Hey Rudy have you tried it? I've seen that you put two caps back to back in palce of C9...
Question: which is the sense of the resistor from the PW_gnd to the 318's 0V trace? Do you feel the necessity of a decoupling? Why? The input is already decoupled...right?
Or the reason is that in this way to use a regulated supply to feed the lm318 is simpler...
Mark
Question: which is the sense of the resistor from the PW_gnd to the 318's 0V trace? Do you feel the necessity of a decoupling? Why? The input is already decoupled...right?
Or the reason is that in this way to use a regulated supply to feed the lm318 is simpler...
Mark
Nice board Rudi!
Not sure about using 20'000uf per rail for most small current transformers but seeing as though you use those nice high VA rated C-cores
How about something like a pair of 4700uf caps with a small choke coupling inbetween the caps
The bi-polar caps for the fb and certainly the dual bridges for Schottkies are certainly worth doing
Not sure about using 20'000uf per rail for most small current transformers but seeing as though you use those nice high VA rated C-cores
How about something like a pair of 4700uf caps with a small choke coupling inbetween the caps
The bi-polar caps for the fb and certainly the dual bridges for Schottkies are certainly worth doing
rudi said:
I was thinking on that board you did, the latest one
It seems people report better sound using lower capactance with these chip amps
for one half rail If we used say 4700uf first then a choke and then another 4700uf would the chip see just 4700uf because of the choke coupling and should result in more mains borne interference being filtered
Probably Bol***s but just an idea
Hi Rudi,
Surprised to see the zeners there....Did you not decide that a separate reg'd supply was better? If so then getting the regs close to the chip might be an idea. You could leave the bidges and trans offboard. Or are zeners really ok again??? They seem alright to me, but I've not tried alternatives yet.
Regarding c9, you're all saying a bipolar here beats a normal elcap?
Glad to see the continuing drive for improvements
Surprised to see the zeners there....Did you not decide that a separate reg'd supply was better? If so then getting the regs close to the chip might be an idea. You could leave the bidges and trans offboard. Or are zeners really ok again??? They seem alright to me, but I've not tried alternatives yet.
Regarding c9, you're all saying a bipolar here beats a normal elcap?
Glad to see the continuing drive for improvements
So, I got a little soldering time tonight. Thought I would would build up a RevA chanel on one of the RevC boards and show some pics.
First, there's a few parts to leave off:
I am also leaving off the 6 Wima 1nF bypass caps, as they are not strictly RevA or RevC, more of an overall improvement.
So, basically just put it together, part for part, except for C10, which is special. The output resistor (R3) was left off so I could take a picture.
So, for C10 (100pF). It should mount under the board between pins 2 and 6 of the LM318. It's pretty easy with the tiny axial C0G cap.
I recommend also adding the 1uF bypass caps as well. Now, to make this a second RevC for a stereo set (my preference)...
First, there's a few parts to leave off:
An externally hosted image should be here but it was not working when we last tested it.
I am also leaving off the 6 Wima 1nF bypass caps, as they are not strictly RevA or RevC, more of an overall improvement.
So, basically just put it together, part for part, except for C10, which is special. The output resistor (R3) was left off so I could take a picture.
An externally hosted image should be here but it was not working when we last tested it.
So, for C10 (100pF). It should mount under the board between pins 2 and 6 of the LM318. It's pretty easy with the tiny axial C0G cap.
An externally hosted image should be here but it was not working when we last tested it.
I recommend also adding the 1uF bypass caps as well. Now, to make this a second RevC for a stereo set (my preference)...
float said:Hi Rudi,
Surprised to see the zeners there....Did you not decide that a separate reg'd supply was better? If so then getting the regs close to the chip might be an idea. You could leave the bidges and trans offboard. Or are zeners really ok again??? They seem alright to me, but I've not tried alternatives yet.
Regarding c9, you're all saying a bipolar here beats a normal elcap?
Glad to see the continuing drive for improvements
Hi FLoat
well funny you should mention that. I published this one just to see the reaction. I started with one with 78xx and 79xx regs last night. bit tight for space. but it is comming
you might have seen my 1.1kVA version not so long ago, and that is running with resistors and zeners and it is certainly sounding very good, but i just can't help myself and i will be giving it regulation VERY SOON, just to try.
I just prooved to myself again that power supply DOES MATTER and alot at that. the 1.1KVA version just sounds so much better than my 2 x 220va version. i haven't heard this much control in a very long time. the amp posesses a magic that i never used to have.
