I have the NTC between PSU Gnd and chassis. This I have on all my builds and all chassis in my system is connected via the Y/G wires.
I have never experienced hum like this. It is like if you mix screen and hot in a signal cable.
I have never experienced hum like this. It is like if you mix screen and hot in a signal cable.
well, fiddle with everything you can think of
what I wrote about take-out point from PSU GND is always valid - caps, even final ones, are juicing current from "upstream"
where there is current, there is pumping, whatever
and takeout point for channel pcbs must be cleanest possible ......
sadly, I'm not here with you to be able to help instantly, but from own experience - something which looks as one problem sometimes is sum of small things
for starters, just short GBR and go from there
what I wrote about take-out point from PSU GND is always valid - caps, even final ones, are juicing current from "upstream"
where there is current, there is pumping, whatever
and takeout point for channel pcbs must be cleanest possible ......
sadly, I'm not here with you to be able to help instantly, but from own experience - something which looks as one problem sometimes is sum of small things
for starters, just short GBR and go from there
Why do some recommend to use "cap-ground" as speaker return?
I remember for M2X it was a recommendation to avoid hum. I did not and used the terminals at PCB without any hum.
It is first time I use speaker return directly at cap. I did this because then wiring was a little bit simpler.
I remember for M2X it was a recommendation to avoid hum. I did not and used the terminals at PCB without any hum.
It is first time I use speaker return directly at cap. I did this because then wiring was a little bit simpler.
take note that NFB path in this amp is severely current heavy, resistances involved are entire or even two ranges lower than in "normal" amps
thus is important that each path to GND is shortest and cleanest and isolated from any other
try doing those few things I wrote few posts back, and then inform
thus is important that each path to GND is shortest and cleanest and isolated from any other
try doing those few things I wrote few posts back, and then inform
Freedom of speech.
Someone on diyAudio wrote that returning the speaker ground to the audio board was wrong, and said that returning the speaker ground to the power supply board was the only correct way. I did not agree with him and suggested that returning the speaker ground to the audio board and twisting the speaker wires together was good for low noise.
He replied, "Nonsense... you are just trying to stubbornly prove something here but lack understanding of the current flows in this case.". And, "All amplifiers in the world do not adhere to what you are suggesting here and I suppose it is a real shame that this culture of incorrect speaker return wiring is ripe in these forums".
Amusing. 🤓
It is not a constructive way to communicate.
In my case I need to do some drastic so my assumption is that just moving the return wire from cap Gnd to PCB is not a drastic change. Maybe I will learn that it is. I feel the most drastic change I can do related to input signal is the GBR resistor. So I will try that first. Also because it should be obsolete for mono block build. In my mind it should enhance signal reference stability to reference directly to PSU Gnd for the input signal. We will see. I have placed the solder iron at the table so it is ready for tomorrow. It was put back on the shelf after I assumed the power amps was ready 🙂
Never store your solder iron until you are 100% sure everything works!
In my case I need to do some drastic so my assumption is that just moving the return wire from cap Gnd to PCB is not a drastic change. Maybe I will learn that it is. I feel the most drastic change I can do related to input signal is the GBR resistor. So I will try that first. Also because it should be obsolete for mono block build. In my mind it should enhance signal reference stability to reference directly to PSU Gnd for the input signal. We will see. I have placed the solder iron at the table so it is ready for tomorrow. It was put back on the shelf after I assumed the power amps was ready 🙂
Never store your solder iron until you are 100% sure everything works!
I am sorry but it was not meant as criticism of your efforts. It was to point out that there are many opinions on the subject, and forums are open to all to express whatever they believe in.
Here is good information by diyAudio member Bonsai on amplifier noise and grounding. It includes information on GBR (he called it HBR), and also trouble shooting:
https://hifisonix.com/wp-content/uploads/2019/02/Ground-Loops.pdf
https://hifisonix.com/wp-content/uploads/2019/02/Ground-Loops.pdf
yeah
note Bonsai's note about proper HBR positioning; by that my GBR is in wrong position
but, as I know Bonsai's document from before and probably was taught by same mistakes as he was
, I was thinking along same lines, but realized it isn't possible to have it in Babelfish F8, exactly because of impedance range of NFB resistors
thus I resorted to old proven arrangement, Hafler DH200, which had 47R (if I remember correctly) soldered ditto on one input RCA shell and GND lead soldered to
both in factory made and in kit form, shown in build guide
so GBR here - safe bet is - start with 0R, then - if problem arose, change one, and if needed change both to - say - 15R
or the other way - start with 15R, then , then
note Bonsai's note about proper HBR positioning; by that my GBR is in wrong position
but, as I know Bonsai's document from before and probably was taught by same mistakes as he was
, I was thinking along same lines, but realized it isn't possible to have it in Babelfish F8, exactly because of impedance range of NFB resistorsthus I resorted to old proven arrangement, Hafler DH200, which had 47R (if I remember correctly) soldered ditto on one input RCA shell and GND lead soldered to
both in factory made and in kit form, shown in build guide
so GBR here - safe bet is - start with 0R, then - if problem arose, change one, and if needed change both to - say - 15R
or the other way - start with 15R, then , then
Interesting......I was about to ask same question about placement of GBR:
I will try tomorrow to first set it to zero and see (hear) what happens.......
I will try tomorrow to first set it to zero and see (hear) what happens.......
as feedback net is comprised 330R divided with (33+33)R, putting anything in between said (33+33)R and GND, would trash things significantly
thus only possible way to implement some optional remedy was to put it as I did, in old "Hafler" way of things
as illustration, say that amp in case is "normal" amp, feedback net would likely have at least 33K/(3K3+3K3), thus 15R in addition leading to GND would mean nothing for gain per se
thus only possible way to implement some optional remedy was to put it as I did, in old "Hafler" way of things
as illustration, say that amp in case is "normal" amp, feedback net would likely have at least 33K/(3K3+3K3), thus 15R in addition leading to GND would mean nothing for gain per se
Thanks again to Zen Mod, my F8 build is done - but done a bit differently to others. I used 2 off Meanwell LRS150F-24 SMPS, one for pos rail and one for neg rail. Outputs fed into Mark Johnson's Amy Alice filter boards (2 were modified for neg voltage as per his thread). So pos/neg filter on each rail on each channel - total of 4 filter pcb's. They then fed dual cap Mx boards, one for each channel with added caps and adjusted to+/-23.00VDC output each. My heatsinks are slightly smaller (400mm x 100mm x 50mm than a 3U chassis), so I toned down the bias on each channel to 1.3Amps - that gave a heatsink temp after 4 hours of 50 degC.
DC offset on each channel adjusted to hover at only 1-2mVDC.
I can honestly say I have never heard a sweeter and more defined sound out of my test bed Yamaha NS-10M speakers driven by ZM's marvel.
My build was with the Semisouths in the neg rail.
Thanks again to Zen Mod for this great design - others should definitely give it a go. Even without the SS's, I think it will be a revelation.
DC offset on each channel adjusted to hover at only 1-2mVDC.
I can honestly say I have never heard a sweeter and more defined sound out of my test bed Yamaha NS-10M speakers driven by ZM's marvel.
My build was with the Semisouths in the neg rail.
Thanks again to Zen Mod for this great design - others should definitely give it a go. Even without the SS's, I think it will be a revelation.
