2nd Gen Build

[dbharris]

Thinking about this more and running some numbers here is what I've come up with. Voltage dividers are linear (though a pot may have a different taper) meaning if you think of the divider as attenuating a certain percentage of signal voltage, it does not matter if that input is 10V or 50V, the ratio is the same. I understand that current flowing through the resistors will dictate different wattage ratings.

In the traditional Dumble design of the ratio control there is a fixed series resistor connect to a pot wired as a voltage divider. The more resistance between ground and the wiper, the more signal that flows (less attenuation). The resistance of the series resistor will impact maximum limits. The resistance between the wiper and the input lug of the pot is essentially added to the fixed series resistor to become the full series resistance in the divider.

In most 4th Gen/5th Gen amps that we know of you will find a 150K series resistor connect to a 100K pot. If you assume 10V input (for easy math), then with the largest amount of resistance between ground and the wiper (100K), 4V will pass. Ratio on "10".
With 75%, 3V; 50%, 2V; and 25%, 1V.

The ratio control on my #102 lives between 25%-50%.

For the stock 2nd/3rd Gen designs, the 1M/180K network would output 1.525V from that same 10V input.

My plan is to keep the 1M resistor as the series resistor and then wire a 250K pot as a voltage divider. This would yield: 100%, 2V; 75%, 1.5V; 50%, 1V; and 25%, 0.5V.

That would spread the useable range over the whole rotation with the stock value being about 3/4 of the rotation.

Does this seem correct in theory?

-Dan

[ijedouglas]

Hey Dan,

Regarding your tone stack being underwhelming, check your slope resistor. You have a 100R instead of a 100K.

Credit goes to Jelle for spotting it

[dbharris]

Hey Dan,

Regarding your tone stack being underwhelming, check your slope resistor. You have a 100R instead of a 100K.

Credit goes to Jelle for spotting it

Thank you Ian and Jelle! I will get that swapped out before any other changes. You guys are awesome!

-Dan

[martin manning]

Thinking about this more and running some numbers here is what I've come up with. Voltage dividers are linear (though a pot may have a different taper) meaning if you think of the divider as attenuating a certain percentage of signal voltage, it does not matter if that input is 10V or 50V, the ratio is the same.

Correct.

In the traditional Dumble design of the ratio control there is a fixed series resistor connect to a pot wired as a voltage divider. The more resistance between ground and the wiper, the more signal that flows (less attenuation). The resistance of the series resistor will impact maximum limits. The resistance between the wiper and the input lug of the pot is essentially added to the fixed series resistor to become the full series resistance in the divider.

Correct.

In most 4th Gen/5th Gen amps that we know of you will find a 150K series resistor connect to a 100K pot. If you assume 10V input (for easy math), then with the largest amount of resistance between ground and the wiper (100K), 4V will pass. Ratio on "10".
With 75%, 3V; 50%, 2V; and 25%, 1V.

The ratio control on my #102 lives between 25%-50%.

So if I split the difference at 37.5%, that would be 37.5k out of 100k + 150k, 15% or -16.5 dB wrt the OD output at the master volume input lug.
This setting will depend on where the OD drive is set since that will affect the absolute level at the OD output.
In all ODS, the Clean output feeds the master volume directly, with no attenuation.

For the stock 2nd/3rd Gen designs, the 1M/180K network would output 1.525V from that same 10V input.

The 2nd Gen has a 180k to ground, and then a 1M series resistor feeding a 1M Level pot, which functions as the master volume. That means the signal hitting the master is at 50% of the OD output level, -6 dB. The load is 180k//2M, or 165k, which is probably important to the 2nd Gen character.

If you were to run the OD output direct to a 250k pot, you could go to from -0 dB attenuation to as much as you want, with the load being 190k at 50% Ratio, where the signal hitting the master is at -6 dB like the fixed divider.

Note all of the above ignores the Zo of the OD second stage (39k), and the Zin of the PI (~2M), but those are common to both volume systems.

[dbharris]

Hey Dan,

Regarding your tone stack being underwhelming, check your slope resistor. You have a 100R instead of a 100K.

Credit goes to Jelle for spotting it

Thank you Ian and Jelle! I will get that swapped out before any other changes. You guys are awesome!

-Dan

Swapped in the proper slope resistor and also switched the outside foil orientation for the bass cap. A little ugly with the RTV now but should work fine. Too late to test here tonight.

Thanks again Ian and Jelle.

-Dan

[dbharris]

Thinking about this more and running some numbers here is what I've come up with. Voltage dividers are linear (though a pot may have a different taper) meaning if you think of the divider as attenuating a certain percentage of signal voltage, it does not matter if that input is 10V or 50V, the ratio is the same.

Correct.

In the traditional Dumble design of the ratio control there is a fixed series resistor connect to a pot wired as a voltage divider. The more resistance between ground and the wiper, the more signal that flows (less attenuation). The resistance of the series resistor will impact maximum limits. The resistance between the wiper and the input lug of the pot is essentially added to the fixed series resistor to become the full series resistance in the divider.

Correct.

In most 4th Gen/5th Gen amps that we know of you will find a 150K series resistor connect to a 100K pot. If you assume 10V input (for easy math), then with the largest amount of resistance between ground and the wiper (100K), 4V will pass. Ratio on "10".
With 75%, 3V; 50%, 2V; and 25%, 1V.

The ratio control on my #102 lives between 25%-50%.

So if I split the difference at 37.5%, that would be 37.5k out of 100k + 150k, 15% or -16.5 dB wrt the OD output at the master volume input lug.
This setting will depend on where the OD drive is set since that will affect the absolute level at the OD output.
In all ODS, the Clean output feeds the master volume directly, with no attenuation.

For the stock 2nd/3rd Gen designs, the 1M/180K network would output 1.525V from that same 10V input.

The 2nd Gen has a 180k to ground, and then a 1M series resistor feeding a 1M Level pot, which functions as the master volume. That means the signal hitting the master is at 50% of the OD output level, -6 dB. The load is 180k//2M, or 165k, which is probably important to the 2nd Gen character.

If you were to run the OD output direct to a 250k pot, you could go to from -0 dB attenuation to as much as you want, with the load being 190k at 50% Ratio, where the signal hitting the master is at -6 dB like the fixed divider.

Note all of the above ignores the Zo of the OD second stage (39k), and the Zin of the PI (~2M), but those are common to both volume systems.

Thank you Martin. I've read this a few times and I think it is starting to sink in. I was not considering the master volume at all since in my thinking it was a constant in both scenarios.

Will test to confirm everything works tomorrow after my changes tonight, then set the PI trimmer, and then measure voltages. After that I will experiment with a ratio control.

-Dan

[dbharris]

Forgot to pay the "tax" with an updated gut shot:

PXL_20240517_023439314.jpg

[Dr d]

Forgot to pay the "tax" with an updated gut shot:
PXL_20240517_023439314.jpg

Your tone controls should work now!

[ijedouglas]

Forgot to pay the "tax" with an updated gut shot:
PXL_20240517_023439314.jpg

Any difference in the tone stack?

[dbharris]

Forgot to pay the "tax" with an updated gut shot:
PXL_20240517_023439314.jpg

Your tone controls should work now!

Forgot to pay the "tax" with an updated gut shot:
PXL_20240517_023439314.jpg

Any difference in the tone stack?

Yes, had a quick test this morning. Treble and middle function as expected and bass too (with the limited range of the James configuration). Will likely lift a lead on that ceramic cap and see if I like it better.

Seemed like I had a better balance between the channels too but that required maxing out the volume and keeping overdrive around 50%. Was a low dB test though and may not be reflective of normal playing levels.

More fun later tonight after work....

-Dan

[martin manning]

Seemed like I had a better balance between the channels too but that required maxing out the volume and keeping overdrive around 50%. Was a low dB test though and may not be reflective of normal playing levels.
More fun later tonight after work....

I ran a simulation including source impedance for OD2, 10n coupling cap, and load impedance for the PI (2M). I found the best match is a 250kB Ratio pot with a 560k across the element. With the pot at noon, the load, frequency response, and level are essentially the same as for the fixed divider. The level/ratio can then be adjusted up or down as intended. The load changes slightly with ratio, but only +/-5K with +/- 20% rotation of the pot from center. This will be useful for anyone wanting to build a 2nd Gen in a 4th/5th Gen chassis having two OD controls, or for adding a ratio trimmer or control to a 2nd Gen build. It's in a very accessible spot on the preamp board, easy to add and to experiment with values

[dbharris]

Seemed like I had a better balance between the channels too but that required maxing out the volume and keeping overdrive around 50%. Was a low dB test though and may not be reflective of normal playing levels.
More fun later tonight after work....

I ran a simulation including source impedance for OD2, 10n coupling cap, and load impedance for the PI (2M). I found the best match is a 250kB Ratio pot with a 560k across the element. With the pot at noon, the load, frequency response, and level are essentially the same as for the fixed divider. The level/ratio can then be adjusted up or down as intended. The load changes slightly with ratio, but only +/-5K with +/- 20% rotation of the pot from center. This will be useful for anyone wanting to build a 2nd Gen in a 4th/5th Gen chassis having two OD controls, or for adding a ratio trimmer or control to a 2nd Gen build. It's in a very accessible spot on the preamp board, easy to add or to experiment with values

Thank you so much Martin, I will make this mod and report back. Need to order a couple things first.

-Dan

[Matt J]

Seemed like I had a better balance between the channels too but that required maxing out the volume and keeping overdrive around 50%. Was a low dB test though and may not be reflective of normal playing levels.
More fun later tonight after work....

I ran a simulation including source impedance for OD2, 10n coupling cap, and load impedance for the PI (2M). I found the best match is a 250kB Ratio pot with a 560k across the element. With the pot at noon, the load, frequency response, and level are essentially the same as for the fixed divider. The level/ratio can then be adjusted up or down as intended. The load changes slightly with ratio, but only +/-5K with +/- 20% rotation of the pot from center. This will be useful for anyone wanting to build a 2nd Gen in a 4th/5th Gen chassis having two OD controls, or for adding a ratio trimmer or control to a 2nd Gen build. It's in a very accessible spot on the preamp board, easy to add and to experiment with values

That's awesome, Martin! Thank you for doing this!
I'm taking a break from my own 2nd Gen hybrid right now to work on something else but I will definitely look at implementing this when I get back to work on it.
Matt J.

[dbharris]

While waiting on some parts to arrive I "balanced" the PI. Scare quotes because with the 5K trimmer I installed and the tubes in the amp, I could get close using the ampeg method but not all the way to a 0v offset. It settled around 1.7. I do have another 12AT7 and may try that to see if it gets me closer. Since my v3 plate resistors are the same nominal resistance, I added the trimmer mainly to achieve the difference there HAD designed into the circuit.

I'm calling that good enough for Rock and Roll in a 2nd Gen. My assumption being that HAD did not yet implement the trimmer in his design and was probably hand selecting tubes to get closer to a balanced power amp than a typical guitar amp of the day.

I do still need to set the FET output for 10dB voltage gain. I think I'm close by ear.

I also pulled voltages.

Line: 120.8Vac
Heaters: 6.68Vac
B+1: 448V
B+2: 447V
B+3: 423V
B+4: 278.3V
B+5: 269.5V

V5: 43.2ma (61%)
V4: 41.4ma (58%)
V3: 1- 289.2V 3&8- 116.5V 6- 276V
V2: 1- 187.3V 3- 1.36V 6-188.4V 8-1.34V
V1: 1- 182.7V 3- 1.26V 6-179.7V 8-1.31V

Haven't really rolled preamp tubes yet.

-Dan

[Dr d]

Hi Dan, If I am reading your voltages correctly, 289.2v and 276v on the PI V3 appear way too far apart. With a balanced 12AT7 and no trimmer you should be getting roughly no more than about 6 or 7 volts apart. Everything else looks good.