Same type and spec producing same B+ does not automatically mean electrically identical. Variations exist and there may exist other unmentioned measurable differences such as dynamic resistance, inter electrode capacitance, asymmetry of geometries (on full wave rectifiers), natural mechanical resonance and RFI pickup.
I come to testify too because I just had to experience it with my 5u4gb. I had a problem due to my stupidity which led to change the power supply
of one of my amp. I first restarted with my last 5u4gb EH and despite the changes, I did not notice a big difference until I receive a 5u4gb RCA black plate that I hastened to intaller and there, I found a big difference, the treble much finer, the more medium forward (especially on the violin Grappelli and the bass a little shorter but faster . so I changed again for the EH and back ... I do not have an explanation, it's just my findings
of one of my amp. I first restarted with my last 5u4gb EH and despite the changes, I did not notice a big difference until I receive a 5u4gb RCA black plate that I hastened to intaller and there, I found a big difference, the treble much finer, the more medium forward (especially on the violin Grappelli and the bass a little shorter but faster . so I changed again for the EH and back ... I do not have an explanation, it's just my findings
What can anyone do with this information? Do you want a list of possible explanations? Why no testing information?
I did find some notes with a GE 6AX5 and Bendix 5852 in the same SE amp, same conditions. The GE 6AX5's I use have 6087 guts, with a 6.3 volt heater. Both of these tubes would have been new. The 5852 is more like a 6X5 so not quite like comparing a 6106 and 6087. The 5852 put out 5 more volts than the 6AX5. The IMD difference at the speaker terminals was negligible.
What parameters would you suggest folks test for? In the above example, static loaded output voltage, and significant construction and material differences, arguably two separate tube types, made no difference to the meter.
I wish I had recorded the 6087 / 6106 measurements so that I would not have to rely on memory, but I could not find that I did.
Win W5JAG
We can select 2 rectifiers from different manufacturer to give same B+ at a certain circuit but will have different ripple amount showing slight difference in the internal dynamic resistance that will also affect transient behavior. The selected 2 rectifiers will give not so similar B+ on other circuits. Off course differences will stay within manufacturing tolerance of say 5% , on the other hand we have people claimed audibility on 0.1% electrically measured difference.
My point is when a difference is noted, more than likely that there is a measurable difference, but measuring only the B+ value may not be sufficient tell the whole story. W5JAG method of measuring IMD should shed more light. A full FFT and distortion residual waveform analysis will tell more.
I kind of suspect that many old timers in the know found that it is better for others to remain oblivious to the difference, keeps the price of the good ones more manageable.
but, but isn't ripple voltage a function of the filtering capacitors? and load currents?
do you have test setup/data that we can replicate?
being unable to quantify, then our claims becomes mere opinions....
i have no problem with anecdotes and opinions, i am not bothered....
Hence… to the question prior quote “different internal resistances, different transient response” (paraphrased), I can only say this: active regulation is a substantially “right” answer for most every of the suggested effects.
Absolute regulation (zeners, gas discharge tubes, etc) have the advantage of setting an operating point, and keeping it there with little regard to line voltage variation, swapping rectifiers, aging of them, out-of-spec (or spec variation) of the primary transformer. Regulation doesn't in its own right make up for transient response though. It just defines a kind of response which is generally quite decoupled from the "stuff before it" in circuit.
Relative regulation (voltage dividers and capacitor LPFs) can offer higher best-case downwind voltage, but relative to line voltage, and so on, maintaining a fixed margin (the voltage divider chain) for the amplifier's stages beyond.
BOTH kinds of regulation can markedly reduce one's expenditures on more expensive, larger value chokes (if you're inclined to use them) and filtering capacitors in larger sizes. I regularly get better than –40 dB of ripple reduction just from an active MOSFET series regulator of utterly simple (-minded?) design. This is not to say that the wonderful goodness of a nice choke in a CLC→regulator→C section isn't still warranted. It is.
Thing is tho', relative regulation WILL couple the rectifiers' idiosycracies to the rest of the circuit as bulk quiescent voltage differences. absolute regulation does not. Almost completely decouples the rectifier swapping and ageing differences over the useful life of the vacuum diodes.
Just saying.
Amazing what a few under-two-bucks-each parts will do.
Yet keep true to the underlying design theory of an amp.
GoatGuy
as a builder of amps, my goal has always been to build them as best as i could from get go....
i will get depressed if by changing rectifiers my amp sounded better, that meant i did a lousy job building my amp...but this is just me.....
i have no problem with others stating their opinions, it doesn't bother me...
i go by what i have...
I must say that I agree 100% with this sentiment. But I would suggest that it is not in the build so much as the design...
I wish it was that simple.
It was not a test, I just remembered something while helping a friend build a simple 01A LC loaded preamp. The noted difference was between a big mesh AZ1 rectifier from Philips and a tall solid plate AZ1 TFK with CRC filter. B+ was measured at 207V and 209V, close enough for me to be called equal, but B+ showed ripple of 3.2 mV for the mesh against 1.8 mV for the solid plate. We then use brute force method and modify to a CLCRC filter to reduce ripple to < 0.5mV. At the finalized build, each rectifier produce subtle but distinct enough difference in presentation, but our curiosity level was not high enough to make further investigation nor measurement, the preamp sounded good enough for us as it was.
I know you have way more experience and better equipment than I do, perhaps you can share your thinking and findings you have noted?