Plate load resistor, coupling cap, bypass cap and resistor impact on tone?

[Mike_J]

In the process of trying different bypass cap and resistor values in the AB763 one-channel build with 6V6s. Goal is to find three values that will work for my Strat with single coils and Gibson with humbuckers. Looking for rhythm and lead values for each guitar. Will put best values on a 3-position switch as recommended by Ed Chambley.

Read somewhere along the way that the plate resistor should be about 67 times larger than the bypass resistor. So a 100K plate would have a 1.5K bypass resistor, 150K:2.2K, 180K:2.7K and 220K:3.3K. However have seen 820R bypass resistors with a 100K plate resistor as well as other values that don't conform with the 67 to 1 rule. Could anyone provide some clarity to this issue?

Trying to determine how the plate resistor, coupling capacitor, bypass resistor and bypass capacitor impact tone. Seems logical that a larger plate resistor would reduce voltage on the tubes plate. Higher plate voltages from my experiences have more headroom whereas lower voltages break up earlier. Read somewhere that larger coupling caps let more bass through and vice versa. Also read somewhere that the bypass resistor biases the preamp tube. Guessing that a larger bypass capacitor lets more bass through as well. Please correct any of these if incorrect or provide further clarity if possible.

Thanks
Mike

[HotBluePlates]

... Read somewhere along the way that the plate resistor should be about 67 times larger than the bypass resistor. So a 100K plate would have a 1.5K bypass resistor, 150K:2.2K, 180K:2.7K and 220K:3.3K. However have seen 820R bypass resistors with a 100K plate resistor as well as other values that don't conform with the 67 to 1 rule. Could anyone provide some clarity to this issue? ...

The "rule" of thumb works backwards to that.

Usually, you pick a plate load resistance 2-5 times bigger than the expected internal plate resistance of a triode, based on the supply voltage available and required output voltage swing. Then as a first-guess, you would expect the cathode bias resistor is about Mu-times smaller than the plate load ("Mu" being the tube's amplification factor). So with a 12AX7 and a 100kΩ resistor, you might expect a cathode resistor around 100kΩ/100 = 1kΩ. A 12AU7 with a plate load of 47kΩ might reasonably have a cathode resistor of 47kΩ/20 = 2.3kΩ or even quite a bit higher.

There is wide latitude in implementing this in practice when you're picking your operating point on the loadline defined by the plate load resistor. As you've observed, anything from 820Ω-3kΩ will probably work with a 100kΩ plate load.

Sticking just with the plate load resistor, there will be some variation of internal plate resistance, plate current and resulting output impedance when you change a plate load resistor value (though the cathode resistor plays a role in this, too). The output impedance has some effect on the treble response of a gain stage, and some effect on the loads the stage can drive.

... Seems logical that a larger plate resistor would reduce voltage on the tubes plate. Higher plate voltages from my experiences have more headroom whereas lower voltages break up earlier. ...

A bigger plate load resistor might allow a bigger output voltage swing and higher gain for the stage, if the operating point set by the cathode resistor allows it.

You can use a big plate load resistor but bias the stage to almost zero idle current, which will leave almost the full supply voltage at the plate. That seems counter to your intuitive reasoning. However, at the same time, the tube will easily cut off with little input signal and will distort heavily with little headroom. That's also counter to the intuitive reasoning.

Granted, people don't usually set up a gain stage this way unless they're looking for distortion. But it also demonstrates that sweeping generalizations often cited have significant problems. Point being, you have to consider the plate load and cathode resistor together, as both are required to set the tube's operating point. Supply voltage also plays a role. Requirements of the gain stage are also important, as amplifying a 20mV signal with a 12AX7 might only give a plate voltage swing of 1.2v so high headroom in that stage may not be important.

... Also read somewhere that the bypass resistor biases the preamp tube. Guessing that a larger bypass capacitor lets more bass through as well. ...

An unbypassed cathode resistor will result in local negative feedback, which will reduce the gain of the stage. Common values with a 12AX7 can be assumed to cut stage gain by about half (this will be different for different tubes & different-sized cathode resistors).

