mV setting, but I can put it on 2V and read .019 at one extreme and .015 at the other. That doesn't seem right.
Your resistors are 1 ohm, therefore, if 1 ampere flows through them, you can measure 1v dropped across each resistor. Since you want tube current in the milliamp range, you will measure millivolts across each resistor.
It's too bad your meter doesn't have a millivolt range, because you could look at the reading and just assume it says "milliamps". As it stands, you're measuring 19 and 15 millivolts, or 19 and 15 milliamps. You current seems lower than I'd expect with 6L6's and -44v bias, if the plate and screen is over 420v or so.
-
Yah. I've got 455V on the plates. My bias range resistor is indeed 470ohms per spec. Kind of baffled.
Sluckey - I moved the bias test point ground to the same PT lug as the other center taps.
I have no mV setting, but I can put it on 2V and read .019 at one extreme and .015 at the other. That doesn't seem right.
Your resistors are 1 ohm, therefore, if 1 ampere flows through them, you can measure 1v dropped across each resistor. Since you want tube current in the milliamp range, you will measure millivolts across each resistor.
It's too bad your meter doesn't have a millivolt range, because you could look at the reading and just assume it says "milliamps". As it stands, you're measuring 19 and 15 millivolts, or 19 and 15 milliamps. You current seems lower than I'd expect with 6L6's and -44v bias, if the plate and screen is over 420v or so.
-
What's the logic regarding the ground?
The idea is to ground the test point (which is really the ground for the cathodes of the 6L6s) at the same point the filter cap that supplies the plates for the 6L6s is grounded, and that is the PT bolt where the CT is connected. The way you had it, the cathode current had to pass thru the tube socket ground lug, thru the bolt/nut for the tube socket, and thru the chassis to get to the negative lead of the filter cap.
Are those 6L6s the same that were redplating earlier? They may not be as strong as they used to be.
My bias range resistor is indeed 470ohms per spec.
Measure the voltage on pin 5 of the 6L6s with the bias pot maxed out each way. What's the swing? Now try a 1KΩ for the range resistor. What's the swing of the voltage on pin 5? What's the swing on the bias test points?
It would be great if you could check the bias with a meter that has a millivolt range. Got a friend nearby?
-
Edit- in all my rushing to leave the house on time, I forgot my manners. Thanks for the help. I'm trying to understand the logic and realized I need to understand each separate circuit within the amp as well as the circuit as a whole. Im going to run through the entire amp later on and take measurements per the schematic.
It is indeed time for a real meter haha! No one with a real one around.
These are my backup tubes and aren't the ones that roasted. I'll get those measurements when i return from our gig. Love getting up at 4:00am for these marathon shows!
What's the logic regarding the ground?
The idea is to ground the test point (which is really the ground for the cathodes of the 6L6s) at the same point the filter cap that supplies the plates for the 6L6s is grounded, and that is the PT bolt where the CT is connected. The way you had it, the cathode current had to pass thru the tube socket ground lug, thru the bolt/nut for the tube socket, and thru the chassis to get to the negative lead of the filter cap.
Are those 6L6s the same that were redplating earlier? They may not be as strong as they used to be.
My bias range resistor is indeed 470ohms per spec.
Measure the voltage on pin 5 of the 6L6s with the bias pot maxed out each way. What's the swing? Now try a 1KΩ for the range resistor. What's the swing of the voltage on pin 5? What's the swing on the bias test points?
It would be great if you could check the bias with a meter that has a millivolt range. Got a friend nearby?
-
6l6GC -
Pin 5 range: -46.3V - -44.4V
Bias Range: 19-15mA
6l6WGB - now bias range of 30-35ma???? WTH! One is at 34ma and one is sitting at 32ma - so much for being matched..didn't touch range resistor..
Thanks
Are those 6L6s the same that were redplating earlier? They may not be as strong as they used to be.
Measure the voltage on pin 5 of the 6L6s with the bias pot maxed out each way. What's the swing? Now try a 1KΩ for the range resistor. What's the swing of the voltage on pin 5? What's the swing on the bias test points?
It would be great if you could check the bias with a meter that has a millivolt range. Got a friend nearby?
-
Pin 5 range: -46.3V - -44.4V
That ain't enough range. What value pot did you use? Hold off on the 1K for now.
-
Pin 5 range: -46.3V - -44.4V
That ain't enough range. What value pot did you use? Hold off on the 1K for now.
10K per schematic - just popped WGBs back in and they are drawing between 30-15ma on extreme ends of sweep lol.
-
I'm talking about the voltage swing on pin 5. Pull the 6L6s out (or kill B+ with STBY switch) and recheck that range.
-
I'm talking about the voltage swing on pin 5. Pull the 6L6s out (or kill B+ with STBY switch) and recheck that range.
Sorry was in a rush when i responded. The bias range is showing an available 30-35 ma which is good for the 6l6wgbs. The other numbers I posted from pin 5 were with b+ applied. When I get home I can measure again with the amp on standby. Thanks man.
-
Ok measurements on pin 5 are the same with no b+
-
I'm talking about the voltage swing on pin 5. Pull the 6L6s out (or kill B+ with STBY switch) and recheck that range.
The bias range is showing an available 30-35 ma which is good for the 6l6wgbs.
No. That's not the -bias range, that's the current draw of the output tubes.
