Hi Folks,
I had a few 5k SE OPTs and a TSE II set up on the bench, so I ran some comparisons with the Analog Discovery 2 and Audio Analyzer Suite. The test setup wasn't fair to all, running around 385V B+ and biasing each at 91 ma where I was getting low distortion on the amp, but not necessarily being nice to the tiny Tamradio. Only one channel driven. 8 ohms only.
The transformers were the Monolith S-11 (new), Antek MS-30W50 (new), Edcor GXSE 15-8-5k (used), and the Sony/Tamradio 5k with 8 and 600 ohm taps from the Sony TC-500a reel to reels (and others) (very used).
I'll be curious to hear what folks think. My relatively uninformed ideas are that that Monolith is a different beast with its extreme HF response. The Antek is quite a thing at the price. Interesting that the square wave isn't quite symmetrical. Lots of power too, making me wonder if it isn't quite 5k. The Edcor is beloved for a reason at the price. The Tamradio, well, not a fair fight.
If I don't get distracted actually finishing some projects, I'm tempted to compare the Antek to the Edcor and the Tamradio in the chassis where both live, for the former, to see what the Antek does with some more power and UL connected in the Tubelab SSE and the latter to see if the Tamradio looks happier in comparison in its normal 2.6W RH84 home. But then, I did hurt a bit 845 amp doing power runs and both the TSE II and a preamp need to find their way into enclosures, so I may not get it done.
Paul
I had a few 5k SE OPTs and a TSE II set up on the bench, so I ran some comparisons with the Analog Discovery 2 and Audio Analyzer Suite. The test setup wasn't fair to all, running around 385V B+ and biasing each at 91 ma where I was getting low distortion on the amp, but not necessarily being nice to the tiny Tamradio. Only one channel driven. 8 ohms only.
The transformers were the Monolith S-11 (new), Antek MS-30W50 (new), Edcor GXSE 15-8-5k (used), and the Sony/Tamradio 5k with 8 and 600 ohm taps from the Sony TC-500a reel to reels (and others) (very used).
I'll be curious to hear what folks think. My relatively uninformed ideas are that that Monolith is a different beast with its extreme HF response. The Antek is quite a thing at the price. Interesting that the square wave isn't quite symmetrical. Lots of power too, making me wonder if it isn't quite 5k. The Edcor is beloved for a reason at the price. The Tamradio, well, not a fair fight.
If I don't get distracted actually finishing some projects, I'm tempted to compare the Antek to the Edcor and the Tamradio in the chassis where both live, for the former, to see what the Antek does with some more power and UL connected in the Tubelab SSE and the latter to see if the Tamradio looks happier in comparison in its normal 2.6W RH84 home. But then, I did hurt a bit 845 amp doing power runs and both the TSE II and a preamp need to find their way into enclosures, so I may not get it done.
Paul
Attachments
It would be nice to see what happens at about 140-160kHz (ringing).
The Antek transformer seems to have a very steep roll-off that sparks my curiosity, I find it quite unnatural/hard to achieve such behavior for an audio transformer.
To make sure I'm clear on this, I'm not criticizing the data, but looking from an enigma view. I love studying transformers. Did you redo the measurement on the Antek? I'd suspect such behavior within a fault with the measurement equipment or a faulty connection during the measurement.
P.S. This might not be a roll-off, my bet is on a steep dip resonance, but it seems due to a lack of measurement bandwidth, only the downfall is visible . So most likely, your measurement equipment is operating normally.
To make sure I'm clear on this, I'm not criticizing the data, but looking from an enigma view. I love studying transformers. Did you redo the measurement on the Antek? I'd suspect such behavior within a fault with the measurement equipment or a faulty connection during the measurement.
P.S. This might not be a roll-off, my bet is on a steep dip resonance, but it seems due to a lack of measurement bandwidth, only the downfall is visible . So most likely, your measurement equipment is operating normally.
For both of these last comments, no, I'm not that confident in my measurements. Though the square wave above certainly helps! I might see how high the rig goes. I got similar results measured several different weekends as I goofed with B+, current, etc. I also tried some different leads and stuff.
On the Antek, the Monolith looks fine up there. The Edcor is rolling off, but not like that. Also note that Antek's own spec sheet shows a pretty steep drop in not an entirely different place.
I'm open to ideas. It'll be interesting to see if it behaves the same way on the other amps. And maybe I should try different loads and the other tap to see if the shoulder moves around?
As you can tell, I was really curious to run the Antek. I've always loved their value in power transformers and every single one I've gotten has behaved exactly as I had hoped. I'm a fan boy.
It'd be interesting if they made more 15W versions as that might really have a good price/performance ratio. This 30W version seems like more than one needs for most SET applications. Likely a great option for UL amps using bigger tubes, which is the ultimate plan for these.
Another interesting option for the price is the Antek 7.5k for an 845, given that these seem to take plenty of power. I just hurt one of my 845 monos doing similar testing. Hopefully something small. When I get it out on the bench, I might try this 5K into the 4 ohm taps with an 8 ohm load on the 845 just for fun.