A cathode bypass cap restores stage gain. But if you make the cap small enough, the impedance through the cap will look big compared to the resistance through the cathode resistor, and stage gain will drop again to half. So sometimes the cathode bypass cap is intentionally sized to trim bass, or in some amps to even make the stage response very treble-heavy. The cap value at which this occurs depends on size of the cathode resistor (how much bass gets trimmed), while the amount of gain reduction (how much bass-trim) depends on the tube type used (and its characteristics at the operating point) & size of the cathode resistor.

There are literally whole books on presenting how preamp tubes work and the tradeoffs in designing a preamp gain stage.

[Mike_J]

There are literally whole books on presenting how preamp tubes work and the tradeoffs in designing a preamp gain stage.

If I continue to read excellent responses like the one you provided here then some day I may be able to read those books and understand them. Much thanks for your time and effort in presenting explanations that are easily understandable.


Thanks
Mike

[HotBluePlates]

You're welcome! It almost requires an in-person class over time with lab homework assignments. 

Then again, maybe you wouldn't mind studying for that class...

[Mike_J]

You're welcome! It almost requires an in-person class over time with lab homework assignments. 

Then again, maybe you wouldn't mind studying for that class...

Where do I sign up.


Thanks
Mike

[Shack]

Im in too 

[Joel]

I remember studying this stuff in my basic electronics training (and not all that long ago either!  Only a few years ago - in the '90's).  We had to work out the load lines and then draw the input and output wave forms.  From this you can see the distortion generated by different biases.  At the time I wasn't into valve amps so it didn't have too much relevance to me and I forgot most of the theory over time.   

Recently, this book has been the single best resource I've read - and it's geared towards the layman!  http://www.amazon.com/Designing-Tube-Preamps-Guitar-Edition/dp/0956154522  Worth every dollar of it's RRP and international shipping to Oz.

Merlin is also active over at AX84 in case you really get stuck on a concept.

[Mike_J]

I remember studying this stuff in my basic electronics training (and not all that long ago either!  Only a few years ago - in the '90's).  We had to work out the load lines and then draw the input and output wave forms.  From this you can see the distortion generated by different biases.  At the time I wasn't into valve amps so it didn't have too much relevance to me and I forgot most of the theory over time.   

Recently, this book has been the single best resource I've read - and it's geared towards the layman!  http://www.amazon.com/Designing-Tube-Preamps-Guitar-Edition/dp/0956154522  Worth every dollar of it's RRP and international shipping to Oz.

Merlin is also active over at AX84 in case you really get stuck on a concept.

Thanks for recommending this book. Reviews are impressive. Will order the book.


Thanks
Mike

[Ed_Chambley]

Over the years I have taken HBP's class and have learned quite a bit.


Most of what we read does not really let us know much until after some experimentation.  Like load lines.  After you hear it, then it begins to make sense. 


For instance, I am trying to learn more of what I see on a scope.  Feed a sine wave through a 5879 pentode and the top is an arc, but on the bottom it is a V shape.  I can change values and see a shift in the scope, but this change doesn't really tell me anything about what the change in sound will be.


After I finally find the tone I am searching for, then the scope reading has value.

[HotBluePlates]

... Most of what we read does not really let us know much until after some experimentation.  Like load lines.  After you hear it, then it begins to make sense. ...

Yep, at the end of the day you just have to try and hear for yourself.

There's lots of theory & "rules" but these encapsulate decades of experimental results. Theoretical rules are also developed to speed design towards an "optimum result" so you don't have to experimentally discover it each time. But "sub-optimal" also results in a functioning amplifier.

[Mike_J]

Most of what we read does not really let us know much until after some experimentation.

Thanks Ed, waiting for the rest of the parts I need for the jig I am making. Will use it for tuning the AB763 one-channel build per your suggestions. Currently planning a jig for tuning the tone circuitry. Never noticed how similar the 5f6a and AB763 tone stacks were. One has the 100K resistor tied to ground the other to power. (See attachment)


Thanks
Mike

[Willabe]

That's a very nice drawing but if you want to compare circuits you would (and so would we) be much better off with an electrical schematic not a layout drawing.

