Help improve pp-18 tone
Moderators: zaphod_phil, Daviedawg, Graydon, CurtissRobin, colossal
-
- Frequent poster
- Posts: 914
- Joined: Mon 12/10/07 2:00 am
- Location: Melb, Aust
- Contact:
Where are you and what shop is your preferred source? Was thinking I could look through a website for you and make a recommendation.
The MJE340 was the best choice for what is available here in Aust through one of our last surviving electronics retail store chains. It's always hard putting part numbers on schematics because different countries have easier access to different parts.
The MJE340 was the best choice for what is available here in Aust through one of our last surviving electronics retail store chains. It's always hard putting part numbers on schematics because different countries have easier access to different parts.
Last edited by katopan on Fri 11/11/11 4:49 pm, edited 1 time in total.
0 x
-
- Frequent poster
- Posts: 914
- Joined: Mon 12/10/07 2:00 am
- Location: Melb, Aust
- Contact:
- asd
- Newbie
- Posts: 75
- Joined: Fri 08/28/09 2:00 am
- Location: Unknown
I can go at the local shop tomorrow. What kind of transistor I should be asking for? Any npn 300V rated transistor?Any luck finding a transistor asd? The MJE340 is overkill. What are all the npn options you have from your local store?
BTW: is it possible to change my nickname or at least my password? I'm still using that account found on the internet, virtually anyone can log in with this account...
0 x
-
- Frequent poster
- Posts: 914
- Joined: Mon 12/10/07 2:00 am
- Location: Melb, Aust
- Contact:
It would be a pain but I don't know if you could get a list of the npn transistors they have in stock, and then we could pick the most suitable.
300V is overkill. This thing sits on the ground side of the coupling caps, so it's all signal voltage with no DC. Looking at a test waveform document I put together ages ago, the PI output / power grid input is no more than 50V peak. Worst case peak current is if you had the VVR turned all the way up and the variable zener all the way down - from the 12AX7 curves it's about 3.5mA peak, but let's allow for 5mA. We're only clipping one side so half the cycle is 5mA peak and the other side our blocking diodes prevent current flow through the transistor. Average power dissipation is going to be something like 50V x 5mA x 0.5 duty cycle = 0.125W. Geez it's looking like this could be done with a small signal transistor like a BC546.
To be honest I would go for something in a bigger case. Typical TO-92 thermal resistance junction to ambient is something like 200 degC/W so our 0.125W is going to result in a temperature rise of 25 degC above ambient. But I'd want to have this thing bulletproof. I think the minimum spec you'd want would be TO-126 case or bigger, and 80V or more rating. Anything that meets that will be able to take the 5ma current. The protection diodes cover everything else you'd be worried about.
Let us know what you find.
300V is overkill. This thing sits on the ground side of the coupling caps, so it's all signal voltage with no DC. Looking at a test waveform document I put together ages ago, the PI output / power grid input is no more than 50V peak. Worst case peak current is if you had the VVR turned all the way up and the variable zener all the way down - from the 12AX7 curves it's about 3.5mA peak, but let's allow for 5mA. We're only clipping one side so half the cycle is 5mA peak and the other side our blocking diodes prevent current flow through the transistor. Average power dissipation is going to be something like 50V x 5mA x 0.5 duty cycle = 0.125W. Geez it's looking like this could be done with a small signal transistor like a BC546.
To be honest I would go for something in a bigger case. Typical TO-92 thermal resistance junction to ambient is something like 200 degC/W so our 0.125W is going to result in a temperature rise of 25 degC above ambient. But I'd want to have this thing bulletproof. I think the minimum spec you'd want would be TO-126 case or bigger, and 80V or more rating. Anything that meets that will be able to take the 5ma current. The protection diodes cover everything else you'd be worried about.
Let us know what you find.
0 x
-
- Frequent poster
- Posts: 914
- Joined: Mon 12/10/07 2:00 am
- Location: Melb, Aust
- Contact:
- asd
- Newbie
- Posts: 75
- Joined: Fri 08/28/09 2:00 am
- Location: Unknown
Is it ok if I leave the leak resistors on board and use only the wiper and ground connection of the pot? Basically I'd remove the 56k resistor and connect the diode at the top of the schematics and the ground accross the leak resistor as the ruby zener+diode would... and could it be possible to use just a single potentiometer so the bases share the same pot wiper? I mean if I swap the 150k resistor with the pot so each transistor base is connected to its 150k resistor and then share the wiper pot to ground? Or am I dreaming? : )
0 x
-
- Frequent poster
- Posts: 914
- Joined: Mon 12/10/07 2:00 am
- Location: Melb, Aust
- Contact:
- asd
- Newbie
- Posts: 75
- Joined: Fri 08/28/09 2:00 am
- Location: Unknown
And the two bases connected together is a big no no I guess?
I'm absolutely not whining, instead I'm greatly thankful that you draw that schematic fot me, I'm just asking because a triple gang pot is hard to find at best (if not impossible), and if I'm going to build the circuit like it is, it's a separate control for the buzz fix on the backpanel of the amp, that I have to deal to the same amount as the VVR control on the front. Not so convenient.
