Tweet
Hi,
I worked on a new amp project and I would like to share the way I design an study the conception of my amps.
At first we abord the preamp design.
I'm looking for a very dynamic preamp but in the same time which can achieve warm and smooth distortion.
It will be a two stage preamp with two 12AX7, four triodes.
It's the opportunity to abord most of the different ways to polarize triode valves (12A*7 / ECC8* ... )
You will recognize in the end, the kind of amp that I'm inspired.
Now, to forward step by step we interest of V2a valve which is the first triode of preamp second stage.
It's a very simple and traditional triode design with a plate resistor of 100k, a cathode resistor of 1,5k supply by a DC voltage of 280V. It's a classe A design with traditional cathode BIAS.
Take a look on the loadline.
What can we see ?
A is the lower point when the triode drain no current, it's a blocking point.
That's the first point I draw, easy to find.
B is the theorical point when the tube drain maximum current.
The math is very simple, we have U and we have Rp and I = U/R = 280 / 100k = 2,8mA
With these two points it's easy to draw the loadline (in red on the picture).
Each crossing line points help to achieve the tranfer caracteristic line.
At the point, it's easy to chose a cathode resistor value.
I'm not looking for the most linear result, in guitar amp design, we are looking for harmonic colourings.
However I choose a "center point" around 1,5V with a current around 1mA. (there are different ways to achieve nice harmonic colourings and distortion, we will see this point belowe)
R = U / I = 1,5 / 0.001 = 1500 Ohms
With the loadline we can easely see that Ua = 180V and Uk = 1,5V for a consumption of 1mA, that's the bias point of V2a.
We can also see that the real max current value drain by our triode is around 1,85mA at Ug1 = 0V.
Until now, we are in a very traditional way.
now, we take a look on the V1 valve of the amp, I have to tell you that it's a double triode design which share the same plate resistor and the same cathode resistor too.
Here are the values we need to know :
U=280V
Rp=220k
Rk=1,5k
We have to understand and to concider that each components are crossing by the current of the two triodes.
Thus each component values can be understood as to be two time more important.
Therefore to draw the loadline we use this two new relative values, Rp=440k and Rk=3k.
Here is the loadline :
The loadline seem to be very "flat" and that's true !
The double triode design is very interesting for a V1 position because of its very good dynamic and headroom.
We most work on voltage variation which is the dynamic spring in tube amp.
We can easely find the gain ratio.
For grid 1V Voltage we find anode Voltage of 95V
For grid 2V Voltage we find anode voltage of 170V
So I can say that for a changing of 1V on the grid that result 170-95 = 75V of amplifcation.
The Facto the gain ratio is 75.
We can compare it with the V2a gain (see above).
Ug1=1V => Ua=150V
Ug1=2V => Ua=210V
210-150=60
So the Double Triode design give us a dynamic of 20% better than the single triode design.
Very good thing for the first stage of our preamp, isn't it !
The next time we will take a look on V2b which is a DC cathode follower site between V2a and the tone stack.
We will try to understand how it's a very interesting design for achieve smooth an warm overdrive.
Let's talk about the DC Cathode Follower which it's the last triode of our preamp design (V2b).
You remember that we already study the first triode of our preamp second stage (V2a see above).
I have to precise that it's the stage which achieve good preamp distortion.
Thus this is a good erea to use a DC cathode follower.
Here is a vue of how it looks like.
We can easely recognize the cathode follower, the triode without plate resistor and its only resistor on the cathode which is the point we will recover the signal.
Here are the values we have to know :
U=280V
Rk=56k
Now take a look on the very interesting loadline.
Because of the no plate resistor design, the tube works differently.
The plate voltage is constent but the cathode works like it was the plate.
This why we have to reverse the voltage on the up side of the graph.
We see that the grid is in a positive voltage and we understand that it's going to tend towards zero volt and drain current.
On the loadline, I colored it in brown, we can see that this current will tend towards 0,5mA.
This current will come from the plate resistor of V2a which will create a SAG and by this way low down the plate voltage of V2a.
U = RI = 100k x 0,0005 = 50V
Thus at first we see that our DC cathode follower will works to achieve more distortion from V2a.
But we can also see on the loadline that the tube is "locked" on 0V therefore all positive voltage will be very nicely compress.
What could be see as a very bad design for an audiophile user is a very good way to achieve smooth and warm distortion for a guitar player.
Now the first part of the preamp design is done.
I can know the preamp consumption which will be necessary for the further power supply design.
Iv2a = 1mA
Iv2b = 3,2mA
Iv1 = 0,75mA
Ipreamp = 4,95mA
It's time to talk about the voicing.
It's important to know the power amp tube brand.
Each tube has its own voice and that a point who has to be take in concideration very seriously.
My power amp will run with EL84 tubes which have a very round and medium tone.
In the same time I'm looking for a very simple tone stack with only a bass and treble pots.
I will adjust the tone for a Vox / Matschless voicing but with my taste.
Vox Top Boost channel tone stack is very accurate.
For best gain capability we use a large and usual cathode capacitors value of 22µf that give us a cut frequency of 4,8Hz.
F= 1/(2pi x Ck x Rk)= 1/(6,28 x 0,000022 x 1500) = 4,8Hz
In the same time, we work on the design of a Rock amp and want to achive very nice distortion tone.
We all know that distortion doesn't like too much bass in the voicing.
We can low down the bass level with the coupling capacitor between V1 and V2a which is going to help us in the same time to protect V2a of blocking distrotion.
Because of the very medium tone of the EL84 and the tone stack simulation I choose a cut frequency of 700Hz.
C = 1/(2pi x F x R) = 1/(2pi x F x R) = 1/(6.28 x 700 x 220k) = 1,034nF
1nF will be my coupling capacitor between the first and the second preamp stage.
Well, now my preamp is becoming almost totaly design.
It's missing the gain pot value, and I choose a 500klog for not overload the V2a grid and adding a 180p bright cap for very nice sparkle clean tone.
I worked on a new amp project and I would like to share the way I design an study the conception of my amps.
At first we abord the preamp design.
I'm looking for a very dynamic preamp but in the same time which can achieve warm and smooth distortion.
It will be a two stage preamp with two 12AX7, four triodes.
It's the opportunity to abord most of the different ways to polarize triode valves (12A*7 / ECC8* ... )
You will recognize in the end, the kind of amp that I'm inspired.
Now, to forward step by step we interest of V2a valve which is the first triode of preamp second stage.
It's a very simple and traditional triode design with a plate resistor of 100k, a cathode resistor of 1,5k supply by a DC voltage of 280V. It's a classe A design with traditional cathode BIAS.
Take a look on the loadline.
What can we see ?
A is the lower point when the triode drain no current, it's a blocking point.
That's the first point I draw, easy to find.
B is the theorical point when the tube drain maximum current.
The math is very simple, we have U and we have Rp and I = U/R = 280 / 100k = 2,8mA
With these two points it's easy to draw the loadline (in red on the picture).
Each crossing line points help to achieve the tranfer caracteristic line.
At the point, it's easy to chose a cathode resistor value.
I'm not looking for the most linear result, in guitar amp design, we are looking for harmonic colourings.
However I choose a "center point" around 1,5V with a current around 1mA. (there are different ways to achieve nice harmonic colourings and distortion, we will see this point belowe)
R = U / I = 1,5 / 0.001 = 1500 Ohms
With the loadline we can easely see that Ua = 180V and Uk = 1,5V for a consumption of 1mA, that's the bias point of V2a.
We can also see that the real max current value drain by our triode is around 1,85mA at Ug1 = 0V.
Until now, we are in a very traditional way.
now, we take a look on the V1 valve of the amp, I have to tell you that it's a double triode design which share the same plate resistor and the same cathode resistor too.
Here are the values we need to know :
U=280V
Rp=220k
Rk=1,5k
We have to understand and to concider that each components are crossing by the current of the two triodes.