X-Calibre
Ok here is the Big Sinner going against all recommendations
I might have to open a new thread for this if i am to pursue it
Regulated supply for the LM318 (still have to check it for correctness) and filtering on all the diodes.
And to top it all, I even included something for Carlos.
Please take note. i am not saying that the resistor/zener sounds bad, I actually like it. We are just pushing the boundaries here and I for one like the regulators better
To list the changes
More main caps.
Bigger input cap (now you can use more exotic caps)
Regulated power supply for the LM318
Bipolar feedback
Discrete diodes for rectifier
LM318 Biasing option (might change this to ccs)
Ok here is the Big Sinner
going against all recommendationsI might have to open a new thread for this if i am to pursue it
Regulated supply for the LM318 (still have to check it for correctness) and filtering on all the diodes.
And to top it all, I even included something for Carlos.
Please take note. i am not saying that the resistor/zener sounds bad, I actually like it. We are just pushing the boundaries here and I for one like the regulators better
To list the changes
More main caps.
Bigger input cap (now you can use more exotic caps)
Regulated power supply for the LM318
Bipolar feedback
Discrete diodes for rectifier
LM318 Biasing option (might change this to ccs)
Re: X-Calibre
You're going to amp-builders hell
All they have there are carbon comps and weedy transformers
Actually the way I see it is this: Mauro did such a good job on the core of the amp that it shows all the benefits of supply tweaking.
Before you go for the boards, did you consider/have you tried Carlos's PS options; i.e the LM338 reg one and the CRCC unreg?
I've no experience of them, but if you have the bits handy...
One other thing- and its probably nothing- could the diode noise affect the output signal, as the diodes are close to the relay?
Keep up the good work mate.
rudi said:Ok here is the Big Sinner
You're going to amp-builders hell
All they have there are carbon comps and weedy transformers Actually the way I see it is this: Mauro did such a good job on the core of the amp that it shows all the benefits of supply tweaking.
Before you go for the boards, did you consider/have you tried Carlos's PS options; i.e the LM338 reg one and the CRCC unreg?
I've no experience of them, but if you have the bits handy...
One other thing- and its probably nothing- could the diode noise affect the output signal, as the diodes are close to the relay?
Keep up the good work mate.
What pcb software are you using Rudi?
I tried feeding the LM318 with ALW super regs last night
Sorry guys but theres no way I.m going back to zeners
I know the zeners are nice and compact but it shows that there is more performance to be had with this amp
A new thread sounds a good idea Rudi
I tried feeding the LM318 with ALW super regs last night
Sorry guys but theres no way I.m going back to zeners
I know the zeners are nice and compact but it shows that there is more performance to be had with this amp
A new thread sounds a good idea Rudi
Hi All,
It would be interesting to get feedback on these mods from the circuit's designer, Mauro. I know that in my experiments with LM78xx and LM317 etc. performance has actually suffered, measurably so.
Lower SNR, and higher THD resulted from my tests with very good regulated supplies on the LM318 circuit. And that with excellent bypassing and seperate trafos.
I would not be too quick to jump to conclusions without solid facts to back them up, at least I would like to see some measurements to see what you are seeing, because I can say without any doubt that I certainly can't hear it. If I could I would speak to Mauro about it, and ask what it is I might be seeing.
Have any of you who enjoy the reg mods actually measured anything on the bench such as Mauro has with all of his revisions? I would very much like to see the results.
My experience has been that while I have experimented with many PS combinations, and some even more exotic then those seen here, I have always come back to the original circuit, as it is very difficult to improve on, at least in any demonstrable concrete way. Though I had fun futzing around.
I would sooner concentrate on alternate current pump configuartions (as I am doing) then worry so much about the LM318. It is not a buffer, you cannot treat it as such, it is truly integrated with the LM3886(both PS and signal), this is not trivial, and not easy to grasp, but operating it on distinct supplies can actually nullify certain design attributes.
One certain way to get very good answers on how certain modifications would effect the circuit is to consult with the circuit's designer. This is something I think has not been done enough.
Of course fiddling is fun too! I do it all the time, but I don't pretend to be authoritative. I am not.
I for one am not in the least disuaded by those simple zeners, they are actually LESS noisy (measured), and not more noisy than most discrete or IC regulators.
So at a minimum a measurement of noise and THD would be very good for future discussions, also it would be good to involve Mauro so as not to appear to be trying to "hijack" his design. This is just common curtousy to the person who showed great kindness in sharing the concept in the first place.
If you wish a radicial departure from the design or one which has design goals which are contrary to those of the original designer, probably a new design/thread would be appropriate, hopefully one with the same level of detail and knowlege presented by Mauro.
