 
|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
|
|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
72.131.105.61
Hello all-
I have been working toward building a preamp and power amp setup that I plan on keeping for the duration of my short life on earth. Cost has minimal affect (maybe $3K for preamp, $6K for power amp), my only requirements are that the power amp be push-pull (I need the power) and that it is completely DIY. More common tubes would be best, and I think I lean towards the 2A3 over the 300B. I would prefer to avoid kits, as well. It's fun to have your friends listen and say "I built the whole thing from scratch, minus tubes and trannies."I'm a power engineer and very competent circuit builder/designer, so voltage does not scare me. I could build a 1500VDC supply if necessary.
Anyway, I like the Lynn Olsen Raven/Amity/Karna designs very much, but would like to get some opinions as to any other readily available designs that compete or excel over these. The documentation should be just as complete as Olsen's, and provide the type of sound quality I will be happy with for years to come. My suspicion is that I will not be building my own amps after this last batch, but may end up building for my friends.
Oh yeah, currently using the Ariel ME-2's, but who knows, that may change in the future.
Thanks for your serious suggestions, and thanks Lynn Olsen for sharing the fruits of your labor.
Follow Ups:
I've been investigating dipoles with 10, 12, and 15-inch drivers and about 97dB efficiency recently. Yes, this is radically different than the Ariels and ME2's, which were based on the best drivers I could get in 1993, when they were designed. Unfortunately, the Scan-Speak 9000 is no longer available, and I have some reservations about the successor drivers. Similarly, as mentioned by another poster, the 5.5" Vifa no longer represents the best available midrange - although most midranges of "exotic" construction still have very ragged response curves.Some of the bigger prosound drivers - 18Sound, Beyma, B&C - have some interesting large-format midrange/midbass drivers that look promising for dipole use. I hear rumors of interesting large-format drivers from Hemp Acoustics in Canada, so I'm looking into those too.
I'm persona non grata in horn circles these days, so my comments about horns will probably count for nothing. That doesn't stop me from liking the products from Azzalino and Cogent, the best I've heard so far, although with very different personalities.
As for my own designs, high-efficiency dipoles are the direction I intend to go - and the recent discovery about DHT's being *very* sensitive to filament voltage (about a 1% wide window) has kind of put more amplifier research off to the side for the moment.
Hi Kurt,I've build both the Amity and Karna, the former being much easier to get right. They are both wonderful amps with good tone and punch, capable of driving almost every speaker out there.
I also have a pair of ME2s which I really like, but prefer to listen to the Quad ESL57 now. The ME2s can be a little too polite sometimes.As others said, if you choose to build one of these amps the iron is going to be a big factor as well as in performance as in costs. I started with all Lundahl iron, which is hard to beat at that price. A couple of months ago I changed the input transformer to a Tribute 1:1+1 and I can advise to get those from the start.
regards,
aj
If the recommendation is to put in the Tribute right away, what do I do with both a Raven and Amity? Do I put in the Tribute as the input of the Raven, with the Lundahl output tranny?Then, with the Amity, another Tribute input, with Lundahl's from there?
Finally, I have looked on their web site, and didn't find much information. What part number am I buying, and is there any spec sheet on it?
Thanks.
Hi,well if you build a Raven + Amity then I guess you could use two pairs. The question is what makes more sense, building a Raven + Amity or a Karna?
About the Tribute input, just ask for "the input phase splitter that Lynn uses in his Karna" and he'll know which one it is.
Hi Kurt,You need to know yourself and your listening needs.
I use a pair of ME2's as well with passive 6dB/oct at about 150Hz and SS amps for the woofers. My amps are DIY and probably about 8 watts. The power amp is a class A triode connected 6AR6 PP. I don't listen extra loud and my room isn't large. I find the volume levels fine for my listening. The only times I wish I had a bit more power is listening to operatic arias where the singers can get very dynamic.
Regards,
John in Seattle
Pretty much the ultimate New Wave tube amp. If 13E1 scares you, a sextet of KT88 might be fun to substitute.
I realize this post is going to cause a lot of folks to think poorly of me, however.Having built all of the magnetics in Lynn's Karna amp, I can attest to their complexity and level of performance. I doubt that anything short of the massive HV SE amps can compete with the Karna in detail, musical expression, micro and macro dynamics and size. The latest version is now using Gary Pimm's SS CCS and separate B+ transformers for 2A3 and 300 B B+ stages. Lynn has also switched his input splitter to Tribute's excellent toroid. If you email me I will provide you with the schematic of this latest version. If you are seriously considering making these amps you must contact Gary Pimm and discuss the unusual ground plane solutions required and I can provide you with his email address.
Lynn's Amity amps, as have been built by Gary Dahl and many others, are somewhat easier but without the 35 watts of clean audio power the Karna provides. more in line with the 20 watts suggested by other posts. They are extremely musical and are well suited to the same level of efficiency you are dealing with.
If you are ready for a challenge there are two other brave souls currently on planet who are either building or planning to build the Karna's, so there will be some folks to discuss solutions with, in addition to Lynn, as he heals his broken leg, Gary Pimm and myself.
Bud
Thank you, for the reply, Bud. You are clearly more than familiar with the construction.To be honest, it is the complexity (and cost) of the Karna that has caused me to shy away from it. Sure, I think I could build it, but the sheer cost of the tubes (45 = $$) alone makes me cringe at the thought of my wife seeing the bill. In addition, my direct experience with DHT's is too limited to tackle the size of Karna and its ac heating. So, yes, I have been leaning the Amity direction. It is my belief and hope that the ME2's will be more than happy with the Amity.
I am a little miffed as to the complexities of the grounding: to me, it all seems fairly straightforward - basic instrumentation grounding. I have read the boring and informative IEEE Emerald book, and would follow those rec's above all others.
If I head this direction, I will certainly contact you regarding more info. I am currently using the schematic shown on nutshell's site. Is that not the best version?
Kurt, good luck on your quest. I built and use Amitys in both 2a3 and 300b versions. The signal iron in both is Lundahl. On the 2a3 version I used Hammond power transformers and chokes. On the 300B, I used all Lundahl iron throughout. I used a modest OT on the 2a3, the amorphous core OT on the 300B. Both amps are very good. It seems to me that once you get to the level of performance of the 2a3 version (even with its more modest OT), the cost to upgrade performance rises dramatically, and you have to decide how much that extra fraction of a percent of accuracy is worth. If you haven't looked at the K&K audio forum, please do - there is a wealth of info on these amps at this site. Additionally, Kevin has a transformer couple KT88 design posted in the projects section that might interest you was well.
Most of the horror facing Gary Pimm, while constructing the Karna's, came from having two chassis, with power mostly on one and signal and the tubes on the other. There are so many tubes and transformers that even one gigantic box would not have held them all. The entire ground scheme was made from Nylon selvage covered Litz wire and if you copy nothing else in your instrumentation grounding scheme, copy that.Gianluca is absolutely correct about the interstage being the key to the amps performance. These amps are just so revealing that even the deservedly well thought of original interstage's, were found to be wanting. Took me several iterations before learning how to balance low permeability gain and this unusual sensitivity to the internal coil capacitance's.
The current units in Lynn's amps, being level three on our performance chart are ruler flat to 20k and then climb to +6 dB at 35 k. Too many turns within the same winding space. So, it will take a larger unit to flatten the highs while keeping the permeability increase to a minimum and so controlling the overall distortion, but specifically where the core becomes permeable, 1 kHz and lower. Level one interstage's do not have this problem with skewed performance, but they are not as revealing either.
I agree with most folks here that a triode connected PP with perhaps a SE driver stage to a voltage splitting transformer coupled interstage will be good enough to soothe the need for music and dynamics from what are some of the best speakers available.
My personal amps are input splitter transformer to PP parallel 7119,s to an 1.414 to 1 PP step down interstage to eight EL 34's in PP paralell to a 1560 z ohm output we originally supplied to Cary for their 6 pac amps. Very noble, powerful, amazingly detailed sound. The only passive components are the grid blockers.
So, don't just look at the DHT amps, as other tubes and topologies are actually equaling that delightful performance and providing power to boot.
Bud
I'd strongly recommend that the OP goes with Lynn Olsen's designs. I'd add that he should use DHTs all through, including the first stage and driver. DHTs are important and so are the interstages. Get those right and you have your amp for life.Bud is being a bit modest here - I have his 6.6k Push Pull OPTs as used in the Olsen designs and they are superb. I'd start there (not sure what the waiting list is like - you may have to get your order in!) and then consider the interstage options - have a look on the K and K site here for the Lundahl options with amorphous cores, or as said consider Tribute and one or two others. To keep costs down you can use Russian or Chinese 2a3 or 300b in the first instance to get it working.
But a word on chassis - isn't it possible to build this fairly easily on two 2u rackmount 19" chassis? My current all DHT transformer coupled balanced Push Pull amp is built on a single 2u rackmount chassis - wouldn't do it again, it's too tight. But a couple of chassis - surely that's OK?
And considering the problems with approving umbilicals in domestic equipment (in the EU virtually everything is not approved for higher voltages DC), two monos rather than a PSU and a signal chassis.
Andy
I was not clear enough in my description of Lynn's amps. They are split into four chassis, sort of a dual mono with respective channels power and output chassis connected with a very serious umbilical. Your estimate of size is just about spot on too and just barely big enough at that.
