Finaly built a subminiature guitar amp... now I need help!
General idea was to scale down a full blown amp design, and build it with subminiature tubes. I used a Trainwreck Express schematic, tubes are 6112 in the preamp and two 5639 in the power (will try later also 5902WA).
Some specs.
Power tubes are cathode biased and not fixed like in the Wreck schematic. for main power I used two transformers back to back, 230v/12v and 230v/18v where the 12v feeds the 18v. this goes into a bridge rectifier (SB607) which gives me DC around 160-175. From there I have the regular cap>resistor>cap chain.
Heaters have a 2A tanny.
Two other changes are a 1.5M vol pot and not 1M, and 0.27uF cap for NFB loop instead of 0.1uF.
what else... Grounding: first power cap (plates) + power cathode bias + minus from bridge rectifier are one star ground, rest of the amp goes to star ground on the chasiss near the input.
NOW, the problem
Amp acts weird. for a few seconds it works fine, but then the sound starts to fade away, distort quickly and goes to total silence. when I measure voltage for the power cathode bias it goes down to ZERO. readings of B+ all around the amp change, go from 60vdc to 170vdc. amp at this stage sounds like a noise gate, meaning no sound comes from the speaker, but when I play the guitar it all of a sudden opens up, gives a sound, and slowly fades back to silence.
Readings of the cathode bias shows that voltage fluctuates.
At this point, readings of the AC shows that it also fluctuates... BUT!.... if I touch some points on the circuit with a prob from my voltmeter, for example right at the output of the last coupling cap (between the PI and power tube), the amp starts to "build up" again, you can hear the natural hiss/hum come alive from the speakers, guitar sounds too, voltage on the cathode bias builds up, AC also comes up
This can hold up for a few seconds, then it deteriorates again to its sleep mode/gate noise behaviour
At this point I honestly have no idea where the problem could be.
Oh, one last thing, when the amp is in its "sleep mode", if I turn the volume all the way down, the amp starts giving a loud noise that goes like "puttt puttt buttt bttttt puttt"
Open the vol a bit and this goes away.
Please help
General idea was to scale down a full blown amp design, and build it with subminiature tubes. I used a Trainwreck Express schematic, tubes are 6112 in the preamp and two 5639 in the power (will try later also 5902WA).
Some specs.
Power tubes are cathode biased and not fixed like in the Wreck schematic. for main power I used two transformers back to back, 230v/12v and 230v/18v where the 12v feeds the 18v. this goes into a bridge rectifier (SB607) which gives me DC around 160-175. From there I have the regular cap>resistor>cap chain.
Heaters have a 2A tanny.
Two other changes are a 1.5M vol pot and not 1M, and 0.27uF cap for NFB loop instead of 0.1uF.
what else... Grounding: first power cap (plates) + power cathode bias + minus from bridge rectifier are one star ground, rest of the amp goes to star ground on the chasiss near the input.
NOW, the problem
Amp acts weird. for a few seconds it works fine, but then the sound starts to fade away, distort quickly and goes to total silence. when I measure voltage for the power cathode bias it goes down to ZERO. readings of B+ all around the amp change, go from 60vdc to 170vdc. amp at this stage sounds like a noise gate, meaning no sound comes from the speaker, but when I play the guitar it all of a sudden opens up, gives a sound, and slowly fades back to silence.
Readings of the cathode bias shows that voltage fluctuates.
At this point, readings of the AC shows that it also fluctuates... BUT!.... if I touch some points on the circuit with a prob from my voltmeter, for example right at the output of the last coupling cap (between the PI and power tube), the amp starts to "build up" again, you can hear the natural hiss/hum come alive from the speakers, guitar sounds too, voltage on the cathode bias builds up, AC also comes up
This can hold up for a few seconds, then it deteriorates again to its sleep mode/gate noise behaviour
At this point I honestly have no idea where the problem could be.
Oh, one last thing, when the amp is in its "sleep mode", if I turn the volume all the way down, the amp starts giving a loud noise that goes like "puttt puttt buttt bttttt puttt"
Open the vol a bit and this goes away.
Please help
Here is a nice layout drawing of a Trainwreck Express circuit:
http://www.ceriatone.com/images/layoutPic/trainWreckLayout/TwreckXpressCeriatone.jpg
Grid resistors?
Lets see, I have a 1M in the input stage, 1.5M volume pot on the second stage. in the third stage I have a switch with 50k/100k/150k resistors to choose from (checked it, it works). 4th and 5th stages are the PI, have 1M on each.
Why did you think about grid resistors?
Could something like that make all the above issues?
Thanks for checking in
http://www.ceriatone.com/images/layoutPic/trainWreckLayout/TwreckXpressCeriatone.jpg
Grid resistors?
