Sheldon said:
Well, S/N ratio is not perfect. In fact, it may be quite bad, but some of the measured noise may be caused by not proper measuring method, so I don´k now how bad it really is.
I still wasn´t able to solve all problems with ground loops around my computer.
In real life the hum is inaudible, until you put your ear 5cm far from woofer (86 dB sens.) in very quiet room.
An externally hosted image should be here but it was not working when we last tested it.
Wronski said:
I'm running a multi-amped set up with a DEQX. The more channels, the more chance for ground problems. I had to experiment some to get everything to behave. Each individual amp was very quiet, but all hooked up I got some noise and Rf interference. The biggest benefit was to work out my mains connections in a star type configuration. I was able to get things very quiet.
I actually have a four way system, but the DEQX only does three way. So I made an active line level phase perfect crossover at 80Hz between the woofer and sub. In the DEQX I treat the pair as a single driver. I made the crossover circuit on a small perf board and installed one in each of two stereo Symasym amps - a lesser version of what you have done. What noise I get now is actually from the crossover circuit. It measures less than 0.5mV and can only be detected no further than 20 or so cm from the woofer (97dB). Nothing detectable from the sub. One of these days, I'll redo the ground connections in the crossovers in the pursuit of perfection.
As a practical matter, if you can't hear the hum any further than 5m away, then it's good to go.
Sheldon
I have redone the cascode reference to my headphone Amp circuit as recomended by MikeB with good results and fantastic PSRR. http://esnips.com/doc/c8d86780-d638...TRANSCENDENTAL-Amp-sction-schematic-v-1.1.pdf
I have some 2SA1216-2SC2922 - would you recommend them for symasym? What value for C14 if so?
bremen nacht said:
yes, but I had ceramics, which impacted bad on the sound. You should use as good C14 as possible.
I was afraid that I had fakes of those transistors, but they worked well, their CB capacitance was very low (~22pF measured).
wronski,
I have found a image hosting site here:
http://imageshack.us/?x=my6&myref=
I haven't tried it yet, but it looks fine.
Let me know if it works well! 😉
Nice ! 🙂 Just curious, what will drive trebles ? Don't forget, the mjl2119x need quite high bias.
To you and wronski, if symasym drives bass/mid only, the 1000uf onboard can be skipped, this helps cleaning up the 50hz residuals from the PSU.
Mike
To you and wronski, if symasym drives bass/mid only, the 1000uf onboard can be skipped, this helps cleaning up the 50hz residuals from the PSU.
Mike
Hi Mike,
I'd leave the caps there. They will help with transients and reduce supply noise. They don't cost much either in the grand scheme of things.
-Chris
I'd leave the caps there. They will help with transients and reduce supply noise. They don't cost much either in the grand scheme of things.
-Chris
MikeB said:
Another symasym 🙂 ; the first one, with mjl15034/5 and mjw0281/0302.
Thank you for the tip. I'll remove all 1k caps.
What do you think about the 10pf that I used now? 😀 It's a very old cylindrical capacitor removed from a russian tubes Rubin TV
Thanks Mike, my Symasym is now driving two way speakers, but it is intended to be a flexible system to help me with setting different speakers.
Considering the slot system and swapable cards of the active crossover, it´s reasonable to keep each channel in perfect condition 😉
Considering the slot system and swapable cards of the active crossover, it´s reasonable to keep each channel in perfect condition 😉
Yes, in this case its better to keep the 1000µF, only if symasym is planned to operate at low freqs only these caps may be omitted.
Chris, its about the 50/60hz current peaks, a grounding issue will give some harmonics of these into the signal. If the 100nf bypass caps stay, the noise should still be filtered, what do you think ?
Mike
Chris, its about the 50/60hz current peaks, a grounding issue will give some harmonics of these into the signal. If the 100nf bypass caps stay, the noise should still be filtered, what do you think ?
Mike
Hi Mike,
I think the 1,000 uF caps help with both filtering the PSU's and also with any current demand. This would be frequency independant.
Possibly a line of 1-470uF, a couple 100 uF and some 10's and 1 uF plus the 0.1 uF. That should present a nice low impedance across the band. Of course, it would eat up board space too!
At any rate, I think there should be something on-board that can supply current in bursts.
Have you tried this yet Mike? I know I haven't.
-Chris
I think the 1,000 uF caps help with both filtering the PSU's and also with any current demand. This would be frequency independant.
Possibly a line of 1-470uF, a couple 100 uF and some 10's and 1 uF plus the 0.1 uF. That should present a nice low impedance across the band. Of course, it would eat up board space too!
At any rate, I think there should be something on-board that can supply current in bursts.
Have you tried this yet Mike? I know I haven't.
-Chris