Four chassis ?I was also thinking about the monoblock design, attempting to keep the PS and amp proper together. Maybe what would make this more possible is additional R-C filtering, removing the CCS? I know many would frown on this, but Lynn seems to indicate both methods worked well.
I would definitely be interested in recommendations for all transformers, and will get advice from K&K. No doubt, though, they will push for all Lundahl throughout. If the tribute is best for input, and James for output, I would absorb all recommendations.
Thanks for your help. Can't seem to locate O-Netics website?
Check the Karna print from Lynn's website, contact information should be on it.Bud
Triode, UL or pentode mode?
Triode only. The original OPT's for Cary(these) did not have the 43% UL tap, not that it would have made much if any difference. All production did have UL taps. I did not like the sound of this amp with UL or in Pentode mode with the needed feedback, so Triode it is.
Even though the love of tubes tends to blind all other parameters, your speakers account for a majority of your sound. If 8 watts are not going to cut it, it is highly doubtfull that 20 watts will do any better. A 300b PP will only put out 20 to 25 watts.Be honest about the load that your speakers present and if tubes other than KT88 or parallel PP KT88 will foot the bill. If your speakers are really current hogs with swinging impedance, you might need a solid state amp with a tube preamp.
You might think about the whole chain and get speakers which support the tube lifestyle.
_______________________________
Long Live Dr.Gizmo
They were also a design by Lynn Olsen,http://www.nutshellhifi.com/Ariel.html
They are a reasonably easy load for any amplifier, especially when you take the bass out. I believe the specs are somewhere around 89-91 dB. They are really designed for tube amps, as Lynn attests to.
I have selected the ME2's due to my room size: 11' x 9' x 80"H. I may move them someday into a larger room, but that's what I have to deal with for now. I plan on crossing over with passive 80Hz 6dB point at the input to the amp, and use a powered subwoofer.
As I listen to anything from classical to oldies to modern, I have read an almost universal recommendation that PP would be the best choice. I really enjoy good midbass (100-300Hz). Currently run a triode strapped EL34.
I am sure I will get resounding disagreement, but I am of the philosophy that a good amp should be relatively immune to load. Yes, I understand matching damping factor to a speaker's mechanics is important, but I don't plan on building a new amp everytime I try out a new set of speakers. DHT's have good plate Z, and PP should be able to drive what I give them.
I guess my point in more watts is that I don't want to pursue the 3W SE path (at least until I hear a system that blows my mind - anyone near Milwaukee, WI?)
and I don't understand, considering your budget, why you would settle for the kind of [audio] compromises you contemplate here.A very good DHT amp can be built with 3C24 driving a GM70 or 845. A 45 driving 845 makes a good amp too...and you can avoid cap coupling. BUT you still need good HIGH EFFICIENCY speakers to get everything this amp can do. Cogent is an extremely good place to start....and end, but there are others too.
Would you use a RS turntable or a WalMart CDP in your system?
I don't think Lynn's amps sound any better than stuff like Rogue. Maybe a slight improvement, but just as squished, constrained and sterile. IMO.
As for the Ariels, I would suggest you adapt and build them with PHY KM30, Fertin or equal drivers. Nothing less than 99db efficient.
Later
D
I believe if Lynn was to design these speakers today, he would use better midbass drivers.
Something like 845? Even singles will get you up to 25 good watts of (IMHO) better sound than PP all things being equal and will drive whatever you through at 'em. Or you could parallel and yes, even push them into A2 on peaks for all the power you'll ever need. 845's are particularly good in the frequency range you like and they aren't the only big DHTs.Naz
Is this not a problem, as I increase the size of the OPT gap I limit the frequency performance?In addition, I understand the cost of a huge SET to be somewhat logarithmic with respect to output power? How is the midbass of a SET this large?
so it can be large even in a lower powered SET. In any case it only affects inductance and therefore very LF response which is compensated to some degree by the designer.In terms of cost SET is simpler so there should be an overall saving ... simpler gain stage(s) with no phase splitter, the PS is similar and perhaps simpler since there is no dramatic current swing but the OP tranny may be more expensive.
Mid bass and mids are excellent with the latter definitely more three dimensional than PP IMO.
Naz
Consider that a speaker with 89dB sensitivity will require 128 watts to play as loudly as a 101dB sensitive speaker furnished with just 8 watts. The efficient speaker sips music from the linear surface of a power amplifier. The 89dB sensitive speaker is a sloppy drunk. Going to more efficient speakers will open up a new world of very favorable possibilities for you.Rgs, JLH
Your post cheers me up - until now, I've been blaming myself ;)My speakers are Yamaha NS8390 8 ohm, 3-way, rated at 70w and 89dB sensitivity. They are OK in a small room with 15w, but any less power than that would not allow enough headroom for transients (I listen to orchestral music quite a lot). In a bigger room, I would prefer 50w.
JLHSorry to be devils advocate here but ...
I have yet to hear a high efficiency speaker
that does not drive me out of the room
because of the harsh/grating sound from them.
I do realize I have more sensitive ears than most.
I have heard about 12 different ones so far and they range
from midly annoying to run-from-the-room-screaming-in-pain.
I have not heard yours, of course, but I am saying
that hi-efficiency speakers don't appeal to everyone.DanL
I am not a big fan of horns, if that's what it takes to get +100 dB efficiency. I absolutely love the soft dome tweeter, which is what is used in the ME2.
Once you are familiar with the sound of a good dome or ribbon tweeter, there is no compomise on that. And those domes could be prety efficient, 95 dB.
It has reasonably flat frequency response, and can be powered by any amp you come up with. Even a preamp might drive it. And one guy who owns an Avalon traded it in after hearing this thing. I can hardly wait to get it.I got the horn bug. When you find yourself to be a horny, nothing else will do. The hair trigger dynamics, the lack of compression, the low distortion, and in this case excellent imaging.
I am getting the BD-Design Orphean coaxial 2-way horn loaded compression drive speaker system, plus SS-powered direct radiators for the bass (the weak link).
I used to own Avalons myself, then ProAcs and my own 3-way design. Finally I got into horns and will not be looking back. I just want better horns.
But I can see why some prefer a more docile sound. I used to be like that. Horns can be docile, if you use a proper SET amp, and don't play too loud. And passive EQ for a bad FR can be constructed as well if you're a designer type guy. I do that sort of thing all the time. I make the horns work for me. It might take a little more coaxing, but it's worth it.
Kurt
aaa
I built a 6H30PP interstage 2A3PP with LLs. I liked it but you really want to use the highest quality interstage trannies. After I replaced the LL1660 with a nickel interstage I found the sound was lifeless before. Probably it is more critical than OPT.I stepped into the deep end of (parafeed) SE ... no turning back for the time being.
Ciao
Gianluca
Replacing Lundahl interstage transformer with UTC HA-107 was a big improvement. Permalloy cores are inherently more linear than silicon steel. They may be not the best choice for output transformer due to saturation issue, but for interstage use they are unbeatable.HA-107:
15+15 K primaries
45+45 K secondaries
20-20,000 Hz 3 dB
Pmax 200 mW
No unbalanced DC
Weight about 4 lbThe core in these transformers is supermalloy (or hypermalloy, as UTC called it). It is 80% Ni alloy annealed in hydrogen atmosphere to achieve the highest mu (about 20,000 compared to 10,000 of 45% Ni permalloy and 5,000 of silicon steel). High permeability allows achieving the same inductance with fewer turns, thus decreasing stray inductance and winding capacitance. HA-107 has two layers of shielding, external mu metal and internal copper. Separate windings allow independent biasing of output tubes.
The Amity/Karna designs rely on a shared cathode resistor with a dual triode PP topology. Assuming the triodes are perfectly matched, I will get no DC imbalance.What kind of real world triode-triode matching can I expect (lets assume JJ ECC99 for now), and what will a "no DC imbalance" transformer allow me to operate under?
Or would I be better off with separate cathode resistors, one of which is adjustable?
HA-107 will take 1-2 mA imbalance in the primary. My driver circuit is a pair of 45 with a common cathode resistor. I matched tubes as best as I could, and then compensated for the remaining difference with additional small resistor in one of the cathodes. Due to drift, there could be 0.5 - 1 mA unbalanced current, but I did not notice any deterioration.
With regards to both the Raven and Amity/Karna, what manufacturer and parts would you recommend?
You got an answer. UTC is a great choice, probably the best for PP. LL winds the same coils also on amorphous cores, that would be an other choice. Tribute or AE in Europe are other choices, they use also amorph cores. Japs are utterly expensive.I picked up a very different beast... the EXO173PPS Ni by Mike at Magnequest. You wont have anymore a PP into PP amp as it works with a SE driver ... but as I said I stepped into the SE world.
If I may, I would suggest you to have a look at the Seth schematics on MagneQuest site just to have a different view on PP amps (it comes with a full and detailed tutorial as well)
http://www.magnequest.com/diy_lessard_2a3pp.htm
A similar schematic was used also by SunAudio
http://www.meta-gizmo.com/Tri/RRM/sun.html
and in almost all the Sakuma PP amps. For example this parafeed interstage splitter
http://www10.big.or.jp/~dh/work/8005.html
Don't ask me to explain why I preferred ni cores ... I can't live without now. That's a fact.
Ciao
Gianluca
Describe your speakers and room, if you would.Poinz
This post is made possible by the generous support of people like you and our sponsors:
Top of Page ]
[ Contact Us ]
[ Support/Wish List ]
[ Copyright © 2025, Audio Asylum, all rights reserved ]
