Burning in

[Richard]

I recently came across the first measureable proof I've seen that something happens with running in of a device(s).

Phono3S has an adjustable capacitor on each channel that allows the output frequency response to be fine tuned. To set it, a square wave is passed through an inverse RIAA filter to the phono amp. The phono amp applies RIAA correction and outputs it as a square wave again. The accuracy of this can be seen visually on a scope monitoring the output. The cap is adjusted for the cleanest sw.

Upon completion of my Phono3S I set the adjustable caps at a mid point and played it to check it worked. All components including valves were new. I hadn't checked sw and it sounded good so it was left a week or so in the system to soak test whilst I constructed the inverse RIAA filter to do the final adjustment.

With maybe 50-100 hours on it, it went back on the bench and was duly adjusted. All valves were kept in the same locations and no other changes made. The sw was excellent with a good sharp corner. Back in the system it sounded fine, not a big adjustment nor change in sound.

After about 5 weeks of daily use it was sounding great. I felt the slight sense of edginess and newness had gone and it all gelled really well. This is the point where I've always wondered if it was me or the equipment "running in"! The Pre3 was also new so there were a lot of new components in the chain.

I decided I'd now try a couple of different types of valves in Phono. The test gear was set up in readiness to adjust sw for the new valves and the existing valve set up checked before being removed; the sw corner was now rounded, noticeably, and equally on both channels. Swapping channels of the inverse RIAA, scope channels, another scope and another sig gen, all gave exactly the same result. High frequency output was definitely lower after those weeks of burning in.

(I should say that I've made these adjustments under varying conditions and they are not particularly sensitive to things like mains voltage variation or the cables I use for hook up etc. I even tried Phono powered from the PSU Pre outlet which changed HT by 50V to no effect.)

Anyway, a tweak on the caps instantly restored the sw and on listening there was a noticeable improvement, simply the best it had been. I left it and didn't re-try the valve rolling again for a couple more weeks. (When I did, incidentally, I tried 2 more types but reverted to the original choice.)

Anyway, a long story as I wanted to convey the circumstances and the certainty I feel about the result. There was a definite "softening" of the sound which coincided with a measurable loss of high frequencies after a month of use. It will be interesting to check again after a while.

Rich

[Ianm2]

I used to be VERY sceptical about this, however, I have noticed things change with time, even my dac starts to sound different to the xport after 5 mins

I suppose its quite obvious with caps, viz the electrostatic forces pulling and pushing the plates and electrolyte heating, etc...

[andrew ivimey]

Yes, yes yes, that's all very well but what kind of fuse are you using!

[Ianm2]

generic, spose i can't hear the difference bcos of that

[NealG]

So what's changing Rich? I strongly suspect it's the valves characteristics changinging slightly...

[Richard]

Originally Posted by NealG

So what's changing Rich? I strongly suspect it's the valves characteristics changinging slightly...

I really don't know, I've reported it all "as happened" so you guys could see the same as me.

It could be the fuse of course then again the audible change is just as many are used to hearing from new caps and, as you say, new valves. They both exhibit this audible change don't they.

What is new (to me anyway) is finally being able to see something on a test instrument. As mentioned, the test is stable and easy to reproduce, not at all hard to interpret or fussy to set up.

The adjustable caps are 50pF. At a guess I might have needed to tweak them 10 or 20 degrees to restore the lost hf, so possibly 5 or 10pF or so. What effect that would have on frequency response I don't know without attempting to work it out against the circuit. The higher frequencies for sure, those making up the higher harmonics of the 1KHz sw.

It may be a very sensitive test in that it shows the change in high frequencies easily. Indeed, I've never been able to differentiate between equipment sound by looking at a simple frequency sweep; most gear just looks ruler flat but still sounds different as many of us know.

It wasn't due to warm-up time on the bench, it had been on a good while before both set ups. I agree there can be a change when equipment warms up but none that is visible here in the sw if you do this test from soon after switch on to hot.

I'm actually using a second Phono3 now, an "S" without MC input and bass filter. I built it too with all new components and the adjustable caps. It was set up with the same type of (new) valves a few weeks ago. So, a complete second check on the first one. It's not been touched since so I'll check it on the bench when time permits and see if it too has changed.

After that could be a further test in a month or two to see if there is a further change. If there hasn't been the test could then be done again with just new valves to see if they are changing.

Anyone else set theirs up this way?

Rich

[Audiognome]

Interesting, Richard, but I don't find it surprising.

I used to work for a company that made (VERY expensive) oscilloscopes and each instrument had to be tested, calibrated and then sit in the "soak" room, powered up, for a couple of days. The soak room was kept at a constant 40degC. They were then re-calibrated, before shipping out.
This was done to allow the calibration to settle and also to catch - horrible expression - "infant mortality", i.e. components which fail in the first few hours of use.

I think it's well accepted that components will drift in value, most especially in circuits which self-heat and this was all hi-tech, surface mount, digital stuff.

So, if we're talking about resistors, caps, valves, etc, then I'd totally agree with the idea of burning-in.

When it applies to bits of wire, etc .... well ....

<ducks and waits for barrage of abuse>

[Richard]

Originally Posted by Audiognome

So, if we're talking about resistors, caps, valves, etc, then I'd totally agree with the idea of burning-in.

When it applies to bits of wire, etc .... well ....

<ducks and waits for barrage of abuse>

Ah, but now you're being selective without giving a reason Nigel . Why accept a change in a cap, resistor or transistor, evident in the scopes you worked on, but not the wiring or pcb?

In the audio system, I felt there was a change after a period of burning in, always towards being more mellow, but never had any evidence as (my) ears from week to week are not absolute in the same way as eyes viewing a line on a scope.

Rich

[acorn]

I would agree with Neal on this one as the most probable cause to be the valves , but sometimes resistors can drift of spec so that the valves work at a different point than expected, when making tests you need to note the cathode and anode voltages each time just to make sure that they tally every time.

Acorn

[Audiognome]

Originally Posted by Richard

Ah, but now you're being selective without giving a reason Nigel . Why accept a change in a cap, resistor or transistor, evident in the scopes you worked on, but not the wiring or pcb?

Ha!

Richard, I suspect you're either -

(a) being deliberately obtuse, or
(b) actually in agreement with me and would like to hear my reasoning

Let's assume it's (b), shall we?

OK, let's take a resistor as an example. No resistor is perfect - it exhibits inductance & capacitance as well as resistance. Due to the physical properties of the material used, these properties will all drift with age/time. The L & C may or may not be significant, at audio frequencies (most of the time not, I would argue). The resistance obviously could have significant effect (see also Acorn's post above). They needn't necessarily drift out of spec, maybe just a few %, may still have an audible effect.
If you have an application that requires a resistor which DOESN'T drift, which is often important, then you need high-stability resistors, standard ones will not do. Likewise, all resistors have a temperature coefficient of resistance, generate various differing noise spectra, depending on their materials & construction, etc, etc, blah, blah. Lots of variables, I suppose I'm trying to say.

Now let's look at the bit of PCB track. Well it might have a resistance of .001 ohms, capacitance to ground (or another track) of < 1pF and inductance of, I dunno, bog all?
Inter-track capacitance can be very siginificant in the RF world, as can dielectric absorbtion of the board material (I have to spend some of my life thinking about 433MHz and 2.4GHz circuits - yuck!). At audio frequencies, all these parameters are negligable. Any drift in the properties of the track (if, indeed, there is any drift) will make no difference at all.
If you think otherwise, then we're back to directional fuses, I'm afraid.
If the latter turns out to be true, then I give up. I'll become a "railway enthusiast" instead

Am I an authority on any of this? No, of course not! Just a bit of common sense. I'm just trying to ensure that my journey into the world of audio does not result in me disappearing up my own a**e