Tools for fine tuning an amp.

[Mike_J]

Looks like I will finally get around to firing up the 5f6a rebuild. Was just thinking about tools to use to fine tune an amp. Have enclosed a box I used to knock the blocking distortion out of the 5E3 build which I now consider a great amp. It also includes a variable plate resistor and coupling capacitor tool that can be used for a variety of things in addition to fine tuning V1, using the coupling cap side and a separate 100K pot I place on a narrow strip of G10 with a thin cardboard backing to mark common slope resistor values for Marshall and Fender. Values of 33K, 56K and 100K are some I can think of. Can also compare different tone caps in the tone section. Have identified 12 pF cap sizes available from 47K to 130K. Will make a tool to knock out ice pick sound and harsh sounding overdrive.


Can't think of anything else right now that would be useful but would be happy to hear suggestions.


Thanks
Mike

[bmccowan]

Ears.

[Mike_J]

Ears.

That is the reason the tools are needed so as values are changed the impact on tone can be evaluated. In the case of the Bassman it has been established as a solid amp for many years. My major concerns are checking the 820r resistor V1-3. Will compare it to a 1.5K which would be a standard Fender value for a single triode cathode resistor. I changed the slope resistor to 100K from 56K. Also changed the bass tone cap from a .02uF to a .1uF. Tools will allow me to hear the difference between these components in a matter of seconds which is about the extent of my memory so i need the help the tools provide.

[PRR]

> My major concerns are checking the 820r resistor V1-3. Will compare it to a 1.5K which would be a standard Fender value for a single triode cathode resistor.

But this is not a single triode. It has a DC-connected cathode follower.

With the 1.5k and a *single* triode, we arrive near 70% across the triode. But with a direct coupled follower, that puts only 30% across the second triode, not enough for full swing. With 820 we get about 55%.

If you do the 1.5k this stage goes dirty VERY early. This is a "popular mistake".

[Mike_J]

> My major concerns are checking the 820r resistor V1-3. Will compare it to a 1.5K which would be a standard Fender value for a single triode cathode resistor.

But this is not a single triode. It has a DC-connected cathode follower.

With the 1.5k and a *single* triode, we arrive near 70% across the triode. But with a direct coupled follower, that puts only 30% across the second triode, not enough for full swing. With 820 we get about 55%.

If you do the 1.5k this stage goes dirty VERY early. This is a "popular mistake".

Thanks so much for the information. By dirty very early does that mean it would have less headroom and if so, could it be tuned somewhere in between to have a good combination of both headroom and overdrive without having to shake the roof off the house? Or is it even worth experimenting with?


Also, are you sure you aren't thinking about V2-3? It also has a 820r resistor to ground. As far as V1-3 is concerned on the original Fender 5f6a schematic both of the cathode resistors, V1-3 and V1-8 were shared by a 820r resistor. When Hoffman created his 5f6a layout he had the single triode use a 820r cathode resistor. Fender in the blackfaced era commonly would use a 1.5K resistor for a single triode which you know better than I do. That is the reason I suggested testing for a good resistor value.


Also, is there a percentage you shoot for to make sure you are getting full swing?


However, regardless, is V2-3 a good place to experiment for the reasons I mentioned earlier?


Have attached the schematic because I know how to do that now.


https://el34world.com/charts/Schematics/files/Fender/Fender_bassman_5f6a_schem.gif


Thanks
Mike

[Mike_J]

Just learned a new trick so had to show it off. Have far to go though I fear.


I am pretty confident the 820r can be played with on the V1a triode cathode. V1-8 has a 1.5Kr-22uF/50v going to ground. Not even a little bit confident in touching the V2a cathode resistor. Doesn't mean I wouldn't try if I thought there might be some value in it.


Thanks
Mike

[66Strat]

You might be interested in this.
https://www.ampbooks.com/mobile/classic-circuits/fender-bassman-5F6-A/

There is a cornucopia of information on Kuehnel's web page.