A layout drawing is for wiring up a new amp build or finding where the parts are in an amp to repair and/or mod it.

 

[Mike_J]

That's a very nice drawing but if you want to compare circuits you would (and so would we) be much better off with an electrical schematic not a layout drawing.

A layout drawing is for wiring up a new amp build or finding where the parts are in an amp to repair and/or mod it.

 

I understand the advantage of a schematic. What I am using the layout for is as a basis to create the jig. Will need three multi-position switches for the cap selections and a pot for the slope resistor. Not sure if the 100K resistor that goes either to ground or power has much impact on tone or not. Reason I did this is to see if a different jig is needed for a cathodyne tone stack versus a plate driven tone stack. Seems like the same jig could be used for either. Let me know if you still think a schematic would be beneficial and I will make one.


Thanks
Mike

[Willabe]

I understand the advantage of a schematic. What I am using the layout for is as a basis to create the jig.

I get that but you posted a layout drawing not a jig drawing, which wouldn't help with comparison of circuits and you said;

Never noticed how similar the 5f6a and AB763 tone stacks were. One has the 100K resistor tied to ground the other to power. (See attachment)

So I open your attachment to see what your referring to and there's a layout drawing. 

This is just me but I would much rather see a schematic than have guys write out what they are referring too and/or post a layout drawing for comparison of circuits.

Much easier, much faster and if drawn right much clearer because 'a picture tells a 1000 words.'

Never noticed how similar the 5f6a and AB763 tone stacks were. One has the 100K resistor tied to ground the other to power.

1 is plate driven the other is K driven.

[sluckey]

Some subtle differences...

The horizontal resistor in the 5F6A circuit should be 56K. Cap values for AB763s are different from the 5F6A also. Mid pots are wired differently.

[HotBluePlates]

Some subtle differences...

The horizontal resistor in the 5F6A circuit should be 56K. Cap values for AB763s are different from the 5F6A also. Mid pots are wired differently.

And reportedly, there are discrepancies between the 5F6-A layout and actual 5F6-A amps. Where the layout shows two 0.022uF caps, supposedly the actual amps came with 0.022uF, one 0.1uF and one 0.022uF, and one 0.1uF and one 0.047uF. Supposedly, the 0.1uF and 0.022uF combination is most common.

I haven't verified these claims myself by looking at old Bassman amps, but had a customer for a 5F6-A board once get very angry with me for not giving him one of the non-schematic/layout cap mixes, which he thought was more authentic.

[HotBluePlates]

That's a very nice drawing but if you want to compare circuits you would (and so would we) be much better off with an electrical schematic not a layout drawing. ...

... Not sure if the 100K resistor that goes either to ground or power has much impact on tone or not. ...

It does. But if you were looking at a schematic instead of a layout, you'd notice that 100kΩ to the power supply/ground is load of the tube stage ahead of the tone stack.

The 5F6-A has a cathode follower (load between cathode & ground) feeding the tone stack, which drives the tone circuit from a low source impedance.

The AB763 has a common-cathode stage (load between plate & power supply) feeding its tone stack, which drives the tone circuit from a relatively-high source impedance.

[Mike_J]

Some subtle differences...

The horizontal resistor in the 5F6A circuit should be 56K. Cap values for AB763s are different from the 5F6A also. Mid pots are wired differently.

Agree to what you are saying about the subtle differences. I made the values the same because they were so subtle and I was trying to focus on the possibility of creating one jig that could be used for either type tonestack. The jig would have different rotary switches for the treble, mid and bass capacitors so six values could be selected for each of them. Kind of a mix and match to tweak the tonestack. Also would have a pot for slope resistor adjustment. Read something somewhere that the slope resistor value matters somehow.


Thanks
Mike

[Mike_J]

That's a very nice drawing but if you want to compare circuits you would (and so would we) be much better off with an electrical schematic not a layout drawing. ...

... Not sure if the 100K resistor that goes either to ground or power has much impact on tone or not. ...

It does. But if you were looking at a schematic instead of a layout, you'd notice that 100kΩ to the power supply/ground is load of the tube stage ahead of the tone stack.

