|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
150.148.218.189
In Reply to: RE: OK now I am totally confused... posted by Ralph on June 05, 2012 at 08:56:55
But I am finding out that not everyone who is excited by this idea has a thorough understanding of what they are doing when fiddling with the crossover and the transformers, and I do not wish to be responsible for the possible negative consequences of that. However, I agree it would sound very bad long before the transformer would actually burn up.
Contrary to one small element in your post, I found that the impedance curve of my speaker is more flat now through the midrange than it was "before". If you look at the measurements made by Kevin Covi about 10 years ago, the parallel resistor created an impedance dip around 500 to 1kHz. This caused a midrange "suck-out" when driving the SLs with your OTLs that was always audible despite all my prior efforts to fix it, including the use of the Zero Autoformers. (That suck-out was probably not evident if one drives the old crossover with a solid state amplifier, as does Sound Lab at their factory.) The midrange dip in the speaker response is completely gone once I converted fully to the Australian transformer (and even just by removing the resistor from the input of the SL toroid), and if you will review the measurements I made, this audible result is concomitant with the fairly stable impedance from around 400 Hz up to 2kHz (using the AU transformer).
Follow Ups:
This is what Dr. West sent me late last year, which is probably a lot more recent.
The one with the peak is with no resistor if I have his notes right. Sorry for the bad resolution but this is how I got it. So I have base a lot of my conjecture on this chart, but it is quite clear that it probably does not represent all speakers- probably just the newer ones. That peak though being up to 40 ohms says to me that the cap can be about 1/4-1/3 of its original value in those models after the resistor is removed.
Since I can see none of the numbers on the X or Y axes, and since the curves are not labeled, the graph means little to me, except to say that my measurements are nothing like that, assuming the bottom two curves are of impedance. There is no such huge peak at whatever frequency that is, with no resistor, using either their toroid or the AU transformer. My speakers were built in late spring of 2011. Anyway, if that peak is up to 40 ohms, then the impedance at lower and higher frequencies would appear to be too low to be ideal for our amplifiers. Looks like 5 to 10 ohms mostly. OK for an MA2 but nothing lesser.Edit: After sleeping on this post overnight, I decided that I over-stated the problem. In fact, an MA1 should also be able to drive the load(s) represented in that graph. (I think the lower two curves show impedance vs frequency, and the upper two curves show amplitude in db vs frequency.) But things get a lot better than suggested by those data, when one removes the parallel resistance in the hi-pass filter, which was my main point.
Edits: 06/06/12
Right!
We have enough people using MA-1s with the Sound Lab to know that they do work :) Doing the mod really helps them out
The big increments on the graph are 5 ohms. You can see on the left that the scale goes from double digits down to one single, and Roger mentioned something to me about a 40 ohm peak, which does correspond to the graph. That puts the minimum impedance of the bass transformer at about 7 ohms. The far right side of the chart appears to be 20KHz. I suspect that with all the other mods (not just the resistor) that 1.5-2 ohm value comes up a little.
I posted my own numbers for impedance vs frequency on SLOG, I think. You can see that my new 845PXs with the AU transformer and no resistor/no capacitor never get anywhere near to being as low as 5 or 7 ohms, except above 2kHz. Having measured two different speakers, my own plus my neighbor's U2PXs, I have come to believe that there are differences among the speakers probably in relation to the many versions of the stators, the diaphragm coating, the mylar, etc, that SL has experimented with over years. (The U2PXs had been upgraded to full PX about 4-5 years ago.) Also, although I did not do any direct measurements to prove it, I believe the impedance will vary according to the setting of the bias voltage. If so, no two speakers will ever be exactly alike.But the important point is that there is NO huge impedance peak in the curve for my speakers OR those U2PXs. (40 ohms is 5X to 8X the baseline of 5 to 10 ohms that is shown in the graph you posted.) The bass transformer, which is the same whether or not one uses the new mod, never gets anywhere near 7 ohms in my speakers (at all frequencies, the impedances I measured were higher), perhaps because I measured without the parallel resistor (which would tend to bring down the impedance as the natural impedance of the spkr is also falling with rising frequency). I did not measure with the parallel resistor in place. In fact, since a resistor would be expected to have a linear effect with frequency upon impedance, I don't see how removing a resistor could give you such a huge peak where there was none WITH the resistor, except to guess that it causes some aberrant resonance in the toroidal treble transformer, which is maybe why RW wants us to use the resistor with the toroid. This reveals an issue with the toroid, not an issue with the speaker itself.
Edits: 06/07/12
I don't see how removing a resistor could give you such a huge peak where there was none WITH the resistor, except to guess that it causes some aberrant resonance in the toroidal treble transformer
The resistor does 'linearize' the impedance curve, without it the transformer expresses its natural inclination to have a high impedance at the lower end of its range. It does this at a different range of frequencies than the bass transformer because its wound differently.
I'm thinking that there are variances in the transformers used as well as variances in panel spacing, materials and the like.
You wrote, "without it the transformer expresses its natural inclination to have a high impedance at the lower end of its range. It does this at a different range of frequencies than the bass transformer because its wound differently."
This is all true, but if you measure overall speaker impedance with both transformers hooked up and the parallel resistor and the series capacitor in front of the treble transformer and the series inductor in front of the bass transformer, the removal of the resistor (only) should not cause that peak, unless there is something else going on peculiar to the toroidal treble transformer. This is my point. How do you "wind" a transformer so it gives you a resonant peak at upper midrange frequencies? This is not something you do on purpose, I would think.
I think it has mostly to do with the turns ratio and the capacitance of the speaker. I'm pretty sure if you removed that transformer from the speaker you would find no such peak :) I'm pretty sure the bass transformer plays a role too.
"The bass transformer, which is the same whether or not one uses the new mod, never gets anywhere near 7 ohms in my speakers (at all frequencies, the impedances I measured were higher),"
I have info from one SLOG member who told me that the bass transformer DCR has changed over the years. I suppose it is hard to ensure uniformity through the years.
DCR would have little to do with large differences in speaker impedance except maybe it's an indication that they also played with the turns ratio over the years. The turns ratio is by far the primary determinant of the impedance seen by the amplifier, but a change in turns ratio would probably be reflected in a small change in DCR.
Right. IOW this has to do with how the driver interacts with the transformer. I'm not sure I'm on board with the idea that the peak represents a resonance. If that were the case then bass would really be a problem with our amps since the impedance is so high in the bass, but its not.
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: