I note from an image in post 3 that the input RCA connectors are very close to the some output cables. The a.c. currents in these outputs due to magnetic radiation have potential to induce currents in the input wires. These may be shielded to prevent this but the shield needs to be earthed at one end
Normal practice is to specify separate equal secondary windings and from colour coding on the insulation tube leads to work out which winding mates with the correct partner so a centre tap can be formed with correct phase for the secondaries. The code information should be on a the sticker on the side of the toroid.
I hope our member is are not using the switch mode supply arrangement for his replacement transformer power supply.
Normal practice is to specify separate equal secondary windings and from colour coding on the insulation tube leads to work out which winding mates with the correct partner so a centre tap can be formed with correct phase for the secondaries. The code information should be on a the sticker on the side of the toroid.
Consumer Unit I used one on before, this was 2008, I moved from here in 2013. I did A-B as I had the existing ring main in place
each line was fused individually
Mjona, hi
No swicthmode for any supplies in the whole set up, apart form a PC that is off when any DSP adjustments have been made and the projector that is off when any measurements taken and is connected to the video processor via a fire cable and powered by an individual spur.
When the transformer is powered remotely the and connected with loose leads to allow some trial dressing no noticeable difference unfortunately
Just to riterate the noise I want to reduce now is predominately 'hiss' or equivalent, very little mains low frequency noise now
Andrew, thanks
In a perfect world I would have used one CU (consumer unit) with a direct or very low impedance supply to the CU then matched low impedance spurs each fused at the CU. Ive done this before a number of times and it was great. Alas where I live now I am unable to do that. The set up I have is a direct fed CU on the second floor with a separate ring for the AV set up, all plugs on the back wall. My only earth to signal GND is at the BR player and all GND connections from units have been improved to reduce any voltage differnces the existing sub unit was quiet (BPSP 500) the BK1K is plugged into the same connections in the same location as I want to replace the 500's feeding my 18" subs.
Yes interesting idea, Ive always fancied trying these items https://www.hificollective.co.uk/catalog/mlhc100030-47000uf-100v-mlytic-electrolytic-p-3981.html. ! But I wood be going down another rabbit hole !
The subs are these
https://bmsspeakers.com/index.php-11.html?id=bms_18n862 in 4Ohm not 8 Ohm
Yes I am ignoring the blue area as this is clearly misleading the green seems to correlate too what I hear. The black is the totally silent BPSP 500 the brown the evolved BKL1000, this is still audible at the seating position.
I did try to complete a loop back test but this wasn't possible, the speaker is useful to 1K then falls away anyway, hence looking at the green are
This is the SPL at the seating position with one of the 18" subs on a test sweep keeping distortion below 5%
The green area is dominated by harmonics of 50Hz.
You can't see the gaps between the harmonic spikes due to the width of the filter.
If that was an RF spectrum analyser plot, I would be reaching to reduce the resolution bandwidth.
There's a lump at ~25Hz. Then there's intermod-esque sidebands on all the harmonics at ~25Hz delta.
The frequency of that 20-something Hz lump changes between the black and brown traces.
It might be prudent to understand that. If you repeat the scan, does it move?
Consumer unit/ Switch Board/ Fuse Box/ Breaker board.
Ta.
Hiss. Hmmm. Where? All equipment has background noise characteristics.
Still, with an extensive system, the crap that now comes through our mains electric plays a larger role than it used to,
even a decade ago it wasn't that bad.
I find the 3am test the easiest one.
If it sounds as good in the wee small hours as it does through the day, you have well designed power supplies.
if not, yeah maybe it's time to look at filtration of the incoming mains.
I recently spent three years on and off looking into this nightmare, building on three decades of observations from experience.
Generally, there are around twelve or thirteen types of mains interference/ problems.
Every locale will have its characteristics, as does the equipment connected.
This is too large a topic at this time for me to 'flesh out' a coherent overview. and frankly not something I feel comfortable sharing over a public forum.
Here's the X-men of power filters: Furman P-2300 IT E
The noise floor cancellation design is as far as i am aware, second to none on the market presently, probably best deployed on all source and pre.
That will get rid of the worst, depending on what your particular contextual 'worst' is!
It will give you that 'black background' though.
or Sinalda systems, pretty good general protection and filtration, better for larger current draw audio such as amps.
Localised (within house) low pass filtration of SMPS is the way to go. Kill it at source.
But of course, the triplins harmonics thru the mains is on the increase.
Try finding 100A Low pass filters, basically.
Ta.
Hiss. Hmmm. Where? All equipment has background noise characteristics.
Still, with an extensive system, the crap that now comes through our mains electric plays a larger role than it used to,
even a decade ago it wasn't that bad.
I find the 3am test the easiest one.
If it sounds as good in the wee small hours as it does through the day, you have well designed power supplies.
if not, yeah maybe it's time to look at filtration of the incoming mains.
I recently spent three years on and off looking into this nightmare, building on three decades of observations from experience.
Generally, there are around twelve or thirteen types of mains interference/ problems.
Every locale will have its characteristics, as does the equipment connected.
This is too large a topic at this time for me to 'flesh out' a coherent overview. and frankly not something I feel comfortable sharing over a public forum.
Here's the X-men of power filters: Furman P-2300 IT E
The noise floor cancellation design is as far as i am aware, second to none on the market presently, probably best deployed on all source and pre.
That will get rid of the worst, depending on what your particular contextual 'worst' is!
It will give you that 'black background' though.
or Sinalda systems, pretty good general protection and filtration, better for larger current draw audio such as amps.
Localised (within house) low pass filtration of SMPS is the way to go. Kill it at source.
But of course, the triplins harmonics thru the mains is on the increase.
Try finding 100A Low pass filters, basically.
Last edited:
Earth, hi
I have experimented a lot with mains filtration over the the years, but alas I've always found there is a trade with dynamics and lowering the noise floor when in series. I borrowed a huge Isotek when first they hit the market and have tried a number since. I have never got on with series filters that add impedance but I have got on with parallel filters focussing on differential noise.
I have a few of these near preampfification and source unit supplies. I ended up making them
I'm going to modify the GND on the other board to 'match' the better board with the mods and see if I can measure the power rails with my new Scope
I have experimented a lot with mains filtration over the the years, but alas I've always found there is a trade with dynamics and lowering the noise floor when in series. I borrowed a huge Isotek when first they hit the market and have tried a number since. I have never got on with series filters that add impedance but I have got on with parallel filters focussing on differential noise.
I have a few of these near preampfification and source unit supplies. I ended up making them
I'm going to modify the GND on the other board to 'match' the better board with the mods and see if I can measure the power rails with my new Scope
Earths,hi
"Where? Input or output of the LMxx? it's easy enough to download the PDF and start with the manufacturers recommendations. In my experience, increasing low ESR caps values on the outputs decreases general noise.
Going beyond the manufacturers recommendations on the input doesn't make much difference, but using a cap there will help keep the output as clean as possible."
The 0.47uF was close to the recommended 0.33uF so I added this, are you suggesting something like a low ESR (Panasonic FR) 100uf cap added as well, I could add more or less ?
+
"Where? Input or output of the LMxx? it's easy enough to download the PDF and start with the manufacturers recommendations. In my experience, increasing low ESR caps values on the outputs decreases general noise.
Going beyond the manufacturers recommendations on the input doesn't make much difference, but using a cap there will help keep the output as clean as possible."
The 0.47uF was close to the recommended 0.33uF so I added this, are you suggesting something like a low ESR (Panasonic FR) 100uf cap added as well, I could add more or less ?
+
Attachments
Wow, I see your prototyping for the Mars Rover!
Tell me they're not polypropylene!
NiZn ferrites cores with the 240AC wires tight wrapped 3 1/2 times around the core
and a couple of 10nF X rated to E
Tell me they're not polypropylene!
Yes, differential mode interference was the one that I found caused most problems, particularly for the reliability of digital devices.
NiZn ferrites cores with the 240AC wires tight wrapped 3 1/2 times around the core
and a couple of 10nF X rated to E
Last edited:
Yes. Maybe go polymer on the 100µF.
Have you asked your self what function that Zener performs? they are noisy.
7815 noisy too, plenty low noise regs out there.
I have a couple of small toroidal Transformers so will feed via W04M rectifier and 2,200uF cap circa 20V to replace the 24V supply to the L7812CV
I already have a small 12V equivalent for the 12V fan that works fine on a temperature switch I assume this will be the best way to wire the transformer ?
It's a 0-115-0-115 and a 2 x15V transformer
I already have a small 12V equivalent for the 12V fan that works fine on a temperature switch I assume this will be the best way to wire the transformer ?
It's a 0-115-0-115 and a 2 x15V transformer
I will check how it operates today, time allowing, I have added a 15VA toroid with a rectifier and 2,200uF cap. I have fed this in to the board at the location of the 24V supply to the anti thump circuit. I can measure voltages. I think I will be near 20V under load to the 7812. I hope any noise will not be hitting the circuit now as no long to the main board I removed the GND link
The 12V fan is on a temperature control so the small 3VA supply is only powered when required. (PS I will ensure everything is electrically insulated when finished)
I could do some measurements to use a resistor instead of the diode to the 7812, but my little supply is unregulated
The 12V fan is on a temperature control so the small 3VA supply is only powered when required. (PS I will ensure everything is electrically insulated when finished)
I could do some measurements to use a resistor instead of the diode to the 7812, but my little supply is unregulated
Well all works as expected, well nearly - it's made the whole problem worse from a feedback and noise perspective. I wired in the feed to the anti thump circuit undependently so the circuit was only fed from the power supply and not linked to the amp circuit. Much noisier, so I experimented and it was better with a connection to signal GND
Overall looks like a 'noisy' 7w voltage dropper is currently better
Black the 'target' the existing 500W amplifier. Silent in operation, (I am unsure about the 50Hz peak as not audible)
light green BK1000W amp direct from DSP with the 'new' power supply to the 12V anti thump circuit
Blue BK1000W and direct feed from DSP with new power supply but also a link from the new anti thump circuit supply GND and signal GND
Dark Green BK1000W with existing voltage drop to supply the anti thump
All the measurements with REW measuring no signal from power amp output
Overall looks like a 'noisy' 7w voltage dropper is currently better
Black the 'target' the existing 500W amplifier. Silent in operation, (I am unsure about the 50Hz peak as not audible)
light green BK1000W amp direct from DSP with the 'new' power supply to the 12V anti thump circuit
Blue BK1000W and direct feed from DSP with new power supply but also a link from the new anti thump circuit supply GND and signal GND
Dark Green BK1000W with existing voltage drop to supply the anti thump
All the measurements with REW measuring no signal from power amp output