I need to find adequate existing power trafo & choke, or order custom made ones. What I don't know is what is the recommended headroom. For instance, for 1 trafo, the actual HT idle current is 50mA, the rectifier filament is 6.3V @ 0.6A and amp tubes filaments (rectified) are 2 x 10V @ 0.45A (each).
What are the recommended trafo and HT choke current ratings in the above example?
What are the recommended trafo and HT choke current ratings in the above example?
Wattage. So if you need 50ma, go for 80 or 90. If your filaments will pull 2 watts, you should probably get at least a 3 watt tx for them. I don't think it's an exact science.
I'm just guessing here, but as far as how much over, it probably depends on the quality of your transformer. Like, if going w/ an inexpensive Hammond, go for 1.8x more current than you need. If getting a higher quality tx, you can probably get away w/ 1.6x headroom.
(Someone correct me if I'm wrong, but that's the way Ive been doing it lately)
I'm just guessing here, but as far as how much over, it probably depends on the quality of your transformer. Like, if going w/ an inexpensive Hammond, go for 1.8x more current than you need. If getting a higher quality tx, you can probably get away w/ 1.6x headroom.
(Someone correct me if I'm wrong, but that's the way Ive been doing it lately)
Hammonds and Edcors seem to function very nicely when sized according to requirements. They may get a little warm, but that's OK.
If you use AC filaments, then you don't need much headroom on the filament windings. If you use DC filaments, you will need heavier windings, maybe 2x because the peak currents will be high.
If you use valve rectification for B+ then you need less headroom. If you use solid state rectification, you need more headroom, because the peak currents from diodes will be higher. You may want to use a current limiting resistor with diodes, similar to how valve rectifiers use a current limiting resistor.
The headroom for chokes are a different matter. Always oversize your choke some to accommodate bias adjustment, tube changes (i.e. changing a 12AU7 for a 12AX7 can quadruple current, or swapping a KT88 for a EL34), and other factors.
Also, when choosing a choke, pay attention to DCR. Smaller current chokes have higher DCR. Large current chokes have significantly less inductance.
I have the power transformer and chokes for my EL84 PP and KT-88 PP are oversized double, because I built them as monoblocks, but used popular (Hammond and Edcor) transformers designed for stereo amplifiers. They use solid-state rectification, too.
The power transformers and chokes (both Hammond) for my Akikdo are much closer to requirements, especially the chokes, because I wanted to maximize inductance as much as possible. This circuit uses valve rectification.
If you use AC filaments, then you don't need much headroom on the filament windings. If you use DC filaments, you will need heavier windings, maybe 2x because the peak currents will be high.
If you use valve rectification for B+ then you need less headroom. If you use solid state rectification, you need more headroom, because the peak currents from diodes will be higher. You may want to use a current limiting resistor with diodes, similar to how valve rectifiers use a current limiting resistor.
The headroom for chokes are a different matter. Always oversize your choke some to accommodate bias adjustment, tube changes (i.e. changing a 12AU7 for a 12AX7 can quadruple current, or swapping a KT88 for a EL34), and other factors.
Also, when choosing a choke, pay attention to DCR. Smaller current chokes have higher DCR. Large current chokes have significantly less inductance.
I have the power transformer and chokes for my EL84 PP and KT-88 PP are oversized double, because I built them as monoblocks, but used popular (Hammond and Edcor) transformers designed for stereo amplifiers. They use solid-state rectification, too.
The power transformers and chokes (both Hammond) for my Akikdo are much closer to requirements, especially the chokes, because I wanted to maximize inductance as much as possible. This circuit uses valve rectification.
Thanks.
I'm going to use specific circuits with no tubes types swapping. There will be 2 different circuits, however each one of them is set and fixed.
For HT it will be tube rectifier only.
For rectifier tube filament I'll use AC and for amp tubes filaments DC only.
So, for tubes filaments I'll need 2x the actual current rating.
What is the recommended headroom for the rectifier filament, for HT and for the choke?
I'm going to use specific circuits with no tubes types swapping. There will be 2 different circuits, however each one of them is set and fixed.
For HT it will be tube rectifier only.
For rectifier tube filament I'll use AC and for amp tubes filaments DC only.
So, for tubes filaments I'll need 2x the actual current rating.
What is the recommended headroom for the rectifier filament, for HT and for the choke?
IMHO, using valve rectifiers and a fixed circuit, your requirements for headroom should be small.
The Hammonds work fine at rated power. So do the Edcors.
"Headroom" can be explained very easily. Small transformers use small-gauge wire for the primary and secondary wingdings. These windings have DCR (DC resistance). If overloaded, the DC resistance creates too much voltage drop, which drops the voltage below rated specifications. The DCR also creates heat. The eddy currents in the metal core will also create heat, resulting in an overheated transformer. In other words, the combined resistance of the wire and core will lose votlage and create heat.
"Headroom" means the wire gauge and core size have lower resistance, so the transformer won't get hot and loses less voltage.
In other words, the transformer's "rating" is the best balance of cost, size, resistance, voltage, and heat. "Headroom" means more cost and size but reduced resistance, voltage loss, and heat.
You don't need any headroom for the rectifier filament or for HT, unless you want to run your transformer cooler (my Harman Kardon Citation III transformers ran so hot, I could touch it only for seconds at a time, and it was a commercial design!)
For your DC filaments, the instantaneous transformer currents will be high, so 2x actual current rating is smart. Be sure to install snubbers on the DC filament diodes, because they are noisy! Use good diodes and lots of power supply filtering on the filaments. In my experience, DC filaments are the noisiest circuit of the amplifier.
Chokes are a different matter, because the core can saturate, and the inductor loses effectiveness. You always want to overspec your choke a little bit to avoid saturation - no exact rule here. Maybe 1.2 as a minimum?
The Hammonds work fine at rated power. So do the Edcors.
"Headroom" can be explained very easily. Small transformers use small-gauge wire for the primary and secondary wingdings. These windings have DCR (DC resistance). If overloaded, the DC resistance creates too much voltage drop, which drops the voltage below rated specifications. The DCR also creates heat. The eddy currents in the metal core will also create heat, resulting in an overheated transformer. In other words, the combined resistance of the wire and core will lose votlage and create heat.
"Headroom" means the wire gauge and core size have lower resistance, so the transformer won't get hot and loses less voltage.
In other words, the transformer's "rating" is the best balance of cost, size, resistance, voltage, and heat. "Headroom" means more cost and size but reduced resistance, voltage loss, and heat.
You don't need any headroom for the rectifier filament or for HT, unless you want to run your transformer cooler (my Harman Kardon Citation III transformers ran so hot, I could touch it only for seconds at a time, and it was a commercial design!)
For your DC filaments, the instantaneous transformer currents will be high, so 2x actual current rating is smart. Be sure to install snubbers on the DC filament diodes, because they are noisy! Use good diodes and lots of power supply filtering on the filaments. In my experience, DC filaments are the noisiest circuit of the amplifier.
Chokes are a different matter, because the core can saturate, and the inductor loses effectiveness. You always want to overspec your choke a little bit to avoid saturation - no exact rule here. Maybe 1.2 as a minimum?
- Status
- Not open for further replies.