The 5F6-A has a cathode follower (load between cathode & ground) feeding the tone stack, which drives the tone circuit from a low source impedance.

The AB763 has a common-cathode stage (load between plate & power supply) feeding its tone stack, which drives the tone circuit from a relatively-high source impedance.

So would you add another potentiometer to adjust for the 100K resistor that goes to either ground or power? I know I am looking at this from a simpleton's point of view. On the AB763 isn't the 100K resistor essentially a plate load resistor? The cathode resistor and cap as well as the plate load resistor look similar to a normal Fender gain stage. Would the AB763 tonestack be a fancy coupling capacitor arrangement for the gain stage?


Going to start a new thread on this and create a schematic showing the two tonestacks. This could be an excellent class.


Thanks
Mike

[Ed_Chambley]

The BF Tonestack cannot really be compared to the Tweed Bassman as HBP mentioned the source for each is different.  The magic to the Bassman was the cathode follower supposedly which uses less gain and is not as lossy as said.  I have been inside a couple of real Tweed Bassmans and the tonestacks  had a 56K slope, 250pf, .02 and a .1 and appeared to be original.


Again, what you prefer will win out.  For instance, in a JTM45 I prefer to use a 33k Slope and .022 in both positions, but I really cannot ever get by putting a larger mid pot installed than 25K.  I just like to have that available.


I prefer for my current playing style a plate driven stack as the loss of gain I will trade for the ability to have a more responsive EQ.  To me a CF tonestack doesn't really change the amp a lot and I usually end up tweaking values and using cap switching over the famed 5F6a tonestack. 


Sure, the slope resistor has a huge impact.  Just open Duncans tone stack calculator and change the slope resistor only in the Marshall stock from 33k to 56K and the 450pf to 250pf  that -10db around 800hz will disappear.  Our ears are easily cluttered by mids and upper mids and generally if your guitar frequencies get too high the result is what is generally referred to as brittle.


I made a statement once that I wish I understood tone stacks better and PRR replied for me to study divider networks.  I began down this rabbit hole about 2 years ago and have found some light.  I used to think more of everything has to be better, but why is the AB763 circuit so popular?  So in addition to divider network study I began looking into how our ears hear sounds and what those sounds translate to.


So in designing I have begun to understand that taming frequencies beginning at around 1.75k and up and letting the natural harmonics carry what we hear as highs sound much more pleasant to my ears and tend to have a singing quality.  Looking at the attachment you will see the fundamental notes, even when playing way up the neck, are not very high.  After a few builds under my belt and tweaking and ending up with boomy and brittle I began to understand why the mid scooped tone got so popular.  It falls into the category where our ears hear warmth and woody tones up to around 400hz.


Anyway, I just thought I would mention a couple of things I really never considered when tweaking for tone.  Seems the more I learn about this the greater appreciation I have for the Champ preamp.

[HotBluePlates]

So would you add another potentiometer to adjust for the 100K resistor that goes to either ground or power? ...

No. Because that 100kΩ resistor is part of the tube gain stage preceding the tone stack, but is not part of the tone stack itself.

That's a very nice drawing but if you want to compare circuits you would (and so would we) be much better off with an electrical schematic not a layout drawing. ...

... Not sure if the 100K resistor that goes either to ground or power has much impact on tone or not. ...

It does. But if you were looking at a schematic instead of a layout, you'd notice that 100kΩ to the power supply/ground is load of the tube stage ahead of the tone stack.

The 5F6-A has a cathode follower (load between cathode & ground) feeding the tone stack, which drives the tone circuit from a low source impedance.

The AB763 has a common-cathode stage (load between plate & power supply) feeding its tone stack, which drives the tone circuit from a relatively-high source impedance.

... On the AB763 isn't the 100K resistor essentially a plate load resistor? The cathode resistor and cap as well as the plate load resistor look similar to a normal Fender gain stage. Would the AB763 tonestack be a fancy coupling capacitor arrangement for the gain stage? ...

See attached.

In the AB763, there are two 100kΩ resistors in that area instead of one as in the 5F6-A. But only one is part of the tone stack (the "slope" resistor). You can find it easiest in your layout drawings if you look for 1 resistor which touches all 3 caps in the tone stack.

The other 100kΩ resistor (inside the red box in both schematic fragments) is not a part of the tone stack, but is the load resistor for the tube gain stage. In the 5F6-A it is a cathode load (because the gain stage is a cathode follower). In the AB763 it is a plate load (because that gain stage is a "common cathode" gain stage, sometimes called "plate-loaded").

+1 to Ed's suggestion to download the Duncan Tone Stack Calculator. The "Marshall" tab in that program is similar to the 5F6-A, but with some changed parts values.

You will see a big midrange notch in the graphs for the Fender & Marshall circuits. Try changing the slope resistor, and you will see the depth and slope of that notch change (which is really the source of the name). But other parts values change it as well, because it's all a balancing act.

A deep understanding of tone stacks would require covering the following topics:
1.  Basic electricity/electronics (how elementary parts work)
2.  Basic High- & Low-Pass Filters (R-C style)
3.  The Bridged-T Filter
4.  Modified/Variable Bridged-T Filter (3-knob Tone Stack)

[Mike_J]

After a few builds under my belt and tweaking and ending up with boomy and brittle I began to understand why the mid scooped tone got so popular.  It falls into the category where our ears hear warmth and woody tones up to around 400hz.

Thanks Ed, exactly why I want to build these jigs. Want to get boomy and brittle out of all my amps and make them sing more. Great information about how to get pleasing tone. Also appreciate the slope resistor information. Think this jig will definitely have some value if only to help set the best slope resistor value for the stack.


Thanks
Mike

[Mike_J]

So would you add another potentiometer to adjust for the 100K resistor that goes to either ground or power? ...

No. Because that 100kΩ resistor is part of the tube gain stage preceding the tone stack, but is not part of the tone stack itself.
+1 to Ed's suggestion to download the Duncan Tone Stack Calculator. The "Marshall" tab in that program is similar to the 5F6-A, but with some changed parts values.

You will see a big midrange notch in the graphs for the Fender & Marshall circuits. Try changing the slope resistor, and you will see the depth and slope of that notch change (which is really the source of the name). But other parts values change it as well, because it's all a balancing act.

A deep understanding of tone stacks would require covering the following topics:
1.  Basic electricity/electronics (how elementary parts work)
2.  Basic High- & Low-Pass Filters (R-C style)
3.  The Bridged-T Filter
4.  Modified/Variable Bridged-T Filter (3-knob Tone Stack)

HBP, much help from what you said. Hopefully these jigs will help me hear the different cap and resistor values. As I recall you mentioned in another reply that my ears are good tools for making a successful build. Looks like I will only need a pot to represent the slope resistor and three six-position switches to hear how different treble, mid and bass caps sound in a circuit. Will have to work on the four areas necessary for learning about tone stacks. In the meantime hopefully my ears will help accomplish getting rid of either brittle or woofy tone.


Thanks
Mike

[tubeswell]

If you want some rules of thumb… (ands assuming a conventional inverting stage)


The 'most efficient' plate resistor is one that leaves the plate sitting at about 2/3 of the HT (supply) voltage for that stage. This gives the most optimum load resistance for the amount of plate current swing in relation to grid voltage (transconductance)


The 'most efficient' cathode resistor is one that will leave the grid idle voltage centre biased (1/2 way between Vg1=0V and cutoff)


The 'most efficient' cathode bypass cap is one that will enable full boost of all frequencies that the stage is intended to amplify. in this regard, the required capacitance is related to the resistance of the cathode resistor.


YMMV

[Mike_J]

If you want some rules of thumb… (ands assuming a conventional inverting stage)


The 'most efficient' plate resistor is one that leaves the plate sitting at about 2/3 of the HT (supply) voltage for that stage. This gives the most optimum load resistance for the amount of plate current swing in relation to grid voltage (transconductance)


The 'most efficient' cathode resistor is one that will leave the grid idle voltage centre biased (1/2 way between Vg1=0V and cutoff)


The 'most efficient' cathode bypass cap is one that will enable full boost of all frequencies that the stage is intended to amplify. in this regard, the required capacitance is related to the resistance of the cathode resistor.


YMMV

Thanks Tubeswell, think I can calculate 2/3rds of HT supply voltage for a stage. Know I couldn't calculate the resistor value that would leave the grid idle voltage centre/center bias 1/2 way between Vg1=0v and cutoff. Would be nice to know that resistor value for a 12AX7, 5751, 12AY7, 12AT7 and 12AU7. As far as the "most efficient" cathode bypass cap goes, if I remember something I read, the 25uF bypass cap can pass all the bass frequencies a guitar can produce and in my experience some you don't want to produce on some amps. Can't remember where the bottom range was for that cap to be able to produce those frequencies but as I recall it was less than 25uF.


Really appreciate the useful information.


Thanks
Mike

[Fresh_Start]

Mike - kudos for starting another really useful discussion. 

I'd pay good money for a class taught by HBP and PRR.  Oh wait!  They've given one for free for years - THANK YOU GUYS!

Ed - thanks for that frequency chart.  I have another around here but not as clear as that one.

Cheers,
Chip

[tubeswell]

...Would be nice to know that resistor value for a 12AX7, 5751, 12AY7, 12AT7 and 12AU7.

Take a look at Merlin Blencowe's load line plotter


Click here to download the Excel version (837kB)


Open Office version (800kB)

[Mike_J]

...Would be nice to know that resistor value for a 12AX7, 5751, 12AY7, 12AT7 and 12AU7.

Take a look at Merlin Blencowe's load line plotter


Click here to download the Excel version (837kB)


Open Office version (800kB)

Thanks Tubeswell, this could be very useful if I can ever figure it out. So far I learned a bigger volume pot will give you more gain (or is that mu), at least I think it will. This one will take some tinkering with.


Thanks
Mike

[Mike_J]

Have been looking at the Load Line Plotter provided by Tubeswell. Can see where HBPs comment describing this as a balancing act comes from. Some of it I don't understand and would appreciate help with if possible.

Does the HF roll-off due to grid stopping (-3dB)(Hz) give the frequency where the high frequencies have been reduced 3dB for a specific grid stop resistor? If so, looks like the lower the resistance of the grid stop resistor the higher the frequencies that are let through. For example, the Fender 68K cuts 3dB at a little over 24K Hz when using a 12AX7. The Marshall 33K cuts 3dB at about 50K Hz when using a 12AX7 or ECC83. The frequency is about twice as high if you use a 12AY7. Isn't the highest audible frequency for a human being around 20K Hz? May explain why Fender used the 68K value. Any reason Marshall or other builders would use a lower value resistor if it lets frequencies through that can't be heard.

Another question. Does the 1M resistor on the input jack have anything to do with the grid stop resistor value?

Looking at the LF frequency roll off due to RI (think this would usually be either a volume pot or pots tied to a tone stack) or Co (cathode cap). Reducing RI from 1M to 500K cuts the low frequency by 3dB at 7Hz with a 1M value versus 13.4 Hz with a 500K value when using a 12AX7. This is while using a .022uF coupling cap. If a .001 coupling cap is used the low frequency reduction by 3dBs is at 295.6 Hz, Does this mean more can be done to tame bass with the coupling cap versus reducing the volume pot value.

Can't figure out what half-boost frequency is? Any help would be appreciated. Looks like the half-boost frequency goes up as the bypass capacitor value goes down. Not quite sure what the difference between bypassed cathode and unbypassed cathode only means. Guessing it has to do with one of them having the bypass resistor across it and the other not but not sure.

Thanks
Mike

[Willabe]

Not quite sure what the difference between bypassed cathode and unbypassed cathode only means. Guessing it has to do with one of them having the bypass resistor across it and the other not but not sure.

No, it means bypassing the K R with a cap.

[HotBluePlates]

Have been looking at ...

Does the ...

Another question. ...

Looking at ...

Can't figure out what ...

And this is why I said, "There are literally whole books on presenting how preamp tubes work...," back in Reply #1.

I guess I'll try to answer some of the questions, though it will take a series of posts (and even then, there's plenty of foundational knowledge you should have already). I don't use the load-line plotter, so I'll avoid answering specific questions about how it displays results. Maybe I'll just hit the questions PRR-style...

Does the HF roll-off due to grid stopping (-3dB)(Hz) ... Fender 68K cuts 3dB at a little over 24K Hz ... Marshall 33K cuts 3dB at about 50K Hz ... Any reason Marshall or other builders would use a lower value resistor if it lets frequencies through that can't be heard. ...

A grid stopper is about attenuating radio frequency (RF) interference; it ain't about tone.

It could be about tone is you purposely massively-oversize a grid stopper to shave treble, but few people do that outside of shred-machines. You will need to learn something about Miller Capacitance to sort out the difference.

Why do people copy values if there's no audible effect? People copy a lot of things when they don't know how it works, because copying a known-working plan will have a better likelihood of resulting in a working-copy. 

... Does the 1M resistor on the input jack have anything to do with the grid stop resistor value? ...

No, but it might impact how much treble you lose due to your cable's capacitance. And also how well a voltage signal is coupled into the tube grid. These two goals tend to direct you to go in opposite direction with that resistor's value.

... Looking at the LF frequency roll off due to RI (think this would usually be either a volume pot or pots tied to a tone stack) or Co (cathode cap). Reducing RI ... cuts ... by 3dB at 7Hz with a 1M ... versus 13.4 Hz with ... 500K ...

Co = "Capacitor, output" or a coupling cap.

Caps don't just magically cuts low frequency, all by themselves. Caps have a rising reactance (frequency-dependent analog to "resistance") as frequency goes down. If the C is in a circuit with an R, it will form a frequency-dependent voltage divider with the R. It might cut bass or treble depending on how the C & R are arranged, relative to each other.

With a coupling cap and a grid-reference resistor (the "RI" or "Resistance, input", which goes from tube grid to ground) making C or R smaller move the -3dB roll-off point higher. Usually, there is some other good reason(s) for making R the value it is, so we typically change the C.

... Not quite sure what the difference between bypassed cathode and unbypassed cathode only means. ...

If you have a cathode resistor running from tube cathode to ground, all by its lonesome, that is unbypassed.

Tube plate current due to applied signal also has to flow through that cathode resistor. But the resistor sets the tube's bias. And you should already know from Ohm's Law that changing current through a resistance results in a changing voltage drop across the resistor. All that means the tube's bias changes somewhat continually when a signal is applied.

Instead of seeing a bobbling voltage with your meter, you will find that the tube's gain is reduced from that which the math would predict for that gain stage. You have "local negative feedback due to an unbypassed cathode resistor". With a 12AX7 when using typical resistor values for plate and cathode resistors, gain will drop by about half (or something-other-than-half if you're not using a 12AX7 & typical parts).

A big-enough cap will bypass audio frequency current variations around the cathode resistor, restoring stage gain. Or you might use a "not quite big enough" cap to give full gain to highs while trimming bass at/below some frequency. The size of C to do this again depends on the size of R in this circuit (as a first-guess "R" is taken to equal the cathode resistor).

[Mike_J]

A grid stopper is about attenuating radio frequency (RF) interference; it ain't about tone.

It could be about tone is you purposely massively-oversize a grid stopper to shave treble, but few people do that outside of shred-machines. You will need to learn something about Miller Capacitance to sort out the difference.

Thanks HBP. Guess I will try to learn what I can about Miller Capacitance over the weekend. Ed mentioned he saw value taming high frequency at around 1.75K. May be looking at this entirely too simplistically but if my calculation is correct you could make the grid stopper 940K and achieve the attenuation starting at about 1.75K. Don't know if it is apples and oranges or not.


Thanks
Mike

[HotBluePlates]

... you could make the grid stopper 940K and achieve the attenuation starting at about 1.75K. ...

Got a tube amp? Got a 1MΩ? Put the 1MΩ between the grid and whatever you already have for the input jack.

I think you'll decide it's way too dark. It will also likely make the amp noisier. But tone is in the ear of the beholder...

