| Electronics > Projects, Designs, and Technical Stuff |
| Calling all vacuum tube pros: A power supply question |
| << < (4/4) |
| cvanc:
To give a little closure here, this repair has been set aside while the search for a suitable replacement power transformer continues. (My theory on why the transformer failed? The original selenium rectifiers were still in there and I bet they finally shorted out) If any of you want to suggest vendors for WELL SHIELDED transformers suitable for a tube preamp, I am all ears. I am having zero luck finding one. Also, here's a different power supply from the same era that gives a good example to the "bias your filaments at some DC voltage" concept. It's another McIntosh, an all-tube FM tuner, model MR-71. Note one string of filaments is biased to sit on top of a +95V rail. Interesting. If it matters, the tubes in this DC-biased string are the oscillator V3, L&R audio out V11/V12, and MPX V14/V15. I wonder why these tubes were 'deserving' of this bias and not any of the others? :-// |
| schmitt trigger:
A common reason is to prevent exceeding the cathode- heater maximum voltage rating. There may be other reasons. |
| floobydust:
Indirectly-heated filaments are DC biased above the heater in order to lower noise. Normally it's reverse in an amplifier circuit - the cathode is at a positive potential compared to the filament and some electrons flow from filament to cathode. Reverse-biasing blocks the charge there, I forgot the physics of it all. It's also a diode so distortion can show up on high-impedance cathode circuits. Capacitive voltage dividers have a short lifetime, the ripple current (peaks) are very high and the first capacitor (i.e. C308 quickly fails, a 40uF cap does not take much ripple current. I would get rid of it, if rewinding the transformer. You need double the secondary turns but thinner wire, not sure how much room there is on the original core. I also think the transformer needs to be double-insulated as this is not earth-grounded. For custom audio transformers I have used Sowter Transformers in the UK. They have pro vintage offerings for Neve, Fairchild, Altec, Pultec, RCA, Ampex etc. They offer no-cost custom transformer design too. Very good quality but a trip across the Atlantic for British steel. Edcor USA sells laminations, bobbins, end bells and parts to build one and are low cost. For shielding, it's just the (steel) end-bells here? I'd have to see a pic of the transformer. Shielding is more commonly done by the tube shields. |
| NiHaoMike:
--- Quote from: floobydust on January 01, 2019, 07:19:21 pm ---Capacitive voltage dividers have a short lifetime, the ripple current (peaks) are very high and the first capacitor (i.e. C308 quickly fails, a 40uF cap does not take much ripple current. --- End quote --- Use a motor run cap as a replacement - those can take a lot more current. |
| duak:
cvanc, the tubes chosen for biased hum reduction are the most sensitive links in the audio frequency chain. Any hum injected there will be amplified the most. Ah, but why the oscillator tube too? Any low frequency AC injected into the oscillator will modulate its output in both AM and FM modes. Since this is a superhet, the input signal is mixed with the oscillator output producing a difference signal that is amplified and then detected by the FM demodulator, Any modulation of the oscillator's frequency will be extracted and then amplified and will appear on the output as hum. |
| Navigation |
| Message Index |
| Previous page |