L0rdGwyn's DIY Audio
Jan 5, 2021 at 5:00 PM Post #1,606 of 6,808
Since I am just killing time here, I put together a 6J5 input gyrator-loaded direct-coupled 300B OTL in LTSpice, running the 300B pretty hard, output impedance at 1kHz is 60ohm, does 2W into 300ohms with a 2Vrms input.
 
Jan 5, 2021 at 5:01 PM Post #1,607 of 6,808
Hey Zachik,

Nothing wrong with toroids, in fact they have many advantages, one of them being low emitted magnetic field, I have one in my phono stage. This is just personal preference, but from an aesthetic standpoint if I have all of the transformers out on top of the chassis, I like them to match, whether that be mismatched brands / types covered in matching shields (which will be the case in my 841 headphone amp), or matched brand / and or build. They just have to look the same or very similar to satisfy my OCD tendencies. So I am unlikely ever to pair a toroid with a EI core output transformers in a SET design unless I can shroud them similarly. However, since there are no OPT in an OTL and the transformer is singular, I would definitely consider using a toroid in an OTL. In fact, I am very likely to do it at some point. I probably have six Antek toroids in a drawer in my garage that I bought for various types of experiments over the past year, many of their off-the-shelf toroids have dual secondaries as well which is useful.

I believe the input tranny in the Darkvoice is a toroid.
 
Jan 5, 2021 at 5:02 PM Post #1,608 of 6,808
Not very hard, it is a 300B OTL (notice no transformers?), I slopped together a LTSpice schematic, not optimized but output impedance of a 300B cathode follower is around 70ohm at 1kHz. As a cathode follower, would need a modest driver tube as the input capacitance is no longer an issue. I would probably direct-couple it since the 300B biases at around -80V on the grid. Just spitballin' here.
Would such design be able to drive both headphones (100-300 ohms) as well as efficient (Omega) speakers? (not simultaneously... :wink:)
What about planar headphones (the downfall of most OTLs)?
 
Jan 5, 2021 at 5:04 PM Post #1,609 of 6,808
Since I am just killing time here, I put together a 6J5 input gyrator-loaded direct-coupled 300B OTL in LTSpice, running the 300B pretty hard, output impedance at 1kHz is 60ohm, does 2W into 300ohms with a 2Vrms input.

Try plugging in a 300BXLS. And then how long before you can build one for me? :joy: :joy: :joy:


1609884144441.png
 
Jan 5, 2021 at 5:10 PM Post #1,610 of 6,808
Would such design be able to drive both headphones (100-300 ohms) as well as efficient (Omega) speakers? (not simultaneously... :wink:)
What about planar headphones (the downfall of most OTLs)?

100% no on the speakers, the output impedance is way too high, need some sort of current gain to drive a speaker load (transformer, transistor, etc.). Like other OTLs, high impedance headphones would be the name of the game, not sure what the market is like for high-impedance planars, but low impedance planars would be a no no.
 
Jan 5, 2021 at 5:23 PM Post #1,612 of 6,808
I mean they are all multi billion dollar enterprises and their employees get paid good money to drive around in a truck and play kick ball with all my packages.

I have already had two packages in the last month fly from the west coast, to the east coast, then back to the west coast, just to make it to me 2 weeks late.

Then they have the nerve to claim covid. I swear some of these companies are standing on the graves of those who have died to save a few bucks off the operating costs.

I'm not defending the system, it is in need of improvement. But crunch time is no fun for any person working a physical job, especially one who is more aware than the average joe and laments the fact that he has to be a tiny cog in a large machine that doesn't work well.
 
Jan 5, 2021 at 5:33 PM Post #1,614 of 6,808
Woah, that thing is a beast!

Yeah, a bit over 21 watts/channel in the Jota amp. And people say SET's are wimpy. Maybe compared to a 2kw/channel Class D amp, but then that's more like comparing fine wine to dog poop. :laughing::laughing:
 
Jan 5, 2021 at 5:51 PM Post #1,616 of 6,808
Jan 5, 2021 at 6:15 PM Post #1,617 of 6,808
100% no on the speakers, the output impedance is way too high, need some sort of current gain to drive a speaker load (transformer, transistor, etc.). Like other OTLs, high impedance headphones would be the name of the game, not sure what the market is like for high-impedance planars, but low impedance planars would be a no no.
So how about adding a 2nd toroid transformer (or using toroid transformer with 2 secondaries) - one for headphone out and one for the speakers? Would that work?
What is the transistor option you've mentioned above as another possible solution for the speaker load? Other than committing blasphemy - any reason not to use a transistor based solution for the speaker taps?
 
Jan 5, 2021 at 6:39 PM Post #1,618 of 6,808
So how about adding a 2nd toroid transformer (or using toroid transformer with 2 secondaries) - one for headphone out and one for the speakers? Would that work?
What is the transistor option you've mentioned above as another possible solution for the speaker load? Other than committing blasphemy - any reason not to use a transistor based solution for the speaker taps?

So the toroid is the power transformer, not the output transformer, since we are talking about a hypothetical OTL design. The 60ohm output impedance is without a transformer, the output is AC coupled via an output capacitor. I was just pointing out that in order to lower the output impedance to speaker-usable levels in any amplifier, you need a way to turn voltage into current, which is typically done via a output transformer in tube designs or power transistors in solid-state designs. It can also be done by paralleling several output tubes in a speaker OTL design. Combining a 300B OTL and adding a power transistor for speakers, you might as well just build a 300B SET IMO.
 
Jan 5, 2021 at 6:43 PM Post #1,619 of 6,808
So the toroid is the power transformer, not the output transformer, since we are talking about a hypothetical OTL design. The 60ohm output impedance is without a transformer, the output is AC coupled via an output capacitor. I was just pointing out that in order to lower the output impedance to speaker-usable levels in any amplifier, you need a way to turn voltage into current, which is typically done via a output transformer in tube designs or power transistors in solid-state designs. It can also be done by paralleling several output tubes in a speaker OTL design. Combining a 300B OTL and adding a power transistor for speakers, you might as well just build a 300B SET IMO.

It's also worth adding to this that no one has cracked the code on how to make really high quality audio toroid transformers quite yet. Their interwinding capacitance is usually huge compared to a cheap EI transformer.
 
Jan 5, 2021 at 6:47 PM Post #1,620 of 6,808
Combining a 300B OTL and adding a power transistor for speakers, you might as well just build a 300B SET IMO.
It's also worth adding to this that no one has cracked the code on how to make really high quality audio toroid transformers quite yet. Their interwinding capacitance is usually huge compared to a cheap EI transformer.

So, if I go this route - I would need 1 toroid for power and 2nd toroid serving as the SET transformer? (i.e. abandoning OTL architecture)
...and this would be quite noisy amp (due to toroid capacitance)?
 

Users who are viewing this thread

Back
Top