John Curl's Blowtorch preamplifier

Status
Not open for further replies.
Jump to Latest
#13,081
Joshua_G said:



Did you make those tests on a resistive load, or actual loudspeakers?
Click to expand...

Resistive load, dummy load and actual speakers.

Though NFB has issues (as everything), the view presented here is often misleading.

Depends on OLG corner, loopgain frequency response (both amplitude and phase) and phase margin, loopgain value in flat portion of amplitude response; and on inherent linearity before the GNFB is applied.

Every unqualified pronounced view is wrong, and there is a lot of totally unqualified, simplified views here, covered in word-pictured long sentences.
 
#13,082
Re: Re: Records tg Die for

Terry Demol said:
[snip]By comparison some of the worst sounding digital ever released
was done on the current de facto standard, 'tool$'.

T
Click to expand...


Terry,

The question then could be: is that the problem of the current tools or the current productions? I vote for the latter. If you compress music in the top 20dB of your medium and clip it 50% of the time, no matter how good your tools, it'll sound horrible.

Jan Didden
 
#13,084
john curl said:
So long as the wet slug tantalums are DC biased, they are very good. No DC bias and they can go bad.
Click to expand...


I never had a wet slug go bad, although I have seen the dry slug types fail as coupling caps on a number of occasions. They were used allot in the 80's.

When using the wet slugs I tweek the DC offset just a bit to make sure they are biased. The sound is quite pure relative to either aluminum or plastic film caps, especially on strings or vocals.

Mike
 
#13,085
Re: Re: Re: Records tg Die for

janneman said:



Terry,

The question then could be: is that the problem of the current tools or the current productions? I vote for the latter. If you compress music in the top 20dB of your medium and clip it 50% of the time, no matter how good your tools, it'll sound horrible.

Jan Didden
Click to expand...

Jan,

'Tools' is jargon, meaning protools. I assumed you were familiar
with what is used in a studio these days.

Any how, it ('tools) has evolved into -the- de facto recording
system, and out of all the players that we had 10 years ago, most
people agree it was probably the worst sounding.

Still, the issue is you were critical of those early digital tape
recorders, and most of what I hear from engineers, who don't give
a damn about specs is that a lot of them sounded pretty good. Not
in the same league as ;wide format analog though.

WRT compression etc etc - agreed.

cheers

T
 
#13,086
Re: Re: Re: Re: Records tg Die for

Terry Demol said:


Jan,

'Tools' is jargon, meaning protools. I assumed you were familiar
with what is used in a studio these days.

Any how, it ('tools) has evolved into -the- de facto recording
system, and out of all the players that we had 10 years ago, most
people agree it was probably the worst sounding.

Still, the issue is you were critical of those early digital tape
recorders, and most of what I hear from engineers, who don't give
a damn about specs is that a lot of them sounded pretty good. Not
in the same league as ;wide format analog though.

WRT compression etc etc - agreed.

cheers

T
Click to expand...

Understood and agreed. The point I wanted to make is that often the technology is blamed for bad production/mixing etc (don't know the right words).

Jan Didden
 
#13,092
Bob Cordell said:



Hi Joshua,

The tests I published were done with resistive loads, pretty much as prescribed by Otala.

I have privately done TIM and PIM tests into loads that were reactive at high frequencies, where these distortions tend to be problematic. In some cases I saw some increases in these distortions into reactive loads, but not really big increases. In these cases I also observed increases in THD-20.

It is usually difficult to do high-power tests into real loudspeakers because they will be damaged. I have, in the past, however, built some large networks that attempted to simulate a loudspeaker load with passive components. This work was mainly to evaluate the behavior of protection circuits when feeding a reactive load.

I've also built I/O compare circuits to evaluate amplifier behavior when driving real loudspeaker loads with real music. This device, as expected, subtracts a scaled version of the amplifier output from the input to achieve a null that is indicative of the faithfulness of the amplifier's behavior. The challenge with this is the need for careful adjustment of amplitude and phase compensation to facilitate the proper subtraction. This is not unlike what I do in my Distortion Magnifier (DM) used for sensitive THD and CCIF IM testing. The I/O compare has not revealed notable I/O differences in high-performance amplifiers except when they clipped. A key thing to bear in mind is that well-recorded music can have a very high crest factor and cause amplifiers to clip more than we would think.

Cheers,
Bob
Click to expand...


It's obvious (or should be obvious) that doing measurements on actual loudspeakers is anything but simple, easy and low-cost.
However, considering the interaction between the loudspeakers and the power amplifier, it seems that such measurements are necessary in order to get the true picture – even when only actual music crest-clipping is considered.
 
#13,095
PMA said:
But it is usually not an issue, in case that the amp is well designed; and I am sure that the Bob's is.
Click to expand...


Yes, however, "well designed amp" is the key phrase here, since there are amplifiers that may measure well on resistive loads, but may not sound that good with music fed to loudspeakers. So, it seems to me that "well designed amp" should be explained. Don't you think so?
 
#13,097
Terry, my first responsibility at Ampex audio was co-designing a phase locked servo for the AG-440, our first audio capstan servo. My second project was designing the next generation's electronics for a future project. I left the audio department once the servo was finished, and moved down the hall to the research department, to develop an advanced video control system, and a very high power motor drive amp.
 
#13,098
IIM is somewhat more complicated that what has been expressed here up to now. There are examples of significant IIM, but most power amps that we normally design usually avoid it. If you look at one of my output stages, or Charles Hansen's latest power amp output stage, you will find them almost IIM free, by design. Other designs on some other threads here, MIGHT be IIM prone because of intrinsically open loop high impedance made low by global negative feedback.
 
Status
Not open for further replies.
Top Bottom