the "splices" in the Return should be done to achieve least possible loop area between Flow and Return.
The wire thickness does not matter. What matters is equal impedance in the two routes of the split Return. Equal impedance and diametrically opposite sets the Physical Centroid in the space that coincides with the middle of the capacitor.
I don't claim this to work.
It's just an idea following from star quad and low loop area around/through very large axial leaded capacitors.
I have not thought through whether this can be applied to radially leaded capacitors.
It would be nice to hear reports of some tests. By measurement, not listening to hum/buzz at the speaker which is generally much more tolerant of noise than a DVM reading to 0.1mVac. Although sensitive higher impedance headphones are a very good test if they can reproduce 50Hz hum. My Red Devils don't but my Pro4aa and hd465 do let me hear the 50Hz hum and the buzzes that generally come out of Power Amplifiers that are not perfect.
BTW, I have never built a multichannel perfect power amplifier. All are noisier than a monoblock that is built properly.
The wire thickness does not matter. What matters is equal impedance in the two routes of the split Return. Equal impedance and diametrically opposite sets the Physical Centroid in the space that coincides with the middle of the capacitor.
I don't claim this to work.
It's just an idea following from star quad and low loop area around/through very large axial leaded capacitors.
I have not thought through whether this can be applied to radially leaded capacitors.
It would be nice to hear reports of some tests. By measurement, not listening to hum/buzz at the speaker which is generally much more tolerant of noise than a DVM reading to 0.1mVac. Although sensitive higher impedance headphones are a very good test if they can reproduce 50Hz hum. My Red Devils don't but my Pro4aa and hd465 do let me hear the 50Hz hum and the buzzes that generally come out of Power Amplifiers that are not perfect.
BTW, I have never built a multichannel perfect power amplifier. All are noisier than a monoblock that is built properly.
Last edited:
There are Threads discussing in detail separate PSU.
There are many downsides to a separate PSU. Some of these are Safety related.
What is "it" in
Andrew I will try this and report any change. I still have a slight audible hiss with my ear next to the tweeter so if this will help great, its worth a try. I do also have hum without the pgnd connection but I am still wanting to solve that too, so some internal work is in order. All of my changes will be observed individually obviously.
A question for those using fastens for signal connections. Have all of you soldered the conecrions all the way across the bottom? I soldered mine only on both of the pins through the pcb not the center
I am trying to figure out where I have strayed from the tutorial on both pcbs equally.
Andrew, is it possible to measure the differences in hiss with a DMM set to mV? Taking a measurement before and after the wiring change? Or do I need a scope and signal?
As it sits though this amp is still outstanding, I just want to make it perfect!
A question for those using fastens for signal connections. Have all of you soldered the conecrions all the way across the bottom? I soldered mine only on both of the pins through the pcb not the center
I am trying to figure out where I have strayed from the tutorial on both pcbs equally.Andrew, is it possible to measure the differences in hiss with a DMM set to mV? Taking a measurement before and after the wiring change? Or do I need a scope and signal?
As it sits though this amp is still outstanding, I just want to make it perfect!
A scope set to 2mV/div is not as sensitive as a 0.1mVac reading on a DVM.
But they are very different instruments.
The scope can be set to look at frequencies from 50Hz upto about twice it's stated bandwidth.
A general purpose DVM will probably not get far above 50kHz and the response will be far from linear over the range 50Hz to 20kHz.
A good 1MHz rms reading DVM (probably a bench top rather than a handheld) will stop around 1 or 2Mhz.
Have a look at the test set in the equipment Threads.
Back to the DVM.
I can easily re-arrange grounding such that output noise including hum is anywhere between 0.0mVac (i.e. <50uVac) to 20mVac.
I never settle for more than 0.3mVac hum and noise at the output of a Power amp.
That is very easily measured with a 199.9mVac scale setting, but is virtually unmeasurable on a 2mV/div scope. 0.3mVac will be audible on wideband sensitive headphones.
Is the hiss from the power amp? Is the hiss the same when the input is shorted, or fitted with a 1k dummy load?
But they are very different instruments.
The scope can be set to look at frequencies from 50Hz upto about twice it's stated bandwidth.
A general purpose DVM will probably not get far above 50kHz and the response will be far from linear over the range 50Hz to 20kHz.
A good 1MHz rms reading DVM (probably a bench top rather than a handheld) will stop around 1 or 2Mhz.
Have a look at the test set in the equipment Threads.
Back to the DVM.
I can easily re-arrange grounding such that output noise including hum is anywhere between 0.0mVac (i.e. <50uVac) to 20mVac.
I never settle for more than 0.3mVac hum and noise at the output of a Power amp.
That is very easily measured with a 199.9mVac scale setting, but is virtually unmeasurable on a 2mV/div scope. 0.3mVac will be audible on wideband sensitive headphones.
Is the hiss from the power amp? Is the hiss the same when the input is shorted, or fitted with a 1k dummy load?
I will look into the instruments more in depth. However to short input I want to be very clear that I am doing it right. Is it as simple as connecting the signal "hot" to the ground at the input? And as for a dummy load, is this as simple as connecting a resistor across the input hot and ground?
Sorry for the beginer questions, I just don't want to burn anything up being hasty
Sorry for the beginer questions, I just don't want to burn anything up being hasty
You have it right.
I keep a stock of dummy RCAs with different resistor values between hot and cold. From 0r0 to 100k.
With pre on and source cd player on there is slight hiss with ear next to speaker. With inputs shorted on amp there is almost no hiss. I did not have the time to set up rcas with the dummy load resistors last night, maybe tonite.
I have used the cd player before with no problem, the pre though is an integrated amp with the amp section disabled. I have used it as an integrated with no problem, but this is the first time I have used it as a pre amp.
I have used the cd player before with no problem, the pre though is an integrated amp with the amp section disabled. I have used it as an integrated with no problem, but this is the first time I have used it as a pre amp.
Hiss from a CD input may indicate you have too much system gain.
Try Pano's test signals to measure the actual output from the CD player and the actual output from the amplifier when the vol pot is set to maximum volume.
The -9dB signal should still be 9dB below clipping at the speaker output terminals when the attenuation is zero.
Have you read the article "System Gain"?
Try Pano's test signals to measure the actual output from the CD player and the actual output from the amplifier when the vol pot is set to maximum volume.
The -9dB signal should still be 9dB below clipping at the speaker output terminals when the attenuation is zero.
Have you read the article "System Gain"?
With just cd player connected the hiss is almost totally inaudible with an ear on the tweeter. As good or better than with the inputs shorted.
I had a thought about the pgnd connection too. With it disconnected and just the cd player connected there is no hiss and just the smallest amount of hum only audible with an ear very close to the speaker, much better, but not quite perfect.
A different pre is definatly in order.
I had a thought about the pgnd connection too. With it disconnected and just the cd player connected there is no hiss and just the smallest amount of hum only audible with an ear very close to the speaker, much better, but not quite perfect.
A different pre is definatly in order.
Will,
Andrew's thought about gain and your test make sense. The MyRef has a gain of 30 which means it doesn't need amplification from the pre. It works just fine with a potentiometer volume control and no amplification in the pre. Normal commercially made amps typically have lower amp gain (10 - 20) and have additional gain in the pre.
Now for the more exotic, Peter Daniels posted relatively recently that he is using a passive pre that is nothing more than a source selection switch and a transformer volume control selected by a discrete multi-pole switch. Not sure how it would work, but it sure is interesting.
SOWTER ATTENUATOR TRANSFORMERS TVC VOLUME CONTROL
Jac
Andrew's thought about gain and your test make sense. The MyRef has a gain of 30 which means it doesn't need amplification from the pre. It works just fine with a potentiometer volume control and no amplification in the pre. Normal commercially made amps typically have lower amp gain (10 - 20) and have additional gain in the pre.
Now for the more exotic, Peter Daniels posted relatively recently that he is using a passive pre that is nothing more than a source selection switch and a transformer volume control selected by a discrete multi-pole switch. Not sure how it would work, but it sure is interesting.
SOWTER ATTENUATOR TRANSFORMERS TVC VOLUME CONTROL
Jac
Jac, thanks for the link, I will look for Peters thread. Its funny I have a lighter note that is 90% complete and a 10k TKD potentiometer that I ordered to make a simple passive pre, but of course I put them both on the back burner and just used the integrated as a pre so I could do some listening. I may try to whip up a quick pre this weekend with the TKD to continue listening.
Dario, this is good news, I was hoping the hum would go away with the integrated pre also, not totally but close. I feel like it is somthing simple, I will keep at it. Though soon I may start making some of the layout changes you suggested. Thanks gents
Dario, this is good news, I was hoping the hum would go away with the integrated pre also, not totally but close. I feel like it is somthing simple, I will keep at it. Though soon I may start making some of the layout changes you suggested. Thanks gents