You know, when you are older, you realize that much that you took for granted, was actually pretty good, maybe even better than 'new and improved'. I found this so with some recordings done in the 1950's that just 'knocked my socks off', in the '80's or 90's.
It is also true with test equipment. Many here think that a calibrated Panasonic mike is as good as a B&K, but I don't think so, except for the most mundane of measurements like frequency response over the audio range. What about distortion, for example? Have you ever had to measure the distortion from a horn loudspeaker, direct radiator? Can you trust a Panasonic?
Other test equipment, much the same. IF I were starting out from scratch, let's say after another firestorm, and all my equipment was destroyed, then I would get what many of you have and use for test equipment. Why not?
For the record, when I first met SY, he had so little in measurement equipment, that I felt sorry for him and 'loaned' him an HP3581 that was sitting in my closet. He seemed overjoyed at the time. Later, he bought it from me, for 1/2 that I paid for it, wholesale.
Now, apparently, we can get pretty good results with a computer and a sound card. Good! However, not necessarily better than the much more expensive measurement equipment made in the past. Perhaps in some measurements a little bit better, but for other measurements perhaps a little bit worse. It does not follow that newer and cheaper is better than older and more expensive.
It is also true with test equipment. Many here think that a calibrated Panasonic mike is as good as a B&K, but I don't think so, except for the most mundane of measurements like frequency response over the audio range. What about distortion, for example? Have you ever had to measure the distortion from a horn loudspeaker, direct radiator? Can you trust a Panasonic?
Other test equipment, much the same. IF I were starting out from scratch, let's say after another firestorm, and all my equipment was destroyed, then I would get what many of you have and use for test equipment. Why not?
For the record, when I first met SY, he had so little in measurement equipment, that I felt sorry for him and 'loaned' him an HP3581 that was sitting in my closet. He seemed overjoyed at the time. Later, he bought it from me, for 1/2 that I paid for it, wholesale.
Now, apparently, we can get pretty good results with a computer and a sound card. Good! However, not necessarily better than the much more expensive measurement equipment made in the past. Perhaps in some measurements a little bit better, but for other measurements perhaps a little bit worse. It does not follow that newer and cheaper is better than older and more expensive.
So Sy was correct, nothing after 1969 ...for you ..........
Except what I understood myself about audio design.
John; what I can get from computer is a bit different than what I could get before: to observe FFT plots almost for free. As I said before, it does not mean some standard measurement, it is more like very good tool for search for flaws and further directions: I can dial in-out input level and see how spectrum changes. I can change regimes and check if I was right. I can change topology in order to avoid discovered flaws. And so on. Tools for designers can be still the same as for QA and marketing people, but purposes of their usage are different.
You can compare for yourself: 4P1L is better for fraction of price.
I saw once a picture of 4P1L in the bulb of RE-604, to keep "Look and feel"
Attachments
Sy, but Sy is older than Dirt ............he's been around way before measuring equipment, hmm
Yes, but when I left Nicolet, they didn't let me take the equipment with me, dammit.
I don't think anyone was actually saying that, John. I don't remember anyone arguing that lower price and newer makes it better. Maybe you are seeing cause and effect where there is none?
OK, even if true the doubling doesn't bother me much. 1000% or better does. That's what I'm thinking about for devices that run hot all the time.
True enough. And a rise in temp is not going to immediately be sunk. But if the device is already hot, just how much hotter (in %) is it going to get with dynamic peaks? I don't think it will get all that much hotter, there won't be a huge change. With devices that run at lower idle currents and temps, there may be a much more dramatic temperature delta.
Hi,
I am familiar with the curves for both tubes, personally I would not call the 4P1L more linear than the RE604, plus both sets curves are heavily averaged from large lots.
But if you like the 4P1L so much, please send me all RE604's and YO-168 you happen to have and I'll send you double the number of 4P1L's back in return and promise not to e-bay the tubes you send me.... ;-)
Ciao T
You can compare for yourself: 4P1L is better for fraction of price.
I saw once a picture of 4P1L in the bulb of RE-604, to keep "Look and feel"
I am familiar with the curves for both tubes, personally I would not call the 4P1L more linear than the RE604, plus both sets curves are heavily averaged from large lots.
