Rerurn to Romy the Cat's Site

Audio Discussions
Topic: Transient effect and DSETs

Page 1 of 1 (15 items)

Posted by Romy the Cat on 01-31-2008

After finishing my  left 6-chennal Melquiades and putting amps and Macondo back as it was intended to be I was expecting that it should be “better” but I did not expect HOW MUCH better it might be. Many aspects of sound are more interesting now, with much defined improvements just across the entire range. However, I would like to do not concentrate on the specific advances that Macondo-Melquiades have in their current state but rather I would like to pay attention to the phenomenal opportunities that a DSET idea offers in context of a multi-way acoustic system.

It is truly phenomenal as active channels open so many creative and expressive opportunities. The balance and the minute satellites of the individual channels might be dialed- in very accurately and without affecting anything else. It changes the entire thinking about acoustic systems and let to view it from a slightly different angle.

Let presume that a center of you attention is MF Channel and you would like to increase not output but “weight” of the MF range’s presentation.  The output volume of MF is the MF Channel’s upper knee and the weight of MF Channel is the MF Channel’s lower knee. So, you have multiple options.

1)  To load the amp that drives MF Channel very slightly harder remapping the output transformer and EQ volume.

2)   To change the amount of current in the tube that driver the MF Channel, effectively changing the plate impedance and consequentially the load damping.

3)   Increase the output of Fundamentals Channel – the next channel toward to LF after MF

Interesting the all 3 ways to deal with the task create slightly different results and mixing them it is possible to shape very-very exact sound you in your mind. Moving across the enter range it possible to control and to structure the transients and harmonic characteristics of an acoustic system with very high degree of correctness. Also it is possible very accurately match the transients/harmonic balance of loudspeakers with the patterns of given acoustic environment of the listening room and with own listening preferences and priorities.  In the end the right questions are popping up and fully active audio installation from the adolescent inquiry ”how does it sound” become to question the right questions: ”who are you” and “what and why you want”.

What is also is very interesting that a change of the Channels’ volume has no impact to the perceived frequency response of the entire acoustic system. It is hard to explain and it should be heard. The change output of ONE channel has a “special” felling. It is in a way remanding the feeling with which the properly working Lamm L1 preamp was able changed volume. Very interesting effect indeed …

Anyhow, the fully active system is a wonderful tool, though a lot of pain in ass to maintain Still, I feel anyone who went DSET and multi-way acoustic system would hardly ever look back… 

The caT

Posted by Romy the Cat on 03-02-2008

With all hugely exuberant welcoming and DSET and multi-way acoustic system approach I have to admit that it is a enormous pain in ass to maintain the system.

Two weeks ago I begin detect that my left channel somewhere in upper MF began to throw some mechanically sounding misrepresentations. I figured that MF drivers picked the dirt and need to be cleaned. I opened it up and cleaned it. Three times, as the alien “sounds” did not go away. Keep chasing it as debugging it further I figure out that one of my… (believe me or not!!!) shock-absorbing tube adapter on 6E6P is responsible for the problem. I plugged the tube direct and the host socket and the problem is gone. Then I opened up the shock-absorbing adapter, looked in it and … closed it back. Then just for fun I decided to try to use the same adapter again – the problem was gone. How fun is sit if you have 12 of those adapters and perhaps the same idiotic behavior? Two weeks I was wondering where the hell THAT sounds was coming from!!!

The caT

Posted by Paul S on 03-02-2008
Just the symptoms caused by the bad electricity, tons of tubes and other parts in various stages of decline, pins/sockets/connections, along with analog and digital source issues is the stuff of madness.  I am so happy when everything works that I start to hyperventilate.  I can only imagine adding all the complexity and interactive issues of 5-way/DSET as you have done.  I recently tracked down a subtle pulse in my left channel to dirty tube pins (phono stage), and I had just cleaned those pins 3 weeks ago, I think.  I guess it might help to keep a log...

Anyway, one question I do not like to answer to myself is how much I get to listen via my system at its best.  I am glad to be finally figuring out how to listen to and enjoy all kinds of music with my system at sub-optimal performance.

In fact, I think a semi-conscious sub-strategy of mine for a while now has been to try to get the seemingly impossible combination of ease of use with optimal performance, with the former trumping the latter at times with respect to implementations.

Good luck with the rest of the gel thing-ees.

Best regards,
Paul S

Posted by Romy the Cat on 07-11-2008
The 1933 articale in Radio-Craft by W.H. Hollister about DSET system

The Cat

Posted by Romy the Cat on 11-14-2008
With all advantages of DSET with multi-ways it is a pain in ass to maintain it. For instance now I am chasing for 2-3 weeks a mean thing. Whoever use 6C33C know how the tube dies – it begin to produces from time to time those dry non- harmonic bursts. They are not dangers for speakers and they are a good indicator that the tube needed to be changed. So, one of the tune of my right DSET begin to “ping” once in few hours but cannot found out what one. Based upon the frequency of the “ping” it might be any of 3 channels, but which one? It would not be a gig deal to say which channel is guilty but I need to be right next to the speaker when the amp pings. Since the damn channels are well-integrated it is very difficult to get which channels ping with “side vision”. It is funny to say but I ended up disconnecting the channels and playing “as is” waiting for the next erratic burst in order to find out which amp cause the problem.

The CaT

Posted by Romy the Cat on 04-28-2011

The people who use a single driver speaker believe the while their driver move back and forth for it also radiates a problem-free MF and HF. Near the same people believe that if an amp has capacitors with precious metals foil then it not only do sound better but cure cancer and  heal consciousness after an abortion. Still amplifiers do not read stupid audio forum and they use own mind to operate.

One of much ugly moment in amps operation is handling complex LF ULF signal.  Even if the amp has no LF restricting magnetic in the signal path then inner modulation of the signal do a lot of damage to sound. While you amp is and it’s power supply play care a fine, fraction of dB flute tone the very same amp at the very moment might care some crash from bass sections. Did you note that most of playbacks, while handling some strong and powerful signals, do vandalize fine minor signals?

A month of so ago a local guy stopped by to hear Macondo and he commented: “for some reason Macondo does not collapse with stress”. Sure 14 channels and 14 amps, 12 of them in Class A1 – why would it be stressed?

Anyhow, of if the secret sin my view is compartmentalization of bass. Make an experiment. Mix at your preamp 50/50 signal from your CD and your TT. Stop the TT spinning, put an LP on platter, lover your tonearm to the stopped record, turn off the LF high-pass filter if you have it on your phonostage. Now let somebody to jump next to your TT. Pay attention to your woofers. You most likely will not hear the sound as it will be too low but you will have very good ULF in-rash. Now, crank up you playback, play a nice and soft music from CD and let your “jumper body” to rock your woofers by phonostage signal. Pay attention that sound of your   Mozart Flute Concerto from CD will be to a large degree depending from how your jumper jumps.

In the multi-amping situation your HF amp will see only the signal from your CD player and bass amp will see the signal from LP and no matter what kind stress the LP amp has it will not be passed to MF amps. I do not say anything new. It is known to anybody but for whatever reasons is not used by anybody in High-End audio. The Morons are accustom to build and to buy power amps that cost comparative to annual income but they do not organize playback with 2 or more cheap  multi-amping amps, dedicated to own respectful frequency range.  In my view it contradicts with the notion of High-Endnes. There is nothing wrong to use one amp but using cost-benefit analyses multi-amping way beats “better amp”.

The Cat

Posted by Paul S on 04-29-2011

Speaking for myself, I went big hoping to power my speakers FR with 1 amp per channel.  I did this partly because I wanted to keep it simple, and partly because I did not really think it through before I got that far.  At this point, while I am not clear about how I will balance gain throughout, I will definitely be adding amps, and I see no reason why they should be big and expensive, if they are matched and limited to specific ranges and loads in the first place.  In the end, there really is no single "best" amp, but only the amp best suited to its task.  Taken at face value, this is a case where "good enough" really is good enough.  Hell, maybe they don't even need to be SET...

Best regards,
Paul S

Posted by N-set on 04-30-2011
I've been trying to theoretically understand this LF - MF,HF interaction for years without much success.
One thing I've found (putting aside the interaction via PS, the LF OPT core saturation, etc) is that frequencies
close to -3dB points get dynamically attenuated for some small period of time.

This is a pure transient effect, not visible via normal Fourier analysis. The idea is that any reactive element
stores energy and hence needs to charge. This charge comes from the signal. The time scale of the charging
is given by the time constant of the coupling=-3dB point. The closer the signal freq. to the -3dB point of the coupling,
the stronger the attenuation. So when a strong LF+MF, HF signal suddenly appears the LF part, being much
closer to the -3dB point gets attenuated for a some small period of time, while MF,HF not.
This applies to RC coupling, output transformers (Lpri needs to charge), LC coupling, DC coupling with inductive load.
I wrote a small note were this effect is calculated. If someone's interested I can send it.
The moral was that if one wants to loose at most 3% of the signal energy, the -3dB point should be
4x lower than the lowest freq. reproduced.