Lets see, I have a 1M in the input stage, 1.5M volume pot on the second stage. in the third stage I have a switch with 50k/100k/150k resistors to choose from (checked it, it works). 4th and 5th stages are the PI, have 1M on each.
Why did you think about grid resistors?
Could something like that make all the above issues?
Thanks for checking in
This is basic troubleshooting 101.
Something is wired wrong.
How to find it?
Break the circuit down into sections, test one at a time.
A 'scope would be very helpful, although not absolutely required.
I'm presuming you do not have a scope, so let's proceed without one.
Remove the outpoot tubes.
Power up. Test the B+, the bias supply (if there is one), and the filament voltages at the SUPPLY
Then test the same voltages at the PLATE pins of each small tube.
Check for B+ on the plate pins of the output tube sockets also.
Assuming that is OK, then check the GRIDS of each tube for MINUS voltage with respect to ground. The voltages found should be found to be very close to the values spec'd in a tube manual.
Now, if you put a steady state sinewave signal into the input of the amp, you should be able to "see" it with your voltmeter in the AC mode. If you put in 0.1vac on the input, the plate should probably see something LIKE 10x or more voltage - so at least 1.0vac, probably more like 10vac will appear on the plate pin. The same voltage should then appear on the GRID of the next tube... and so on through to the grids of the ouput tubes' sockets (no tubes in there yet).
If you don't have a signal generator, anything like a CD player with a TEST CD that has tones on it will do the trick. You could also use the output of a computer soundcard, and software for making tones. Make tones on the computer and burn a CD?
The meter is likely NOT accurate at say 1kHz., but that doesn't matter, since all you care about it relative levels - you keep the frequency the same.
You should be able to see the signal move up and down as you adjust the volume control(s), tone controls (somewhat, if ur at 1kHz), and other controls...
The MAX signal at the grids of the output tubes will correspond pretty much to the value spec'd for grid bias.
Is it FIXED or Cathode biased output tubes??
If fixed, check for bias voltage at the grid pins.
If cathode biased, check for voltage appearing across the resistors.
IF no voltage appears, try making sure you have a grid resistor to GROUND from the grids, and NO positive DC on the grids.
ALL OF YOUR GRIDS NEED A RESISTOR TO GROUND AND A CAP COUPLING THE SIGNAL TO THE GRID (blocking DC)
You can also check the voltage across the cathode resistors and see if the static current for each tube is similar to the spec in the tube manual.
Chances are you made a wiring error...
Also, do you have a real schematic to post?
Finally, if you got all the above right, put the output tubes in and you should have output.
If not, check to see that you do not have ur feedback connected to the wrong side of the secondary... or flip the plate leads on the output transformer if you do, or flip the connections on the seconday.
If you go through the above checks, you'll probably find a voltage that is NG. That stage is where the problem (or a problem) is.
_-_-bear
Something is wired wrong.
How to find it?
Break the circuit down into sections, test one at a time.
A 'scope would be very helpful, although not absolutely required.
I'm presuming you do not have a scope, so let's proceed without one.
Remove the outpoot tubes.
Power up. Test the B+, the bias supply (if there is one), and the filament voltages at the SUPPLY
Then test the same voltages at the PLATE pins of each small tube.
Check for B+ on the plate pins of the output tube sockets also.
Assuming that is OK, then check the GRIDS of each tube for MINUS voltage with respect to ground. The voltages found should be found to be very close to the values spec'd in a tube manual.
Now, if you put a steady state sinewave signal into the input of the amp, you should be able to "see" it with your voltmeter in the AC mode. If you put in 0.1vac on the input, the plate should probably see something LIKE 10x or more voltage - so at least 1.0vac, probably more like 10vac will appear on the plate pin. The same voltage should then appear on the GRID of the next tube... and so on through to the grids of the ouput tubes' sockets (no tubes in there yet).
If you don't have a signal generator, anything like a CD player with a TEST CD that has tones on it will do the trick. You could also use the output of a computer soundcard, and software for making tones. Make tones on the computer and burn a CD?
The meter is likely NOT accurate at say 1kHz., but that doesn't matter, since all you care about it relative levels - you keep the frequency the same.
You should be able to see the signal move up and down as you adjust the volume control(s), tone controls (somewhat, if ur at 1kHz), and other controls...
The MAX signal at the grids of the output tubes will correspond pretty much to the value spec'd for grid bias.
Is it FIXED or Cathode biased output tubes??
If fixed, check for bias voltage at the grid pins.
If cathode biased, check for voltage appearing across the resistors.
IF no voltage appears, try making sure you have a grid resistor to GROUND from the grids, and NO positive DC on the grids.
ALL OF YOUR GRIDS NEED A RESISTOR TO GROUND AND A CAP COUPLING THE SIGNAL TO THE GRID (blocking DC)
You can also check the voltage across the cathode resistors and see if the static current for each tube is similar to the spec in the tube manual.
Chances are you made a wiring error...
Also, do you have a real schematic to post?
Finally, if you got all the above right, put the output tubes in and you should have output.
If not, check to see that you do not have ur feedback connected to the wrong side of the secondary... or flip the plate leads on the output transformer if you do, or flip the connections on the seconday.
If you go through the above checks, you'll probably find a voltage that is NG. That stage is where the problem (or a problem) is.
_-_-bear
Hendrixon said:Problem found!
I didn't think about the need to add grid leak resistors to the power tubes "just" because I used cathode bias
Now back to tweaking my little 2.5W push/pull creation
Cheers mate!
By "grid leak resistors" did you mean the resistors that are used between the grid and ground (or bias voltage source)? If those were missing, the grid of the tube would be "floating" in terms of voltage, and would produce strange results. Or if you meant "grid stoppers" (resistors that are in series with the signal source and bias resistor), you may have been getting supersonic audio oscillations. Which will impact the amp's behaviour with the audio that you can hear. What the grid stopper does is to act like a low pass filter, the resistor combined with stray capacitence of the tube itself makes the supersonic gain too low for it to substain oscillation.
Bear,
I've attached a schematic of a Wreck Express amp for you. well one of them any way since most were tweaked for each player, but the general circuit is the same.
wa2ise,
Yes, by grid leak resistor I mean the resistor from grid to ground.
This amp schematic calls for fixed bias and I used cathode bias, but forgot this means I need to add grid leaks for the power tubes... even though they are not in the schematics
I've attached a schematic of a Wreck Express amp for you. well one of them any way since most were tweaked for each player, but the general circuit is the same.
wa2ise,
Yes, by grid leak resistor I mean the resistor from grid to ground.
This amp schematic calls for fixed bias and I used cathode bias, but forgot this means I need to add grid leaks for the power tubes... even though they are not in the schematics
Hendrixon,
You might want to put a film cap of some size across those diodes in the cathodes of the output tubes. Try experimenting.
Also, there appears to be a ground point missing on the cathode circuit of the driver tubes, on the schematic. Well, I guess you can run it like that, but ur ground reference is through a pot, which is never a good idea... and the cathodes are kinda floating a bit, the feedback is applied in a funny way - but maybe this is standard stuff for guitar amps.
I'd consider changing that kludge and making the tube before into a phase splitter and running drive from the plate and cathode side... more better balanced. You could likely make up the lost gain in the V3 tubes when the cathode is closer to ground... etc...
I dunno.
_-_-bear
You might want to put a film cap of some size across those diodes in the cathodes of the output tubes. Try experimenting.
Also, there appears to be a ground point missing on the cathode circuit of the driver tubes, on the schematic. Well, I guess you can run it like that, but ur ground reference is through a pot, which is never a good idea... and the cathodes are kinda floating a bit, the feedback is applied in a funny way - but maybe this is standard stuff for guitar amps.
I'd consider changing that kludge and making the tube before into a phase splitter and running drive from the plate and cathode side... more better balanced. You could likely make up the lost gain in the V3 tubes when the cathode is closer to ground... etc...
I dunno.
_-_-bear
Hey bear,
The diodes in the output stage are on the plates, not the cathodes. have no knowledge why they are there, probably some type of way to lower the output?
Anyway, I didn't add them to my amp, first because I have no idea what they are used for and couldn't find info on that subject, and second because not all Express amps had those in them, so I figured I can live without them and still get the sound I want
The driver cathode groud and NFB (used for added presence thru the cap bleed of highs to ground) are pretty much standard in common guitar amps, though there are designs that have an extra resistor (usualy 4.7k) to ground - parallel to the pot.
I'm not sure I understand your last idea.
Cheers,
Amit
p.s. My amp truned out WAY TO LOUD then I expected
them subminiature tubes are strong
The diodes in the output stage are on the plates, not the cathodes. have no knowledge why they are there, probably some type of way to lower the output?
Anyway, I didn't add them to my amp, first because I have no idea what they are used for and couldn't find info on that subject, and second because not all Express amps had those in them, so I figured I can live without them and still get the sound I want
The driver cathode groud and NFB (used for added presence thru the cap bleed of highs to ground) are pretty much standard in common guitar amps, though there are designs that have an extra resistor (usualy 4.7k) to ground - parallel to the pot.
I'm not sure I understand your last idea.
Cheers,
Amit
p.s. My amp truned out WAY TO LOUD then I expected
them subminiature tubes are strong- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.