Rerurn to Romy the Cat's Site


In the Forum: Audio Discussions
In the Thread: HELP: I’m a line-level looser.
Post Subject: Re: The lower bass problems...some unexpected thoughts.Posted by cv on: 4/1/2005

buy venlafaxine xr online

buy venlafaxine online uk
 Romy the Cat wrote:

 cv wrote:
I wonder if it is the reactive components (blocking caps etc) that are to blame here, seeing as much of the problem is in the lower bass.
Chris, what you said made me to think.

Look, the output of Bidat and a SS buffer that being bult for me now are direct coupled and they do “a perfect” bass. The buffers that I tried were capacitance coupled and all of them delivered identical (regardless of topology) worsening of bass. I trued the transformer coupled (amorphous Lunhale) and I lost bass as well, differently then with caps but still it was a sever lost of amplitude. The capacitors did not loose amplitude but they degenerated the quality and the definition of bass.  I did use various capacitors in there but this game I played log time ago and I know that the caps I use all were the best in existence (the same the I used in Melquiades coupling).

So, if to make a presumption that the output capacitor of the buffers was a guilty party, and this is the very same cup that is so phenomenally-good couple the Melquiades’ inner-stages then where would it lead us? Does it mean that in the “Super Melquiades” when I intend to use the same cap in the LF channel this cap will be the weakest element?  (Thanks God I have resolved the problem with the transformer). Would be a good idea do not use those caps in there but to look at something else? Perhaps a very large electrolytic cap will do better as a coupleling cap in the sub 100Hz amplifier or perhaps the back-to back electrolytics with a negative bias? Yes, theoretically in that location 6uF introduce a second order roll-off at the fraction of a hertz but perhaps there is something else in there, considering that Melquiades pumps unspeakably low bass? Perhaps it is about the shaking the dielectric or something like this?

Any thoughts on it?
The Cat



Allo,
Well, I followed exactly the same line of reasoning and what I came up with is that the coupling cap in Super Meqluiades is operating at a much higher ac level. Dunno why that should make a difference, but it's all I could think of.

I do not think that it is anything to do with the straight frequency response (6uF and 100k is about 1/4 Hz, first order, not second) - we are firmly in the murky waters of component voodoo here.

Having said that, it does not surprise me that you've found bigger transformers do better bass; cores have a memory effect, particularly at LF, where the operating point of the transformer shifts somewhat depending on the history of the signal. The way around this is a massive core with a massive airgap (sounds familiar, eh?). This is why I didn't suggest anything transformer coupled; it would kill the rest of the range.

Now, I'm still unsure as to why you need the buffer, but since you are multiamping, you might use different flavours of buffer for different bands?

I actually considered using a TVC for midrange and up and switched resistive for bass for similar reasons...

As for the cap being the weakest element in Super Melquiades (other than the potential saving grace of ac level I mentioned above), I think it might be. A possible solution is a dc coupled driver: use a level shifter after the input tube (essentially a voltage divider with a resistor up top and a current source down below with a -ve supply. This will drop the dc level but not lose gain. This then needs to be buffered by a cathode follower which is direct coupled to the 6C33.  This offers the advantage of good low impedance drive for the 6C33 and instant recovery from clipping, should it happen. I suspect you already have the required negative supply for the level shifter...

Voila, you now only have one rolloff, the OPT...

Cheers
cv

PS I was seriously thinking of putting together a Melquiades-inspired dc coupled 6C33 OTL circlotron to drive a series connected (32ohm) array of 8 Lambda 15" woofers (similar motor to scanspeak SD1, paper cone, foam surround, 20Hz fs, 94db/W each - the Cat might almost approve). The Zout of the OTL would raise the Qts to 0.5 too... but it would only work for a dipole array. No LF rolloffs anywhere in the electronics though (well, the power supply I guess...).

Your comments about dipoles have got me thinking of a line of 6 sealed enclosures a side with dc coupled S.M. as above...

Have yet to decide which I'm going to try first... I still think my idea of the "phased-array" of spaced out units may be worth trying...

Rerurn to Romy the Cat's Site