All they (new generation) have to do is copy or reverse engineer it, right? I'm new at this game so please bear with me.
I am very suspicious of all the burn-in talk as well. The engineer in me insists on having some type of measurable parameter that can be linked back to the changes that take place in the audible realm. Essentially, the physics should explain the audibility. If the physics do not explain it, either we have not measured the correct parameter, or our mind is playing games on us.
That being said, I can attest to my experience with burn-in, which is very little. But one in particular stands out that I have problems explaining away. Built a Raven style preamp with Lundahl output transformers. Tubes were 6n6p. Upon initial startup and testing, things sounded horrible. Like very much unlistenable after just a few minutes. The fatigue and headaches set in quickly. Out of frustration, I went to K&K's forum, requesting help. A few things were attempted (thanks Kevin and Dave for your patience), but to no avail. I would play this amp to my friends, who all agreed it sounded bad; shrill and harsh. Then I went into test mode (not making any changes), just playing around with the scope and meter. Sometimes with dummy loads, sometimes open circuited, test tones, music, etc. Over the course of a few days, I noticed a marked improvement when returning to speaker loading. Thought to myself, no way, couldn't be. Ran it similarly to tubelab's recipe above, with Sheffield test CD's and various music. Listen again- yes, no question things were improving.
Brought my friends back over to listen again, and they loved what they heard. Psychology, possibly. But I very much doubt it. Now, to be fair, most all the components were new, so the burn-in could have just been the tubes, and nothing else. The physics attesting to (acknowledging) the reality of tube break in has been mentioned already, and I think most agree new tubes do need a little burn in time. So I can't say the xfmrs changed, but I can say something changed.
But on a different note, I can provide some physics behind transformer testing. We routinely test power transformers for insulation quality. Sizes range from 30 kVA to 30,000 kVA, voltages from 208V to 34,500V. Ya, these ain't audio transformers, but they are nonetheless silicon grain oriented steel cores with kraft paper, transformerboard, maple supports, vacuum impregnated varnish, copper/aluminum windings, and sometimes various epoxies. Some dry type, some filled with mineral oil, high molecular weight hydrocarbon, silicone, askarel, or vegetable oil. At the end of the day, a transformer is a transformer; just a few coils wound on a common magnetic core.
We test a number of things, but the most notable are DC and AC insulation tests. The AC insulation test is the most revealing, as it is highly sensitive to temperature (there are correction factors), moisture, winding tension, and fluid (if any) quality. Cutting edge technology is frequency response characterization through impulse injection, but that's a factory test, not a field test.
Anyway, I can attest to measurable changes that occur on a transformer, whether after a few days in service or a few years. There are, of course, many many variables that affect the degree of these changes, most notably environmental conditions and service conditions. I don't claim to provide any statement about whether the effects are audible, but I will with confidence attest to the fact that transformers DO 'burn-in'. I suppose, to be most reliable, I should perform some of these tests on audio transformers, but hey, I got better things to do, and I'm not trying to pick a fight, here. But no question about it, transformers change, measurably. Both within a few hours, and after many years.
Power transformers are also run very near saturation, and experience a great deal more magnetostriction than your average audio transformer. So I am not trying to apply exact comparisons between audio and power xfmrs, but from a simple material standpoint, they are close brothers. It's quite common for us to tighten the blocking on transformers from year to year, there is so much mechanical stress.
Whether it be moisture content of the paper or temperature, I guarantee you a power AND audio transformer will measure differently after being in use for a few hours. Something is changing here.
*puts helmut on*
That being said, I can attest to my experience with burn-in, which is very little. But one in particular stands out that I have problems explaining away. Built a Raven style preamp with Lundahl output transformers. Tubes were 6n6p. Upon initial startup and testing, things sounded horrible. Like very much unlistenable after just a few minutes. The fatigue and headaches set in quickly. Out of frustration, I went to K&K's forum, requesting help. A few things were attempted (thanks Kevin and Dave for your patience), but to no avail. I would play this amp to my friends, who all agreed it sounded bad; shrill and harsh. Then I went into test mode (not making any changes), just playing around with the scope and meter. Sometimes with dummy loads, sometimes open circuited, test tones, music, etc. Over the course of a few days, I noticed a marked improvement when returning to speaker loading. Thought to myself, no way, couldn't be. Ran it similarly to tubelab's recipe above, with Sheffield test CD's and various music. Listen again- yes, no question things were improving.
Brought my friends back over to listen again, and they loved what they heard. Psychology, possibly. But I very much doubt it. Now, to be fair, most all the components were new, so the burn-in could have just been the tubes, and nothing else. The physics attesting to (acknowledging) the reality of tube break in has been mentioned already, and I think most agree new tubes do need a little burn in time. So I can't say the xfmrs changed, but I can say something changed.
