Beyond the Ariel

Going back to the discussion of SE:pP vs PP:pP interstage driver transformers, I would be interested if primary inductance for the SE:pP IT can ever be high enough to satisfy requirements for phase angle of the load laid out in RDH4 5.3 Table 2 p213. This is the great thing that PP:pP offers. SE:pP offers the power peaks of an AB PP output stage but with the limitations of a SE amp in this regard. Inbalance of the secondaries can be overcome, but primary inductance is limited by the gap.
 
If it's not a secret , what kind of output transformers Thom used in his Flesh & Blood 300B amps? I admit I made these amps several years ago trying to closely follow the original recipe just used custom OPT's since specified Tangos were unavailable anymore.

It's not a secret...Thom's amps use MagneQuest FS-030's. I have owned more than one pair of these transformers and hold them in high regard. I also used several Tangos back when they were still available -- NC18, NC20, XE20S and X5-S. All excellent, but different strengths.

It's been fun to hear about Thom's successes in taking the Reichert design to its next steps. Herb Reichert gave me a lot of help (and inspiration) when I was building my own Flesh and Blood amps back in the 90s.

Gary Dahl
 
Last edited:
4p1l is the best tube I've heard in SE with filament bias; it sounds very clean and detailed. However it has quite a bit of harmonics beyond the third, I'm not sure if it would be as clean as a 45 in PP driver. I'm pretty much relegating myself to new production 45s when it comes to finding a matched quad.

According to at least one of Ale Moglia's measurements of the 4P1L, it can be quite clean beyond the 3rd harmonic. I 'm sure that you could find certain specimens of this type that could present an increased amount of higher order harmonics at specific operating points, but as indicated here this tube has some of the cleanest distortion spectra around. It's also worthwhile noting, that in this graph the second harmonic is down 70 db while the third is down by more than 103 db in triode SE mode with no feedback!

4?1?/4P1L THD sweet spot | Bartola Valves
 

Attachments

  • 4p1l distortion.JPG
    4p1l distortion.JPG
    169.5 KB · Views: 512
Last edited:
It's not a secret...Thom's amps use MagneQuest FS-030's. I have owned more than one pair of these transformers and hold them in high regard. I also used several Tangos back when they were still available -- NC18, NC20, XE20S and X5-S. All excellent, but different strengths.

It's been fun to hear about Thom's successes in taking the Reichert design to its next steps. Herb Reichert gave me a lot of help (and inspiration) when I was building my own Flesh and Blood amps back in the 90s.

Gary Dahl
Oh , than there was something else to blame . It happened that my friend who measured the custom OPTs lent me a pair of FS-030 to try and yes they had some bass but overall were not the most refined nor especially great sounding in that design Admittedly it was a rather old run of those trans and Mike might have improved them. I kind of concluded that without original Tango units or some custom Audio Note trans the circuit is not worth consideration. At that time I was using K-horn bass bins and Lowther Dx4 in Martin's Azura 202Hz horns. Too bad I sold all those Black gate , Jensen copper caps and Tantalum resistors from the amps.
Last time we exchanged messages with Thom was many years ago when he was working on his budget Serac table since I had PVC platter and the bearing. I assume he is into another "budget" table project?
 
>>but back to the main topic of the target speaker design: where can one see the polar plots for the end design on this thread?

I don't think the design is done, so I wouldn't hold your breath for comprehensive measurements which are rarely done for DIY projects. However, I can suggest checking out some of my measurements of a similar speaker I have been working on (LC midrange w/ large format driver + 15" woofer + LC super tweeter). Here are some measurements of the system as I am listening to it now:
http://www.diyaudio.com/forums/multi-way/100392-beyond-ariel-614.html#post4173924

And here is some simulation based on that measured data for what I want to try for a crossover in the future to smooth out the power response:
http://www.diyaudio.com/forums/multi-way/100392-beyond-ariel-635.html#post4386135
 
Oh , than there was something else to blame . It happened that my friend who measured the custom OPTs lent me a pair of FS-030 to try and yes they had some bass but overall were not the most refined nor especially great sounding in that design Admittedly it was a rather old run of those trans and Mike might have improved them. I kind of concluded that without original Tango units or some custom Audio Note trans the circuit is not worth consideration. At that time I was using K-horn bass bins and Lowther Dx4 in Martin's Azura 202Hz horns. Too bad I sold all those Black gate , Jensen copper caps and Tantalum resistors from the amps.
Last time we exchanged messages with Thom was many years ago when he was working on his budget Serac table since I had PVC platter and the bearing. I assume he is into another "budget" table project?

