[rmihai0]

I have a huge respect for Dave and Frank Van Alstine. But  I doubt that a Dyna ST-70 is able to drive the 1801bs. Just because the question oo how much power needs 1801b was just keep coming back in my head in the last couple of months, I decided to test the, So, I gave them to a friend that is engineer in Paradigm factory. He took them to the factory and tested them under the most demanding conditions.

And, the answer to ethernal question is now clear: Ellis 1801b need 176.77 watts per channel in 8 ohms. This would be the ideal. More watts it will be equally harmfull as less watts.

[David Ellis]

He took them to the factory and tested them under the most demanding conditions.

Objectively... exactly what condidtions were tested?  ... you really don't need to answer.  I do realize a complete dissertation on this issue would be extremely long.  Nonetheless, I'd be hapy  to entertain this if desired.

Subjectively... I have heard Paradigm speakers up to the $1500 commercial level and 3-5 1801 customers have happily replaced their Paradigm speakers up to the $3k level with 1801 speakers.   The 1801 hasn't "encountered" a more expensive Paridigm speaker. It appears the criteria & focus of the folks at Paradigm produces a set of results very different than my results.  I obviously believe my results are correct.  Paradigm believes their results are correct.  Even after a very lengthly discussion of criteria and methodology we would agree to disagree.

On the amplifier realm, I can address a few issues, but remain very subjective in my decision regarding amplifier recommendations.  Have you ever listened to the 1801b through an El34 push pull amplifer?  Has the paradigm engineer every listened to the 1801b through a Push Pull tube amp.

Digressing... my very first a/b experience with a tube amplifier was at Mike Bengfort's home in Minneapolis MN.  Mike had 3 amplifiers.  All of them played the same music on some Vifa/SEAS 2-way speakers with @85db sensitivity and no impedance compensation.  My subjective thoughts were:

1.  8wpc Cary 300b SET.  The worst of the group on Mike's speakers.  The saxophone was wonderfuly holographic on a very simple Jazz track, but there were no dynamics.

2.  15wpc El84 push pull Pilot 240 tube amplifer.  Holy buckets... only 15wpc from that cheap looking @1964 amplifier??!!  It was sweet, smooth, fairly detailed and pushed solid sound pressure.  This experience totally astounded me!

3.  VanAlstine 120wpc solid state amplifier @2002.  This amplifier was also very good, and right for the system.  However, at normal listening levels I didn't subjectively find it any better than the 15wpc Pilot 240.

Ellis 1801b need 176.77 watts per channel in 8 ohms.

Summary... I disagre with these findings.

HOWEVER, amplifier "ratings" vary.  Certainly a 176.77wpc Home Theater (HT) amplifier would be just right for the 1801 since the per channel agreed upon (not enforced) standard for HT amplifiers is 1 channel driven at a time.  Given a 2 channel load, the power supply output will drop to about 88wpc.     This would seem about correct.  

There are further issues about amplifier testing with regard to burst tones versus extended loads and their testing.  There is also the issue of even order versus odd order harmonic distortion.  There are actually many issues herein.  I cannot comment at length regarding this issue, but can certainly enlist the help of someone who can if the issue heads this direction.  As such, I am very open to entertain this discussion too.

I will certainly allow that my old Kenwood 125wpc receiver clipped significantly sooner than my Bryston 3B-ST 120wpc amplifier.  My subjective opinion is that the Kenwood audibly clipped at the same SPL as my Jolida 302b.  The Bryston 3B-ST and AVA 550EXR had significantly more power, but this wasn't necessary.  IMO, these amplifiers represent too much added amplifier wattage circuitry the 1801 simply doesn't require from an SPL perspective.

I must be completely honest about my motives in this discussion.  My interest in all of this discussion is purely selfish.  It's in my personal self-interest to ensure that my 1801s are fed the best possible signal possible.  I want the 1801s to sound their best.  This requires wise desicisions in the amplifier realm.