But if you like the 4P1L so much, please send me all RE604's and YO-168 you happen to have and I'll send you double the number of 4P1L's back in return and promise not to e-bay the tubes you send me.... ;-)
Ciao T
Hi,
Let us take a "100W" transistor with 1K/W thermal rise into an infinite heatsink, running at 50W dissipation (say 50V/1A Iq) at 25 Degrees ambient so it is at 75 degrees.
We now double the dissipation briefly and the chip will near instantaneously rise to 125 degrees, and as we remove the extra load slowly drop back to 75 degrees as it gets rid of the extra heat into the heatsink.
Meanwhile the transistor that has all the power removed will drop it's temperature, depending on the thermal resistance between it and the heatsink and the thermal mass of it's heatspreader.
Now the same in Class AB running at 50V/0.06A dissipating three watt. It now is at 28 Degrees and increasing it's conduction by 1A to deliver 1A load current will almost instantaneously raise it's temperature by 50 Degrees, the same relative amount of change as in "Class A".
As for the absence of LF distortion rise, try running the Amp open loop, most amplifiers have very high levels of NFB at DC and very low frequencies, dropping quickly as frequency goes up, plus as remarked the local degeneration of the output stage using emitter resistors will significantly reduce the effect. For example if we operate at 1A quiescent current and use 0.33 Ohm Emitter resistors the output transistor is degenerated around 20dB...
Hence use minimum value emitter resistors and open loop operation to observe...
Ciao T
PS, just because the effect is swamped in the sense of traditional measurements by a number of mechanisms does not mean it may not have any impact somewhere. We have a system who's behaviour will change as a result of signal history in ways that do not correlate directly with the signal...
Let us take a "100W" transistor with 1K/W thermal rise into an infinite heatsink, running at 50W dissipation (say 50V/1A Iq) at 25 Degrees ambient so it is at 75 degrees.
We now double the dissipation briefly and the chip will near instantaneously rise to 125 degrees, and as we remove the extra load slowly drop back to 75 degrees as it gets rid of the extra heat into the heatsink.
Meanwhile the transistor that has all the power removed will drop it's temperature, depending on the thermal resistance between it and the heatsink and the thermal mass of it's heatspreader.
Now the same in Class AB running at 50V/0.06A dissipating three watt. It now is at 28 Degrees and increasing it's conduction by 1A to deliver 1A load current will almost instantaneously raise it's temperature by 50 Degrees, the same relative amount of change as in "Class A".
As for the absence of LF distortion rise, try running the Amp open loop, most amplifiers have very high levels of NFB at DC and very low frequencies, dropping quickly as frequency goes up, plus as remarked the local degeneration of the output stage using emitter resistors will significantly reduce the effect. For example if we operate at 1A quiescent current and use 0.33 Ohm Emitter resistors the output transistor is degenerated around 20dB...
Hence use minimum value emitter resistors and open loop operation to observe...
Ciao T
PS, just because the effect is swamped in the sense of traditional measurements by a number of mechanisms does not mean it may not have any impact somewhere. We have a system who's behaviour will change as a result of signal history in ways that do not correlate directly with the signal...
Uh.... You've lost me. How is going from 75º to 125º the same relative amount as going from 28º to 78º? It's the same number of degrees, but not the same increase relative to idle temperature, is it? I see an increase of 66% vs 178%.
Hi,
That is because you are calculating in Celsius with one arbitrary point of zero as the temperature of freezing water. If you where to calculate in Kelvin the percentage difference would be much closer, however the thermal changes in bipolar transistors are pretty linear "XX per Kelvin" changes related, so absolute temperatures are not really relevant to the topic of relative thermal parameter modulation with signal.
Ciao T
That is because you are calculating in Celsius with one arbitrary point of zero as the temperature of freezing water. If you where to calculate in Kelvin the percentage difference would be much closer, however the thermal changes in bipolar transistors are pretty linear "XX per Kelvin" changes related, so absolute temperatures are not really relevant to the topic of relative thermal parameter modulation with signal.
Ciao T
Yes it does ..and try Photon torpedoes, wide spread....... Now !!!!
I don't know Sy, i hear it in the midrange, thermal tracking does pull up the bass, but it is also present in the midrange , the distortion is very audible and can be repeated , i have done so on 5 different a/ab SS amplifiers.
Sorry, I have no RE604 tubes. But I would send them to you if you love and can use them, like recently I sent to one guy on Russian forum LED indicators that he needed, and to another guy some samples of opa1612 I got from TI when they were fresh new, but never used.
- Status
- Not open for further replies.