Greetings ...

I don't have time to track back into this thread (of for audio forums in general these days), so I'll be brief and try to address your transformer question.

Firstly, using using the Flesh & Blood as a departure point was an easy decision, since I've pondered basic, 2 stage, RC coupled designs for years (while toying with transformer coupled concepts as well).

Secondly, the FS030's and a few other parts were rescued from a pair of "garage" amps of John Tuckers (his words), and they needed to justify themselves in order to stay in the house after sitting on a shelf for ten years. I implemented my own design inside the original chassis, but the limited real estate didn't make for a happy prototyping experience and I parked them.

I don't understand your logic about the circuit not being worth building if Tangos or A-N's can't be had. Based on this logic, no amplifier with high aspirations should be built. Will I try a pair of Noguchi's or Monolith Magnetics (or Intact/Slagles) in these amps? You bet! My sense is that people are about to realize the renaissance going on with iron - much as they did with vacuum tubes.

I assume that since you have a Serac (PVC) platter and bearing (not a complete Galibier turntable?), that you were part of the Randal Museum project in San Francisco? Your story line is unfamiliar, as I'm unaware of any Serac owners who dabble in the DIY tube world.

The new 'table is called the Eiger. It's a departure from what I've done before. Stay tuned ...

[edit] Upon pondering this a bit further and surfing back a page in this thread, perhaps what you were trying to say is that the reason the A-N amp "smoked" your design was because of their iron? Certainly, this could have contributed. Taking a schematic and making a "winner" out of it has more to do than with parts selection however. Perhaps I'm preaching to the choir, but things like layout and lead dressing can have a major effect. I have a guitar amplifier made by a master down in Phoenix (Redplate Amplifiers). Henry says he could care less if his schematics are circulated over the internet, because no one will duplicate his sound.

Cheers,
Thom
 
Last edited:
Going back to the discussion of SE:pP vs PP:pP interstage driver transformers, I would be interested if primary inductance for the SE:pP IT can ever be high enough to satisfy requirements for phase angle of the load laid out in RDH4 5.3 Table 2 p213. This is the great thing that PP:pP offers. SE:pP offers the power peaks of an AB PP output stage but with the limitations of a SE amp in this regard. Inbalance of the secondaries can be overcome, but primary inductance is limited by the gap.

I'm without the Designers Handbook and guess that this is to do with Rp Vs Lpri and phase shift at LF (?)

If thats the case, theres a set of compromises which need to be balanced and that might include LF limiting the front end. This would reduce the complications WRT the IT and also make life quite a bit easier for both the output stage and the loudspeaker.

This assumes the system requires >~50Hz from the amplifier (thats a happy amplifier).

There could still be the phase shift issues WRT how you LF limit the design, RC coupling, C bypass for Rk etc and I'm unsure how this is any different from Lpri being the determining factor. Yet, a drive-rack might mitigate this altogether, but then I don't know how active XO affect phase at the turn point.



Another question might be; what are the impacts of phase shift at LF and above which frequencies does this become non-issue?


L.H
 
Last edited:
According to at least one of Ale Moglia's measurements of the 4P1L, it can be quite clean beyond the 3rd harmonic. I 'm sure that you could find certain specimens of this type that could present an increased amount of higher order harmonics at specific operating points, but as indicated here this tube has some of the cleanest distortion spectra around. It's also worthwhile noting, that in this graph the second harmonic is down 70 db while the third is down by more than 103 db in triode SE mode with no feedback!

4?1?/4P1L THD sweet spot | Bartola Valves
From this it does not really need PP to cancel the 2nd order harmonics for this and better performing valves. I like the SE sound with the DHT, whether it is 300b 845 or whatever. But if I built up or modded another I might do the PP but not with 845's. Naturally, the 100dB plus 3rd order harmonics down mentioned in the posting is desirable. My 1987 revised PP SS amps are way below 100dB 2nd Harmonic. Well below any DAC output.

By the way, the SE 845 offers 20 watts per channel, so it has a decent range of speaker possibilities. Alternative 845's can be biased up to 40 - 50 watts.
 
Greetings ...

I don't have time to track back into this thread (of for audio forums in general these days), so I'll be brief and try to address your transformer question.

Firstly, using using the Flesh & Blood as a departure point was an easy decision, since I've pondered basic, 2 stage, RC coupled designs for years (while toying with transformer coupled concepts as well).

Secondly, the FS030's and a few other parts were rescued from a pair of "garage" amps of John Tuckers (his words), and they needed to justify themselves in order to stay in the house after sitting on a shelf for ten years. I implemented my own design inside the original chassis, but the limited real estate didn't make for a happy prototyping experience and I parked them.

I don't understand your logic about the circuit not being worth building if Tangos or A-N's can't be had. Based on this logic, no amplifier with high aspirations should be built. Will I try a pair of Noguchi's or Monolith Magnetics (or Intact/Slagles) in these amps? You bet! My sense is that people are about to realize the renaissance going on with iron - much as they did with vacuum tubes.

I assume that since you have a Serac (PVC) platter and bearing (not a complete Galibier turntable?), that you were part of the Randal Museum project in San Francisco? Your story line is unfamiliar, as I'm unaware of any Serac owners who dabble in the DIY tube world.

The new 'table is called the Eiger. It's a departure from what I've done before. Stay tuned ...

[edit] Upon pondering this a bit further and surfing back a page in this thread, perhaps what you were trying to say is that the reason the A-N amp "smoked" your design was because of their iron? Certainly, this could have contributed. Taking a schematic and making a "winner" out of it has more to do than with parts selection however. Perhaps I'm preaching to the choir, but things like layout and lead dressing can have a major effect. I have a guitar amplifier made by a master down in Phoenix (Redplate Amplifiers). Henry says he could care less if his schematics are circulated over the internet, because no one will duplicate his sound.

Cheers,
Thom

Thom
Yes , I'm probably a lousy builder. The amps were quiet , measured (static operating points ) exactly to specs. Parts selection followed Herb choices since he stated that if one will substitute beyond his recommendations it won't be F&B anymore sort of things . I had 1.5 H 16 Ohm chokes and 300VA power trans , Cerafines PSU and black gates and jensens , tantalums. Anyway I made clone of Yamamoto 08 with Bud's level one Opts and it also sounded like a distant cousin from the real thing. There are so many considerations when building.
Regarding the turntable I should have specified that it was not your PVC platter although I bought it from your Aussie customer who upgraded to one of your high spec platters (stelvio or gavia) and the bearing I got from Chris Brady. I don't want to build any more DIY tables....ever:D
 

Attachments

  • table 008.jpg
    table 008.jpg
    606.2 KB · Views: 568
I don't think the design is done, so I wouldn't hold your breath for comprehensive measurements which are rarely done for DIY projects. However, I can suggest checking out some of my measurements of a similar speaker I have been working on (LC midrange w/ large format driver + 15" woofer + LC super tweeter). Here are some measurements of the system as I am listening to it now:
http://www.diyaudio.com/forums/multi-way/100392-beyond-ariel-614.html#post4173924

What program did you use to plot the polar map?
 
@ John S: thanks for the reference. nice work!

btw if you feel like sharing some of your matlab scripts I could use something to plot the polars, DI and summed responses. I am about to start working on my OB XOvers and for the measurements I am only using a generic sftw from work which lacks speaker design plot features; so I cannot do "fancy" :rolleyes:. but I could export data in matlab format and do the rest there.
 
It's all pretty straight forward programming. For the polar map, just look at the surf function. Turn on interpolation so you get a smooth plot, orient the plot so you're looking down at it so you're not seeing the height, just the color, and I plotted a contour of the -6dB points as well. For the response plots, all the calculations are in typical acoustics books like Beranek's or Olson's. For the simulation stuff I did, just take your measured data, normalize it to the on-axis curve, then apply theoretical filters to all the normalized transfer functions for both drivers you're working with and add the results together to see the summed response using different filtering strategies. Then you would do as above to calculate the summed polar map, and power response, etc. I don't do a lot with passive crossovers, but obviously you could easily extend this to include driver impedance and passive networks. Or you could go about it another way - calculate the required filter voltage transfer function to get the desired acoustic response you simulated for good polar response and then design a filter using the measured driver impedance to produce that voltage transfer function. You could get fancy and write an optimizer for that as well - lots of possibilities.
 
Having listened to some Bosendorfer speakers, they seem well suited for certain types of music. Reading the white paper, I think I have to read it a few times to get a good feeling what is trying to be expressed.

Did not realize I missed so much of the discussion...
 
Last edited: