Thank you for the advice. I pondered removing windings from the toroidal, but I thought I'd either completely ruin it or make it uglier and somehow degraded. Best kept for another project, or sold.
I eventually purchased that dual 25VAC, 225VA toroidal I was considering, with some additional parts in my order. I expect something like +-36 to +-37 VDC post rectification. So I will still be pushing it a bit, but hopefully not too much. I'll let you guys know when stuff arrives.
Yes. Have a look at the link I included in Post #5.
Once the supply voltage exceeds ±32 V the internal protection in the LM3886 will start to kick in on the signal peaks with a 4 Ω load. So the max output power plunges drastically with a 4 Ω load once you exceed that voltage. That's why I recommend ±30 V for 4 Ω load.
Tom
Once the supply voltage exceeds ±32 V the internal protection in the LM3886 will start to kick in on the signal peaks with a 4 Ω load. So the max output power plunges drastically with a 4 Ω load once you exceed that voltage. That's why I recommend ±30 V for 4 Ω load.
Tom
Hi there, I just ran across this AD doc on shielding last night, I am sure it will be of use for DIYer's as this topic comes up a lot.
https://www.analog.com/media/en/technical-documentation/application-notes/41727248an_347.pdf
Cheers !! 🙂
jer🙂
https://www.analog.com/media/en/technical-documentation/application-notes/41727248an_347.pdf
Cheers !! 🙂
jer🙂
22-0-22VAC gives us approximately +/-34VDC unloaded when line voltage is high(variates) & i found that lm3886tf (insulated version) which is very popular among diy circle is perfectly ok with that, even with 4ohm load (power maybe reduced due to SPIKE but i played loud enough). I mean it's a robust chip, hard to kill !!
New 225VA / 2x25VAC toroidal is here:
It's a bit heftier than the 160VA / 2x30VAC, but I think I won't have trouble fitting it in my case.
Yesterday I joined the primary coils and connected it to mains (measured 237VAC) and with the secondary coils floating, each measured around 26.4VAC.
Today I connected it to the power amp board and measured voltages at rectification input and filter capacitor output:
Oddly enough, both sides measured almost exactly 27VAC, very balanced with each other. I presume this has something to do with the rectification circuit that follows. I found the slight increase from 26.4VAC a bit odd, I'm attributing it to either mains fluctuation, some temperature change, or its just an expected phenomenon at play I do not understand. And now for post rectification, which is the important part of all this:
One of the sides is always a bit higher than the other no matter how I feed this board. I don't think this is a major issue.
Voltage sits around 36VDC which if I understand correctly is on the high side, but not in really unsafe territory any more.
All these measurements are without any input or speakers connected. I have not yet tried to measure the amp while working
at normal or high listening levels to see if there's any voltage drop. It should be an interesting observation to make.
Meanwhile more parts have arrived for the amp so I should be able to continue this weekend and see how this thing goes.
It's a bit heftier than the 160VA / 2x30VAC, but I think I won't have trouble fitting it in my case.
Yesterday I joined the primary coils and connected it to mains (measured 237VAC) and with the secondary coils floating, each measured around 26.4VAC.
Today I connected it to the power amp board and measured voltages at rectification input and filter capacitor output:
Oddly enough, both sides measured almost exactly 27VAC, very balanced with each other. I presume this has something to do with the rectification circuit that follows. I found the slight increase from 26.4VAC a bit odd, I'm attributing it to either mains fluctuation, some temperature change, or its just an expected phenomenon at play I do not understand. And now for post rectification, which is the important part of all this:
One of the sides is always a bit higher than the other no matter how I feed this board. I don't think this is a major issue.
Voltage sits around 36VDC which if I understand correctly is on the high side, but not in really unsafe territory any more.
All these measurements are without any input or speakers connected. I have not yet tried to measure the amp while working
at normal or high listening levels to see if there's any voltage drop. It should be an interesting observation to make.
Meanwhile more parts have arrived for the amp so I should be able to continue this weekend and see how this thing goes.
Quick question: I have 4 leads from the two secondary coils. I see people twisting couples of cables to neutralize produced magnetic fields. In my first attempt with the 30V toroidal, what I did was twist one center tap lead with one "outer" lead. Is that a good idea? I wonder if twisting the two outer leads would be somehow superior in cancelling magnetic fields. Due to the absolute symmetry of twisting, I don't think so, but I thought to ask, mostly out of curiosity.
Since it is very relevant to noise issues, allow me to introduce you to my enclosure:
As the specs indicate, inner useful space is 21cm (W) x 28cm (D) x 8cm (H).
Oddly enough, the inner side of the side panels looked much better to me, much "cleaner". So I decided to install them inner side out:
I think the inner side looks much better from the outside. However, this forced me to install the internal base plate 2cm higher. Therefore, I have a "ground floor" with a height of 6cm that luckily can contain all parts, and a "basement" of 2cm under it. Note that the "basement" might be useful for routing cables, for the sake of tidiness and perhaps also some negligible noise improvements, I am not sure.
This is my back plate, the 3d model I sent to the case manufacturer and the actual thing with connectors installed:
For now, connectors of interest are power socket, speaker connectors and Right/Left RCA input.
So now that I have both a transformer and a power amp board, the plan to install them looks something like this:
I chose this placement in order to minimize cable distances from RCA input to board and also from speaker output to back panel speaker connectors. The only "trick" I am doing here is taking the two twisted pairs of speaker output cables and routing them under the base plate, in a circular motion towards the speaker connectors. I'm not sure if taking them to the connectors in a straight line would be better. With this arrangement they are covered under the base plate, but they also pass directly under the power supply.
Does this arrangement look sensible to you? Any ideas/recommendations would be welcome, especially concerning cable routing and perhaps utilizing the space under the base plate even more.
As the specs indicate, inner useful space is 21cm (W) x 28cm (D) x 8cm (H).
Oddly enough, the inner side of the side panels looked much better to me, much "cleaner". So I decided to install them inner side out:
I think the inner side looks much better from the outside. However, this forced me to install the internal base plate 2cm higher. Therefore, I have a "ground floor" with a height of 6cm that luckily can contain all parts, and a "basement" of 2cm under it. Note that the "basement" might be useful for routing cables, for the sake of tidiness and perhaps also some negligible noise improvements, I am not sure.
This is my back plate, the 3d model I sent to the case manufacturer and the actual thing with connectors installed:
For now, connectors of interest are power socket, speaker connectors and Right/Left RCA input.
So now that I have both a transformer and a power amp board, the plan to install them looks something like this:
I chose this placement in order to minimize cable distances from RCA input to board and also from speaker output to back panel speaker connectors. The only "trick" I am doing here is taking the two twisted pairs of speaker output cables and routing them under the base plate, in a circular motion towards the speaker connectors. I'm not sure if taking them to the connectors in a straight line would be better. With this arrangement they are covered under the base plate, but they also pass directly under the power supply.
Does this arrangement look sensible to you? Any ideas/recommendations would be welcome, especially concerning cable routing and perhaps utilizing the space under the base plate even more.
H
HAYK
I advise you to add this EMI noise suppressor, only $1. It is sold cheap because the writing input and output needs to be swept.
https://www.aliexpress.com/item/100...!sea!TH!4284450371&curPageLogUid=iivPUyxxMz0q
https://www.aliexpress.com/item/100...!sea!TH!4284450371&curPageLogUid=iivPUyxxMz0q
Thanks for your reply @HAYK. I'm not entirely sure what this device does, I presume it does some filtering on the AC power coming to the toroidal transformer. I'm initially trying to find a reasonable way to install my components, in order to be able to assess and troubleshoot noise issues from a specific starting position. Then I 'll see if additional components could help. I do have some room to spare in my aluminum case if the need arises.
I just realized I made another enormous newbie blunder. My speaker connectors are metallic and not insulated, so they short circuit the speaker outputs of the power amp. It's a good thing I tried to connect the speaker outputs and then checked the connections before powering this thing up. The board has certain protection features but who knows if they would have worked.
Now I need to either insulate my existing binding posts or switch to a different insulated speaker connector altogether. First solution has the problem that the rear panel routes are very precise, therefore too snug to use some heatshrink tubing for example.
Now I need to either insulate my existing binding posts or switch to a different insulated speaker connector altogether. First solution has the problem that the rear panel routes are very precise, therefore too snug to use some heatshrink tubing for example.
So, some progress today. I replaced the metallic binding posts that made contact with the non-finished holes of the back panel, making all four of them electrically connected. Again, thankfully I did not power up the amp in that configuration. New binding posts are 100% insulated. I had to widen the 5x4mm square holes to 6.5mm circes and create some inner notches but all went well without any marking or other damage to the panel.
I decided I wanted to also insulate the ground of the RCA inputs although at first this does not seem like an absolute requirement. Since I didn't want to mess with the shape of those holes, all I did was use a couple of turns of transparent office tape and it did the job well. At the front and back the RCA terminals have plastic insuators, although they have little work to do given the non conductive finish of the back panel. So everything is electrically isolated from the back panel now.
A couple of pics:
With this done, I have a couple of questions that could possibly pertain to noise:
a) Speaker + (red) should most certainly be insulated from the back plate, and therefore not grounded. But could it be a good idea to ground the speaker - binding posts (black)?
b) As mentioned, I insulated RCA ground from the back panel for both left and right input channels. Is it a good idea or bad idea to lift them from ground? What is your preference if you know nothing about the rest of the build?
I decided I wanted to also insulate the ground of the RCA inputs although at first this does not seem like an absolute requirement. Since I didn't want to mess with the shape of those holes, all I did was use a couple of turns of transparent office tape and it did the job well. At the front and back the RCA terminals have plastic insuators, although they have little work to do given the non conductive finish of the back panel. So everything is electrically isolated from the back panel now.
A couple of pics:
With this done, I have a couple of questions that could possibly pertain to noise:
a) Speaker + (red) should most certainly be insulated from the back plate, and therefore not grounded. But could it be a good idea to ground the speaker - binding posts (black)?
b) As mentioned, I insulated RCA ground from the back panel for both left and right input channels. Is it a good idea or bad idea to lift them from ground? What is your preference if you know nothing about the rest of the build?