Mikante wrote: ↑Sun Jun 11, 2023 11:10 pm
Thank you very much for these infos.
New schematic.
That looks better. I presume what looks like a resistor with no value next to the grid leak is a leftover from the previous version.
You can make the 470k grid leak smaller if needed when you have figured out suitable ratios for the voltage dividers, to avoid having very large grid stoppers.
"Hey mister, turn it on, turn it up, turn me loose!"
Mikante wrote: ↑Sun Jun 11, 2023 11:10 pm
Thank you very much for these infos.
New schematic.
That looks better. I presume what looks like a resistor with no value next to the grid leak is a leftover from the previous version.
You can make the 470k grid leak smaller if needed when you have figured out suitable ratios for the voltage dividers, to avoid having very large grid stoppers.
Yes, to make space in the drawing for the 470k grid leak resistor i had to move the trim pot a little away and I forgot to also move the resistor symbol in the schematic trace.
About the resistor value, how much lower than 470k can I actually go, so I can have a range to make my mind around.
I need to study this part of the design that is for sure.
Mikante wrote: ↑Mon Jun 12, 2023 4:15 pm
About the resistor value, how much lower than 470k can I actually go, so I can have a range to make my mind around.
I need to study this part of the design that is for sure.
Kind of a tricky question. Someone with more expertice might want to chime in on it.
Generally, grid leaks and resistance to ground should not be too small. Less than 220k is not often seen.
On the other hand, a gain or volume pot performs the same duty in many amp designs (like in basically any Fender amp), making the resistance to ground any value between zero and 1M.
"Hey mister, turn it on, turn it up, turn me loose!"
Mikante wrote: ↑Mon Jun 12, 2023 4:15 pm
About the resistor value, how much lower than 470k can I actually go, so I can have a range to make my mind around.
I need to study this part of the design that is for sure.
Kind of a tricky question. Someone with more expertice might want to chime in on it.
Generally, grid leaks and resistance to ground should not be too small. Less than 220k is not often seen.
On the other hand, a gain or volume pot performs the same duty in many amp designs (like in basically any Fender amp), making the resistance to ground any value between zero and 1M.
I understand, maybe it is because the volume pot usually sits in between the first and second stage of the very first tube?
I am only supposing cause i have no knowledge about this.
Anyway, i will tweak the right value once the amp is up and running, at least i have a rough idea now.
Note that the grid leak is also the load seen by the previous gain stage. If you make this too small then the previous gain stage will not be happy. It will have trouble driving this load and loose gain.
A volume pot does *not* vary the load seen by the previous stage as it looks and acts just like a simple resistor whose value is the pots max value. The variable part is only really seen by the grid of the next stage.
A tubes data sheet will list the maximum value for the grid leak resistor. For the 12ax7 I *think* this may be 1m. If you go above that value then you risk the tube loosing bias (electrons build up on the grid and eventually the tube runs away to max current). There is no minimum grid leak resistance. But obviously zero kills the signal entirely and small values load down the previous stage. So, it is a balancing act. Chances are good that Leo already found a good compromise for the grid leak in question here .
I think the 5M pots will be replaced by resistors when he finds the right values, that's why I suggested he may end up with unnecessarily large values and might want to lower the grid leak.
On the other hand, Fender uses a 3.3M resistor so "large" is relative.
"Hey mister, turn it on, turn it up, turn me loose!"
Thank you guys i understand the purpose of the grid leak resistor better now.
By the way, the two 5m trim pot are there just for fine tuning, chances are i can get away with much lower values for each channel and then replace the two pots with a resistor. I understand in this situation the load on the first two tubes is also related to these two mixing resistors witch is going to make things even more complicated.
Maybe it is better to leave the two 220k mixing resistor in place, have the 470k on the grid leak and use only one 5m trim pot to adjust the gain for both channel?
In the original design the mixing happens after the third stage.
I think a better approach for tuning the mix resistors would be to insert a 500k linear pot, which when centered with the output taken from the wiper would look like 250k mix resistors. Finding the right balance would be simpler, and close to typical values used for that purpose.
the idea of using 5m trim pots instead of the common 220k resistors, if i am not wrong, to have the proper gain fed into the third stage for each channel. When by mistake i erased the 3.3m resistor from the schematic this seemed to be the solution.
I didn t notice that the mixing was happening after the third stage, so those two 220k resistors had noting to do with the third gain stage, setting the ratio between the grid leak and the load was in this environment a much simpler task.
For me at least.
i don t want the two channels to interact with eachother.
Get rid of the extra stage, and add a separate supply branch for the second channel. The difference between the mix resistors will depend on the losses from the two different tone stacks (on the "normal" channel you are only showing a volume control).
You do not have the required permissions to view the files attached to this post.
Mikante wrote: ↑Tue Jun 13, 2023 12:32 pm
the idea of using 5m trim pots instead of the common 220k resistors, if i am not wrong, to have the proper gain fed into the third stage for each channel. When by mistake i erased the 3.3m resistor from the schematic this seemed to be the solution.
I didn t notice that the mixing was happening after the third stage, so those two 220k resistors had noting to do with the third gain stage, setting the ratio between the grid leak and the load was in this environment a much simpler task.
For me at least.
i don t want the two channels to interact with eachother.
The resistors (=temporary trimpots) are the mixing resistors, while also forming voltage dividers with the grid leak for the 3rd stage.
Martin's suggestion is of course the easiest solution. You'll still have enough gain to push the output tubes into overdrive without the 3rd stage. It makes the "vibrato" channel basically identical to the Normal channel in a blackpanel AB763 Deluxe.
You will have more gain from the upper channel than from the lower, but that might be just fine.
"Hey mister, turn it on, turn it up, turn me loose!"
Mikante wrote: ↑Tue Jun 13, 2023 12:32 pm
the idea of using 5m trim pots instead of the common 220k resistors, if i am not wrong, to have the proper gain fed into the third stage for each channel. When by mistake i erased the 3.3m resistor from the schematic this seemed to be the solution.
I didn t notice that the mixing was happening after the third stage, so those two 220k resistors had noting to do with the third gain stage, setting the ratio between the grid leak and the load was in this environment a much simpler task.
For me at least.
i don t want the two channels to interact with eachother.
The resistors (=temporary trimpots) are the mixing resistors, while also forming voltage dividers with the grid leak for the 3rd stage.
Martin's suggestion is of course the easiest solution. You'll still have enough gain to push the output tubes into overdrive without the 3rd stage. It makes the "vibrato" channel basically identical to the Normal channel in a blackpanel AB763 Deluxe.
You will have more gain from the upper channel than from the lower, but that might be just fine.
Isn t the vibrato channel the most loved from users and not just because of the effects? I had a silver face deluxe reverb, long time ago, I remember the vibrato channel to be more usable, both clean and overdrive sounded better and more lively then the normal channel.
As I said it is not mandatory but I d like to keep the third stage.
It has been suggested to have one channel not going into the third stage and hooked up to the lower side of the phase inverter circuit, In this case I could use jumpers and the two channel would also sound very distinct from each other.
DABD58F0-859F-4B33-ADF2-D0EDA9109D1C.png
You do not have the required permissions to view the files attached to this post.
Mikante wrote: ↑Tue Jun 13, 2023 10:48 pm
Isn t the vibrato channel the most loved from users and not just because of the effects? I had a silver face deluxe reverb, long time ago, I remember the vibrato channel to be more usable, both clean and overdrive sounded better and more lively then the normal channel.
As I said it is not mandatory but I d like to keep the third stage.
It has been suggested to have one channel not going into the third stage and hooked up to the lower side of the phase inverter circuit, In this case I could use jumpers and the two channel would also sound very distinct from each other.
DABD58F0-859F-4B33-ADF2-D0EDA9109D1C.png
I tend to agree. To keep the BF sound we all love as close as possible to stock, maybe you should have your "Vibrato" channel going through the 3.3M/10pF and the 220k gridleak into a 3rd stage?
Yes I suggested going into the lower half of the PI, but you can't have NFB in that case.
"Hey mister, turn it on, turn it up, turn me loose!"
Damn i forgot about that.
At this point, i think it is better to leave the 220k mixing resistors and the 5m trim pot with the 10pf cap in parallel right after that.
Could this be a solution?
Now that i think about, adding a separate tube for each channel, i mean for the third stage.
There should be enough juice from the power supply.
.