I'm going to build it in anycase to prove the concept, but I don't know if it will stay unless a reliable 3 gang pot source is found or the circuit can be done with a single ganged pot.
What do you think?
PS. I'm already contacting some custom pot dealer to find out if they require a minimum order or not.
I'm absolutely not whining, instead I'm greatly thankful that you draw that schematic fot me, I'm just asking because a triple gang pot is hard to find at best (if not impossible), and if I'm going to build the circuit like it is, it's a separate control for the buzz fix on the backpanel of the amp, that I have to deal to the same amount as the VVR control on the front. Not so convenient.
I'm going to build it in anycase to prove the concept, but I don't know if it will stay unless a reliable 3 gang pot source is found or the circuit can be done with a single ganged pot.
What do you think?
PS. I'm already contacting some custom pot dealer to find out if they require a minimum order or not.
0 x
-
- Frequent poster
- Posts: 914
- Joined: Mon 12/10/07 2:00 am
- Location: Melb, Aust
- Contact:
Custom pot.... wow. I'd want to test that it works first before shelling out for a custom pot. As I said, I haven't tested this circuit myself yet but it should do what it's supposed to.
All the ways I looked at it I couldn't work out how to do it with one pot, and especially a 1M pot which would allow you to use a dual gang for the VVR and variable zener. It always came down to anything shared between the two sides provided a conduction path between the two power grid inputs and stuffed everything up. But I might be missing something. I'll have another look over it. When I drew it up I kept thinking "how can I simplify this" and nothing came of it.
All the ways I looked at it I couldn't work out how to do it with one pot, and especially a 1M pot which would allow you to use a dual gang for the VVR and variable zener. It always came down to anything shared between the two sides provided a conduction path between the two power grid inputs and stuffed everything up. But I might be missing something. I'll have another look over it. When I drew it up I kept thinking "how can I simplify this" and nothing came of it.
0 x
-
- Frequent poster
- Posts: 914
- Joined: Mon 12/10/07 2:00 am
- Location: Melb, Aust
- Contact:
- Bronsboi
- Newbie
- Posts: 96
- Joined: Sat 03/06/10 2:00 am
- Location: Bradford, England
- Contact:
-
- Frequent poster
- Posts: 914
- Joined: Mon 12/10/07 2:00 am
- Location: Melb, Aust
- Contact:
-
- Frequent poster
- Posts: 914
- Joined: Mon 12/10/07 2:00 am
- Location: Melb, Aust
- Contact:
Someone on The Amp Garage was having this problem where the Ruby mod fixed the fizz problem at high VVR settings but became ineffective as the VVR was turned down. After trying out a number of different suggestions I was going to post the variable zener circuit which I'd put up in this thread, saying it was very much unproven. Before getting that far he tried zeners with much smaller threshold voltages and seems to have found a very simple answer.
He used 3V 5W zeners in the standard Ruby mod and says that it now doesn't fizz across the whole VVR range. Using zeners of too low a value risks clipping the negative side of the power grid signal before the power valve is cut off. This would introduce unwanted distortion. Paul Ruby's document suggests picking zeners a couple of volts higher than the cathode bias voltage. He aimed to make the clipping around symmetrical with the grid clipping on the positive side. But in theory there's no reason the zeners can't be lower than that as long as they don't clip before the power valve reaches cutoff.
I've never tested the cutoff voltage of EL84s in 18W amp conditions. From the data sheet curves it might be something like -20V. Bias is typically 10.5 - 12.5V, but that cools off under signal even before output clipping. So the 3V isn't that unreasonable. Anyway, something for people to try if the Ruby mod is the only fix to their fizz problems and they want VVR as well.
Here's the thread for those with a login at TAG:
http://ampgarage.com/forum/viewtopic.php?t=18157
He used 3V 5W zeners in the standard Ruby mod and says that it now doesn't fizz across the whole VVR range. Using zeners of too low a value risks clipping the negative side of the power grid signal before the power valve is cut off. This would introduce unwanted distortion. Paul Ruby's document suggests picking zeners a couple of volts higher than the cathode bias voltage. He aimed to make the clipping around symmetrical with the grid clipping on the positive side. But in theory there's no reason the zeners can't be lower than that as long as they don't clip before the power valve reaches cutoff.
I've never tested the cutoff voltage of EL84s in 18W amp conditions. From the data sheet curves it might be something like -20V. Bias is typically 10.5 - 12.5V, but that cools off under signal even before output clipping. So the 3V isn't that unreasonable. Anyway, something for people to try if the Ruby mod is the only fix to their fizz problems and they want VVR as well.
Here's the thread for those with a login at TAG:
http://ampgarage.com/forum/viewtopic.php?t=18157
0 x
- Bronsboi
- Newbie
- Posts: 96
- Joined: Sat 03/06/10 2:00 am
- Location: Bradford, England
- Contact:
-
- Frequent poster
- Posts: 914
- Joined: Mon 12/10/07 2:00 am
- Location: Melb, Aust
- Contact: