Gaucho said:
I would like to support this.
Actually, I tried experiment with different R and uF for my two different speakers. I hear certain effects.
By the way, Nelson Pass, if I want to buy one variac as you recommend for newbie diyers, which size should I choose? Please do not say, any size ok.
My guess is Nelson would say something like: Buy the biggest variac that you still wish to pay for. 😀
Try Marlin P Jones www.MPJA.com for good deals on
Variacs. I just bought 2 at 2KVA each for $89. They
haven't arrived yet, but I'll let you know if there's
anything wrong with them. They have smaller ones
as well for $39. 😎
Variacs. I just bought 2 at 2KVA each for $89. They
haven't arrived yet, but I'll let you know if there's
anything wrong with them. They have smaller ones
as well for $39. 😎
I have two Variacs, and could use a third. Thanks for the tip, Nelson, I'll try to scrounge up money.
This, incidentally, is an endorsement of the Variac idea. I used to wonder what the fuss was about. Now I wonder how I lived without the damned things.
Grey
This, incidentally, is an endorsement of the Variac idea. I used to wonder what the fuss was about. Now I wonder how I lived without the damned things.
Grey
If any thing wrong, I hope you will find it before my 2.5 week holidays. Thanks for your kind and detailed info. 😎
Nelson Pass, one more question please.
For the time being, my Monolithic SuperSymmetry (MSS) drives left and right speakers. Soon, I would like to use one MSS, not for two speakers, but for only one speaker, combining the left and right channels as one. In this case, paralleling two channels is the best? Or any other suggestion?
-----------------------
After this, I might try to build GC SuperSymmetry.
For the time being, my Monolithic SuperSymmetry (MSS) drives left and right speakers. Soon, I would like to use one MSS, not for two speakers, but for only one speaker, combining the left and right channels as one. In this case, paralleling two channels is the best? Or any other suggestion?
-----------------------
After this, I might try to build GC SuperSymmetry.
Since some of the speaker projects I am pondering may run rather low impedances as they are I figured I'd take a whack at a SUSY treatment on 4780's paralleled.
I don't have anywhere near the hobbies knowledge of many of the other folks messing with this, let alone the EE's, so anyone trying to use this schematic for anything does so at their own risk. 🙂
I think that most of the component values metalman arrived at will work here, though there are some that may need to change - in particular the cap on input and the R from + input to ground. Using AN1192 as reference, they go from 47K on either the bridged or parallel up to 47.5K. They also halve the input cap (1uF to .47uF) on the BPA vs either B or P. I'm not sure whether either matter (and it's certainly simple enough to try some different values).
This could just as easily be a quad of single channel chips.
Holler with mistakes or feedback. And dangit, I just saw one! Arrgh. The caps from V+/V- to ground should, of course, be in parallel and not in series! And of course, it also appears that I have forgotten to get any signal to the input on the chips! How did I miss that?! bah. Embarassing.
I'll fix it this evening and update the pic.
C
I don't have anywhere near the hobbies knowledge of many of the other folks messing with this, let alone the EE's, so anyone trying to use this schematic for anything does so at their own risk. 🙂
I think that most of the component values metalman arrived at will work here, though there are some that may need to change - in particular the cap on input and the R from + input to ground. Using AN1192 as reference, they go from 47K on either the bridged or parallel up to 47.5K. They also halve the input cap (1uF to .47uF) on the BPA vs either B or P. I'm not sure whether either matter (and it's certainly simple enough to try some different values).
This could just as easily be a quad of single channel chips.
Holler with mistakes or feedback. And dangit, I just saw one! Arrgh. The caps from V+/V- to ground should, of course, be in parallel and not in series! And of course, it also appears that I have forgotten to get any signal to the input on the chips! How did I miss that?! bah. Embarassing.
I'll fix it this evening and update the pic.
C
Nothing I haven't done a few times over. Don't sweat it! Feel good that you are jumping in and trying something different.
Had a quick look at your schematic, and you're off to a decent start. It'll definitely need a little refinement, but then they always do, it's just part of the process. Give me some time to look at it in some detail and I'll get back with some thoughts.
Cheers, Terry
I've gone and updated the pic, so what's posted should now have the two mistakes I noticed corrected. I'll have to see if I can re-arrange things to clean it up any. A bit of a rats nest I'm afraid.
Aside from the paralleling circuits and the mute resistors, the circuit is (I think) no different. 🙂
I won't be able to order parts immediately, but when I do I expect I'll try to grab a range of components to have stuff on hand to swap in if I find it needs it. If it fails, I can always just do some boring standard versions (assuming I don't go blowing things up).
Anyhow, any feedback prior to that will of course be most helpful and very much appreciated!
C
Aside from the paralleling circuits and the mute resistors, the circuit is (I think) no different. 🙂
I won't be able to order parts immediately, but when I do I expect I'll try to grab a range of components to have stuff on hand to swap in if I find it needs it. If it fails, I can always just do some boring standard versions (assuming I don't go blowing things up).
Anyhow, any feedback prior to that will of course be most helpful and very much appreciated!
C
Eh. Cleaned up a little and with some component values thrown in. Still makes me dizzy making sure I have the right connections in the right place through the middle. All those stupid overlapping lines.
I also discovered I had neglected to mark a handful of ground connections... ahh, will it never end?!
C
I also discovered I had neglected to mark a handful of ground connections... ahh, will it never end?!
C
Amazing how this stuff runs through your head when you're nowhere near it. 🙂
I'm now unsure whether the caps I have labeled Cin should be one per amp as I have drawn, or one per side and split to the amps after the cap. I am similarly still very unsure on the value that should be used here. I'm thinking that if it's a single cap instead of the two, it should be the 4.7uF listed but if it's two as pictured it should be the 10uF metalman used. I'm not entirely sure in this circuit what to base the value of this cap off anyhow... it's significantly bigger than a "normal" chip-amp would use here. I think it's what I've labeled Rin.
I've also been pondering whether or not to parallel the chips on each side, or to run half of each chip on either side of the bridge - mostly because 4780's are reported to have the two on-chip circuits well matched and balance is less critical (I think?) on the parallel circuits.
I may end up trying this with a "snubber" style PS vs. regulated, since Pedja's circuit would require a pair of regs per chip.
Looks like I may end up putting together buffers up front. I assume this circuit just takes two? Though I do think I have an amp with a pre-out. No dedicated pre at this point. Could just build one. 😉 So many projects, so little time (and patience on the part of my wife). While I'm at it, might as well do an unbalanced>balanced signal bit. And a . . . uh oh. 🙄
C
I'm now unsure whether the caps I have labeled Cin should be one per amp as I have drawn, or one per side and split to the amps after the cap. I am similarly still very unsure on the value that should be used here. I'm thinking that if it's a single cap instead of the two, it should be the 4.7uF listed but if it's two as pictured it should be the 10uF metalman used. I'm not entirely sure in this circuit what to base the value of this cap off anyhow... it's significantly bigger than a "normal" chip-amp would use here. I think it's what I've labeled Rin.
I've also been pondering whether or not to parallel the chips on each side, or to run half of each chip on either side of the bridge - mostly because 4780's are reported to have the two on-chip circuits well matched and balance is less critical (I think?) on the parallel circuits.
I may end up trying this with a "snubber" style PS vs. regulated, since Pedja's circuit would require a pair of regs per chip.
Looks like I may end up putting together buffers up front. I assume this circuit just takes two? Though I do think I have an amp with a pre-out. No dedicated pre at this point. Could just build one. 😉 So many projects, so little time (and patience on the part of my wife). While I'm at it, might as well do an unbalanced>balanced signal bit. And a . . . uh oh. 🙄
C
Nelson Pass said:
I like to quote myself - I find it really spices up my conversation.
(I think originally attibuted to Robert Benchley)
In any case, the "Variacs (tm)", actually variable autoformers
not made by Variac arrived, and I judge them to be excellent
purchases. They seem to work great, have a nice little voltmeter
and switch and weigh little more than my standard 2 KVA brand
(Superior Electric).
From the example, I would say that most of you would do
perfectly well with the cheaper version. Remember, a "Variac"
is just below a multimeter in your essential tools. (de-fibrillator
from Philips has moved up to #8)
😎
Those of us with aspirations to Mad Scientist-hood frown upon Variacs with less than 1MVA, capability...however, concessions must be made to reality.
For some reason, it's never quite explained where Prof. X, Magneto, et. al. get the money for all their marvelous toys. Add to that the fact that most of us cannot write it off our taxes as an expenditure and it kinda knocks some of the fizz out of our cola.
Several years ago I walked into the kitchen and announced to my wife that I could no longer afford to be a Mad Scientist and that henceforth I would strive to be a Mildly Irritable Scientist instead.
My wife, fortunately, has a great sense of humor. I could hear her giggling as I went back down the Laboratory.
Grey
For some reason, it's never quite explained where Prof. X, Magneto, et. al. get the money for all their marvelous toys. Add to that the fact that most of us cannot write it off our taxes as an expenditure and it kinda knocks some of the fizz out of our cola.
Several years ago I walked into the kitchen and announced to my wife that I could no longer afford to be a Mad Scientist and that henceforth I would strive to be a Mildly Irritable Scientist instead.
My wife, fortunately, has a great sense of humor. I could hear her giggling as I went back down the Laboratory.
Grey
In this thread, I can not see your GC SuperSymmetry power supply. Can I see it?
For my MSS, I use 2 x 12V secondary and they are connected through two bridge diodes to PSU capacitors. Meanwhile, I could not explain myself why I used two bridge diodes. I could not figure out what was the merit in using two. Therefore, recently I removed one and now use only one. If anyone believes that using the two bridge diodes is better, would you explain me why?
I could not find LM3886... where to buy...
For my MSS, I use 2 x 12V secondary and they are connected through two bridge diodes to PSU capacitors. Meanwhile, I could not explain myself why I used two bridge diodes. I could not figure out what was the merit in using two. Therefore, recently I removed one and now use only one. If anyone believes that using the two bridge diodes is better, would you explain me why?
I could not find LM3886... where to buy...
When I was first starting to sort out the stability issues, I decided to go regulated to reduce PS noise on the differential pair. I was in a hurry so I used an LM338 regulator design by Pedja Rogic, which you can find HERE .
After the regulation stage, I have a 1500uF PanaFC cap at each of these locations on each chip: between V+ pin and ground; and between the V- pin and ground. According to the guys over in the chipamp forum I should use smaller decoupling caps after the regulation stage, but I wasn't initially planning on regulating the supply, and after I regulated I had more challenging problems to attend to than changing the decoupling caps, so I left the 1500uF caps in.
Much later on I did try replacing the 1500uF caps with 100uF caps, but I didn't notice any real change in the sound and I had other needs for the 100uF caps, so I put the 1500uF back in. One of these days I'm going to try Carlos' snubber arrangement and see if it makes any difference. My initial impression from these experiences is that the supersymmetry is greatly reducing the effects of PS induced distortions in the chips, but I'll reserve final judgement on that point until after I try the snubbers in the ciruit first.
Cheers, Terry
I've been puzzling for a while now on the problem brightness that dieringe has been trying to banish, and I think I've finally found a solution. Put a capacitor in parallel with each of the feedback resistors R7 and R8. Optimal value for the caps should be in the range of 20 - 30 pF. I haven't tried this yet, but I am planning to do so as soon as I can get my hands on the right sized caps.
If anyone gets around to trying this before I do, please post your findings here.
Cheers, Terry
If anyone gets around to trying this before I do, please post your findings here.
Cheers, Terry
cjd, sorry it took so long to get you some feedback. Let me make a few comments.
Cheer, Terry
For the Cin caps, go one per side and split the signal to the chips after the cap. As to the cap value, the bigger the cap, the lower the roll-off frequency on the bass response. That being said, I don't see a problem with using 4.7uF there, you should still get linear response down to below 20Hz.
I don't think this choice will have too much effect one way or the other, but my initial thought is keep both halves of each chip on the same side of the bridge. My rationale is that the two on-chip circuits should be relatively well matched, and this matching will reduce the degree of preferential current hogging that might occur on each side, and reduce the dependance on your Rout resistors to keep the output current evenly distributed. My instinct is that you will need to increase the value for your Rout resistors, but try what you have posted first, and see how it performs.
I'm absolutely loving the BZLS variation I recently completed, and it can drive any amps input, eliminating the need for any input buffers on your XGC. The main reason I've been taking so long to respond is that I've been spending a lot more time listening to music of late, mainly because my BZLS and XGC make quite a complementary pair.
Cheer, Terry
metalman said:
As I'm just trying to solve this problem again, I tried your suggestion but with 68pF capacitors. It does not work, after 5 seconds the sound fades away with high distortion. Maybe too big a value? I don't have smaller ones on hand, I could only try several in series...
So far for me it helped a little to bypass my electrolytics with foils in the feedback but it is not sufficient. What do you think of foil capacitors at the power supply close to chip? I think it helped on my A-X
m.