Please give the -bias voltage swing. In _ -dcv _ not mA's.
The -bias swing will give you what you need to set the output tubes current.
Brad :icon_biggrin:
-
-46.3 to -44.4 my man.
-
-46.3 to -44.4 my man.
That's not right. If you have -46.3V on the top of the bias pot, then you should have -31.8 on the bottom of the pot. And that is a decent voltage swing for 6L6s. But the numbers you have indicate either a wrong value pot or wrong value 22KΩ resistor, or a bad connection somewhere.
Here's a simple test to determine the problem. Turn the amp off. Measure resistance from chassis ground to each outer lug of the bias pot. What are the two numbers?
I've drawn your bias circuit to make it easier to follow...
-
You've got to be frustrated by now- thanks for your patience. You must be a drinking man :icon_biggrin:
All solder joints are great and resistors and the pot are of the proper value.
My meter is poop, but with the ohm setting of 20K I am getting 1.80 and 1.86 from each respective lug to ground.
-46.3 to -44.4 my man.
That's not right. If you have -46.3V on the top of the bias pot, then you should have -31.8 on the bottom of the pot. And that is a decent voltage swing for 6L6s. But the numbers you have indicate either a wrong value pot or wrong value 22KΩ resistor, or a bad connection somewhere.
Here's a simple test to determine the problem. Turn the amp off. Measure resistance from chassis ground to each outer lug of the bias pot. What are the two numbers?
I've drawn your bias circuit to make it easier to follow...
-
My meter is poop, but with the ohm setting of 20K I am getting 1.80 and 1.86 from each respective lug to ground.
I assume that means 1.80KΩ and 1.86KΩ? That ain't much swing. It should be 22K to 32K. Before I beat my head on the floor much more, do me a favor. Take a spare resistor, something like 10K or 22K or anything similar. Measure the resistance of that resistor on the 20K scale. What does your meter say?
There is still something wrong in that simple bias circuit. I do drink beer but I've never had an AB763 make me drink. Yours is no different. We'll get to the bottom of this yet.
-
I'm goin to lowes to pick up a proper meter before I respond. Thanks Steve.
-
OK - with my fancy new meter, here's what I get at the lugs of the bias pot - 29.65k/21.36k
My meter is poop, but with the ohm setting of 20K I am getting 1.80 and 1.86 from each respective lug to ground.
I assume that means 1.80KΩ and 1.86KΩ? That ain't much swing. It should be 22K to 32K. Before I beat my head on the floor much more, do me a favor. Take a spare resistor, something like 10K or 22K or anything similar. Measure the resistance of that resistor on the 20K scale. What does your meter say?
There is still something wrong in that simple bias circuit. I do drink beer but I've never had an AB763 make me drink. Yours is no different. We'll get to the bottom of this yet.
-
OK - with my fancy new meter, here's what I get at the lugs of the bias pot - 29.65k/21.36k
That's more like it!
Now adjusting the bias pot end to end, recheck the negative voltage swing on pin 5. Then recheck the millivolts swing at the bias testpoints. What do you have?
-
ok -
-v swing @ pin 5 is -53.5 to -55.7
mv swing @ bias test points is 31.1mv to 35.3
This is me punching myself in the throat if my problem turns out to be my old meter! But! I'm learning so much about the bias circuit on these amps, so I'm very grateful for the direction!
-
-v swing @ pin 5 is -53.5 to -55.7
That's still no good. That voltage swing should be 10 to 15 volts difference.
Your bias cap is not connected as it is on a real AB763 or Doug's AB763 layout either. Make this simple change and recheck those voltage swings...
Disconnect the red wire from the center lug of the bias pot.
Reconnect that red wire to the lug that has the green wire connected.
Now you will have a red and green wire on the righthand lug (as seen in your photos).
Now recheck the negative voltage swing on pin 5. Then recheck the millivolts swing at the bias testpoints. What do you have?
-
Done...WOW -
pin 5 swing: -33.7 to -56.3
bias test points: 32.8 to 77.9
-v swing @ pin 5 is -53.5 to -55.7
That's still no good. That voltage swing should be 10 to 15 volts difference.
Your bias cap is not connected as it is on a real AB763 or Doug's AB763 layout either. Make this simple change and recheck those voltage swings...
Disconnect the red wire from the center lug of the bias pot.
Reconnect that red wire to the lug that has the green wire connected.
Now you will have a red and green wire on the righthand lug (as seen in your photos).
Now recheck the negative voltage swing on pin 5. Then recheck the millivolts swing at the bias testpoints. What do you have?
-
Sluckeys the best. :icon_biggrin:
-
Sluckeys the best. :icon_biggrin:
I wish I could be an apprentice to someone with knowledge of this stuff! SO grateful to you guys here. Lol, may I now direct your attention to my other thread? :l2:
-
Finally! I knew if we stomped on it enough it would give up it's secret. For a better looking fix, just totally remove that red wire. Now put a jumper between the diode and the bias cap (edge of the board).
If you're concerned that the bias range may be too hot for you comfort, change the 22K/3W to a 27K/3W or even 33K/3W. One watt resistors are plenty big enough.
Otherwise, I'd say we're done with the bias circuit.
I learned something today... The position of the bias cap makes a difference. I've always used two caps so I never witnessed that before.
-
I learned something today...
Glad to be there when it happened :icon_biggrin:
Thanks SO much!