Any idea what the primary inductance is on that OPT?
The more inductance, the better for bass response, and the higher rp the output triodes can have. With roughly 8 to 10 mA for Ip:
rp for 6SN7 is about 10k ohms
rp for 6DJ8 is about 3k to 4k ohms (or so)
rp for 6N6P is about 2k to 3k ohms
rp for 12B4A is about 1k ohms
The lower the rp and the higher the pri L, the lower in the bass the amp will reproduce.
The higher the rp and the lower the pri L, the higher the bass cutoff will be in the amp's output.
Since we want to reproduce frequencies down to -3dB at about 15Hz, we want enough Lpri mated with a low enough triode rp.
--
A 6sl7 shouldn't have any issues with slew rate driving 20VPP to the 6sn7 grids. The larger issue is with the parasitic capacitance of the output transformers. I haven't measured it yet, but in my experiments with these transformers, you'll need about 30mA idle current in the output stage to get over 20KHz. One of these days I will get around to building this circuit and do some measuring. My earliest amps were all 6sn7 output.
30mA would be perfect for a push-pull pair of 6N6P or 5687. 15mA each triode at 230V plate-cathode would be 3.45 watts plate dissipation, which is about the max for each triode using the two triodes in one envelope.
Plate voltage about 240V if self-biased.
Grid bias would be about -11V.
OR
Plate voltage 210V self-bias
Grid bias -9V
Plate voltage about 240V if self-biased.
Grid bias would be about -11V.
OR
Plate voltage 210V self-bias
Grid bias -9V
I've used 6n6p in push pull with the 230v:5v 10VA Anteks. I was very pleased with the sound. I ran them in an active loaded parafeed configuration to avoid any DC imbalance. Never measured it though. Toroidal power transformers can sound very good if implemented properly. Keep the output primary swing below 250 Vpp at 20 Hz to avoid core saturation and run enough current through the stage to overcome parasitics. Of course you can squeeze more power out of it by increasing your low frequency cutoff to say 40 Hz where you can safely swing 515Vpp. Just keep DC out of the things. Experiments also seem to indicate that you can wire two transformers in series to double your voltage capability without increasing the parasitic load... unfortunately, my bench supply has limited power to test it properly. I'll get back to playing with these things someday.
If you've collected a bunch of old junked amps, there's no reason you have to use a teenie-weenie OPT for a headphone amp. Why not use a PP EL84 OPT for a pair of 6N6P triodes? It wouldn't have to be the greatest 8k or 10k to VC OPT ever made because you'll be way under its max ratings, right?
Or use one of the Edcor PP OPTs with 10k primary
Does using an Antek toroidal power transformer as an OPT yield that much better results? Or is the objective to use new parts that are as inexpensive as possible?
Or use one of the Edcor PP OPTs with 10k primary
Does using an Antek toroidal power transformer as an OPT yield that much better results? Or is the objective to use new parts that are as inexpensive as possible?
Last edited:
It's hard for me to pass up on a 10 dollar transformer that performs as well as they do. They are particularly well suited for headphone use where the voltage swing is quite small. I have made spud headphone amps using toroids and 6p15p and 6p36s with the screen as the anode. The mu of the 6p15p allowed me to use a much larger stepdown ratio giving a reflected load of 116K ohms (55 ohm headphones) making the coupling capacitor quite small for 20Hz cutoff frequency (.47uf for curious minds). I am a big fan of parafeed and getting quality wound transformers is costly.
Hard to beat a $10 transformer that works well for a new build. Find me a similarly priced real tube output transformer for the same price and I'll try it, I have plenty of 6V6, 6p14p, and 6p43p-e to try with such a transformer.
PP OPT for headphone use, not for loudspeakers:
EDCOR - XSM Series - the 10K:150 one (8.2:1) - $20.59/ea
For more money, bigger power (rated 10W), wider bandwidth:
EDCOR - XPP10-10K - 10k:8 $23.24/ea
Those would be designed for some plate current, so not parafeed. Bandwidth isn't great (the 10W one is rated 70Hz - 18kHz <1dB) but with lower than rated output it should have wider bandwidth than that, right?
If the 8 ohm secondary is loaded with 32 or 50 ohms, the primary will reflect back higher impedance, so that XPP10-10K OPT could be good for PP 6N6P to 32 ohm headphones (Grado, Audio Technica, etc.).
Yes, the Antek toroid PTs are cheaper, but you need to add the cost of caps, CCSs or plate chokes, etc. for parafeed. Does that make it a wash?
I built a SE 12GN7A-triode 'spud' headphone amp using an Edcor 8k:50 OPT, which has worked out very well. I have a pair of the 10k:300 PP OPTs which I need to get to work on. They were $15/ea when I got them.
I have no idea how the bandwidth was measured, and Edcor is pretty notorious for not giving out specs and details like that. They offer a figure for primary inductance for some of their OPTs, but only a few, and they don't say how they measure that either. Oh well.
However, I figure if the OPT in question is rated for 10W with a nominal bandwidth of 70Hz-18kHz -1dB, then it should perform better if only asked to deliver 1W max output. Or am I wrong about that?
However, I figure if the OPT in question is rated for 10W with a nominal bandwidth of 70Hz-18kHz -1dB, then it should perform better if only asked to deliver 1W max output. Or am I wrong about that?
Lingwendil : I am intending to build the amp you are describing in your two Flea Amplifier User guides (1) or/and (2).Hard to beat a $10 transformer that works well for a new build. Find me a similarly priced real tube output transformer for the same price and I'll try it, I have plenty of 6V6, 6p14p, and 6p43p-e to try with such a transformer.
Reading all over this thread, I lost track what your latest modifications are on your final design.
Can you give me some concise info about this? Would be much appreciated!
Joe.