Hmm.... after reading my remarks above, they look slightly agressive.  They look like a very solid "slam dunk".   This sentiment is NOT my intent.  My intent is to simply convey my thoughts on the matter in a very thorough manner.  I welcome any feedback regarding these thoughts, and remain very open to further discussion.

[avahifi]

Years ago I made a comment in Audio Basics that technophiies measuring "watts" was somewhat like weighing meat on a balance scale.  I suggested that the old Stereo Review people could put a 20 pound prime rib on one side of the scale and 20 pounds of White Castles on the other (paper wrapper, soggy bun and boiled onions and all) and say, "these balance, therefore they are exactly the same, so have a nice White Castle banquet".

Unfortunately I have not heard Dave's 1801 speakers yet (that will change this summer) but I do know that the Ultimate 70 does a pretty astonishing job driving my old B&W 801 speakers and these speakers were considered really hard to drive even with most high powered solid state amplifiers.

I can report that most of the time, even with inefficient speakers, the amp power being used is less that 10 watts.  I would suggest that its probably more important to get the first 20 watts really right than anything else.

After you hear a Ultimate 70, I think you too would agree that very few are getting that first 20 watts even close to right.

Frank Van Alstine

[rmihai0]

Dear Dave,

was far from me the intention to compare 1801b with Paradigm speakers. And was far from me the intention to criticize in a way or another Ellis speakers (that I own). My pure intention, as yours, is and was to find what amplifier(s) are best fitted with 1801b crossover and drivers.

I cannot tell you about the testing conditions, because  Iwasn't there. I can assure you tough that no one have seen the crossover or other internal parts. Just because my speakers are closed by Jim Salk and I would never let someone open them other then Jim and you.

What  I can say is that they have been tested for 3 days in a raw. Including tube amplification (VTL and Rogue as long as  I understood).

Regarding the optimum of 177W that is RMS, both channel driven. I am sorry - but this are the results. I didn't posted this to argue, but rather to let other Ellis owners know that they speakers can do much better with more powered amps. I didn't hear yet Frank's Fet Valve 350EXR, but, I guess, that will do justice to 1801b.

Other amps that I didn't listen yet, but  I guess will be a good fit are Plinius 8200P and SimAudio W5.

I know Bryston stuff very well - sorry to say it, but they are nothing else then rudimentar in providing sweet dynamics and "musicality". Bryston and Jolida are not in the league that they claim to be.

[rmihai0]

I just talked with my friend. He specified that any power between 70w-175w it will be just fine. He said 175W is just the optimum, but the results are excellent with 70W also.

[David Ellis]

I can assure you tough that no one have seen the crossover or other internal parts. Just because my speakers are closed by Jim Salk and I would never let someone open them other then Jim and you.

Thanks very much for this consideration.  I did think about it, but it doesn't really worry me.  Professionals in hifi loudspeakers construction are VERY aware of the cost of the components in a loudspeaker.  As such, there is no room for anyone in commercial hifi to profit from the 1801.  Now... if the 1801 sold for $10k I'd have to encapsulate the crossover in a potting compound.  At $5k I might encapsulate the crossover.  At $1500 there is no need for significant protection.

He said 175W is just the optimum, but the results are excellent with 70W also.

I wonder if he used thermal compression, linear X-Max, or maxiumum X-Max to derive his numbers.    

Anyhow, I must admit that my estimation is solely based on necessary SPL for listening, and then a little bit "more".  The "more" depicts an SPL that is simply too loud for listening to anything for longer than about 3 minutes.  These levels are fully viable with 35wpc tube or a good 60wpc SS.  My 120wpc Bryston  and the AVA 550EXR DID push more spl from the 1801, but this level of sound pressure was totally unnecessary IMO.  It was damaging.  At normal listening levels smaller amplifiers make more sense, and sound better - in my limited experience.

I know Bryston stuff very well - sorry to say it, but they are nothing else then rudimentar in providing sweet dynamics and "musicality". Bryston and Jolida are not in the league that they claim to be.

I haven't listened to a completely commercial (i.e. unmodified) amplifier for about 3 years in my living room.  The unmodified Jolida 302b sounded subjectively better than my Bryston, but after mods the sonic discrepancy was profound!  6 years ago I though the Bryston ST sounded pretty good, but really cannot address anything more recent.  Surely there might have been some changes in years past.  I also have never a/b compared any Plinius or Sim gear in my system.  

I auditioned a version of Franks 550EXR with the 1801 manufactured about 2 years ago.  This version fell significantly behind the Jolida 302b, but Frank claims the latest version of his SS amps are the cats-meow.  I cannot attest to this.  I can convey that his Ultimate 70 is profoundly better than I ever expected.  It betters all other amps in my system.  However, the 55wpc AKSA Nirvana + is very close.

Including tube amplification (VTL and Rogue as long as I understood).

Please excuse my memory, but... what are you currently using?  And, did you ever have the opportunity to try a lower (i.e. El34 push pull) powered tube unit?

[rmihai0]

I wonder if he used thermal compression, linear X-Max, or maxiumum X-Max to derive his numbers.

I have NO idea what they used, but they tested them to maximum in anachoic chamber.

Regarding the SPL - here I am fully with you. I never ever listen to more then 90db, and that for very short period time. Even 70-80 db is a little bit too loud for my taste. But it depends - if there is demanding material, I might need to go in this range.

To conclude, most of the time I listen around 60-70db.

I have to agree that for listening experiences under 90db, probably 35W are just enough in medium rooms.

But you will have to agree also that in VERY large rooms (as some of today basements are) 35W might not be enough.

I am currently using a Thule IA60B - 65W per channel, stable to at least 4 ohms, recognized for its POWERFULL bass and slam. And matched with more sensitive speakers, indeed - Thule IS impressing. With 1801b is just another amplifier.

And to clarify a bit - not the loudness is what I need/missing. The quickness and the slam. With more sensitive speakers, Thule was just impressing in proving transients, dynamics and slam. To paraphrase Frank - was a TIGER. Now, with 1801b is a litlle nice cat.

I guess Frank will have another customer for his latest Dyna mod.

I just have to find one in good condition and ship it to him.

[David Ellis]

My current listening room is 14.5' deep 36' wide with 10' ceilings.   While this isn't huge, it does amount to about 5000 cubic feet.  I think the old 35wpc amp does just fine.  My basement will be bigger volumetrically, but it's not finished yet.

More sensitive speakers will have less thermal compression and sound more dynamic.  The only objective caution here is that 4 ohm speakers (most are 4 ohm nowdays) will pass double the current of an 8 ohm speaker at the same sensitivity rating (2.83 volts).  Another way to objectively compare this is 85db @ 8 ohm speakers are equivalent to 82db @ 4 ohm speakers with regard to thermal compression.  There are some other variables (i.e. thicker voice coil wire) that will factor into the degree of thermal compression, but a 3db differential between 8 ohm and 4 ohm speakers is objectively fair when comparing the degree of thermal compression.

There are many areas where distortion happens in hifi.  Thermal compression IS one of them.  However, due to economics, trends and in-store perception there is a continual trend towards LESS sensitive speakers.  

If you are planning to send a Dyna amp to Frank, don't worry about the condition.  My ST70 looked pretty ugly on ebay, but wonderful when it arrived from AVA.  I think Frank has some fancy rust removing paste that he uses on the ST70 chassis.  He also touched up the cage of my ST70 with some matching paint.  The appearance is VERY nice!  I don't know if Frank does this for everyone, but my amp sure looked good.

However, if you have the inclination to solder, I suggest you buy the parts (or just the board) from Frank and accomplish the work on your own.  The work looks VERY minimal.  

Dave

[rmihai0]

Thanks a lot David. I am expecting this week-end a Hafler 9290. (Yes, onr of the first Jim Strickland designs in Hafler co.) It is rated 145W and very few people that I know are saying abut it that is very tubish and musical. (MOSFET output, J-FET inputs). Someone changed for me the caps and diodes (with Blackgates and IXYS).

I will se how I will like it.

If I won't like it I will definetelly become one of Mr. Van Alstine customers.

[David Ellis]

Someone changed for me the caps and diodes (with Blackgates and IXYS).

You might have a winner in that amp.  I have heard from a few gents that the design of power amplifiers hasn't changed significnatly in about 30 years.  However, the parts quality has changed.  Them'r some good diodes and capacitors!

The problem with many capacitor swaps for older amps is real-estate.  Simply, there just isn't enough physical space to use good capacitors in most SS gear.  For some reason better electrolytic capacitors seem to be bigger than lesser quality units.  

I planned to accomplish a gut and replace for the capacitors in my Marantz 2270 receiver, but took one peek under the hood.  After about 3 minutes of debate I decided... no way    .  It would be much easier to start from scratch with a new case and circuit board.

Dave

[rmihai0]

Sunday I sold the Thule IA60B and I plugged in my "new" Hafler 9290 (modded - changed caps and bridge rectifiers). For the first 20 min, nothing to mention. But after, I decided to watch Jesus Christ Superstar. What a pleasent surprise. Acctually HUGE surprise! Hafler, although rated as "lower" hifi then the Thule, is doing a perfect job with the Ellis 1801b. Hafler is rated 145w per channel (Thule is 60W). Is more then I expected: of course Hafler is manging the speakers with great ease and non-chalance. You cannot feel the hidden power of the Hafler until is required, but when there are fast transients and sudden lows, then Hafler delivers. I am surprised also about the beautiful NATURAL way of delivering the mids and the highs - absolutely nothing tiring or strident. It sound just simply natural. IS not having that fancy "BLACK SILENCE" that so many audiophiles are looking for, but is just simply natural sounding. I guess that HAfler guys knew what they were doing when they designed this amp. is not fatiguing, and is very natural.

Finally I can say that my Ellis 1801b came to live. Now the Ellis speakers are delivering so much better then before. Until now I didn't appreciated them that much. Now, indeed - in this combination they sound very natural, a little bit romantic and beautiful. Is like a breeze.

As I expected, low sensitive speakers like these are loving power. Of course, the right power.

[David Ellis]

modded - changed caps and bridge rectifiers

Yep, better components can make a huge impact in the quality of an amplifier/preamp/CD player.  

I think you said there are now primarily Black Gate caps and IXYS Hexfred's - correct?

I my humble opinion, the changes in component quality in the past 40 years FAR outweigh the changes in design quality.  

Who modded your Hafler?

Dave

[rmihai0]

Hi David,

Hafler was modded, as all other electronics that I used by a local technician from Toronto (actually Hamilton). A local dealer who is selling VERY expensive and fancy stuff is using the same guy for repairs. So, I said to myself, if Rob (the dealer) trust this guy to repair for him electronics that cost several thousands dollars, why shouldn't I trust him to mod my electronics.

He replaced the caps with Blackgates and the diodes with IXYS.

He is a nice guy, and now is all over in my Coda Continuum Window II that is having the right channel dead (due to shipping problems - the package was opened at customs and NEVER closed after!!!). So, couple of days came as an open box, and God knows what else happened with it.

[David Ellis]

Unfortunately you aren't the first guy to have something damaged in shipping.    

Dave

[rmihai0]

Now. I have done my own measurements and tests regarding the power needed for Ellis 1801b and SPL. In my room (18x16x10), MAXIMUM that I will ever need are 50W per channel. Supposing I will leave a little bit of headroom for not having distortation at maximum volumes, will mean everything needed are 70W per channel.

Now, very interesting, I found out that most of the time, at normal listening levels (SPL 75-85db), most of the time are used ONLY 7-8 watts!!!

Now is making sense why Jolida 302b (50W) and AKSA 55N+ (55W) are good matches for Ellis 1801b.

But, if you would like to cover ALL the bases (the lows), without introducing distortation, then 70W should be chosen.

[rmihai0]

I used this document to due my calculations: http://www.jblpro.com/pub/manuals/pssdm_1.pdf

[David Ellis]

Argh!!  I just typed a big long response, but it got dumped.  I think the library computer doesn't connect with the audiocircle server very well.  this response will be much shorter.

Anyhow...

Now, very interesting, I found out that most of the time, at normal listening levels (SPL 75-85db), most of the time are used ONLY 7-8 watts!!!

I am not sure if you are referring to using 7-8 watts average or 7-8 watt peaks.  However, there is a very easy way to measure the average power dissipated across you speakers.  It applies the following very simple forumlae.  I won't get into any of the thicker stuff here, because average power consumption is only 1/5th or 1/10th of the peaks.  This application is, howver, very useful to determine how much power you are really using when you listen.

1.  Watts = Amps * Volts

2..  Amps = Voltage/Resistance in ohms

The method for application is thus.

1.  Measure the Resistance across your loudspeakers with NO speaker cable attached.  This resistance should be about 6.2 ohms with the 1801

2.  Turn on your stereo at a comfortably loud level with MUSIC being applied through the speakers.

3.  Measure the Voltage drop across the speaker terminals.  For simplicity sake, we will assum 3.1 volts of drop across the terminals.

Applying these numbers is simple.

Amps = 3.1 volts/6.2 ohms resistance

There are .5 amps of current flowing across the terminals.

Watts = 6.2 volts * .5 amps

There are 3.1 watts being dissipated across the terminals.  

A few of my customers have accomplished this task and all have been VERY surprised.  

Eventually I'll get around to some very real satire regarding how the marketing game of "high power" amplifiers came into vogue.  For now... I'll continue to convey that in the realm of hifi compromise, the smartest set of amplifier compromises for the 1801 points toward about 35wpc push pull tube (commonly El34), or a 60wpc SS amplifier.  More power isn't better.

[rmihai0]

Just to let everybody know - I got a Dynaco ST-70 and will be modded by Mr. Van Alstine. I will let everybody know my opinion when it will be here (about 2 months).

[David Ellis]

Groovy,

It may not be completely clear from my comments elsewhere, but Mrs. VanAlstine and I agree on the choice of coupling capacitors.  However, only she has actually performed a solid a/b test.  We prefer sonicaps.

Frank will likely grumble a bit if you ask for them.  I suggest simply reminding Mr. VanAlstine that his wife is correct.    

The capacitors will cost about $70 from Soniccraft.  You could certainly do this work later.

BTW, Frank made my old grungy ST70 look very nice  He's got some really slick polish and some paint that matched the cage perfectly.

[skrivis]

Groovy,

It may not be completely clear from my comments elsewhere, but Mrs. VanAlstine and I agree on the choice of coupling capacitors.  However, only she has actually performed a solid a/b test.  We prefer sonicaps.

Frank will likely grumble a bit if you ask for them.  I suggest simply reminding Mr. VanAlstine that his wife is correct.    

The capacitors will cost about $70 from Soniccraft.  You could certainly do this work later.

BTW, Frank made my old grungy ST70 look very nice  He's got some really slick polish and some paint that matched the cage perfectly.

I just thought I'd add that AVA has taken what they've learned from the U-70 and applied it to a new model, the T-8 preamp. All tube, and they say it outperforms all of the older preamps, even more expensive ones.

I do agree that AVA goes above and beyond in terms of cleaning up customer's equipment. Frank told a tale of a customer who sent in an amp, and evidently the cat had been peeing into it... They had to take it out and leave it in the sun for a while...