So I'd like to put together an amp with no electrolytic caps. This is easily doable until it comes to the filament supply which I'd like to be DC. Trouble there is the first cap after rectification is usually very large and electrolytics are the only available choice.
Is a DC filament supply possible (regulated or not) without electrolytics or am I dreaming?
Is a DC filament supply possible (regulated or not) without electrolytics or am I dreaming?
Sure. Twenty 100uF motor run caps in parallel will make a very nice 2000uF reservoir. Hope you've got some extra space on that chassis.
Is it that you want the audio path clear of electrolytic for sonic quality, or you have some bragging that the unit has none? In a DC filament supply, electrolytic is an excellent choice. It would be possible to build an inductive filter instead, big iron and only a couple of the above mentioned starter caps, but I sure can't see the point. I would be afraid you are chasing magic.
tvrgeek: 'Not trying to chase magic. Hell, I probably can't hear the difference between AC and DC on the filaments! No, I was just wondering if there was a way it would be feasible that I'm not aware of is all.
poinz: tube lineup is 3.3A worth of 6.3V octals, finishing with 6L6s
poinz: tube lineup is 3.3A worth of 6.3V octals, finishing with 6L6s
Properly implemented AC heaters will not hum and if you 'lift' the supply, only require a modest cap. See Morgan Jones 'Building Valve Amplifiers' to see the techniques required. Mind mentioning why you are looking at DC and no electrolytics? I think you are probably better off spending the extra $ on iron.
Why not
1. Properly damp resonances in the secondary of the filament transformer
2. Use electrolytics
3. Obliterate ripple with a regulator
?
I assume your aversion to electrolytics is due to the fact that you believe them to have poor performance at high frequencies and HF noise will just ride right through. (1) takes care of the input to the 'lytics. HF noise won't come out if you don't put it in.
1. Properly damp resonances in the secondary of the filament transformer
2. Use electrolytics
3. Obliterate ripple with a regulator
?
I assume your aversion to electrolytics is due to the fact that you believe them to have poor performance at high frequencies and HF noise will just ride right through. (1) takes care of the input to the 'lytics. HF noise won't come out if you don't put it in.
Supose you needed a 5v 1A supply you could use a choke then a 6v 20w zener then a .2 ohm 2H choke. Filling in the value for the input voltage and first choke is
up to you. Steve Bench even built a tube amp with no caps in the hivoltage supply
using several stages of inductors and voltage regulator tubes. So it could be done
but probably would be a wast of good parts.
up to you. Steve Bench even built a tube amp with no caps in the hivoltage supply
using several stages of inductors and voltage regulator tubes. So it could be done
but probably would be a wast of good parts.
Carly, if your circuit uses all indirectly heated tubes, why do you want a DC heater supply? I mean, knock yourself out if you're manic for the pain and suffering, but such a supply will need on the order of 20,000μF of capacitance in the final position, and why? I use AC heater supplies in all my circuits and they produce 20 - 40μV (that's two to four one hundredths of a millivolt) of total hum+noise on the output, this dominated sonically by tube rush. A DC heater supply that's anything like actual DC is an expensive and real estate-consuming beastie.
Also, the presence or absence of electrolytic capacitors in such a supply will have no effect on the sound of the circuit. Promise.
Aloha,
Poinz
AudioTropic
Also, the presence or absence of electrolytic capacitors in such a supply will have no effect on the sound of the circuit. Promise.
Aloha,
Poinz
AudioTropic
Sorry I couldn't respond to the contributions to my query sooner. One of the reasons I was entertaining DC was all voltages will be travelling down an umbilical cord and I was concerned that since twisting the AC wires inside the multiconductor cable was not possible, an electromagnetic field would be generated within close proximity to other supply lines. Is this worth worrying about?
Here's what I would do:
1. Use AC but leave some extra room in the chassis.
2. Listen and test to see if it is a problem.
3. If it is a problem, try to solve it.
You will have extra room to build a DC power supply if you think that is the solution, but if you are not careful you can cause more problems than you solve. Damped oscillations caused by diode switching in your DC supply come to mind. I am betting that the AC supply will not be a problem. Oh, and film caps just won't be necessary here.
1. Use AC but leave some extra room in the chassis.
2. Listen and test to see if it is a problem.
3. If it is a problem, try to solve it.
You will have extra room to build a DC power supply if you think that is the solution, but if you are not careful you can cause more problems than you solve. Damped oscillations caused by diode switching in your DC supply come to mind. I am betting that the AC supply will not be a problem. Oh, and film caps just won't be necessary here.
Why in the world would you want to do that?
Capacitor selection in the actual audio path is important, but as far as power supplies go, electrolytic is the ideal choice.
I doubt it. The filament circuit is extremely low impedance, so coupling from the AC is not going to be a problem.
- Status
- Not open for further replies.