[totti1965]

This topic is new. Dave asked me, if i would post my question I asked him via email into this audiocircle.

Please don´t be angry about my bad english. I am the first german customer of Dave I think, and I will get my pair perhaps in September; perhaps I can make some of Daves Ideas popular in Germany also.



I "tortured" Dave with many many many questions the last 18 months and he answered any of them with great patience. So I became an expert of speakers also, but more in a theoratically manner......
If other users in the circle wanted to contact me, here is my email: thorstenwieszniewski@onlinehome.de



Here is the Question I asked Dave:



Hi Dave,
 
Is it possible to get a measuring of the resistance of your speaker over the Frequency range 20 Hz   to 20 000 Hz.
 
My  english is awful but I think the name is Impedance Plot or so.....
 
I don´t use vacuum tube amps till now,
but I have read, that it is easier to drive speakers for tube amps, when the speakers have not so much range in Impedance variations...
 
So a Variation between 4 Ohm and 6 Ohm is better for the Tupe amp than the variation betweet 4 Ohm and 40 Ohm for example.
 
Do you or did Dennis Murphy do any impedance Correction in the crossover?
 
Well, I know that it doesn´t bring any adventiges in case of using transistor amps, but........
 
 
so , till next time,
 
 
yours
 
Thorsten

[David Ellis]

This is a very valid concern that I had a few years ago too.  It started with an experience with a Cary 300B tube amp and a 86db Vifa/SEAS loudspeaker.  The other amp in the system was a 120wpc unit from www.avahifi.com . Quite simply, the Cary 300b fell on it's face.  Initially I thought this was due solely due to the power output.  I later learned the same information that you are learning.  While tube amps CAN sound very good, there are many ways to err in their implementation.  I am not an expert, in the realm of SET/Tube amps.  If you still have questions about my post, please ask.  When my answers are incomplete I encourage you to ask Frank VanAlstine these questions in his forum.  Frank has lived with and understands all ends of the amplifier spectrum intimately.

I believe the shortfalls of tube amps lie in their low power output, and also the dampening factor.  I have heard how sinful ALL tube amps are from a few folks, but none of these folks have heard my setup or measured amps on speakers with questionable impedance.  There is considerable discussion about power tube amplifier dampening factor and common loudspeaker impedance.  As you notice, there is more to a speaker than whether it has high/how impedance.  This would seem to apply to the 1801.  There are swings.  To this my response is simple.  

I attend the symphony regularly.  After hooking my el34 push pull tube amp in my system I never re-attached my Bryston 3B-ST   I don't know that I am a tube-guy.  I am an accuracy guy.  The SS Bryston was quite gritty.   I have subsequently learned the VanAlstine amps are much cleaner, but this was after I purchased my little tube amp.  I remain very happy despite the few critics who convey POTENTIAL problems relating to dampening factor and impedance swings.

As an aside.  There is a very valid argument about the capacitance of Goertz speaker wire.  Critics convey that it will surely cause oscillations problems in amplifiers.  Well, my EL34 push pull SHOULD have problems, but it doesn't.  The Geortz wire has the lowest characteristic impedance of any wire (to my knowledge), and this matters.



If you need it, I can take comparative frequency response measurements of my SS test amp versus my tube amp too.  However, this will take a couple hours of work (many wires), and I'd rather be building speakers.  Please let me know if you have a strong "itch" for this graph.

[totti1965]

Thank you for your quick answer!  Thank you for the offering of more measurement but of course there is no strong "itch" for this graph!

My thoughts to that theme are the following, and I hope there are not to much details:

1. I have heard the name Frank Van Alstine first in 1988 in Germany! The world is very small!!!! Frank did some Modifications and I have heard one of them at Torsten Roscher in Wuppertal, the legendary German High Fidelity "GURU" where people went 500 miles and more to hear music in his 38 ft by 16 ft large room. Torsten was the guy who takes the well tempered lab of Bill Firebaugh to Germany and combined it with Decca Systems (Garred Modificated). It was a
total Linn Killer!
I bought there my little Monitors wich gaves me so much understanding in "hearng right".......

2. Yes there are swings in the Impedance plot. The fact is:
some tube amps works nevertheless better with a speaker who produces swings in the Impedance plot than some transistor stuff.
That is because some tubes are not so awfully sensitive for impedance swings than other.  Some other tubes are so much higher in their "ground Quality" that the transistors who have an easier work can´t beat them nevertheless.......

3.  The dominant Question is:

Wouldn´t it be better to make the tube the work easier nevertheless???
I think that it needs acoustical comparence of one and the same amplifier
(for example a very nice small tube) at two versions of the 1801b.
One standard and one with an impedance correction.

4. Need of and  Advantages and Disadvantages of an Impedance correction.

4.1.
The 1801 b could use one or two circuits: One circuit for the 53 Hz drop and one for the 1600 Hz drop.

The 1600 Hz correction would be more important.
Both circuits needs 3 Pieces: One resistor one capacitor and one conductor.

4.2.
The advantage would be: easier work for many tubes. (This is why magnepan speakers works so beautyfull with tubes, although they have an ugly need of power!!!)

4.3.
The disadvantage:

4.3.1.
There is absolutly no need of this if you uses transistor amps.

4.3.2.
It needs high Quality parts (resistor, capacitor and conductor!) although the two circuits are parallel. (The substantiation of this is a little bit complicated and my english is not perfect, there is an Artikle about this theme in German Speaker Builder Magazin "Hobby-Hifi" 4 / 2004)
If you want to make it a little cheaper than it is better to save the mony for the parts /part-quality of the 53 Hz circuit, the 1600 Hz circuit would need absolute quality!

The greatest disadvantige is of course that it would increase the price of one pair of the 1801b for about 100 or 200 US $ (Depends to one or two circuits- more soldering, good quality of the parts).



5.
The dimension of the Parallel circuits:

I can´t give you the exact dimensions of the  R C L   Circuits but here is a
estimate I can give you because I have read many many Circuit plans in speaker Builder newspapers and I have found two similar 25 Liters constructions with 6 inch  -8 Ohms drivers (bass) an 8 Ohms tweeters:


The correction circuit for the 1600 Hz Peak could be about

R =  3,9 Ohm  20 Watt MOX

C = 47 µF Sonicap

L = 0,22 mH  16AWG Foil conductor

(These are the exact dimensions of the correction network for the Exel E 18 W 001 as Joe D´Apollito uses it in the German modification of the Thor)


The correction circuit for the 53 Hz peak is not so important but it could be about:

R = 1,0 Ohm MOX 5 Watt minimum

C= 330 µF  (no need to buy most expensive here!)

L = 15 mH (no need to buy most expensive here)



6. The goertz cable theme:

Yes Goertz speaker wires have an extremely high capacitence. The fact that you doesn´t have any problems is easy to found: 1. these kind of problems are no problems specially for tube amps! Only transistor amps with their High current will get problems (theoretically) For tube amps there won´t be Problems!  2. the short ways the current has to go will not gave any problems (it is just about one or two ft isn´t it?) - even with transistors:

The alternative for goertz cables (the real world record holder in low inductance!!!)    is nordost flatline
 http://www.nordost.com/Cables/speaker-flatline.htm     or superflatline in cross connection linking (1; 3; 5; 7    and 2; 4; 6; 8 etc.)

Sometimes I think, that the Nordpst People doesn´t know how good their stuff is, because they doesn´t use the trick with the cross linkink!
The inductance is Lower than the goerts one and the capacity is 6 times as low.



The measurings in stereoplay (2/ 1994) proved it  /it is world record till now, Goertz is Number Two I Think!

7. What, if my suggestions of the dimensions are stupid and false?

There will be absolutely no big change in the frequency plot of the 1801b, because the suggested circuits are parallel.

So the differences are extreme little and more a kind of helping the amp to to his work good.  There will be an improvement, even if my dimensions are not 100 %.   Ask Dennis Murphy for precise Dimensions.



8. Abstract:

So, why don´t make a test, Dennis Murphy will just need half an hour for checking out the exactly dimensions of the units and printig a slightly (1600 Hz !!) modificated crossover circuit. It takes only one Resistor one Capacitor and one conducter and I think that the dimensions I gave You for the 1600 Hz Peak are perfect and the Dimensions for the (much less Important) Correction of the 53 Hz Peak are pretty all Right.


So perhaps there will be an alternative kind of 1801b,
a kind of 1801b  ic     (ic for impedance correction) for about 100 US $ per compleated pair more and a Correction of the 1600 HZ peak?


Yours,

Thorsten

[David Ellis]

I must be completely frank about my opinion on this matter.  I believe there are many backyard "laywers" with knowledge of electronics relaying hear-say about the matter of tubes and impedance swings.  Simple folks, like me, intially run at the fear of any tube amp with any speaker not having impedance compensation.  As you have said:
 

2. Yes there are swings in the Impedance plot. The fact is:
some tube amps works nevertheless better with a speaker who produces swings in the Impedance plot than some transistor stuff.
That is because some tubes are not so awfully sensitive for impedance swings than other. Some other tubes are so much higher in their "ground Quality" that the transistors who have an easier work can´t beat them nevertheless.......

In this matter I must admit the only testing I accomplished with with my humble tube amp and my humble speakers.  The tube amp is an El34 push pull, and my speakers have fairly significant impedance swings.  I cannot substantially comment on the effect of a SET 300B tube amp or any other that might have more profound dampening problems.  I can purvey the remarks of Joe D'Appolito on this matter.  He might be the most knowledgable person in the realm of loudspeaker design.  

The topic was the Usher loudspeakers and impedance compensation.  He mentioned that his initial design had impedance compesation.  After extensive listening the folks at Usher decided not to use impedance compensation. His non-verbal expression was interesting.  He conveyed that impedance compensation didn't matter.  As a scientist, I am sure he understands the theoretical impact.  As a listener he also is able to judge with his ears.  

3. The dominant Question is:

Wouldn´t it be better to make the tube the work easier nevertheless???
I think that it needs acoustical comparence of one and the same amplifier
(for example a very nice small tube) at two versions of the 1801b.
One standard and one with an impedance correction.

So to clarify your assertion:  Does the impedance swing cause feedback that will effect the output of the amplifier due to the amplifiers poor dampening?  

I will answer the rest of your questions following a clear statement of your assertion.

[David Ellis]

I just drafted the rest of my response to Thorsten's assertion, then my internet froze - ARGH   I lost the entire post.  Nontheless, I remain eager to close this discussion with defining feedback for Thorsten's worries.

I accomplished a test using the 1801b.  I performed a frequency response using my 1988 Kenwood receiver and my Jolida 302b.  The speaker and the microphone were not moved during this test.  

The goal of this test was to measure the impact of the 1600hz impedance in the 1801 on the Frequency Response.  The hypothesis presented is that the tube amp would suffer in this region due to the poor dampening and feedback/phase angle of the 1801's impedance load.  The results are fairly simple:



My Jolida 302b, a push-pull EL34 tube amp has no measurable problems with the 1801s impedance load.  I can also convey the Jolida 302b has no audible problems with the 1801b's impedance load.  IMO, the 302b sound significantly better than any SS unit around $1k.  It certainly sounded better than my Bryston 3B-ST too (retail $1700).

So, while there is good theory about tube amps haiving problems with difficult impedance loads, I don't see or hear the problem here.  I have no intention of purveying any kind of impedance compensation with the 1801.

I must admit that I have NOT tested the 1801s with any SET amp, and cannot vouch for this setup subjectively or objectively.  I also have no intention of accomplishing this test.  This would require me to buy a SET amp and test the setup.  I don't plan on accomplishing this.  I have only heard 1 good SET setup.  The speaker used was the Tannoy Churchill.  The setup sounded very good, but the 85db sensitivity of the 1801 is a far-cry from the 100+db sensitivity of the Tannoy Churchill.  Further, a very good tube guy who winds transformers for a very prestigous tube amp company told me that a good SET will run in that first 1 watt of power.  The 1801s require more than this.

So, why don´t make a test, Dennis Murphy will just need half an hour for checking out the exactly dimensions of the units and printig a slightly (1600 Hz !!) modificated crossover circuit. It takes only one Resistor one Capacitor and one conducter and I think that the dimensions I gave You for the 1600 Hz Peak are perfect and the Dimensions for the (much less Important) Correction of the 53 Hz Peak are pretty all Right.

Dennis tested the audible impact of impedance compesation when origionally designing the 1801. There wasn't any impact.  The circuit is fairly simply but very unnecessary.  

I realize that I am missing a prime opportunity for marketing.  I certainly could ride the band-wagon of guys professing the need for flat impedance with all tube amplifiers.  I could complicate my product, derive some very good hype, and raise my prices accordingly.  The lack of my aggressive nature in these matters might well be my future demise.  If I can't hear it, and I can't measure it then I won't sell it.

Digressing... Maybe all the folklore about these problems did indeed come from the old days when SETs were the only good amplifier available.  This problem certainly could have very good historical validity.  

Hopefully I addressed your concerns and questions.  They were numerous, and show some education.  Also, If you have any graphs that disprove what I presented above, send them to me via email, and I'll post them for you.  This is good meaty material!

[BeeBop]

I never re-attached my Bryston 3B-ST  I don't know that I am a tube-guy. I am an accuracy guy. The SS Bryston was quite gritty.

Apologize for taking you off topic but have you tried the 3BSST? I have one which I am running with ACI Sapphires and do not find it gritty at all.

[totti1965]

<<<<<<<<So to clarify your assertion: Does the impedance swing cause feedback that will effect the output of the amplifier due to the amplifiers poor dampening?
 
I will answer the rest of your questions following a clear statement of your assertion>>>>>>
 

Yes, my assertion is, there are small differences in the Phase and the Frequency response between tube and transistor- amps if there is no compensation of the Impedance swing (about 1,0 dB or less, depending on the variations of the impedance)
 
There were many analysis on that points in the last 15 years I guess.
The last one was comparative frequency response measurements on an Klipsch RB 75 with transistor and tube amps in stereoplay 5 / 2004.



I must try to find someone with a scanner in funktion!!! Than I send it to you!

 
The point is:
 
There are two ways to look at these thinks:
in an theoretical manner and in an acoustical manner
 
The theoretical:
 
If the frequency (voltage at the output on every frequency) would be perfect (but of course no speaker in the world is as perfect as the voltage output of a transistor amp!!!!) there would be a small difference between reaktion of a tube and a transistor.

If the aim is an identity in the frequency reponse between a transistor with e.g. a damping factor of 100 and a tube with a damping factor of 10 or so. This aim is not to reach without the impedance Correction.

But you are right in the result:

The Question is: Is in the concrete case of use this aim is the right one?

Your alysis for 1988 Kenwood receiver and the  Jolida 302b is great and remarkable!    I need the exact name of the receiver to answer in a analytical way..... there are some words to say speazially to Kenwood!!!!

Please let me know!


Now the acoustical manner to look at these things (also psychoacoustical components are part of that theme!):

The frequency response of any very good construckted speaker has a perfection on about + -  2,5 dB  more Accuracy is not realistic.

The influance of a speaker with very large peaks in the impedance plot is  about minus 1 dB!

The differences are (Minimum!) 5 times as small in the result as the normal ups and downs in the frequency plot!!!

So the effect is at first:

VERY VERY LITTLE!

The small effect is at second in the result (the sound that reaches the ear in the room!) not judgeable in the manner of better or lusher. It can be
a small advantage to use the tube amplifier and the result can be a flatter frequencyresponse as if you use the transistor!!!!!

The result can be also a frequency response with a little bit less accuracy.

But the results are NORMALY not so easy to divine. And NORMALY there is no IDANTICALLY Result. So I am astounding a little bit of your results.

A Speculation of the Results with other Tube stuff and other Transistor stuff than you have analysed (It is a little bit more complicated than just a loss of 0,5 dB or so at 1600 Hz with a tube):

Grundton The undertone range is the one with most sensitivity for the ear.

The impedance minimum of most speakers is in that range:
The impedance minimum of the 1801 b is in the range of 100 Hz to 400 Hz also.

The musik reaches the speakers with a little bit smaller voltages in that range with tube amps.
Normally the sound pressure is adapted for that range, so the button is swiches a little bit to the right (direction 12 o´clock).
The result is: at any other frequencys there is a plus between 0 and 0,5 dB perhaps sometimes 1,0 dB  a verry very very small "loudness effect"
(nearly not audible!) - one (of many!) reasons for the sweeter sound of some tubes!

Because of hearing tubes in that ranges (for example 1600 Hz) a little bit louder (not lower!!!! -because of more impedance will increase (!!!!) the voltage in low damping situations (!!!))

The results for the 1801b might be:  a little little little bit more "punch" with tubes in the area under 100 Hz to 53 Hz (the maximum impedance).

A little little little bit more drive in the area 1000 Hz to 2000 Hz with the tubes.

You sended me the graphs for my speakers. Thank you very much!

The graphs shows nearly absolute perfection in the frequency responses under 0 degrees and 15 degrees off axis. Perfect work! But if you will get a plus of about 0,5 dB to 1,0 dB plus in the range of 1000 Hz to 2000 Hz (by using tube amps) it becomes absolute perfection.

So that perhaps is the reason why the 1801 b sounds  slighthly slightly slightly better with tube amps just in the manner you build it!!!!

This is the reason why your answer of my comments is in the result absolutely correct!   The 1801b needs no impedance correction in the area
of 1600 Hz.   But the foundation of that is a little bit more complicated as you thought.



2:

One word to your measurings with the kenwood!
I need the name of the receiver!  The Kenwood is perhaps a very very very ugly example for a modern high dampening Konstruktion of the year 2004!
Modern good transistors have good dampening (low output voltage at EVERY FREQUENCY BETWEEN 20 Hz and 20000 Hz   e.g.   0,04 Ohms output resistance from 20 Hz to 20000 Hz.

The Kenwood stuff in the 80 ´s was sometimes a catastrophy in Constance, the results were Crazy (like negative damping and so on)

Perhaps that is one reason why your measurings seems at first sight so astounishing!

Normaly there must be some measurable differences. If these differences would be acoustically there, they became, as I sayed a little little little Plus in the sound reality (flatnes of frequency response) when using tubes with the 1801! So you are right.







The normal tube   vs.  SET   thing:

the small effects must be there with normal tubes also but perhaps a little bit larger with some SET's.  It depends on the construction of the tube and not NECESSARRY on the fact SET or other tube.  



The Joe D' Apollito thing:

First:

I have a good link for a D' Appolito Calculator for those guys who understands a little bit German language:
http://www.visaton-bausaetze.de/tools/index.htm?http://www.visaton-bausaetze.de/tools/t_dappolito.htm

Second:

Joe build the Usher  Dancer 8571.   That speaker has perhaps no Correction of the Impedance. But the Plot that I can see in the Papers shows a beautifull flattness nonetheless: The range is just between 2,5 and 8 Ohm between 40 Hz to 40000 Hz.  So Joe doesn´t need the correction anyhow!!!     If the peaks would be larger, I don´t know what he would say.... The only serious answer could be:   it depends................................




The measurings of stereoplay:   I must find someone who knows how to use a scanner.  I have one, but it doesn´t work....
Then I will post it to you, if you want it with some translation from the German language.... (but it takes time)

Hope this posting is not to long and you are not bored.....

[David Ellis]

Yes, this is a very complicated issue.  I am certain those who know huge quantities about this information could discusse the matter of impedance feedback for amplifiers for hours, days, months, years... .  I suppose this issue is similar to Theology.  Those who understand it best can discuss it endlessly.  I have become a very simply guy in my Theology, and will likely remain a very simple guy in my amplifier understanding.  While there is considerable depth to this issue, if I can't measure it and I can't hear it then it's a non-factor.

I can barely hear a 1 db difference in amplitude.  I only tested this in 2 settings a/b with different amplitiude resistors on the 1801 tweeter.  In Dennis's living room I couldn't discern any difference.  Dennis and another gentlemen agreed.  In my living room I heard a slight change in balance with different tweeter resistors.  Hence, I don't think any amplitude shift in the 1801s a valid consideration when examining this issue.  The Jolida 302b is simply a cleaner sounding amplifier than the Bryston 3B-ST.  

I must remark that the bass of a few tube amps is very flabby.  While the 302b is not world class, it's also no slouch.  The 302b bass is very nice.

My Kenwood is a KR-V 126R .  Surprisingly, it puts out a pretty clean 35wpc (my guess) before it starts to clip.  The receiver certainly isn't a hifi $ unit, but it sounds marginally better than any NAD/ADCOM that I have heard.  It also sounds better than my other testing amp - a Technics SA8.  This one is about the same vintage and sold for about $200 more.  Unfortunately, it's very gritty.

Sure, scan those docs.  I'll post em.  When we are both at the end of our knowledge ( this should happen shortly), we can ask a smart guy for insight.  Frank VanAlstine will know.  Hopefully he will be willing to share a couple paragraphs with us.
[/code]

[rmihai0]

bump