If you just want to lower the screen voltage without the degenerative feedback associated with the variable voltage drop across a large screen dropping resistor, or the power wasted in a voltage divider and the associated large filtering requirements, simply use a zener diode of the appropriate voltage in series with the feed to the screen "stopper" resistors, with the cathode of the zener oriented towards the supply voltage.
This will give you a fixed voltage drop equal to the zener voltage and independent of the screen current drawn (provided there is enough idle current to keep the zener in breakdown, otherwise it will just kick in during signal peaks, which may be good enough anyway).
Put a capacitor across the zener if you want to swamp out any noise spikes that may be there. Use a zener rated for the max power you expect to see plus a safety margin (power dissipated in the zener is equal to the zener voltage multiplied by the max RMS current measured at full cranked output).
Randall Aiken
G2 mod gigs well
Moderators: pompeiisneaks, Colossal
Re: G2 mod gigs well
That sounds like a fairly simple solution, Randall, thanks a lot for nice explanation. Perhaps, this may worthwhile for you to test as well, Andy?
-
Andy Le Blanc
- Posts: 2582
- Joined: Sat Dec 22, 2007 1:16 am
- Location: central Maine
Re: G2 mod gigs well
The 50% was an arbitrary choice, I had a PT that was rated to 140ma
so I went forward with the notion on a whim and biased every tube type
I had on the bench that would fit in the sockets to 70ma total dissipation
just see what would happen.
The amp had a remarkably similar tone no matter the tube type or
dissipation figure that came of it, I was surprised, I figured why not make
the arbitrary decision based on the current rating of the PT, why not?
It could be any figure you choose, It gave a common reference point.
It disregards the focus on tube bias and makes the power supply the central design concern
Its a statement of method, this is what I did when I set up power side
I'm not arguing that there is any thing revolutionary with this, it's simply what I did...
The harsh tone at the onset of clip, non musical distortion, distortion not
apparently related the signal applied to the amp... fizz
I'm pretty sure you could post a request for forum users to paint you a picture.
I noticed this in one of my amps that to best of my knowledge was "correct"
And what I noticed was that the screens were slightly positive and the
non musical distortion was significantly less when the grids were negative
in relation to the plate when I measured the difference the screens and the plate at the tube socket.
Its a statement of observation, nothing more.
I got on to trying to control the screens voltage with an extra resistor and
filter stage with bleeders before the regular 470r screen grid resistors
and wanted to see what would happen with measurements I could grab
simply and see what differences would come of the conditions I was
setting up, I used the conduction angle because it fit the measurements
that I could take.
Gaz is very correct, I'm very well aware of the practical limits of the approach
My playing observations are not full over drive but soft clip.
And the presentation of the amp was less harsh, more musical in its context
I've said before, I'm jazzed that one can measure the difference between
the screens and the plate at the tube socket, it gets around the difficulties
measuring over a series screen resistor, and once you can measure it
you can set the voltage of the screens how ever you wish, to where ever
you want. I heard an issue in a fender circuit and that's where I started.
The resolution appeared to be to use the measured difference between
G2 and Plate, and set the G2 voltage less than the plate...
hello there mr aiken, you can also use a 12au7 dressed as a diode before
the screens, the plate dissipation is close to the screen dissipation,
and there's no sand in the circuit
so I went forward with the notion on a whim and biased every tube type
I had on the bench that would fit in the sockets to 70ma total dissipation
just see what would happen.
The amp had a remarkably similar tone no matter the tube type or
dissipation figure that came of it, I was surprised, I figured why not make
the arbitrary decision based on the current rating of the PT, why not?
It could be any figure you choose, It gave a common reference point.
It disregards the focus on tube bias and makes the power supply the central design concern
Its a statement of method, this is what I did when I set up power side
I'm not arguing that there is any thing revolutionary with this, it's simply what I did...
The harsh tone at the onset of clip, non musical distortion, distortion not
apparently related the signal applied to the amp... fizz
I'm pretty sure you could post a request for forum users to paint you a picture.
I noticed this in one of my amps that to best of my knowledge was "correct"
And what I noticed was that the screens were slightly positive and the
non musical distortion was significantly less when the grids were negative
in relation to the plate when I measured the difference the screens and the plate at the tube socket.
Its a statement of observation, nothing more.