[Mike_J]

You might be interested in this.
https://www.ampbooks.com/mobile/classic-circuits/fender-bassman-5F6-A/

There is a cornucopia of information on Kuehnel's web page.

Based on that I should use a 1.64K resistor for V1A-3 since the cathodes have been split and a 125uF capacitor for that matter. Fender probably went with 1.5K because the common resistors were 1.5K and 1.8K if I remember correctly. Would probably need to spend $3.03 for a precision resistor if you wanted a 1.64K resistor. Fender wasn't that crazy. He wanted to make a profit. Nothing I am worried about. Don't build amps to make money. Build them because I hope to make something exceptional whether any of my amps attain exceptional status or not at least it won't be due to a lack of effort by me.
Didn't get to the V2B-3 resistor but will continue reading. Hope what he says makes sense to me.


Thanks
Mike

[Mike_J]

Something very interesting came up. He has reference for value calculations for a 12ay7. Depending on the voltage coming into the plate load resistor and the value of the plate load resistor it provides the cathode resistor rating if I am reading it correctly. Will definitely try that when I get the amp up and running.

[66Strat]

You may want to play around with this calculator. It shows the effect on voltage gain and frequency response of cathode resistor/and bypass capacitors for a variety of tubes. As I said before, Kuehnel has a ton of good stuff on his website.
https://www.ampbooks.com/mobile/amplifier-calculators/cathode-capacitor/

[Mike_J]

You may want to play around with this calculator. It shows the effect on voltage gain and frequency response of cathode resistor/and bypass capacitors for a variety of tubes. As I said before, Kuehnel has a ton of good stuff on his website.
https://www.ampbooks.com/mobile/amplifier-calculators/cathode-capacitor/

Think I may have used it. You put in the cathode resistor and capacitor values and RG which is the one meg potentiometer connected to the coupling cap and the plate load resistor value into the calculator. What it calculates is the gain at 82Hz (guitar low E) and the gain at 10KHz. Observation from playing with it is if you have a 250uF cathode cap, 1 meg pot and 100K plate load resistor then whether you use a 820r, 1.5K or 1.68K cathode resistor it makes no difference at all to the gain. So don't think I am going to spend any more time thinking about that. Lower cost 820r or 1.5K resistor will work just fine. Since the 820r is already in it I don't feel too inspired to change it at this time. However, when the plate load resistor is increased from 100K to 110K both gains go from 34.51 to 35.14, respectively.


If anyone has followed my ramblings I have mentioned that some tech on the internet has observed that amps that come into his shop with plate load resistors in the 110K range sound better to his ears than amps with 91K plate loads. I have some 110K Allen Bradley carbon comps which I will experiment with. They supposedly add sweetness to your tone a component of which is the amount of voltage dropped across the plate load resistor. Well a 110K plate load is going to drop a little more voltage than a 100K or 91K resistor would. Not by a lot I will admit but maybe the increase in gain and drop in voltage across the carbon comp resistor may explain what that tech has learned with his ears.

[Mike_J]

On the subject. Whether a 250uF or a 25uF cap is used for the V1-3 bypass cap it makes only a .01dB difference at 82Hz and no difference at all at 10KHz. That doesn't look like something to spend much time worrying about either.

[Mike_J]

Well, looks like we pretty much nailed down V1a and V1b and I haven't used my amp tool once. Would like to understand V2 and how the cathode follower circuit impacts that tube and the corresponding tone of the amp. Think based on an earlier post the PI is somewhat understood but I am guessing is how well the tubes are matched and other factors may be unknowns in what I know about balancing the PI circuit.

