Hi folks,need help by some project that Im on in this days.I have become some tube power amp that uses for input tube a ecc83 and for output a pair of EL84 pro chanel.So now I have thinking to use some Dact stepped attenuator 2x100kOhm that I have at home and put this in front of the ecc83 tubes,so I could attenuate the volume of this little tube amp.For now in the original design there where a first in series input capacitor of 0,015uF than in paralel the grid leak resistor of 1Mohm and than in series the grid stopper of 47Kohm.So I have decide to omited the capacitor and grid leak resistor becouse I will use the stteped attenuator.IS this ok in this design???But I m now thinking of the grid stopp resistor,what for value I must use if I use the 100Kohm stteped attenuator on input that the frequency response still will be OK in the range of 20Khz.
Plese for some help,becouse Im a little new in this tube hifi stuf.
Plese for some help,becouse Im a little new in this tube hifi stuf.
Last edited:
1k is easily good enough for a grid stopper. Have the resistor body as close to the grid pin as possible.
Re: volume controls, the max value will depend on stage gain. Let's look at worst case: gain of 100. The input plus Miller capacitance will be about 170-180pF or so (Cgk + mu*Cgp). A 100k volume control will have a maximum source impedance of 25k (at midpoint). The rolloff frequency will then be 1/(2piRC) = 35kHz, which should be OK. If you increased the volume control resistance much higher, the rolloff will start right in the audio band and change depending on control setting.
Re: volume controls, the max value will depend on stage gain. Let's look at worst case: gain of 100. The input plus Miller capacitance will be about 170-180pF or so (Cgk + mu*Cgp). A 100k volume control will have a maximum source impedance of 25k (at midpoint). The rolloff frequency will then be 1/(2piRC) = 35kHz, which should be OK. If you increased the volume control resistance much higher, the rolloff will start right in the audio band and change depending on control setting.
Grid stopper on ecc83
Hi SY,
...thanks for the answer.Yes thats that I have thinking about,to use some smaler resistor for grid stopper becouse the bigger the value of that resistor the bigger the hiss that will go than in the power stage.I have think first of some 10Kohm grid stopper but if the 1Kohm is enough than this is fine by me.Thanks again
Hi SY,
...thanks for the answer.Yes thats that I have thinking about,to use some smaler resistor for grid stopper becouse the bigger the value of that resistor the bigger the hiss that will go than in the power stage.I have think first of some 10Kohm grid stopper but if the 1Kohm is enough than this is fine by me.Thanks again
Hi Anders,
...thanks for your replay.Plese tell me why I can not put out the grid leak resistor.I have see from many designes,that if there where used stteped attenuator there were no input cap and no grid leak resistor,becouse the attenuator is the grid leak resistor.Only with simply potentiometer there is always the grid leak resistor in his place.
Plese for futher info about your idea.
...thanks for your replay.Plese tell me why I can not put out the grid leak resistor.I have see from many designes,that if there where used stteped attenuator there were no input cap and no grid leak resistor,becouse the attenuator is the grid leak resistor.Only with simply potentiometer there is always the grid leak resistor in his place.
Plese for futher info about your idea.
An ECC83 has a low mu so it does not need a grid stopper. The only reason for including it is for a guitar amp where you deliberately want to reduce the high frequency response. As others have said, you should retain the grid leak.
Cheers
Ian
I assume you mean low gm rather than low mu, but even so, this does not preclude it from self oscillation. Yes, you usually 'get away with it' without a grid stopper, but it is not worth the risk, IMO. Grid stoppers save lives.
No, definitely not. You can get away with omitting these with a real pot (if you don't mind it going noisy after some years) but for a switched attenuator it is important to keep all DC away from the switch. You need the cap and the grid leak resistor.audiorado said:
47k is too big for a grid stopper in a hi-fi amp. Ok for guitar, where they don't worry about frequency response and need to provide scope for 'tube rolling' as a form of tone control and poor layout practices.
You can not allow the grid to loose its reference for even a fraction of a second, and that will happen if you remove the grid leak. I too have seen this done and it is simply just bad design.
BR,
Anders
Hi Anders,
...thanks for the answer.Yes now I know why the grid leak must be there.But plese if you have time,can you tell me if I need screen grid stoppers for the power tube.The question is ,while one of my good friend will hook up the screens by his ELL80 tube(they are now runing in pentode mode)to the new output transformers that have the UL taps on it.So his idea is to run the tube now in ultralinear mode with this transformer,but would hi need the screen grid resistor or not
...thanks for the answer.Yes now I know why the grid leak must be there.But plese if you have time,can you tell me if I need screen grid stoppers for the power tube.The question is ,while one of my good friend will hook up the screens by his ELL80 tube(they are now runing in pentode mode)to the new output transformers that have the UL taps on it.So his idea is to run the tube now in ultralinear mode with this transformer,but would hi need the screen grid resistor or not
You're welcome
Well, it depends...the ell80 is a rather low gm pentode so you might get away with a small or no stopper on the screen grid. The stopper is more important for stability in UL than in pentode mode. I would start with a lowish value like 100-200Ohms and look for signs of oscillations at the onset of clipping.
//Anders
Well, it depends...the ell80 is a rather low gm pentode so you might get away with a small or no stopper on the screen grid. The stopper is more important for stability in UL than in pentode mode. I would start with a lowish value like 100-200Ohms and look for signs of oscillations at the onset of clipping.
//Anders
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.