UltraPhonix with a few twists

[norburybrook]

Fred, that is indeed a huge difference. On one of my Ods amps with a 12at7 in the PI position I measured the signal on the input of the PI and the Speaker output. Right before clipping a 8,5v PP sine wave signal generated a clean sine wave at the output.
A bluesmaster PI had less headroom.

I love a good mystery so the BM channel at 5v is in the ball park and the tweed is obviously hitting the PI to 'hard' . Can you use a voltage divider somewhere to tame that?

https://www.allaboutcircuits.com/tools/ ... alculator/




M

[erwin_ve]

Fred, that is indeed a huge difference. On one of my Ods amps with a 12at7 in the PI position I measured the signal on the input of the PI and the Speaker output. Right before clipping a 8,5v PP sine wave signal generated a clean sine wave at the output.
A bluesmaster PI had less headroom.

I love a good mystery so the BM channel at 5v is in the ball park and the tweed is obviously hitting the PI to 'hard' . Can you use a voltage divider somewhere to tame that?

https://www.allaboutcircuits.com/tools/ ... alculator/

M

We love pots, dont we?

[norburybrook]

Or a split plate I found this having a google

take the plate resistor on probably the last stage of the preamp, then use two resistors instead of one. If it's 100k, use 2 resistors that add up to 100k or close. Then the coupling cap is now soldered to the junction of those two resistors. Whichever resistor ia the one on the plate side is the one that sets how much gain is dropped. For example, is you used a 22k and a 82k with the 22k being the one that's connected to the plate, that will drop gain much less that if the 82k is the one attached to the plate. So whatever the resistor thats on the plate side, the higher it's value the more gain dropped.

So the tube is still working with a 100k plate, but the signal leaving that stage can be knocked down depending on the resistor values just a tad, or a lot. You could also use a pot to see where you like it then measure it and put resistors of those values in. Or just start with say a 68k and a 33k. That way you can try it with the 33k towards the plate and if thats not enough turn it around. That should give you any idea what value you need if neither of those is quite right.

[erwin_ve]

Marcus didnt you do that on one of your trainwreck/komet builds?

[norburybrook]

Marcus didnt you do that on one of your trainwreck/komet builds?

exactly, worked really well , you have a good memory.

M

[fred.violleau]

So that would do something like this:

IMG_20190719_103556.jpg

And I may use a pot instead of one of the resistor so I can adjust the output value.

Fred.

[fred.violleau]

Woops nope series vs parallel.
I will put the resistor in series.

Fred.

[fred.violleau]

More like this:

IMG_20190719_104602.jpg

Fred

[fred.violleau]

Assuming I am using a pot, do I need to bleed the leftover signal to ground?

Fred

[norburybrook]

More like this:

IMG_20190719_104602.jpg

Fred

that looks right . Erwin will probably be able to qualify that better than me.

be interested in your findings. I'm sure it will change the tone slightly as well as all these things tend to do

split_load_switch.JPG

https://ampgarage.com/forum/viewtopic.php?t=1202

m

[erwin_ve]

More like this:

IMG_20190719_104602.jpg

Fred

For the pot position I would use it after the coupling cap.
A higher wattage trimpot as a plate is possible, a chassis mount pot with these high voltages is stressing your guardian angel Imo.
When doing a trimpot as a partial plate you can wire it like a variable resistor.

[fred.violleau]

Gotcha Erwin_we, security first
Thanks for the drawing Marcus!

I will probably add a pot after the coupling cap.
AFTER this extra pot (variable resistor) are the mixing resistors (220k on each side).This is where the TD and BM channel mix to feed the D-lator.
It would be fairly easy to change one of these resistors.
My question is : do the resistor need to be balanced on both sides as they are also acting as a voltage divider right?

Fred.

[norburybrook]

Gotcha Erwin_we, security first
Thanks for the drawing Marcus!

I will probably add a pot after the coupling cap.
AFTER this extra pot (variable resistor) are the mixing resistors (220k on each side).This is where the TD and BM channel mix to feed the D-lator.
It would be fairly easy to change one of these resistors.
My question is : do the resistor need to be balanced on both sides as they are also acting as a voltage divider right?

Fred.

if the resistors are the same value then you get a 50% drop in gain with a voltage divider I seem to remember.

M

[talbany]

Fred
A few thoughts. Check the (bias) current draw on the tubes on both Tube and SS rectifier settings and make sure they don't red plate when you go to the SS setting.
If you don't ground out the channel you are not using the signal will bleed through the mixers and ghost note on you..
If you already know this carry on!

Tony

[fred.violleau]

Fred
A few thoughts. Check the (bias) current draw on the tubes on both Tube and SS rectifier settings and make sure they don't red plate when you go to the SS setting.

The PT tubes used to red plate when I was running all tubes on empty (no board to draw current).
I have not checked since, but the smell of the PT tube red plating has disappeared
I will take measures to make sure there is no problem on that side too.

If you don't ground out the channel you are not using the signal will bleed through the mixers and ghost note on you..
If you already know this carry on!

At first I was running the two channels feeding the D-lator without mixing resistors, and the amp was oscillating, especially if I did try to shut down one channel (all pots at 0). As soon as I would open the knobs just a bit, the oscillation would go away.. I put in 220k mixer resistors, but it seems like the tweed channel still has to much output. Marcus and Erwin suggested that I tried to lower the output of the tweed channel with either another resistor in serie after the V2b coupling cap, and before the mixer resistors.

It is amazing how two channels output can be that different! I checked on the Charlie Wilson's Tweedle Dee layout and voltages and output measures do fit with what I get. but then there is a 1meg resistor before the signals enter the PI..

I added a 220k 2w resistor after the coupling cap of V2b for the tweed channel, and it made a little improvement, but not enough. The channel still has some distortion going on.
I will give a try at the 470k for both mixing resistors and see how it goes.

I though about using another relay in order to ground the signal of the channel that is not being used instead of using mixer resistors.
It would be synchronized with the input relay.
I know @Sluckey is using a switch on his dual marshall amp to ground the unused signal but only after the first stage of amplification.
Would this avenue allow me to tweek the two channels output independently?

Thanks for the help and suggestions!

Fred.