Thus each component values can be understood as to be two time more important.
Therefore to draw the loadline we use this two new relative values, Rp=440k and Rk=3k.
Here is the loadline :
The loadline seem to be very "flat" and that's true !
The double triode design is very interesting for a V1 position because of its very good dynamic and headroom.
We most work on voltage variation which is the dynamic spring in tube amp.
We can easely find the gain ratio.
For grid 1V Voltage we find anode Voltage of 95V
For grid 2V Voltage we find anode voltage of 170V
So I can say that for a changing of 1V on the grid that result 170-95 = 75V of amplifcation.
The Facto the gain ratio is 75.
We can compare it with the V2a gain (see above).
Ug1=1V => Ua=150V
Ug1=2V => Ua=210V
210-150=60
So the Double Triode design give us a dynamic of 20% better than the single triode design.
Very good thing for the first stage of our preamp, isn't it !
The next time we will take a look on V2b which is a DC cathode follower site between V2a and the tone stack.
We will try to understand how it's a very interesting design for achieve smooth an warm overdrive.
Let's talk about the DC Cathode Follower which it's the last triode of our preamp design (V2b).
You remember that we already study the first triode of our preamp second stage (V2a see above).
I have to precise that it's the stage which achieve good preamp distortion.
Thus this is a good erea to use a DC cathode follower.
Here is a vue of how it looks like.
We can easely recognize the cathode follower, the triode without plate resistor and its only resistor on the cathode which is the point we will recover the signal.
Here are the values we have to know :
U=280V
Rk=56k
Now take a look on the very interesting loadline.
Because of the no plate resistor design, the tube works differently.
The plate voltage is constent but the cathode works like it was the plate.
This why we have to reverse the voltage on the up side of the graph.
We see that the grid is in a positive voltage and we understand that it's going to tend towards zero volt and drain current.
On the loadline, I colored it in brown, we can see that this current will tend towards 0,5mA.
This current will come from the plate resistor of V2a which will create a SAG and by this way low down the plate voltage of V2a.
U = RI = 100k x 0,0005 = 50V
Thus at first we see that our DC cathode follower will works to achieve more distortion from V2a.
But we can also see on the loadline that the tube is "locked" on 0V therefore all positive voltage will be very nicely compress.
What could be see as a very bad design for an audiophile user is a very good way to achieve smooth and warm distortion for a guitar player.
Now the first part of the preamp design is done.
I can know the preamp consumption which will be necessary for the further power supply design.
Iv2a = 1mA
Iv2b = 3,2mA
Iv1 = 0,75mA
Ipreamp = 4,95mA
It's time to talk about the voicing.
It's important to know the power amp tube brand.
Each tube has its own voice and that a point who has to be take in concideration very seriously.
My power amp will run with EL84 tubes which have a very round and medium tone.
In the same time I'm looking for a very simple tone stack with only a bass and treble pots.
I will adjust the tone for a Vox / Matschless voicing but with my taste.
Vox Top Boost channel tone stack is very accurate.
For best gain capability we use a large and usual cathode capacitors value of 22µf that give us a cut frequency of 4,8Hz.
F= 1/(2pi x Ck x Rk)= 1/(6,28 x 0,000022 x 1500) = 4,8Hz
In the same time, we work on the design of a Rock amp and want to achive very nice distortion tone.
We all know that distortion doesn't like too much bass in the voicing.
We can low down the bass level with the coupling capacitor between V1 and V2a which is going to help us in the same time to protect V2a of blocking distrotion.
Because of the very medium tone of the EL84 and the tone stack simulation I choose a cut frequency of 700Hz.
C = 1/(2pi x F x R) = 1/(2pi x F x R) = 1/(6.28 x 700 x 220k) = 1,034nF
1nF will be my coupling capacitor between the first and the second preamp stage.
Well, now my preamp is becoming almost totaly design.
It's missing the gain pot value, and I choose a 500klog for not overload the V2a grid and adding a 180p bright cap for very nice sparkle clean tone.
Comment