let's find out whether attenuating the RF reduces the noise problem, before we start looking at which components could be better or worse.
SC ULD2
"Hi,
your existing RF filter is set very high.
Currently the RC = 100 * 820 * 10^-12 = 0.082us.
The has F-3dB @ ~ 1.9MHz.
Andrew
As usual you are correct in your reply, however the reasons for this are covered in p.25 of the original article that you have access to.
Alex
P.S.
When I suggested reducing the 820pf capacitor to 100 pf a while back,I mistakenly assumed that it was a 1K resistor, not a 100 ohm resistor. This input filter is completely different to normal SC practice of 1K+1n2 or 820pf.
"Hi,
your existing RF filter is set very high.
Currently the RC = 100 * 820 * 10^-12 = 0.082us.
The has F-3dB @ ~ 1.9MHz.
Andrew
As usual you are correct in your reply, however the reasons for this are covered in p.25 of the original article that you have access to.
Alex
P.S.
When I suggested reducing the 820pf capacitor to 100 pf a while back,I mistakenly assumed that it was a 1K resistor, not a 100 ohm resistor. This input filter is completely different to normal SC practice of 1K+1n2 or 820pf.
jp_howard said:
And the RCA out's? are the negatives(grnd) earthed to chassis, pcb grnd, or star grnd?
Don't know if you found this site before.
But always useful info.
http://sound.westhost.com/earthing.htm
"P.S.
When I suggested reducing the 820pf capacitor to 100 pf a while back,I mistakenly assumed that it was a 1K resistor, not a 100 ohm resistor. This input filter is completely different to normal SC practice of 1K+1n2 or 820pf."
ps
Alex I did want to ask you about that, another thing I forgot.
1k0+100pF and 100r+820pF are not that much different (<22%).
I don't like the 100pF idea when considered alone. But let's leave two stage filtering until we know more about this particular amplifier and the way it interacts with the sources available.
I don't like the 100pF idea when considered alone. But let's leave two stage filtering until we know more about this particular amplifier and the way it interacts with the sources available.
I don't have the amps in a case at all yet, so there's no star earth or chassis earth per se. The RCA negatives are simply connected to the PCB's 0V.
jp_howard said:
Jeremy
The reason for the use of a 47uF capacitor is given in the original article. If available, a 47uF Blackgate should be even better for the reasons given in the article. However a 2uF quality polypropylene will sound better than one of those 47uF NP electros ,but it's physical dimensions make it awkward to implement. Hugh's choice of a 2uF Blackgate will certainly sound better than the original.
Alex
Re: SC ULD2
The 100Ohm input resistor and 820pF capacitor constitute a low-pass filter with a -6dB/octave roll-off above 1.9MHz.
This is a much lower impedance network than our previous designs, in order to provide the lowest impedance for the signal source.
Hey cool - until today that bit of the article was all Greek to me... but now it actually makes sense! Thanks guys for educating me somewhat. 🙂
So, in order to keep the impedance low, but filter some of the hiss from the input, should I increase the cap instead of the resistor? What's the practical implication of changing one vs the other (in the filter equation of course they appear together as R*C)?
For those without the article, here's the relevant bit:sandyK said:
The 100Ohm input resistor and 820pF capacitor constitute a low-pass filter with a -6dB/octave roll-off above 1.9MHz.
This is a much lower impedance network than our previous designs, in order to provide the lowest impedance for the signal source.
Hey cool - until today that bit of the article was all Greek to me... but now it actually makes sense! Thanks guys for educating me somewhat. 🙂
So, in order to keep the impedance low, but filter some of the hiss from the input, should I increase the cap instead of the resistor? What's the practical implication of changing one vs the other (in the filter equation of course they appear together as R*C)?
jp_howard said:
Ok
I can relate to that😀
I gather standard psu pcb and cable connections?
If yes, keep it at that and try AndrewT's 1k suggestion.
You can always vary the resistor up or down for preference.
That's one of the many bits in the article I still don't understand. I'll get there one day!...sandyK said:
Actually there's heaps of room on the PCB there - by sticking the poly vertically (extending one of the legs with a bit of extra wire) it fits no problem.
I discovered that Jaycar has something called a "resistor wheel". Might be a useful little shortcut for tweaks like this - wind up the resistance until the hiss goes away, and you've got an amp tuned for your particular source's noise. Or even stick a pot in and let the user do it themselves - is that something that some commercial amps provide? (Or is that a dumb idea for some reason?)awpagan said:
SC ULD2
Jeremy
If the amplifier is performing correctly,and has as low a distortion figure as claimed, then what you are hearing should be pretty close to the sound of what is going into the amplifier,only amplified.
This means that you would have to have a savage input filter, and that means that your treble will be very dull sounding indeed.
Alex
Jeremy
If the amplifier is performing correctly,and has as low a distortion figure as claimed, then what you are hearing should be pretty close to the sound of what is going into the amplifier,only amplified.
This means that you would have to have a savage input filter, and that means that your treble will be very dull sounding indeed.
Alex
Re: SC ULD2
Yes understood. I think it'll be an interesting listening exercise, at least. If I were using the SB Duet without an external DAC, the hiss would be driving me crazy - and in that case I think I might prefer dull treble to horrible noise. In fact, I might try it and see what the SQ is like at the point where I filter out the Duet DAC's noise (just as an experiment - there's no way I'd choose to listen to the Duet's DAC directly now that I have better options!)sandyK said:
SC ULD2
jeremy
You could also try comparing the input and output waveforms with your CRO. I am pretty sure that Hugh wouldn't mind giving you a crash course on how to use one.