I have no idea if all that has any detectable effect. I'd also blame the PS intercation and, more importantly,
OPT core, which for strong LF signal becomes less linear than for a weak MF,HF signal. Or all together.

Posted by Paul S on 04-30-2011
Very interesting, N-set.  Of course we have been setting typical RC out at, say, 2 Hz for many years; but still I have puzzled over a sort of sonic "sump" and/or bungee effect that seems to increase as the sheer mass of parts begins to pile up.  On the flip side, what sounds better than a small SET and raw drivers playing Girl With Balalaika?  The question is, how can we get what we need from music by the least possible means?  Here we are, trying to get GWB sound quality from a full orchestra with chorus at full tilt.  Going backwards, looking first at the air that must be moved, all those motors, and the scale and magnitude of the cycling involved, makes me wonder what I was thinking when I went to single amping to begin with.

Yes, I agree that the first place to put a second amp is at LF, once the working requirements of the LF motors are known.

Best regards,
Paul S

Posted by Romy the Cat on 04-30-2011
Here is another advantage of multisampling and DSET in particularly. There are many moments in amp that in one or another degree affects transients.

The capacitive coupling for instance works fine but as long the amp sits in class A1. As soon amp hit the A2 all hell breaks loose for transients. Some people built “expensive” SETs. This makes me laugh. Let take for instance Lamm ML3. Lamm uses DHT tube that can work well in A2. In addition ML2 has huge second stage buffer (8 anodes) to drive the output stage grid currents. Looks like “expensive” design, does it? Well, ML3 use capacitance coupling –it works is fine as lone the capacitor is biased. What from one side is DC then the dialectic is polarized and the AC is floating fine. However, then 32W Lamm ML3 begin to drive those funny powered woofers of those Wilsons and woofer’s impedance drops then ML3 dive might hit A2. There is nothing wrong for THIS amp to work in A2, the GM70 can do it and the driver can provide sufficient current, but at the very time when amp enter A2 the coupling caps begin to change AC polarity and dielectric begin to recharge. This is not good Sound generally and it is VERY bad for HF. So, at the time when this type of amp have the LF stress, even all supplies can supply everything fine but the upper-range of the amp become transiently compromised due to the re-polarization of capacitors.

The transformer coupling is not much better. Transformer has inductance and act as choke in a way Try to measure current from a PS choke and turn the supply down. The current will be still flowing from the choke AFTER the supply is down. The choke will be storing energy and try to act like a gyroscope, maintaining the cruse current. Now pretend that the same thing sits between as a coupling devise between the stages…

I do not draw the pictures of misery about amp, I juts saying that in multisampling this effect would not exes tans you in multisampling will be able to select for any given channel the preferred coupling options. It is not to mention that any multisamp (DSET) works in a limited frequency range where all signals are of the same similar type, thus the optimization is very much possible.

The Cat

Posted by N-set on 05-01-2011
Paul, unfortunately this charge/discharge effect applies to ALL reactive components. Not only
to the interstage but also to the amp-speaker coupling.
2Hz is easy for an RC coupling, now go ahead and try to design a full-range 
OPT with such a low -3dB! I've once tried and ended up buying 10kg amo cores....and this for a
Ra=1.5k triode with a target -3dB at 30Hz!

What is the bungee effect you mention?


Posted by Paul S on 05-01-2011
N-set, yes, the more you know, the more unlikely "FR" sounds.  And the fact that small SETs with tuned "raw" drivers playing GWB can avoid the most detremental of these effects is largely because (duhhh...) the cited combination is not seriously FR. Now, how few un-stressed, "minimal" amps/drivers does it take to make serious FR?  Over the past few months, I have taken a few steps back to re-think these  and related matters.

As for the "bungee" effect, this is a visual/visceral metaphor that conjures my circum-aural sense of the "charge/discharge" effects during playback.

Best regards,
Paul S

Posted by Romy the Cat on 05-20-2011

I would like to point out at one of many-many advantage of multiamping playback. If you read my ULF thread then you will see that with relative ease I was able to get deal with my lower bass and set it up in the way how I would like it to Sound.

This would be absolutely impossible if I used some kind of “best speaker” with “best amplifier” and it is possible ONLY in context of multiamping, what you can many any change you wish with the channel of you attention and to be assured that it will be no or little impact to the rest of the channels.

The Cat

Posted by Romy the Cat on 10-28-2011
Click on the horses if you have flash installed in your machine

The Cat

Posted by haralanov on 10-28-2011
Hahahhaha :-)

It also shows how does the time alignment work (by clicking the different horses at different points in time).

Page 1 of 1 (15 items)