I got on to trying to control the screens voltage with an extra resistor and
filter stage with bleeders before the regular 470r screen grid resistors
and wanted to see what would happen with measurements I could grab
simply and see what differences would come of the conditions I was
setting up, I used the conduction angle because it fit the measurements
that I could take.
Gaz is very correct, I'm very well aware of the practical limits of the approach
My playing observations are not full over drive but soft clip.
And the presentation of the amp was less harsh, more musical in its context
I've said before, I'm jazzed that one can measure the difference between
the screens and the plate at the tube socket, it gets around the difficulties
measuring over a series screen resistor, and once you can measure it
you can set the voltage of the screens how ever you wish, to where ever
you want. I heard an issue in a fender circuit and that's where I started.
The resolution appeared to be to use the measured difference between
G2 and Plate, and set the G2 voltage less than the plate...
hello there mr aiken, you can also use a 12au7 dressed as a diode before
the screens, the plate dissipation is close to the screen dissipation,
and there's no sand in the circuit
lazymaryamps
Re: G2 mod gigs well
Apologies, I was a bit mean in my previous posts.
So would you say this method is equivilant to 'setting to particular no-signal cathode current' where that current is the same for all tube types?
It would be nice to be able to quantify this method against another but I can't think how that could be done easily.
Ok thanks for the additional details - I wasn't sure previously whether you meant PT primary power or the B+ winding power.The 50% was an arbitrary choice..The amp had a remarkably similar tone no matter the tube type or
dissipation figure that came of it, ..Its a statement of method, this is what I did when I set up power side
I'm not arguing that there is any thing revolutionary with this, it's simply what I did.
So would you say this method is equivilant to 'setting to particular no-signal cathode current' where that current is the same for all tube types?
It would be nice to be able to quantify this method against another but I can't think how that could be done easily.
I know what you mean, but I don't associate it with a properly functioning amp. If I hear it, I look for instability of some kind, eg parasitics that only appear on high level waveforms. May be a bad tube or screen grid resistor. SF Fenders do seems to exhibit this a lot, but it's generally due to poor lead dress, failing caps/tubes/resistors.The harsh tone at the onset of clip, non musical distortion, distortion not apparently related the signal applied to the amp... fizz
Cliff, my understanding is the opposite, ie as screen voltage reduces so does transconductance. eg as an extreme, if screen voltage = 0, the tube is non operational so transconductance = 0. Pete.Lower screen voltage equates to a higher transconductance
-
Andy Le Blanc
- Posts: 2582
- Joined: Sat Dec 22, 2007 1:16 am
- Location: central Maine
Re: G2 mod gigs well
no worries...
That bias approach, setting the zero-signal current to some % of the PS rating
and not basing it on a tube assumption was a happy mistake.
You get poking around looking at the voltage relationship between the plate
and G2, at zero signal, and their relationship is dependent upon the bias condition of the device.
I had text book circuit and circuit conditions and there was a funny little distortion
that I've dealt with when setting up fenders, after going through them and
verifying their circuit.
the distortion, anomaly, fizz went away with a colder bias, I wanted to know why.
I was poking around... screen grid resistors are frustrating, you cant measure
across them without interference, and nobody has more than qualitative
statements surrounding screen grid resistor values...
I was measuring plate voltage between pins 3 and 8 at the socket for bias
calculation, and it fit the rational that it was a voltage difference across the valve,
instead of referencing the reading to ground, etc...
so I applied the same rationale to pins 3 and 4... and bingo...
I could observe the voltage relationships between the plate and screen...
1) with changes in bias condition
2) with different screen grid resistor values
3) under static and dynamic conditions
surprise, surprise... surprise.
Instead of making assumptions, you can test and measure.
I know this very basic, but to DIY with simplistic tools and limited design resources
its fantastic... no more screen grid resistor BS
Its just more observational information, but the G2 is central to the voltage
gradients through the device, its operation, and in the end reflects upon
qualitative assessments of tone, performance and presentation.
I needed a stable known bias condition to make observations...
I asked the question what happens when you change tube type but keep
the same circuit conditions, looking at bias and g2/p relationship
when I swapped around the tubes, at a known bias there was an observable
change between tube types and the g2/p relation ship, but qualitative tone
assessments ended up being remarkably similar, it was a surprise.
I didn't matter whether it was a 6v6 or a kt88, same tone...
I believe that it was the result of stable voltage condition through out the circuit,
in the pre amp stages, inverter, and power side, opens the door
to a better design approach, the problem is the focus on tube bias
you cant ignore your power supply...
Its been fun, and has raised some interesting arguments
That bias approach, setting the zero-signal current to some % of the PS rating
and not basing it on a tube assumption was a happy mistake.
You get poking around looking at the voltage relationship between the plate
and G2, at zero signal, and their relationship is dependent upon the bias condition of the device.
I had text book circuit and circuit conditions and there was a funny little distortion
that I've dealt with when setting up fenders, after going through them and
verifying their circuit.
the distortion, anomaly, fizz went away with a colder bias, I wanted to know why.
I was poking around... screen grid resistors are frustrating, you cant measure
across them without interference, and nobody has more than qualitative
statements surrounding screen grid resistor values...
I was measuring plate voltage between pins 3 and 8 at the socket for bias
calculation, and it fit the rational that it was a voltage difference across the valve,
instead of referencing the reading to ground, etc...
so I applied the same rationale to pins 3 and 4... and bingo...
I could observe the voltage relationships between the plate and screen...
1) with changes in bias condition
2) with different screen grid resistor values
3) under static and dynamic conditions
surprise, surprise... surprise.
Instead of making assumptions, you can test and measure.
I know this very basic, but to DIY with simplistic tools and limited design resources
its fantastic... no more screen grid resistor BS
Its just more observational information, but the G2 is central to the voltage
gradients through the device, its operation, and in the end reflects upon
qualitative assessments of tone, performance and presentation.
I needed a stable known bias condition to make observations...
I asked the question what happens when you change tube type but keep
the same circuit conditions, looking at bias and g2/p relationship
when I swapped around the tubes, at a known bias there was an observable
change between tube types and the g2/p relation ship, but qualitative tone
assessments ended up being remarkably similar, it was a surprise.
I didn't matter whether it was a 6v6 or a kt88, same tone...
I believe that it was the result of stable voltage condition through out the circuit,
in the pre amp stages, inverter, and power side, opens the door
to a better design approach, the problem is the focus on tube bias
you cant ignore your power supply...
Its been fun, and has raised some interesting arguments
lazymaryamps
-
bluesfendermanblues
- Posts: 1314
- Joined: Tue May 22, 2007 12:57 pm
- Location: Dumble City, Europe
Re: G2 mod gigs well
Andy, thanks for bringing up this issue.
I might be considered as one of the diehard Dumble perfectionists in the Dumble section of this forum and after having build a number of Dumble clones, I've learned a lot about voicing amps to my taste, so I figured it would be fun to try something very rudimentary - like building a 5F1 champ.....ONLY to realise that tweed amps are not 'easy'er' than Dumble amps. Surprise
I spent a LOT of time trying to figure out just how to get those classic grail tones from a 5F1. I read you post a couple of times and figured 'why not?' and tried measuring g2 and plate. After a little trial and error with a couple of resistors, I settled on a 1K/5w which gave the 0.5v-screen-voltage-below-plate-voltage.
The result was much nice breakup OD and subtle compression, that sounds convincing and without the nasty top-end fizz, which I previously had to dial out using the tone knob of my guitar (or with snubber cap across the 220k between V2 input and ground)
Having focused mainly on tweaking Dumble preamps, I wasnt aware of the importance of tweaking screen voltage, but I've learned something new. Thank you!
I'll check my Dumble clones for right screen voltages - even if the Dumble tones mainly stem from the preamps.
I might be considered as one of the diehard Dumble perfectionists in the Dumble section of this forum and after having build a number of Dumble clones, I've learned a lot about voicing amps to my taste, so I figured it would be fun to try something very rudimentary - like building a 5F1 champ.....ONLY to realise that tweed amps are not 'easy'er' than Dumble amps. Surprise
I spent a LOT of time trying to figure out just how to get those classic grail tones from a 5F1. I read you post a couple of times and figured 'why not?' and tried measuring g2 and plate. After a little trial and error with a couple of resistors, I settled on a 1K/5w which gave the 0.5v-screen-voltage-below-plate-voltage.
The result was much nice breakup OD and subtle compression, that sounds convincing and without the nasty top-end fizz, which I previously had to dial out using the tone knob of my guitar (or with snubber cap across the 220k between V2 input and ground)
Having focused mainly on tweaking Dumble preamps, I wasnt aware of the importance of tweaking screen voltage, but I've learned something new. Thank you!
I'll check my Dumble clones for right screen voltages - even if the Dumble tones mainly stem from the preamps.
Diva or not? - Respect for Mr. D's work....)
-
Andy Le Blanc
- Posts: 2582
- Joined: Sat Dec 22, 2007 1:16 am
- Location: central Maine
Re: G2 mod gigs well
Wonderful....
Was the 5f1 a kit? Even it was or wasn't, do you see what I mean about design assumptions.
I was surprised...
Kustom amps that I've seen with 6l6 set the screens at -5
blues jr with el84 I've seen .... -23
old traynor's are a mess, big improvement minding the g2 in these
its worth paying attention to, that rig that just seems to have "it".
or the rig that sounds like a duck fart, even though everything looks ok.
Its right up there with having the right loading for the tube type
I bet with the blues jr, they were after a more class "A" like tone.
Its a slightly different rig than the pro jr, I haven't been in one to measure but
I bet its a smaller difference, going for a more "AB" tone.
another thing that I've noticed is that this where the differences in a
matched pair of tubes jumps out, at a known bias point (and circuit voltages)
you with see differences in the measured g2/p relationships.
Which stirs up a whole other set of issues.
even with a matched set you may need to set the screens to the needs of
the valve that has the greatest mismatch.
or selecting tubes based on this criteria, or match checking
makes being able to set g2 voltage look very appealing
especially with the garbage current production tubes.
Was the 5f1 a kit? Even it was or wasn't, do you see what I mean about design assumptions.
I was surprised...
Kustom amps that I've seen with 6l6 set the screens at -5
blues jr with el84 I've seen .... -23
old traynor's are a mess, big improvement minding the g2 in these
its worth paying attention to, that rig that just seems to have "it".
or the rig that sounds like a duck fart, even though everything looks ok.
Its right up there with having the right loading for the tube type
I bet with the blues jr, they were after a more class "A" like tone.
Its a slightly different rig than the pro jr, I haven't been in one to measure but
I bet its a smaller difference, going for a more "AB" tone.
another thing that I've noticed is that this where the differences in a
matched pair of tubes jumps out, at a known bias point (and circuit voltages)
you with see differences in the measured g2/p relationships.
Which stirs up a whole other set of issues.
even with a matched set you may need to set the screens to the needs of
the valve that has the greatest mismatch.
or selecting tubes based on this criteria, or match checking
makes being able to set g2 voltage look very appealing
especially with the garbage current production tubes.
lazymaryamps
Re: G2 mod gigs well
Hi Andy,
Did you happen to try Randall's zener idea and how that sounded compared to the voltage divider you tried?
Is it simply a matter of having the screen supply voltage just a bit
less than the plate voltage?
Also, does this mainly pertain to the 6L6GC?
I see the Blues Jr. mentioned which is a EL84 amp.
Let's take my 100w Dumble over drive special clone for example.
4 x 6L6GC, 5K1 grid stoppers and 500R screen resistors.
When I measure voltages at idle in this amp, the plate and screen voltages are almost always at the same voltage.
The plates are fed from 160uF filter, then a Fender choke which feeds the screen supply.
So the choke is the only thing besides the individual screen resistors dropping voltage.
When I asked about this before somebody said that once the amp is being used to amplify a guitar,
the choke works more to oppose current change.
Will the screen voltage go down a bit when operating?
I guess I never measured that...
Did you happen to try Randall's zener idea and how that sounded compared to the voltage divider you tried?
Is it simply a matter of having the screen supply voltage just a bit
less than the plate voltage?
Also, does this mainly pertain to the 6L6GC?
I see the Blues Jr. mentioned which is a EL84 amp.
Let's take my 100w Dumble over drive special clone for example.
4 x 6L6GC, 5K1 grid stoppers and 500R screen resistors.
When I measure voltages at idle in this amp, the plate and screen voltages are almost always at the same voltage.
The plates are fed from 160uF filter, then a Fender choke which feeds the screen supply.
So the choke is the only thing besides the individual screen resistors dropping voltage.
When I asked about this before somebody said that once the amp is being used to amplify a guitar,
the choke works more to oppose current change.
Will the screen voltage go down a bit when operating?
I guess I never measured that...
Tom
Don't let that smoke out!
Don't let that smoke out!
-
Andy Le Blanc
- Posts: 2582
- Joined: Sat Dec 22, 2007 1:16 am
- Location: central Maine
Re: G2 mod gigs well
I believe it pertains to any beam tube.
every maker emulated fender at one time or another so we've ended up
with the same prevalent design in many forms. Fender used 6v6 and 6l6,
both beam pentodes, marshal started as a knock off, el34.. 6550 ....
beam tetrode form an electrostatic lens, with the screen and grid alignments
and potential, the screen intercepts very little current, and you get a high
electron density between the plate and screen that acts like a suppressor grid.
so you call it a beam pentode, it fixes the dynatron issues with tetrodes
and get a sharper knee in the plate curves, and better efficiency..
this is what your tweaking with the g2 potential.
I'd first look at the measured difference between the plate and g2 for
what ever tube type, right at the pins of the tube socket, at what ever
bias you've already arrived at. You can grab a plate voltage between
the plate and cathode pins. and a ma. current over a 1r to ground from
cathode...
look at static or idle tube current (ma.), and the difference measured between the g2 and plate
change the bias... and observe the changes in plate voltage and the g2/p
relationship
as you increase the bias voltage (reducing the current flow) the plate voltage should rise...
as you reduce the bias voltage (increasing the current) the plate voltage
should drop...
but the screen voltage will stay relatively fixed
when you apply signal the plate goes positive relative to the screen
so if the screen is slightly positive at static conditions, it must first go
to zero as the plate swings positive under signal, there's the fizz...
keeping the screen slightly positive idealizes the location of the virtual
suppressor in a beam tube, and sharpens the knee in the plate curves
but you end up with other issues when you push the power side into clip.
And the g2/p relationship changes with bias, so have to bias colder to avoid
tone issues
Have you ever seen a tube data sheet where the screens were positive
relative to the plate?
Measure the difference g2/p at the tube tube socket, it tells you whats
going on, It shows up when you run Iq and Imax numbers for conduction angle.
The more positive the plate is at static, the more towards the
"A" in AB the power side operates, its a little less efficient but it avoids other issues
when you push the amp.
Its a matter of observing the g2/p and choosing what to do about it.
You can measure every amp that goes across your bench, see how it addressed by other makers and designs.
I went after the simplest way to fix the issue in my amp, scab in a couple bleeders.
If you can measure the g2/p difference you can choose to fix/modify it
any way you want, set it with a different screen grid resistor value knowing what you've done...
bleeders at the g2 pin... a voltage divider
you can use diode to drop the screen supply a couple volts...
regulated supply..
you can get a resolution down to just a few ma/mv with even a basic DMM
and be fairly precise, I was thinking on the way home from the gig last
night that you might even be able to balance mismatched tubes right in
the circuit using the g2/p measurement
haven't had the chance to fiddle with the zener yet, I like the simple fix
less worries when it bombs.
every maker emulated fender at one time or another so we've ended up
with the same prevalent design in many forms. Fender used 6v6 and 6l6,
both beam pentodes, marshal started as a knock off, el34.. 6550 ....
beam tetrode form an electrostatic lens, with the screen and grid alignments
and potential, the screen intercepts very little current, and you get a high
electron density between the plate and screen that acts like a suppressor grid.
so you call it a beam pentode, it fixes the dynatron issues with tetrodes
and get a sharper knee in the plate curves, and better efficiency..
this is what your tweaking with the g2 potential.
I'd first look at the measured difference between the plate and g2 for
what ever tube type, right at the pins of the tube socket, at what ever
bias you've already arrived at. You can grab a plate voltage between
the plate and cathode pins. and a ma. current over a 1r to ground from
cathode...
look at static or idle tube current (ma.), and the difference measured between the g2 and plate
change the bias... and observe the changes in plate voltage and the g2/p
relationship
as you increase the bias voltage (reducing the current flow) the plate voltage should rise...
as you reduce the bias voltage (increasing the current) the plate voltage
should drop...
but the screen voltage will stay relatively fixed
when you apply signal the plate goes positive relative to the screen
so if the screen is slightly positive at static conditions, it must first go
to zero as the plate swings positive under signal, there's the fizz...
keeping the screen slightly positive idealizes the location of the virtual
suppressor in a beam tube, and sharpens the knee in the plate curves
but you end up with other issues when you push the power side into clip.
And the g2/p relationship changes with bias, so have to bias colder to avoid
tone issues
Have you ever seen a tube data sheet where the screens were positive
relative to the plate?
Measure the difference g2/p at the tube tube socket, it tells you whats
going on, It shows up when you run Iq and Imax numbers for conduction angle.
The more positive the plate is at static, the more towards the
"A" in AB the power side operates, its a little less efficient but it avoids other issues
when you push the amp.
Its a matter of observing the g2/p and choosing what to do about it.
You can measure every amp that goes across your bench, see how it addressed by other makers and designs.
I went after the simplest way to fix the issue in my amp, scab in a couple bleeders.
If you can measure the g2/p difference you can choose to fix/modify it
any way you want, set it with a different screen grid resistor value knowing what you've done...
bleeders at the g2 pin... a voltage divider
you can use diode to drop the screen supply a couple volts...
regulated supply..
you can get a resolution down to just a few ma/mv with even a basic DMM
and be fairly precise, I was thinking on the way home from the gig last
night that you might even be able to balance mismatched tubes right in
the circuit using the g2/p measurement
haven't had the chance to fiddle with the zener yet, I like the simple fix
less worries when it bombs.
lazymaryamps
-
Markusv
- Posts: 393
- Joined: Wed Dec 21, 2005 3:16 pm
- Location: Toronto, Canada (yes it's friggin cold!)
Re: G2 mod gigs well
Andy
I may have missed reference to this in the long and interesting thread but I would like to ask if there is any merit in using this technique with el84's?
I'm sure they're true pentodes and not beam power tubes as you are referencing here. (are BPT's Tetrodes?)
Does that matter?
I am running a 4x el84 amp and I wonder is a few volts drop on the screens would be
a) gentler on the tubes when run for hours on end
b) useful for softer onset of clipping in these tubes
Thanks
Markus V
I may have missed reference to this in the long and interesting thread but I would like to ask if there is any merit in using this technique with el84's?
I'm sure they're true pentodes and not beam power tubes as you are referencing here. (are BPT's Tetrodes?)
Does that matter?
I am running a 4x el84 amp and I wonder is a few volts drop on the screens would be
a) gentler on the tubes when run for hours on end
b) useful for softer onset of clipping in these tubes
Thanks
Markus V
.........Now where did I put it?
-
Andy Le Blanc
- Posts: 2582
- Joined: Sat Dec 22, 2007 1:16 am
- Location: central Maine
Re: G2 mod gigs well
they are grouped together for description as pentode, screen-grid, and beam tubes.
tubes having four or more electrodes.... yep a beam tube is a tetrode
but it functions as a pentode, the suppressor grid is virtual
yes a 6bq5/el84 is a power pentode, and thats a great question.
and yes you really need to mind g2, here's why.
A condition happens in power pentodes called virtual cathode
If conditions occur where the plate V is low relative to the screen
the suppressor starts behaving like another cathode, and the plate current drops
while the screen current jumps, a condition to avoid if your really pushing the amp.
I can see this as a rational in a small amp to set the g2 - a good deal
if its a small amp your gonna really push it just to keep up on a loud stage.
putting that aside, its more a worry of extreme operation or a bad design assumption
I dont think that different g2 v would be gentler on the tube, el84 has a small envelope.
"gentler" operation would be a function of plate voltage, lower plate v makes less heat.
respect the published design center limits etc...
I got after this to clean up a issue I heard in a amp. I didn't want to change
the amp, I got a better tone and transition into clip for my playing taste.
The amp had slightly positive screens, Ive found anecdotal statements to
support why an amp with beam tubes would be designed that way and why it might be an issue.
I was really surprised at the improvement 1/2 v made.
I'm jazzed there's another measurement in the power side that can be used to set up a amp,
you dont have to fumble around with screen grid resistor values. you can
set a g2/p difference and audition for your taste.
All of my tinkering has been centered on keeping the plate positive by
a few hundred ma. to a few volts, at the bias point that I choose for the tube
I think your going to find that like with previous conversations about
screen grid resistor values you will find a preferred range to fall upon,
except that you'll be able to nail your tone results to a valid circuit measurement.
I did notice that the g2 voltage can be used to "shut off" the tube just like
the bias voltage can at g1. So... there must be a V range to play with
for audible effect, but the g2/p relationship also affects the bias, if you
change the g2 voltage it changes the bias too, so your gonna want to
maintain a bias setting as your playing with screen grid voltage.
A few post back, a fellow mentioned a marshal floating resistor...
it was an additional resistor of the same value feeding the usual screen
resistors, a bleeder at that junction allows that resistor to vary the g2 V
instead of using the bleeder to drop the V, you can get away with a lower
wattage bleeder, and its easy to swap around its value with out really
modding irreversibly.
tubes having four or more electrodes.... yep a beam tube is a tetrode
but it functions as a pentode, the suppressor grid is virtual
yes a 6bq5/el84 is a power pentode, and thats a great question.
and yes you really need to mind g2, here's why.
A condition happens in power pentodes called virtual cathode
If conditions occur where the plate V is low relative to the screen
the suppressor starts behaving like another cathode, and the plate current drops
while the screen current jumps, a condition to avoid if your really pushing the amp.
I can see this as a rational in a small amp to set the g2 - a good deal
if its a small amp your gonna really push it just to keep up on a loud stage.
putting that aside, its more a worry of extreme operation or a bad design assumption
I dont think that different g2 v would be gentler on the tube, el84 has a small envelope.
"gentler" operation would be a function of plate voltage, lower plate v makes less heat.
respect the published design center limits etc...
I got after this to clean up a issue I heard in a amp. I didn't want to change
the amp, I got a better tone and transition into clip for my playing taste.
The amp had slightly positive screens, Ive found anecdotal statements to
support why an amp with beam tubes would be designed that way and why it might be an issue.
I was really surprised at the improvement 1/2 v made.
I'm jazzed there's another measurement in the power side that can be used to set up a amp,
you dont have to fumble around with screen grid resistor values. you can
set a g2/p difference and audition for your taste.
All of my tinkering has been centered on keeping the plate positive by
a few hundred ma. to a few volts, at the bias point that I choose for the tube
I think your going to find that like with previous conversations about
screen grid resistor values you will find a preferred range to fall upon,
except that you'll be able to nail your tone results to a valid circuit measurement.
I did notice that the g2 voltage can be used to "shut off" the tube just like
the bias voltage can at g1. So... there must be a V range to play with
for audible effect, but the g2/p relationship also affects the bias, if you
change the g2 voltage it changes the bias too, so your gonna want to
maintain a bias setting as your playing with screen grid voltage.
A few post back, a fellow mentioned a marshal floating resistor...
it was an additional resistor of the same value feeding the usual screen
resistors, a bleeder at that junction allows that resistor to vary the g2 V
instead of using the bleeder to drop the V, you can get away with a lower
wattage bleeder, and its easy to swap around its value with out really
modding irreversibly.
lazymaryamps
-
Markusv
- Posts: 393
- Joined: Wed Dec 21, 2005 3:16 pm
- Location: Toronto, Canada (yes it's friggin cold!)
Re: G2 mod gigs well
Phenomenal answer
Thank you Andy. I will then audition this in the future but really limit my expectations to sound rather than tube life considerations
Markus V
Thank you Andy. I will then audition this in the future but really limit my expectations to sound rather than tube life considerations
Markus V
.........Now where did I put it?
Re: G2 mod gigs well
I was looking through my notes because I knew I had seen that Marshall mod before.
Here is a picture of a JTM 45 with the 1K "floating" screen resistor.
Here is a picture of a JTM 45 with the 1K "floating" screen resistor.
You do not have the required permissions to view the files attached to this post.
Tom
Don't let that smoke out!
Don't let that smoke out!
Re: G2 mod gigs well
That's not what Andy was originally referring to though.
He would take that resistor back to ground not the screen supply node.
He would take that resistor back to ground not the screen supply node.
-
Andy Le Blanc
- Posts: 2582
- Joined: Sat Dec 22, 2007 1:16 am
- Location: central Maine
Re: G2 mod gigs well
looking at scheme of the floating resistor...
I'd put bleeders right at the tube socket, I started with a fender circuit.
I've also tried very similar to the floating arrangement with a bleeder and
extra filter cap to ground, extending a extra node exclusively for the screens
and also placing the the bleeder and cap right at the tube socket.
it all amounts to stabilizing the screen voltage at a known spot
getting the screens V more stable than with a series resistor in a simple
manner really helps, it can also provide a bypass to ground from the screens.
I'd put bleeders right at the tube socket, I started with a fender circuit.
I've also tried very similar to the floating arrangement with a bleeder and
extra filter cap to ground, extending a extra node exclusively for the screens
and also placing the the bleeder and cap right at the tube socket.
it all amounts to stabilizing the screen voltage at a known spot
getting the screens V more stable than with a series resistor in a simple
manner really helps, it can also provide a bypass to ground from the screens.
lazymaryamps