Alex
jeremy
You could also try comparing the input and output waveforms with your CRO. I am pretty sure that Hugh wouldn't mind giving you a crash course on how to use one.
Alex
jp_howard said:
I hope this is not the case and the ground wire goes to the (audio) gnd input terminal which should sit 10R above power ground.
You may wish to consider the other end of the LTP which is connected to the speaker output should your speaker cables not be shielded and they run in close proximity to you source of the terrible noise, moving these may help.
My suspicion is the source is one of those computer person designed things that does not take into account audio, and may be poorly designed inside, poor grounding, interconnect cables not shielded, running output in close proximity to the clock, dry joint, dry filter caps and the list goes on.
I agree with SandyK, the amp works as is, box and star grounds may tidy it up.
The logic is if there is no hiss connected to the Headphone Amp, then the amplification factor is revealing something. It did for me, revealed a ground loop in my DAC, but humm is different to hiss, so I think the input filter should be held over and you get closer to the source.
Can you pick up the source and pass it by the amp?
Can you put a shield over the amp, plastic basin and some cooking foil, windsceen shield, ground this to your power ground.
Some vacant thoughts, from the peanut gallery
Alfred
Sorry - that's exactly right. I didn't realise that the audio gnd input terminal wasn't 0V, and wasn't sure of the right jargon to describe it. But yes, that's where it's connected.alfredrofe said:
I've made the speaker cables short, and moved them so they are not close to anything else, but that didn't help.
Yes the amp is working fine - since the Transporter connected directly has no noise at all. Hopefully a visit to Hugh's today will cast more light on the noise coming in to the input.
jp_howard said:
I doubt that the "Hiss"is coming from the Amplifier as i have found that the SCULD is one of the quietest amps that i have listened to , at least mine is , much quieter than some of the commercial amps that i have heard. At no signal conditions i cannot hear any "Hiss or Noise" with my Ear placed right against the Tweeter, this is running from my Shanling CD player thru a Rotel Pre Amp.
The fact that i run all my system on "Balanced Power" may help this or course.
The Amp should be in a Metal Case with correct 'Star Grounding, another thing that
may help is a Good Quality Filtered IEC connector on the mains AC Power Input.
MickyH
Micky, the hiss is definitely not coming from the amp. It is coming from the DAC, and I'm noticing it for the 1st time because the amp has very high gain.
After listening to the hiss today, Hugh told me that it was a normal level of noise for a DAC. His technical suggestion to resolve the issue was: "sit a bit further away"! 😉 His Onkyo CD player and my Logitech Transporter have no audible hiss at all - perhaps it's just unreasonable for me to expect my $200 DAC to have the same characteristics as a $2000 source... Still, if anyone has any ideas on reducing noise from a DAC which might apply, I'm all ears! 🙂
BTW, I've decided to build a few different amps in the next few weeks to compare and contrast them. I've ordered a basic chipamp kit, and I'm thinking of also ordering a 'Truepath' from 41Hz. I don't think I want anything with valves - I can always later use a valve pre-amp if I want the 'warm' option, I guess.
After listening to the hiss today, Hugh told me that it was a normal level of noise for a DAC. His technical suggestion to resolve the issue was: "sit a bit further away"! 😉 His Onkyo CD player and my Logitech Transporter have no audible hiss at all - perhaps it's just unreasonable for me to expect my $200 DAC to have the same characteristics as a $2000 source... Still, if anyone has any ideas on reducing noise from a DAC which might apply, I'm all ears! 🙂
BTW, I've decided to build a few different amps in the next few weeks to compare and contrast them. I've ordered a basic chipamp kit, and I'm thinking of also ordering a 'Truepath' from 41Hz. I don't think I want anything with valves - I can always later use a valve pre-amp if I want the 'warm' option, I guess.
Update: in the last 2 days I've done the current mirror stuff (including trimpot adjustments) and cap upgrades, with ~3 hours listening tests after each. I'm really enjoying the sound, but I find it very hard to say definitively whether there's a clear SQ improvement. It's very difficult to remove exogenous variables - particularly perception bias (e.g. after spending a day battling problems and finally completing an upgrade, I'm feeling very positive and therefore I'm very receptive to the listening experience).
Another issue is that the quality of my source hasn't been that great, and I'm not sure how good my speakers are. Regarding the source, I've decided to move my Logitech Transporter into the sitting room where the amp is (previously I was using it with my headphone amp) - this introduces yet another variable, but I think it's a positive step. Regarding the speakers, a kind person on this forum has agreed to let me listen to some of their nice speakers on Sunday (including some Plutos, which I'll be very interested to hear) - I'll bring around mine as well and compare. If my speakers don't perform, I'll look at upgrading.
Another issue is that the quality of my source hasn't been that great, and I'm not sure how good my speakers are. Regarding the source, I've decided to move my Logitech Transporter into the sitting room where the amp is (previously I was using it with my headphone amp) - this introduces yet another variable, but I think it's a positive step. Regarding the speakers, a kind person on this forum has agreed to let me listen to some of their nice speakers on Sunday (including some Plutos, which I'll be very interested to hear) - I'll bring around mine as well and compare. If my speakers don't perform, I'll look at upgrading.
- Status
- Not open for further replies.