[kagliostro]

This is the only capacitor trimmer I remember it was used on an amp
note the trimmer connected on node A

 (OK, not an external unit to perform tuning)



https://el34world.com/Forum/index.php?topic=19970.msg208766#msg208766

Franco

[Mike_J]

This is the only capacitor trimmer I remember it was used on an amp
note the trimmer connected on node A

 (OK, not an external unit to perform tuning)



https://el34world.com/Forum/index.php?topic=19970.msg208766#msg208766

Franco

What do you think that trimmer is composed of? Doesn't seem like a resistor because that changes resistance. Don't know of a part that will change capacitance values outside a six or twelve position rotary switch with different value capacitors connected to it.


Thanks
Mike

[Mike_J]

These trimmers PRR referenced a number of years ago could be a good tool. They could take up above the 130pF max cap in the 12 position rotary switch. Could cover the pF range using a combination of these.


 https://www.rfparts.com/capacitors/capacitors-certrimmer/capacitors-46series.html

[HotBluePlates]

... Whether a 250uF or a 25uF cap is used for the V1-3 bypass cap it makes only a .01dB difference at 82Hz ...

IMO, the bypass capacitor value used for the input stage was not for Gain/EQ, but to guarantee that faulty tubes could not inject hum at the input stage of this bass amplifier.

I talked with an amp manufacturer who found they had a crate of preamp tubes with heater-to-cathode leakage, which resulted in audible hum if a 25µF bypass cap was installed.  Using a 250µF bypass nix the hum.

     25µF cap impedance at 60Hz: 106Ω
     250µF cap impedance at 60Hz: 10.6Ω

The bigger cap was a more-effective bypass to ground at the problem frequency (below a guitar's Low E).

[Mike_J]

... Whether a 250uF or a 25uF cap is used for the V1-3 bypass cap it makes only a .01dB difference at 82Hz ...

IMO, the bypass capacitor value used for the input stage was not for Gain/EQ, but to guarantee that faulty tubes could not inject hum at the input stage of this bass amplifier.

I talked with an amp manufacturer who found they had a crate of preamp tubes with heater-to-cathode leakage, which resulted in audible hum if a 25µF bypass cap was installed.  Using a 250µF bypass nix the hum.

     25µF cap impedance at 60Hz: 106Ω
     250µF cap impedance at 60Hz: 10.6Ω

The bigger cap was a more-effective bypass to ground at the problem frequency (below a guitar's Low E).

Thanks HBP. Useful information for sure. This would probably apply for the AB763 amps I would think. Perhaps most amps. None of us want to build an amp with excessive hum.


Thanks
Mike

[Mike_J]

You might be interested in this.
https://www.ampbooks.com/mobile/classic-circuits/fender-bassman-5F6-A/

There is a cornucopia of information on Kuehnel's web page.

I went to the 12ay7 calculator on the site above. Put in 325VDC as the voltage coming in to the plate resistor which comes off the schematic and the plate resistor as 100K and pressed calculate. Changed the AC load next stage to 1 meg which is the value of the volume pot and come up with a 2.174K cathode resistor. Shows DC plate current of 1.38mA and DC grid voltage of -3. Kuehnel calculated the values needed to obtain a 1.636K cathode resistor rating using the same voltage to the plate resistor of 325VDC and a 100K plate load resistor. In order to obtain that value using the same one meg for the AC load he used 1.65mA for the DC plate current and a DC grid voltage of -2.7. He chose the 1.636K because it approximates doubling of a 820r resistor.


My question is where did he get these values? It is my understanding the objective in determining a cathode resistor is to bias the triode in the middle of the spectrum. But I don't understand how these DC plate currents and DC grid voltages were derived and how they help us to bias the triode. Can someone enlighten me on this matter? Read the max values for each on the datasheet and they were considerably higher for each.


Thanks
Mike

[PRR]

> 325VDC and a 100K plate load resistor. ...he used 1.65mA for the DC plate current.... My question is where did he get these values?

Half of 325V is 162.5V. 162.5V in specified 100K resistor is 1.65mA (near enough).