Overall, and in partial response to a comment in the marketplace, I continue to think we've got OPT options about as good as we ever had. Despite getting very low distortion, the square wave off the big Magnequest on the 845 wasn't the prettiest. With what I'm seeing from the GXSE 15, I wouldn't be the slightest bit surprised if the Edcor GSXE25-10k gave the Magnequest a good run at a great price. And I'm suspecting that there haven't been too many things available before that can do what that Monolith is showing.
Paul
On the Antek, the Monolith looks fine up there. The Edcor is rolling off, but not like that. Also note that Antek's own spec sheet shows a pretty steep drop in not an entirely different place.
I'm open to ideas. It'll be interesting to see if it behaves the same way on the other amps. And maybe I should try different loads and the other tap to see if the shoulder moves around?
As you can tell, I was really curious to run the Antek. I've always loved their value in power transformers and every single one I've gotten has behaved exactly as I had hoped. I'm a fan boy.
It'd be interesting if they made more 15W versions as that might really have a good price/performance ratio. This 30W version seems like more than one needs for most SET applications. Likely a great option for UL amps using bigger tubes, which is the ultimate plan for these.
Another interesting option for the price is the Antek 7.5k for an 845, given that these seem to take plenty of power. I just hurt one of my 845 monos doing similar testing. Hopefully something small. When I get it out on the bench, I might try this 5K into the 4 ohm taps with an 8 ohm load on the 845 just for fun.
Overall, and in partial response to a comment in the marketplace, I continue to think we've got OPT options about as good as we ever had. Despite getting very low distortion, the square wave off the big Magnequest on the 845 wasn't the prettiest. With what I'm seeing from the GXSE 15, I wouldn't be the slightest bit surprised if the Edcor GSXE25-10k gave the Magnequest a good run at a great price. And I'm suspecting that there haven't been too many things available before that can do what that Monolith is showing.
Paul
I've got some old 845 monoblocks with Magnequest FS-010's sitting right here. I built them for friend 25 years ago and got them back when he passed away last fall. I put them on a scope, which I couldn't do back then, and no, the 10k square wave is nothing to write home about. Beautiful, rich midrange, though. I'm mulling over whether I want to rebuild them. As they are they don't hold a candle to my Williamsons. And yes, the Monolith looks impressive.
Sorry for the abbreviation. The Tubelab TSE II. The schematic is over on that forum. 5842 with a CCS driving a 300B through a Mosfet follower. 5AR4; regulated DC filaments. The levels are shown in the screen shots. Note that the THD graphs are showing output level, frequency response is input, and the square wave shows both.
Paul
Paul
Regarding being unable to put them on a 'scope back then, part of my point with posting this stuff is to encourage those who haven't grabbed gear like this to do it. I haven't spent nearly as many years and as much time on this stuff as others, but compared to sourcing dilapidated old 'scopes and signal generators, testing one frequency at a time, and guessing at distortion by looking at wave shapes, this stuff is just remarkable.
This was my first time using these devices, so I goofed around a bit to see if my results were repeatable. But now, I can carry a little tablet and USB attachment to an amp without even removing it from the system and run all of these measurements in fewer than five minutes. For $150. (I bought the Analog Discovery 2 used and had probes, other cables, and the load resistor.)
As Skunkie points out on YouTube, this not only makes trouble shooting so much speedier, but it allows stuff like quick comparisons of tubes. Sure, no substitute for listening, but a quick way to narrow it down.
And finally, regarding the FS-010s, it would be really interesting to do is compare these Monoliths to older, high end OPTs with actual listening. Isn't the 030 the famous 300B option? And big Electraprints, as I know those are probably more technically advanced. Because from the waveforms I've seen over the years, the high frequency behavior of these Monoliths seems pretty remarkable for big SE transformers.
And finally, finally, as you know Grover, I've got a non-working two chassis version of your amps on the shelf with good AES transformers--the monsters for those Cary amps that went into A2. Given the power I was getting, I think you suspected they were more like 6.5k. So another "to do" I might never get to is testing those against the 010s. Because while the test rig is easy, lugging around and safely setting up the monsters still takes time.
Paul
I have spoken with the Monolith folks while building and testing my amps, and they say they purposely designed the OPTs to have that small overshoot with a couple cycles of well-damped ringing, because it sounds better that way.
Over the years, I have also found, on several amps and preamps, that a little square wave overshoot sounds better than a completely damped square wave response.
By the way, Monolith sent me a picture of my OPTs during construction. That's an ECC83 in the picture for reference...
Over the years, I have also found, on several amps and preamps, that a little square wave overshoot sounds better than a completely damped square wave response.
By the way, Monolith sent me a picture of my OPTs during construction. That's an ECC83 in the picture for reference...
Attachments
This sounds contradictory. Ringing means undampened. Dampened should mean no overshoot. Not that I consider HF resonance behavior of much importance outside the audio range, when n-feedback is not used.
maybe there is a hope for a GU-81m amp... 🙂 I was contemplating parting with the set of tubes and sockets I have been kicking around for 2 decades...LOL
Douglas
Douglas
Damped does not necessarily mean zero overshoot. It can also mean an overshoot that quickly decays to zero. There are degrees of damping.