It is OK to be contrarian, but I would find it so much more interesting to read a greater percentage of facts/theory and a lesser percentage of opinion/conjecture.
So in the future I submit it would be best to be gracious to our benefactor in this case, and consider his educated views on the circuit before we seek to color the thread with too much of our own design. If it is our own, it should be distinct, so as not to unfairly attribute the work of the originator on our own devices.
Cheers!
Russ
It would be interesting to get feedback on these mods from the circuit's designer, Mauro. I know that in my experiments with LM78xx and LM317 etc. performance has actually suffered, measurably so.
Lower SNR, and higher THD resulted from my tests with very good regulated supplies on the LM318 circuit. And that with excellent bypassing and seperate trafos.
I would not be too quick to jump to conclusions without solid facts to back them up, at least I would like to see some measurements to see what you are seeing, because I can say without any doubt that I certainly can't hear it. If I could I would speak to Mauro about it, and ask what it is I might be seeing.
Have any of you who enjoy the reg mods actually measured anything on the bench such as Mauro has with all of his revisions? I would very much like to see the results.
My experience has been that while I have experimented with many PS combinations, and some even more exotic then those seen here, I have always come back to the original circuit, as it is very difficult to improve on, at least in any demonstrable concrete way.
Though I had fun futzing around.I would sooner concentrate on alternate current pump configuartions (as I am doing) then worry so much about the LM318.
It is not a buffer, you cannot treat it as such, it is truly integrated with the LM3886(both PS and signal), this is not trivial, and not easy to grasp, but operating it on distinct supplies can actually nullify certain design attributes.One certain way to get very good answers on how certain modifications would effect the circuit is to consult with the circuit's designer. This is something I think has not been done enough.
Of course fiddling is fun too!
I do it all the time, but I don't pretend to be authoritative. I am not.I for one am not in the least disuaded by those simple zeners, they are actually LESS noisy (measured), and not more noisy than most discrete or IC regulators.
So at a minimum a measurement of noise and THD would be very good for future discussions, also it would be good to involve Mauro so as not to appear to be trying to "hijack" his design. This is just common curtousy to the person who showed great kindness in sharing the concept in the first place.
If you wish a radicial departure from the design or one which has design goals which are contrary to those of the original designer, probably a new design/thread would be appropriate, hopefully one with the same level of detail and knowlege presented by Mauro.
It is OK to be contrarian, but I would find it so much more interesting to read a greater percentage of facts/theory and a lesser percentage of opinion/conjecture.
So in the future I submit it would be best to be gracious to our benefactor in this case, and consider his educated views on the circuit before we seek to color the thread with too much of our own design. If it is our own, it should be distinct, so as not to unfairly attribute the work of the originator on our own devices.
Cheers!
Russ
A board that allows room for alternative components seems a good idea, though it mightl be too late for me.
I couldn't find a good quality bipolar,eg.Black Gate, for C9, that would fit. So I settled for Rubycon ZA 220uF 35V, bypassed with a Vishay Roederstein MKP 0.1uF 100V.
Can't see how you guys fit Freds without a messy board.
Best I could do was follow anothers lead and get a 25A Bridge Rectifier. It's a 400V Vishay for 280V RMS Input, Part No. GSIB2540.
Audie.
I couldn't find a good quality bipolar,eg.Black Gate, for C9, that would fit. So I settled for Rubycon ZA 220uF 35V, bypassed with a Vishay Roederstein MKP 0.1uF 100V.
Can't see how you guys fit Freds without a messy board.
Best I could do was follow anothers lead and get a 25A Bridge Rectifier. It's a 400V Vishay for 280V RMS Input, Part No. GSIB2540.
Audie.
Agreed, we should start a new thread for any mods or changes so this will be my last post on this one
BTW I agree not all regulators are great and I'm not fond of the fixed types, all I do know is that the ALW super regs brought more than a small improvement in my setup, even the missus noticed
The main reason I decided to build Rudi's board is so I can try out various things, if various changes sound bad its removed, this is diy, Mauro's amp is very good as it is and I'm grateful to him for sharing it but I'm sure others too like to experiment with any amp.
BTW I agree not all regulators are great and I'm not fond of the fixed types, all I do know is that the ALW super regs brought more than a small improvement in my setup, even the missus noticed
The main reason I decided to build Rudi's board is so I can try out various things, if various changes sound bad its removed, this is diy, Mauro's amp is very good as it is and I'm grateful to him for sharing it but I'm sure others too like to experiment with any amp.