I dunno about Richard, but Leo Fender clearly took numbers from tube-maker R-C Amplifier tables. Mostly G.E. (not RCA). Though sometimes he changed tubes without changing the resistors. (This is true for most 12AY7 amplifiers because the suggested conditions suck.)

[Mike_J]

> 325VDC and a 100K plate load resistor. ...he used 1.65mA for the DC plate current.... My question is where did he get these values?

Half of 325V is 162.5V. 162.5V in specified 100K resistor is 1.65mA (near enough).

I dunno about Richard, but Leo Fender clearly took numbers from tube-maker R-C Amplifier tables. Mostly G.E. (not RCA). Though sometimes he changed tubes without changing the resistors. (This is true for most 12AY7 amplifiers because the suggested conditions suck.)

Going by the 1.65mA with the other values being equal the 12AX7 should use a 1.2K cathode resistor according to my calculations. The 1.5K is somewhat safely in the middle I guess.


Thanks
Mike

[Mike_J]

> 325VDC and a 100K plate load resistor. ...he used 1.65mA for the DC plate current.... My question is where did he get these values?

Half of 325V is 162.5V. 162.5V in specified 100K resistor is 1.65mA (near enough).

I dunno about Richard, but Leo Fender clearly took numbers from tube-maker R-C Amplifier tables. Mostly G.E. (not RCA). Though sometimes he changed tubes without changing the resistors. (This is true for most 12AY7 amplifiers because the suggested conditions suck.)

Have been studying this a little more. Is it correct to presume you halved the 325V because it is feeding two triodes? I understand you then took the halved voltage divided it by the plate load resistor value to determine the amperage. That must be multiplied by 1,000 to arrive at the number of mA. Do you know how the DC Grid Voltage was calculated at -2.7V? Apparently the Fender measured voltage was -2.5V. The reason I am asking this is I want to calculate the grid voltage for the cathode presuming a 12AX7 was used. If I just put a 12AX7 in could I measure the DC voltage across the cathode resistor to determine the DC Grid Voltage?

[PRR]

> Is it correct to presume you halved the 325V because it is feeding two triodes?

? ? ? ? Two triodes? How is it splitting laterally?

You HAD the concept. "bias the triode in the middle of the spectrum." Maybe I only assumed you meant "halfway up the power supply".

> If I just put a 12AX7 in could I measure the DC voltage across the cathode resistor to determine the DC Grid Voltage?

An interesting question. You have to ask if the grid is at zero (usually is), and do voltages add-up somehow. Think of it a while.

[Mike_J]

> Is it correct to presume you halved the 325V because it is feeding two triodes?

? ? ? ? Two triodes? How is it splitting laterally?

You HAD the concept. "bias the triode in the middle of the spectrum." Maybe I only assumed you meant "halfway up the power supply".

> If I just put a 12AX7 in could I measure the DC voltage across the cathode resistor to determine the DC Grid Voltage?

An interesting question. You have to ask if the grid is at zero (usually is), and do voltages add-up somehow. Think of it a while.

Was and still am confused about how you went from 325v to 162.5v. Read the 5f6a schematic and the negative voltage is read before the cathode cap and resistor even come into play. Should have looked at the schematic before I posted. Will look at the calculator again to let you know what the advantage is, which may be a disadvantage based on the tone you want, of balancing the cathode section.


Thanks
Mike

[Mike_J]

> Is it correct to presume you halved the 325V because it is feeding two triodes?

? ? ? ? Two triodes? How is it splitting laterally?

You HAD the concept. "bias the triode in the middle of the spectrum." Maybe I only assumed you meant "halfway up the power supply".

> If I just put a 12AX7 in could I measure the DC voltage across the cathode resistor to determine the DC Grid Voltage?

An interesting question. You have to ask if the grid is at zero (usually is), and do voltages add-up somehow. Think of it a while.

Was and still am confused about how you went from 325v to 162.5v. Read the 5f6a schematic and the negative voltage is read before the cathode cap and resistor even come into play. Should have looked at the schematic before I posted. Will look at the calculator again to let you know what the advantage is, which may be a disadvantage based on the tone you want, of balancing the cathode section.


Thanks
Mike

By saying bias the triode in the middle of the spectrum I was crudely referring to biasing the triode between cutoff and the other end, whatever they call it.

[Mike_J]

> Is it correct to presume you halved the 325V because it is feeding two triodes?

? ? ? ? Two triodes? How is it splitting laterally?

You HAD the concept. "bias the triode in the middle of the spectrum." Maybe I only assumed you meant "halfway up the power supply".

> If I just put a 12AX7 in could I measure the DC voltage across the cathode resistor to determine the DC Grid Voltage?

An interesting question. You have to ask if the grid is at zero (usually is), and do voltages add-up somehow. Think of it a while.

Now I am reading I can't balance my PI without a matched tube in V3, the PI tube. Just what I need another thing to figure out.


Thanks
Mike

[acheld]

Now I am reading I can't balance my PI without a matched tube in V3, the PI tube.

Where'd you read that?   New to me -- I never heard of this.

[Mike_J]

Now I am reading I can't balance my PI without a matched tube in V3, the PI tube.

Where'd you read that?   New to me -- I never heard of this.

I did today and it was on the internet so you know it had to be true. I don’t even know what a matched tube is.

[Mike_J]

It was balanced triodes? The Gear Page where I heard about needing balanced tubes in the PI. New to me as well. Seems like we would be better off with a 10K pot and balance the 100K and 82K resistors after we get close with the tail resistor and replace them with the modified values. Would be a close cousin of what Dumble did with his trimmer pots but with fixed resistor values. If the PI tube theory is correct it would have to be redone every time the PI tube gets changed.

[HotBluePlates]

It was balanced triodes? The Gear Page where I heard about needing balanced tubes in the PI. ...

"Balanced Triodes" is highly overrated.  Tube vendors charge more money for "balanced triodes" because not every tube they get will have balanced triodes.  Buyers see a higher price and assume "costs more = more better."

I mostly spend time at The Gear Page now.  The voices there range from "Very Well Informed" (including designers & manufacturers of excellent amps & other gear) to... ahem...  "Less-Well Informed."  The reader can't assume all voices are equally credible.

At the bottom is a set of triode curves and some plotted loadlines.  The tube stage investigated had a supply of 360v.  Knowing nothing else beyond "it's probably good to idle somewhere in the middle of the loadline," we could guess 360v/2 = 180v as our plate voltage at idle.

When we know a bit more, we find that grid-to-cathode voltage of 0v is the guaranteed "onset of heavy distortion" absent a special circuit driving the tube-stage being investigated.  The Red loadline in the image seems to cross the 0v grid curve (where "grid-to-cathode voltage" equals 0v) at about 60v on the plate.

360v supply minus 60v (at the 0v curve) = 300v.  300v / 2 = 150v (above the minimum plate voltage of 60v).  150v + 60v = 210v.

But the grid curves get bunched up in the area near high plate voltage & low plate current.  That would mean distortion, and some will tend to stay away from that region.  Talking about where the Red line goes above ~300v.

Okay, so now we have a working range bounded by 60v on the plate at one end, and ~300v on the other.  That's 300v - 60v = 240v, and half of that (to bias in the middle) is 120v above the bottom of the range ===> 60v + 120v = 180v


We started with throwing a dart at "one-half the supply voltage" (180v), refined that by noting a lower-bound (210v), and refined further by noting an upper bound (180v) to arrive at the same place we started (mostly by coincidence than any "rule").

Generally, after accounting for any bounds on "acceptable operating area" when you want the largest possible output swing, and the stage in question is Class A, you shoot for something near half the available voltage-swing, or half the available current-swing.

These are some foundational concepts, but they're easy to breeze past when going down the rabbit hole with every element of an amp circuit.

[Mike_J]

It was balanced triodes? The Gear Page where I heard about needing balanced tubes in the PI. ...

"Balanced Triodes" is highly overrated.  Tube vendors charge more money for "balanced triodes" because not every tube they get will have balanced triodes.  Buyers see a higher price and assume "costs more = more better."

I mostly spend time at The Gear Page now.  The voices there range from "Very Well Informed" (including designers & manufacturers of excellent amps & other gear) to... ahem...  "Less-Well Informed."  The reader can't assume all voices are equally credible.

At the bottom is a set of triode curves and some plotted loadlines.  The tube stage investigated had a supply of 360v.  Knowing nothing else beyond "it's probably good to idle somewhere in the middle of the loadline," we could guess 360v/2 = 180v as our plate voltage at idle.

When we know a bit more, we find that grid-to-cathode voltage of 0v is the guaranteed "onset of heavy distortion" absent a special circuit driving the tube-stage being investigated.  The Red loadline in the image seems to cross the 0v grid curve (where "grid-to-cathode voltage" equals 0v) at about 60v on the plate.

360v supply minus 60v (at the 0v curve) = 300v.  300v / 2 = 150v (above the minimum plate voltage of 60v).  150v + 60v = 210v.

But the grid curves get bunched up in the area near high plate voltage & low plate current.  That would mean distortion, and some will tend to stay away from that region.  Talking about where the Red line goes above ~300v.

Okay, so now we have a working range bounded by 60v on the plate at one end, and ~300v on the other.  That's 300v - 60v = 240v, and half of that (to bias in the middle) is 120v above the bottom of the range ===> 60v + 120v = 180v


We started with throwing a dart at "one-half the supply voltage" (180v), refined that by noting a lower-bound (210v), and refined further by noting an upper bound (180v) to arrive at the same place we started (mostly by coincidence than any "rule").

Generally, after accounting for any bounds on "acceptable operating area" when you want the largest possible output swing, and the stage in question is Class A, you shoot for something near half the available voltage-swing, or half the available current-swing.

These are some foundational concepts, but they're easy to breeze past when going down the rabbit hole with every element of an amp circuit.

Thanks HBP. I may be able to understand this if I go over it a few times. I spend a lot of time in rabbit holes. Find it hard to expand your knowledge if some time isn't spent there. Thinking of just using some of the tools I have developed to find what sounds best to me by changing resistor and capacitor values and finding what sounds best to me. Then play the amp at night in the dark to see if any plates are getting red.

Thanks
Mike

[Mike_J]

Here is the latest tool I am planning to use. It simply has 11 different pF capacitors in it. Developed it because I believe Aiken's said ice pick and harsh overdrive can be cured by increasing the cap across the PI plate caps from 47pF to something below 100pF hopefully because at that point it becomes audible. Also, Rob Robinette mentioned a 100pF cap could be put across the resistor going from the output jack tip and the CW of the presence pot in the 5f6a Bassman to accomplish the same thing. Merlin said amp hiss can be reduced by tying a pF cap from the preamp tube grid to ground so this is another application although this tool may not have all values needed for the last item mentioned.


It is a very simple tool to be used for tuning purposes. Neighbors roofers haven't started yet so I hacked up an old hurricane shutter that never gets used for the faceplate.


Thanks
Mike

[Mike_J]

Ordered polyester 250VDC radial caps from .01 to 1uF to put in my next tool. The 250VDC value should be good to check for different values in the tone stack as well as for coupling caps. Plan on using the last 250VDC switch and a sour cream container to make it. Will also use the variable resistor tool pictured below. If I can successfully replace the 25KL presence pot it will be added to the tool along with a 10KL pot. Mainly designed this to test various slope resistor and presence pot values. Can combine them to test the .02/56K versus .1/100K arrangement for the bass control in near real time.