Hi,
Yeh a resistive divider with some HF bypassing might be an excellent idea for this thing. I haven't really been able to connect a wire to it without bringing on the hiss. Additional bypassing at the connector didn't really seem to do much either.
With no input wire connected it's reasonably silent, can still hear it from 2' away so it's not dead silent, but good.
I went ahead and removed the main ground lift resistor of the PSU, it's a 10 ohm and with it gone you effectively float the output stage (but the input side of the bridge rectifiers are still earthed and can't be avoided unless you further isolate the mounting plate from the chassis, including the bolts you mount it with).
There are two more ground lift resistors as well, one per module, they're 2R2. I pried those off with a rusty butter knife and 2oz ballpeen hammer, then shorted the remaining connection with some silver plated copper string.
It didn't really affect the hiss at all. What it did do, was allow me to ground the chassis without having it get worse. So now my chassis is earthed.
I recommend you leave the 10R PSU ground lift resistor in place unless you really have a good idea what you're doing and don't mind your typical ground loop troubleshooting nightmare.
As per the 2R2 resistors, which lift the signal ground, in my view, you can also look at those as "noise amplifying" resistors, since any signal/shield noise picked up will drop a voltage across them which then gets amplified. In the same way, they don't really help with channel seperation -soundstage/imaging.
I find it better able to resolve complex passages clearly with them removed, as in the finer quieter parts aren't lost amongst more powerful ones, with a slight improvement in holography, soundstage/image precision and depth. It isn't night and day but it is there.
Previously the amp struck me as being extremely smooth sounding with excellent resolution and a bit too silky smooth.. while I found fairly neutral it wasn't transparent, imparting its own signature, which was actually a very pleasing one.
It still seems kind of silky, yet less smooth.. or maybe smoothed over. I'd credit the removal of the 2R2 signal ground lift resistors being removed. Maybe an idea to make it a little more versatile in grounding/connection arrangments would be to add provisions for jumping them out.
With all ground lift resistors in place your connection options are a little bit limited, fact is though there's no single method of connecting things that won't result in a problem for someone.
I'm most likely going to leave that ground lift resistor setup as it stands now, including when I take it fully balanced next month. I fully trust that will reduce the audible hiss.
Please note when I mention the hiss that my speakers are 102db efficient.
I kind of think HF chokes/ferrite clamps on the inputs may help it as well.
Another thing I'd like to caution new users of this module against... do not get the bright idea of a sealed enclosure. Plan on giving it alot of air! There's enough idle current flowing to keep the filter inductors hot enough that you really wouldn't want to keep your finger on it very long, and I can feel the heat rising from six inches above them. Vent your case well!
About those green bypass caps I said I'd try removing... true to my threat (...since I had the rusty butter knife out already) I sawed them off just on the one module. I guess they were doing no harm, but to my hear, they weren't doing any good either. So maybe you can save yourself a few pennies there Lars?
I'm currently enduring some FLAC recordings of Telarc Jazz...
No it's just not my music, but, the amp is doing an admirable job it.
Yeh a resistive divider with some HF bypassing might be an excellent idea for this thing. I haven't really been able to connect a wire to it without bringing on the hiss. Additional bypassing at the connector didn't really seem to do much either.
With no input wire connected it's reasonably silent, can still hear it from 2' away so it's not dead silent, but good.
I went ahead and removed the main ground lift resistor of the PSU, it's a 10 ohm and with it gone you effectively float the output stage (but the input side of the bridge rectifiers are still earthed and can't be avoided unless you further isolate the mounting plate from the chassis, including the bolts you mount it with).
There are two more ground lift resistors as well, one per module, they're 2R2. I pried those off with a rusty butter knife and 2oz ballpeen hammer, then shorted the remaining connection with some silver plated copper string.
It didn't really affect the hiss at all. What it did do, was allow me to ground the chassis without having it get worse. So now my chassis is earthed.
I recommend you leave the 10R PSU ground lift resistor in place unless you really have a good idea what you're doing and don't mind your typical ground loop troubleshooting nightmare.
As per the 2R2 resistors, which lift the signal ground, in my view, you can also look at those as "noise amplifying" resistors, since any signal/shield noise picked up will drop a voltage across them which then gets amplified. In the same way, they don't really help with channel seperation -soundstage/imaging.
I find it better able to resolve complex passages clearly with them removed, as in the finer quieter parts aren't lost amongst more powerful ones, with a slight improvement in holography, soundstage/image precision and depth. It isn't night and day but it is there.
Previously the amp struck me as being extremely smooth sounding with excellent resolution and a bit too silky smooth.. while I found fairly neutral it wasn't transparent, imparting its own signature, which was actually a very pleasing one.
It still seems kind of silky, yet less smooth.. or maybe smoothed over. I'd credit the removal of the 2R2 signal ground lift resistors being removed. Maybe an idea to make it a little more versatile in grounding/connection arrangments would be to add provisions for jumping them out.
With all ground lift resistors in place your connection options are a little bit limited, fact is though there's no single method of connecting things that won't result in a problem for someone.
I'm most likely going to leave that ground lift resistor setup as it stands now, including when I take it fully balanced next month. I fully trust that will reduce the audible hiss.
Please note when I mention the hiss that my speakers are 102db efficient.
I kind of think HF chokes/ferrite clamps on the inputs may help it as well.
Another thing I'd like to caution new users of this module against... do not get the bright idea of a sealed enclosure. Plan on giving it alot of air! There's enough idle current flowing to keep the filter inductors hot enough that you really wouldn't want to keep your finger on it very long, and I can feel the heat rising from six inches above them. Vent your case well!
About those green bypass caps I said I'd try removing... true to my threat (...since I had the rusty butter knife out already) I sawed them off just on the one module. I guess they were doing no harm, but to my hear, they weren't doing any good either. So maybe you can save yourself a few pennies there Lars?
I'm currently enduring some FLAC recordings of Telarc Jazz...
No it's just not my music, but, the amp is doing an admirable job it.Chris: I noted that the hiss went down, when you remove the input connections.
What do you have connected to the input exactly? Is there a passive potentiometer or is it directly hooked up to the signal source?
What do you have connected to the input exactly? Is there a passive potentiometer or is it directly hooked up to the signal source?
"You can earth it of course, but then you should also connect the enclosure to proper GND."
"If you want to earth it anyway, fine, but then you should just add the necessary GND connections to the enclosure."
"That is why i as a producer cannot uncritically connect things to earth, GND etc at free will."
I always thought that "earth" was just the British way of saying "ground", buit you seem to be distinguishing between them.
What's the difference?
Thanks
"If you want to earth it anyway, fine, but then you should just add the necessary GND connections to the enclosure."
"That is why i as a producer cannot uncritically connect things to earth, GND etc at free will."
I always thought that "earth" was just the British way of saying "ground", buit you seem to be distinguishing between them.
What's the difference?
Thanks
Is there only hiss when you have the inputs hooked up? Or is it there without the inputs hooked up as well?
Lars Clausen said:
Hi,
Direct to the source, of course of course.
Still using the same input wire/connectors/connections as I got it with, other than the other changes I've noted.
For the most part the problem is due to my source being pseudo earthed through the monitor. Damn EMU soundcard, if I give it a less than ideal (dirty) earth, and mine is, it pumps out spikes to the input, even if muted. So you can see why I was pushed to float the sucker.
Oddly enough if I unplug the monitor there's increased noise from the SMPS of the computer, whereas for the UCD (balanced inputs), it's noisier with the monitor plugged in.
So I'm not reaaaally going to worry about it until I get the inputs on it balanced, though I've already tried all I can.
Here is one thing that's really bugging me though, this little problem has remained constant throughout all and any changes in wiring from the very beginning. I have recently gone through swapping inputs, and even more recently, disconnected the inputs (quietest mode possible) and swapped the outputs.
The result, no matter what, one module is noisier than the other, by a good margin too. It's the one that's bugging me when sitting across the room trying to listen to it.
Swapping all the wires as I have has narrowed it down to the one module. It's also the one I removed the green gob bypass caps on, in fact it is why I chose that particular module to remove them on.
Guess what... it's the module with the stock SAL128 bypass caps. It's worse with no inputs connected, and it's much worse with them connected.
All I did was take a cheap 1/8 stereo cord, cheaap cheap cheap, spliced in a good 6' length of mic cable, a few inches from the connector, and at the end of that I "Y"'d it with more mic cable, the length of which I tie the shield to the inverting input for left/right.
I'm just guessing it beats a coat hanger 🙄 Connections are good and it seems like a decent cable, not microphonic at all, I'm sure it's not the best though.
I don't imagine someone with a more typical source/grounding arrangment and less efficient speakers would be bothered by it at all.
If you have a recommendation I'll give it a shot though.
In the meantime I plan on having some decent balanced connections next month, and I'm also looking for ideas to fix up my source power supply and grounding issue. I have yet to see a creative type soundcard that didn't have this problem, and I have to have a perfectly clean ground to supply it with. Nope, I don't want to hear signal transformers either.
Cheers,
Chris
soongsc said:
For the sake of argument I'll say it's not there with them not hooked up. Hearing it two feet away on my speakers isn't bad at all. With less efficient speakers you could say it's dead quiet.
If it is not there when disconnected,
isn't it from the source?
Chris, what amp where you using before?
How do the gain differ between that one and the NCD's?
And maybe, get some good ol' english 85 dB speakers,
that'l solve the problem... 😀
/Erland
PS
Yes, my transformator translation was "overly simplistic"
DS
isn't it from the source?
Chris, what amp where you using before?
How do the gain differ between that one and the NCD's?
And maybe, get some good ol' english 85 dB speakers,
that'l solve the problem... 😀
/Erland
PS
Yes, my transformator translation was "overly simplistic"
DS
eleson said:If it is not there when disconnected,
isn't it from the source?
Chris, what amp where you using before?
How do the gain differ between that one and the NCD's?
And maybe, get some good ol' english 85 dB speakers,
that'l solve the problem... 😀
/Erland
PS
Yes, my transformator translation was "overly simplistic"
DS
My other amp is a UcD400AD stereo amp. It has a stock gain of 26dB. NCD has a reported gain of 45dB.
Most of the noise is from the source, and some is from the interconnect itself which is a piece I just slapped together. It still seems like the input could benefit from some HF bypassing. I've been able to use the source end of it as an RF sniffer around the amp... though it was veery faint and the interconnect is about 16' long, speakers of 102db efficiency.
I have used it with some low efficiency speakers and yes it was more or less inaudible then. Of course I had to turn the volume up alot more to get the same enjoyment out of it, and yes, the amp really did drive the living hell out of them. Those speakers are Sound Dynamics "Reference Series" (ooooooh... aaaahhh) R616, that someone made the gross mistake of leaving here for too long. They're a disgusting 12ohms nominal, 8 ohms min. of efficiency unknown, and "max power 150W".
The amp gave them a great beating (I'd imagine the wire in the voice coil is a nice shade of purple and blue now). Once turned up enough to overcome the poor efficiency of them, they actually sounded really good, for which I credit the amp. Turn it up a bit beyond that and the speakers started to cry in pain.
Anyway to better quantify the level of hiss, on my speakers of 102dB efficiency w/m, without interconnects, hiss is audible from about 2.5 to 3' away. I think that's pretty good. The UCD's without interconnects, hiss is only audible with your ear litterally glued to the tweater, and becomes inaudible more than 3" away. I think that's insanely good. You could probably relate that to the difference in gain though.
On the other speakers without interconnects it's totally inaudible, with them, hiss (from the cable and source grounding issues) is audible from 2.5 to 3 feet away.
So I do think the inputs could benefit from a little RF filtering, but, I don't think it's the kind of problem most people are going to encounter with it. I'm maybe just being too honest in saying that my setup is obviously less than ideal. Hey, maybe some Shakti stones inside my ATX case would do the trick
with 20dB extra gain ...
... I personally don't worry about the hiss.
It could well be from the source.
I'm looking for 5.1 setup, and having all five speakers
that close makes if hard to follow the film as well.
Yes, stones could do it , properly prepared
/Erland
transformator - transformer , translations, translations ...
... I personally don't worry about the hiss.
It could well be from the source.
I'm looking for 5.1 setup, and having all five speakers
that close makes if hard to follow the film as well.
Yes, stones could do it , properly prepared
/Erland
transformator - transformer , translations, translations ...
So it really seems that the gain of the NCD is high to minimize feeback and let the sound be more dynamic, as a result, you will hear some hiss if the speakers are very effecient. I think the sensitivity is way to high letting you reach close to full power with less than 0.4V. Not sure what the trade-off considerations were.
The output coils heating up means that there is quite some idle current being drawn, I wonder how much.
The output coils heating up means that there is quite some idle current being drawn, I wonder how much.
Gain and hypex
The top document here was new to me (Late July)
http://www.hypex.nl/applicationnotes.htm
Pure coincidence that Bruno posts some info on choice of gain?
( Well I think it is)
/Erland
The top document here was new to me (Late July)
http://www.hypex.nl/applicationnotes.htm
Pure coincidence that Bruno posts some info on choice of gain?
( Well I think it is)
/Erland
It was new to me earlier today as well, however he's gone over all that in the forum ages ago. 🙂
input caps... dc-coupling?
So it seems that there are input decoupling caps stock on the PCB, if I understand correctly?
And these caps be bypassed? I am puzzled that they are there at all on one of your constructions, Lars... why? Not possible to make the amp DC-coupled all the way through for some reason or what...?
Anyway, some of us feel that the only good capacitor to have in the signal path is NO capacitor... after going DC-coupled all the way a couple of years ago, I don't really feel like going back. Will we have to if we want an NCD amp?
Thanks again to Lars and Chris for letting us read about real-world experiences with the prototype 🙂
/cdl
So it seems that there are input decoupling caps stock on the PCB, if I understand correctly?
And these caps be bypassed? I am puzzled that they are there at all on one of your constructions, Lars... why? Not possible to make the amp DC-coupled all the way through for some reason or what...?
Anyway, some of us feel that the only good capacitor to have in the signal path is NO capacitor... after going DC-coupled all the way a couple of years ago, I don't really feel like going back. Will we have to if we want an NCD amp?
Thanks again to Lars and Chris for letting us read about real-world experiences with the prototype 🙂
/cdl
Lars Clausen said:
Sounds familiar, thats why I use them also😉
With high gain you cannot do without them(I mean the groundlift res.).😀
Re: input caps... dc-coupling?
The amp is DC coupled. This could be a bit of a problem for some folks with high offset at their source. There's a DC offset adjustment but Lars doesn't want it messed with 🙂
I was simply thinking it could use some RF decoupling /filter caps at the input which I don't think it has now. If you want DC coupled inputs though you're all set. If you want AC coupled, you'll probably need a pretty honking big cap, because the input impedance is on the low side.
Really? I took them out, and it sounds much better 😀 Proved my point? Ground lift baaaaaaaad... yucky poo poo (technical term).
BTW, I reworked the RCA ends of my interconnects just now, and found the magic ground lay pattern (by that I mean where the wires hit the floor).. just seperated the channels a bit. Result? High frequency hiss is now very little worse than with no interconnects at all. Almost not audible from across the room 😀
Just in time to wake the neighbours up too. May the listening begin!!!!
cdl said:
The amp is DC coupled. This could be a bit of a problem for some folks with high offset at their source. There's a DC offset adjustment but Lars doesn't want it messed with 🙂
I was simply thinking it could use some RF decoupling /filter caps at the input which I don't think it has now. If you want DC coupled inputs though you're all set. If you want AC coupled, you'll probably need a pretty honking big cap, because the input impedance is on the low side.
Bgt said:
Really? I took them out, and it sounds much better 😀 Proved my point? Ground lift baaaaaaaad... yucky poo poo (technical term).
BTW, I reworked the RCA ends of my interconnects just now, and found the magic ground lay pattern (by that I mean where the wires hit the floor).. just seperated the channels a bit. Result? High frequency hiss is now very little worse than with no interconnects at all. Almost not audible from across the room 😀
Just in time to wake the neighbours up too. May the listening begin!!!!
Re: Re: input caps... dc-coupling?
Phew!!! Happy to hear that. I almost got scared there for a moment...
/cdl
classd4sure said:
Phew!!! Happy to hear that. I almost got scared there for a moment...
/cdl
Lars: I saw on your website that you will be on Hifi & Surround Copenhagen Expo Sep 23rd and 24th 2006.
What will you bring to the expo?
ClassD4Sure: Keep the impressions coming, its nice reading at work... 😀
What will you bring to the expo?
ClassD4Sure: Keep the impressions coming, its nice reading at work... 😀
Re: Re: input caps... dc-coupling?
It all depends on the way you ground the amp.
Without the groundlift resistors my amp. sounded distorted and was humming in a 1 transf. setup. But in your setup it is maybe better to take them out.🙂 There is no "1 off standard/magic" solution for everyone. And your ears/scope will tell you.
classd4sure said:
It all depends on the way you ground the amp.
Without the groundlift resistors my amp. sounded distorted and was humming in a 1 transf. setup. But in your setup it is maybe better to take them out.🙂 There is no "1 off standard/magic" solution for everyone. And your ears/scope will tell you.
Hi Noah,
Ok, the more I read this forum, the madder I get. First of all, we are way off topic. We are a bunch of DIYers, out to have some fun, put together some good sounding gear, and tell others about our experiences. Frankly, who gives a rip about blind testing. I have talked to Chris at length, (hey where you been) and I'll trust his judgment.
To get a little personal, but not to offend, you stated you couldn't hear the difference between a Bedini and an Atcom amp. So, what does it matter to you if it's blind testing or not.
Hey, Chris, I'll be happy to come up there, and perform that blind test for you. I am blind, and I don't need no blind test to tell me a difference between an Atcom and a Bedini. By the way, if you have that Bedini, I might consider taking it off your hands.
Ok, folks, let have some fun. All this patent crap, and blind testing is boring to say the least.
Regards to all,
Ray
Ok, the more I read this forum, the madder I get. First of all, we are way off topic. We are a bunch of DIYers, out to have some fun, put together some good sounding gear, and tell others about our experiences. Frankly, who gives a rip about blind testing. I have talked to Chris at length, (hey where you been) and I'll trust his judgment.
To get a little personal, but not to offend, you stated you couldn't hear the difference between a Bedini and an Atcom amp. So, what does it matter to you if it's blind testing or not.
Hey, Chris, I'll be happy to come up there, and perform that blind test for you. I am blind, and I don't need no blind test to tell me a difference between an Atcom and a Bedini. By the way, if you have that Bedini, I might consider taking it off your hands.
Ok, folks, let have some fun. All this patent crap, and blind testing is boring to say the least.
Regards to all,
Ray
- Status
- Not open for further replies.