Tube Rectifier filtering
Moderators: pompeiisneaks, Colossal
Tube Rectifier filtering
What is the effect (sonically) of having high filtering on the rectifier tube? I have been reading about Ken Fischer having 80uf of filtering on the tube rectifier on Rocket amps. Why would he go so high when people don't recommend going over 60uf?
Re: Tube Rectifier filtering
Rob
I'll try and give you my 2 cents . The caps filter the pulsating DC from the recto and creates a smoother DC. It does this by storing energy. I won't try and talk for Ken and why he used the value he did is a mystery to me and I could only speculate. The reason it's not good to go with a high value of capacitance is that when you first turn the amp on it first has to charge the filter caps and they can draw a lot of current initially after the Standby switch is flipped off. A recto tube is limited in the amount of current it can provide somewhere between 100 and 300 ma depending on the recto tube you use. If the current needed to charge a large cap exceeds the rating of the max current that the recto is capable of delivering then you run the risk of blowing the recto tube. Since most Solid State rectos (diodes) are rated at least 1 amp (1n4007) then the chance of blowing the solid state diode is small. All tube manufacture have specs for rectifier tubes for max capacitance for the 1st filter cap on their data sheets.
Sonically, the caps in the power supply store energy and larger caps can store more energy than smaller ones. Lower frequency notes require more power to produce than higher frequency notes. By using bigger caps you can tighten up your bass because the caps can supply the extra energy needed to produce these notes. Oddly enough high frequencies tend to ride on top of the low frequency wave forms and when you reach a point when the output exceeds the amount of power available the highs are usually the first thing to distort. (this is more noticable with solid state diodes) With a tube recto if the amount of power needed to produce a certain note exceeds what the PS is capable of delevering then the tube recto will let the voltage start to drop. (because it can supply no more current) This is called sag.
I have only touched on the basics and I am sure you could write a whole book on just this subject. Maybe someone else will have a couple of cents to throw in on the matter.
I'll try and give you my 2 cents . The caps filter the pulsating DC from the recto and creates a smoother DC. It does this by storing energy. I won't try and talk for Ken and why he used the value he did is a mystery to me and I could only speculate. The reason it's not good to go with a high value of capacitance is that when you first turn the amp on it first has to charge the filter caps and they can draw a lot of current initially after the Standby switch is flipped off. A recto tube is limited in the amount of current it can provide somewhere between 100 and 300 ma depending on the recto tube you use. If the current needed to charge a large cap exceeds the rating of the max current that the recto is capable of delivering then you run the risk of blowing the recto tube. Since most Solid State rectos (diodes) are rated at least 1 amp (1n4007) then the chance of blowing the solid state diode is small. All tube manufacture have specs for rectifier tubes for max capacitance for the 1st filter cap on their data sheets.
Sonically, the caps in the power supply store energy and larger caps can store more energy than smaller ones. Lower frequency notes require more power to produce than higher frequency notes. By using bigger caps you can tighten up your bass because the caps can supply the extra energy needed to produce these notes. Oddly enough high frequencies tend to ride on top of the low frequency wave forms and when you reach a point when the output exceeds the amount of power available the highs are usually the first thing to distort. (this is more noticable with solid state diodes) With a tube recto if the amount of power needed to produce a certain note exceeds what the PS is capable of delevering then the tube recto will let the voltage start to drop. (because it can supply no more current) This is called sag.
I have only touched on the basics and I am sure you could write a whole book on just this subject. Maybe someone else will have a couple of cents to throw in on the matter.
Re: Tube Rectifier filtering
hello, i read somewhere that ken said he wouldn't consider making a rocket unless he could find nos mullard 5ar4's of a specific time and place of manufacture. perhaps this is the reason------he had to find specific mullards he knew from experience could take the punishment. rh
Re: Tube Rectifier filtering
Could an inrush current limiting device like a thermistor be used to allow a larger 1st capacitor with a tube recto?
Broz
Broz
Re: Tube Rectifier filtering
I may have answered my own question, a 200 ohm (cold) ntc thermistor seems to be at the large end for the devices. This would knock down the peak inrush a bit, but is probably not enough to allow a large increase in capacitance.
Broz
Broz
Re: Tube Rectifier filtering
I have a 5AR4 tube rectifier which was new in the Mullard box, but has Philips written across it.
I have compared it to a new 5AR4 and I can't hear any difference in tone between the two tube rectifiers. Mind you the amp that I did the comparsion with wasn't a high wattage amp, I imagine the higher wattage amp would show more differences.
The point that I'm trying to make is that I think Ken Fischer probably used the Mullard 5AR4 because it is more robust than other brands of 5AR4's.
Somewhere here in old threads there is something by one of the guys about a Rocket clone that he built using a new 5AR4 and a 80uF first filter cap. I believe that this individual did say he didn't have any problems with the amp. However whether he is still saying that is a mystery.
I have compared it to a new 5AR4 and I can't hear any difference in tone between the two tube rectifiers. Mind you the amp that I did the comparsion with wasn't a high wattage amp, I imagine the higher wattage amp would show more differences.
The point that I'm trying to make is that I think Ken Fischer probably used the Mullard 5AR4 because it is more robust than other brands of 5AR4's.
Somewhere here in old threads there is something by one of the guys about a Rocket clone that he built using a new 5AR4 and a 80uF first filter cap. I believe that this individual did say he didn't have any problems with the amp. However whether he is still saying that is a mystery.
Yours Sincerely
Mark Abbott
Mark Abbott
Re: Tube Rectifier filtering
tubedogsmith said
I'm not saying there bad tubes, but I've replaced more of them than any other rectifier ever.
I built a JTM/Plexi combo for a friend with Fane Alnico's mmmmmmmmmm
Anyway the first filter was 100uF and I used a JJ GZ34. I died after a few months of HARD use. I lowered the first filter to 64uF with no noticable change in sound - and it's out on the road with another JJ GZ34 and doing fine.
My issue with the sovtek 5AR4 is that I've been told they are not a true GZ34. In using them I've noticed they drop 20V or so more than a JJ or an RCA which makes me think they are closer to a 5U4GB.
I've had similar experiences with the JJ GZ34's. I've replaced multipules.The first filter stage is 80uf. You need to use good rectifier tubes and even then their lifespan is shortened. I've been runnin the same sovtek 5AR4 in one of these circuits for over three years. A Mullard is even better. JJ's don't seem to work.
I'm not saying there bad tubes, but I've replaced more of them than any other rectifier ever.
I built a JTM/Plexi combo for a friend with Fane Alnico's mmmmmmmmmm
Anyway the first filter was 100uF and I used a JJ GZ34. I died after a few months of HARD use. I lowered the first filter to 64uF with no noticable change in sound - and it's out on the road with another JJ GZ34 and doing fine.
My issue with the sovtek 5AR4 is that I've been told they are not a true GZ34. In using them I've noticed they drop 20V or so more than a JJ or an RCA which makes me think they are closer to a 5U4GB.