John Curl's Blowtorch preamplifier part II

Status
Not open for further replies.
the super

Rolls Royce, the best car in the world.
Horsepower : enough
0-60 acceleration : sufficient

It's sufficient for the client to think to have the top of the line, even if it's not the absolute best.
99 out of 100 want more (than others), basic human nature.

Discussing the con of cables in public is not a wise man's act, they're out of the critic's league, but expressing criticism is right up the vendor's alley.
Right or wrong, as long as the brand name goes around.
 
And to put things into balance. A complete JC designed parasound set (pair of JC-1, JC-2 and JC-3 if you need phono) comes to $16000. In the world of high-end where pre-amps can cost twice that for the right magic flooby suddenly seems to be good value. Especially as you could probably jump start your car with the power amp.
 
BF862 Idss

I have once tested maybe 80 or 100 devices and they were nearly all 12-13 mA, very few outliers by more than 2 mA when I wanted some high Idss ones.

.......

regards, Gerhard

The last time I measured a batch of ~60 in 2009.
See attached.
All over the place, I would say.


Patrick
 

Attachments

  • BF862 Idss Distribution.png
    BF862 Idss Distribution.png
    21.8 KB · Views: 224
No just misread. We agree on the causes but you seem to think I confused the resonance peak with the inductance rise.

No. As I stated, you pose incredibly ambiguous questions and then pounce on any responder with things like "misread". I did not misread, I spoke about BOTH rising impedance areas, because you did not identify which one in your ambiguous question.

You also got the issue of why the inductance rise is not a straight line.

You mean this time, last time, in the damping thread, or multiple times over the past 10 years?

Now the impedance peak will of course combine with any impedance variation from the theoretical power amplifier output to produce a deviation in power delivered to the loudspeaker.
You mean, exactly as has been stated over the last ten years at least that I've read on this forum, describing response to low DF tubers, and 24 awg speaker runs.
Common practice is to correct frequency response deviations with equalization. Yet that will not work on this type of resonance.
I recall that also being stated multiple times on this forum.

Perhaps the search engine would be of assist here.

Ed it would help if you had at least enough respect for the technical community at large that you would collect the literature on this and go through it, it is substantial, before presenting issues like no one has ever thought of them before.

I do this every week the "peanuts" go off like nothing before has ever happened and I reel them in with hours of readings from the past. Just call me Buffalo Bob.
Yo..hey there Bob..

John
 
JN

I was going over the basics before presenting what I think is new. But it seems that there is no interest.

Then there is also the gap between where the common ground is found.

I haven't found correct bits about why the equalization doesn't work as mentioned. I did find claims that it does.
 
Ah well, the system upgrade money has been blown again. This time on the decca 'Dutoit: The montreal years' box set. Interesting thing is that some of these are dawn of digital recordings where the Engineers were marvelling at what could be done. I haven't done any audacity checking for bass roll off, but the dynamic contrasts suggest that little or no gain riding was done.

These recordings from the early 80s are, so far , more natural than a lot of stuff the big labels put out today!
 
JN

I was going over the basics before presenting what I think is new. But it seems that there is no interest.

Going over basics is great when presenting stuff that may be new. However, you really have to re-think your 20 question schtick. You always seem to ask horribly ambiguous questions, then slap everybody who answers upside the head. It was tiring a year or two ago.

I haven't found correct bits about why the equalization doesn't work as mentioned.
You will have to work the search engine harder. Even I've gone over that, and I'm sure I was not the first, nor the best. Reflex, 4th order, 6th order, t-line, they can all be eq'd flat for steady state spl, but not for transients like kick drums...

Do not ask me how to search better, I'm fairly clueless there...sorry.

edit: here's the last time I mentioned some of this stuff.

http://www.diyaudio.com/forums/solid-state/287072-damping-factor-explained-not-13.html#post4627078


quote
""Also of note is, what the amp sees at the terminals is a consequence of the speaker coil resistance, inductance, as well as all the acoustic storage mechanisms, from sealed baffle, horn, out to 4th or 6th order bandpass. Speaker measurements which are of a steady state nature do NOT indicate what will really happen when non periodic signals are applied. One such example is the impedance peak of the woof at LF. That resonance requires multiple cycles to reach that specific impedance.""
eoquote
John
 
Last edited:
JN

I was going over the basics before presenting what I think is new. But it seems that there is no interest.

Then there is also the gap between where the common ground is found.

I haven't found correct bits about why the equalization doesn't work as mentioned. I did find claims that it does.

I did come across something interesting, someone claimed the conjugate network method to make the impedance seen by the amplifier a constant at zero phase was ill-conditioned. I was surprised by that.
 
And the sound was pretty much bad...

. Here's a review that doesn't seem to agree.

Hmm I was enjoying that link until he decided magic feet under amps was clearly audible in a peeking test and that the lampizator stuff was good. But I still suspect I would concur with most of the systems he said were bad.

Last show I went to, around 94 I think there was one room where I liked the sound.
 
Interesting. He missed what I thought was the best sound at the show, but he was unhappy with what I thought was the worst. So I'll rate that review at 50%. 😀

I generally concur, though- there was damn little there that sounded anything vaguely like live music. And where I was able to play good recordings of musicians whose voices and instruments I know well, I was likewise flabbergasted at how unrecognizable the sound was, coming from some pretty hyped brands.
 
Interesting. He missed what I thought was the best sound at the show, but he was unhappy with what I thought was the worst. So I'll rate that review at 50%. 😀

I generally concur, though- there was damn little there that sounded anything vaguely like live music. And where I was able to play good recordings of musicians whose voices and instruments I know well, I was likewise flabbergasted at how unrecognizable the sound was, coming from some pretty hyped brands.

What was your "best sound" at the show?
 
I brought up a question about the conjugate network matching as commonly done in a now closed thread last week. The problem as I see it is that normally the conjugate network is created at a constant 1 watt @ 8 ohm or whatever the speaker impedance rating may be and this would seem to solve the conditions for a steady state tank circuit for that speaker device. The problem I see is that as you increase the power all the values are shifting that your conjugate is not now tracking such as the rise in impedance in the voice-coil, the change of inductance with position of the voice-coil in an overhung or equal hung coil and the change that causes to the matched capacitance in the LCR tank circuit. So as I see this there is no way to actually have an accurate conjugate network between the device and the amplifier that keeps a flat impedance curve that does not change with level.

What say you much smarter gurus here than I am? Do we just do the best we can and match at a low level or do we have some way to make the external LCR tank track what is happening in the dynamic driver over the operating power levels?
 
I brought up a question about the conjugate network matching as commonly done in a now closed thread last week. The problem as I see it is that normally the conjugate network is created at a constant 1 watt @ 8 ohm or whatever the speaker impedance rating may be and this would seem to solve the conditions for a steady state tank circuit for that speaker device.

Different problem, by ill-conditioned I mean there are local minima where no incremental change of any component gives a better solution but there exists a better global solution, i.e. you can't write a set of equations and solve for the answer. I don't know if this is true.
 
I brought up a question about the conjugate network matching as commonly done in a now closed thread last week. The problem as I see it is that normally the conjugate network is created at a constant 1 watt @ 8 ohm or whatever the speaker impedance rating may be and this would seem to solve the conditions for a steady state tank circuit for that speaker device. The problem I see is that as you increase the power all the values are shifting that your conjugate is not now tracking such as the rise in impedance in the voice-coil, the change of inductance with position of the voice-coil in an overhung or equal hung coil and the change that causes to the matched capacitance in the LCR tank circuit. So as I see this there is no way to actually have an accurate conjugate network between the device and the amplifier that keeps a flat impedance curve that does not change with level.

What say you much smarter gurus here than I am? Do we just do the best we can and match at a low level or do we have some way to make the external LCR tank track what is happening in the dynamic driver over the operating power levels?

You could use resistors designed with a high and semi-calibrated temperature coefficient to try and track the voice coil heating, but of course not the inductance changes.

Besides the non-linearity issue is the time issue. When the loudspeaker is at resonance it is ringing, so the apparent level/impedance will vary with the test or actual signal duty cycle.

Now it is possible in theory to map the driver's performance and do a DSP based correction. You may remember when there were loudspeaker controllers that also required a power amplifier output connection for such level sampling. There is of course a reason why such approaches seem to have disappeared.

One of the points that I didn't get to was an underdamped response cannot be corrected electronically, while an overdamped one can to much greater degree be compensated for.
 
Status
Not open for further replies.