[Mike_J]

... you could make the grid stopper 940K and achieve the attenuation starting at about 1.75K. ...

Got a tube amp? Got a 1MΩ? Put the 1MΩ between the grid and whatever you already have for the input jack.

I think you'll decide it's way too dark. It will also likely make the amp noisier. But tone is in the ear of the beholder...

Got a tube amp? Can't turn around without tripping over a tube amp. Isn't life grand. Thanks HBP, will probably try it one day to see what happens. Easy enough to reverse.


Thanks
Mike

[Mike_J]

Here is a picture of the completed jig for sampling plate load resistors, coupling caps, bypass caps and resistors. Hopefully will help me hear what the numbers are saying. Hope to get to Ed's level of understanding in the not too distant future.


Thanks
Mike

[Ed_Chambley]

Mike, my comments about frequency were not intended to be accomplished at one stage.  Again, I have already found what HBP stated and you will not like adding a 1 meg.


In a nutshell, here is how I work at tuning an amp.  First I decide what type of tone I want.  Clean and stark and I can do this with 1/2 if a 12A_7, Clean and swirly normally I will use a cathode follower if using a TMB.  If using what Duncan Calls a James I will use the second half of the 12Ax7 and use the first as a simple gain stage.  Then the other Half of the tube will usually be used to make a verb circuit with another tube.


Where HBP and PRR lead me was looking at the WHOLE circuit including the power section.  So for a nice clean full amp I normally steal from old Ampeg amps but I use cathode bias on the power tubes.  I find that boom is enhanced.  I really love to use a 6Sj7 in V1 (smokey tone) followed by the 12A_7 family for the second gain stage and to the tone stack.  Also I am careful of too much mids and to cover that I normally make either a Raw switch when using a James stack.  If using a TMB I simply use a larger mid pot.


Using the formula for frequency to determine a starting point for for bypass cap and building the amp in the manner usually when I fire it up it is very close.  I don't have too many amps I have built that do not have a cut control as I have found this very valuable in taming the highs.


Finally what I have found to really tailor an amp and finalize are different methods of Negative feedback and side chain compression.  I never considered the Negative Feedback much until I studied divider networks and ones I began to understand the extreme value in designing a good NFB circuit it changed my whole way of looking at amplifiers.


Since every room is different I have simply taken to using a locking multi-turn pot on every build I have.  I even figured out a way to do this on my Vintage amp collection so I have one on my 1968 Marshall.


Mike, my friend, I finally quit building a lot of different amps and just began ripping apart and reworking amps I am not happy with.  I just received a set of Iron from Edcor that I plan to build what I have been prototyping for well over a year.  I am finally happy with the results.  This is the dual 5879 preamp to KT66's.  Yes it will grind and scream, but it is the clean I am the most proud of.  There are characteristics in this tube and the EF86 that if you can tame it will make the little E and B strings have a brilliance without being harsh.


Also read up on using snubber caps.  you will see it across the PI in a log of LTPI designs, but it can be used many other places like on the plate of a plate driven tone stack.  Here you can shave some highs prior to tonestack and it will make the treble seem to have more usable sweep.


A little bit in a lot of places to me works much better that trying to kill frequencies at one stage.


All this and make sure you use the proper power cord!!!!    And above all, do one thing at a time and listen so you can understand what is happening.  I used to want to hurry up and get it built, now I really enjoy the process.


Also, I forgot to mention if I am building as shred machine I go in a completely different way and even use (sushhhhhhhhhhhhhh!) diodes.

[Ed_Chambley]

Here is a picture of the completed jig for sampling plate load resistors, coupling caps, bypass caps and resistors. Hopefully will help me hear what the numbers are saying. Hope to get to Ed's level of understanding in the not too distant future.


Thanks
Mike

My level is low, but thanks for the compliment.  I only did what you are doing.  Having the ability to switch caps and resistors will really get you going, but also can cause problems when you find 4 different settings you really like at the same stage.


It will help you a lot to understand what the changes really do.  You can read about it all day and still not know.  It is like someone telling you about an orange.  They can spend days telling you how they are grown, harvested, distributed and all kinds of things, but biting into it will tell you more.