But on a different note, I can provide some physics behind transformer testing. We routinely test power transformers for insulation quality. Sizes range from 30 kVA to 30,000 kVA, voltages from 208V to 34,500V. Ya, these ain't audio transformers, but they are nonetheless silicon grain oriented steel cores with kraft paper, transformerboard, maple supports, vacuum impregnated varnish, copper/aluminum windings, and sometimes various epoxies. Some dry type, some filled with mineral oil, high molecular weight hydrocarbon, silicone, askarel, or vegetable oil. At the end of the day, a transformer is a transformer; just a few coils wound on a common magnetic core.
We test a number of things, but the most notable are DC and AC insulation tests. The AC insulation test is the most revealing, as it is highly sensitive to temperature (there are correction factors), moisture, winding tension, and fluid (if any) quality. Cutting edge technology is frequency response characterization through impulse injection, but that's a factory test, not a field test.
Anyway, I can attest to measurable changes that occur on a transformer, whether after a few days in service or a few years. There are, of course, many many variables that affect the degree of these changes, most notably environmental conditions and service conditions. I don't claim to provide any statement about whether the effects are audible, but I will with confidence attest to the fact that transformers DO 'burn-in'. I suppose, to be most reliable, I should perform some of these tests on audio transformers, but hey, I got better things to do, and I'm not trying to pick a fight, here. But no question about it, transformers change, measurably. Both within a few hours, and after many years.
Power transformers are also run very near saturation, and experience a great deal more magnetostriction than your average audio transformer. So I am not trying to apply exact comparisons between audio and power xfmrs, but from a simple material standpoint, they are close brothers. It's quite common for us to tighten the blocking on transformers from year to year, there is so much mechanical stress.
Whether it be moisture content of the paper or temperature, I guarantee you a power AND audio transformer will measure differently after being in use for a few hours. Something is changing here.
*puts helmut on*
This is a lot easier in theory that it is in practice. Most of the real innovation in the vacuum tube world was in serious decline by the early 60's. The remaining effort was directed toward cost reduction. Many long time electronics enthusiasts (including myself) abandoned those glowing bottles and went to work in a Silicon (or GaAs) world. This hiatus was about 10 years in my case, much longer for any real manufuacturing. A lot of knowledge, experience, equipment, and tooling was lost, scrapped, or destroyed. There are a few transformer winders that have been around for many years and their transformers do work very good, but they are not cheap.
As to reverse engineering or copying, this too is a learned art. When I first got back into tubes (15 years ago) the Chinese had just started cloning the KT88. These original KT88's tend to remind you that the Chinese invented fireworks. I had one blow so violently that the glass broke. Now 15 years later the Shuguang KT88's work pretty good. I havent found any Chinese OPT's that I like yet. Maybe they will get better over the next few years.
having worked in semiconductor industry, 10 years of which was with Advanced Micro Devices, Phils. Inc., i know for a fact that we do "burn-in" our chips before they leave the door....
i have no problem understanding burn-in done by tubelab, it admirable that he does that on his amps...
what i have problem understanding is why the consumers have to undergo 100hours burn-in, when this burning in is clearly the responsibility of the seller, not the end user.......
i have no problem understanding burn-in done by tubelab, it admirable that he does that on his amps...
what i have problem understanding is why the consumers have to undergo 100hours burn-in, when this burning in is clearly the responsibility of the seller, not the end user.......
Tony said:
I agree..... I too have worked many years in the semiconductor world as a chip designer for a big company i can't mention...(HINT: begins with a N and I have worked with Bob Pease on ocassion).....
The chips would go through "Burn-In" process before test.... But this was primarily for looking at the process variation in the Band-gap reference circuit, thus centering the Band-Gap at room Temp ....
With transformers, it is for stabilizing the di-electric material, from the windings initialy getting hot..
The big mis-conception is the term DIELECTRIC CONSTANT ..... It is not a Constant and never was.... It changes from batch to batch and changes acording to the applied voltage gradient whther it be DC or AC audio.... Di-electrics have memory like some batteries... Initial break-in of di-electrics will have an affect on the di-electric constant.... The bigger the applied signal, such as a transformer, the quicker the break-in...small signal amplitudes such as on interconnects have a much slower time breaking in the di-electric... It is verifiable through measuring the capacitance prior and after such applications...when the material is new...
It is a very small change....in most cases..
I agree the manufacturer "should" do the break-in.....
Chris
zigzagflux said:
Me too. But why does it have to be the transformer? What about caps and other passive parts?
I also use Lundahl and have never noticed any particular break-in effects. And i would love to know how break-in can occur with an MC step-up under normal use. My Tribute step-up sounded absolutely glorious right out of the box and the Lundahl 9206 was not bad at all.
The only transformer i've genuinely disliked (TVC) didn't improve significantly even after weeks of forced burn-in.
Not saying the effect doesn't exist, merely it's beyond what i consider important.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.