PSU boards :
amp boards at first :
I soldered resistors, small caps, big caps, and then the LM's..
For LM's, I use a piece of wood for have a little separate (2 mm) between pins and pcb, like this :
You can see C1 C2 (snubber), and Cf under pcb. Sold two pins and check squareness, then sold all..
LM close up :
In this case, LM's are in the right center of heatsink :
Clean up under pcb and eliminate flux solderinng.
Phil.
An externally hosted image should be here but it was not working when we last tested it.
amp boards at first :
An externally hosted image should be here but it was not working when we last tested it.
I soldered resistors, small caps, big caps, and then the LM's..
For LM's, I use a piece of wood for have a little separate (2 mm) between pins and pcb, like this :
An externally hosted image should be here but it was not working when we last tested it.
You can see C1 C2 (snubber), and Cf under pcb. Sold two pins and check squareness, then sold all..
LM close up :
An externally hosted image should be here but it was not working when we last tested it.
In this case, LM's are in the right center of heatsink :
An externally hosted image should be here but it was not working when we last tested it.
Clean up under pcb and eliminate flux solderinng.
Phil.
Looks good and tidy too.
Except. . . those light green screw terminals are the type that arc, static, slip and spontaneously release the cable. If power cables get loose, the result could be unpleasant and sudden. It would be better to solder the cables.
Except. . . those light green screw terminals are the type that arc, static, slip and spontaneously release the cable. If power cables get loose, the result could be unpleasant and sudden. It would be better to solder the cables.
Hi Daniel,
I agree, green terminals are not the best, but very practical to mount or dismount boards, for tests or changes..
Then, wires was solder.
Just a little detail, at first I wanted mount a Schaffner IEC inlet, but I can't.
So I need to build a simple main filter. Schematic below is Schaffner..
I have componants except coils.
I pick up two coils from an broken TA2020 amp, but I don't knaw value ?
Phil.
I agree, green terminals are not the best, but very practical to mount or dismount boards, for tests or changes..
Then, wires was solder.
Just a little detail, at first I wanted mount a Schaffner IEC inlet, but I can't.
So I need to build a simple main filter. Schematic below is Schaffner..
I have componants except coils.
I pick up two coils from an broken TA2020 amp, but I don't knaw value ?
An externally hosted image should be here but it was not working when we last tested it.
Phil.
10uH (0.01mH) to 22uH (0.022mH) is the range usable with TA2020 (see datasheet or see mfg replacement parts listings), so those coils are probably too small for your power filter application.
Now, I'm ever so curious how Schaffner fit 12mH into an IEC socket. 😀 Could it be a CM choke?
An IEC socket serves mainly to reduce noise emissions of computer switchmode power supply.
However, with your linear supply for audio amp, the ideal place to filter is at your transformer's secondaries, with RC's across the secondaries.
Now, I'm ever so curious how Schaffner fit 12mH into an IEC socket. 😀 Could it be a CM choke?
An IEC socket serves mainly to reduce noise emissions of computer switchmode power supply.
However, with your linear supply for audio amp, the ideal place to filter is at your transformer's secondaries, with RC's across the secondaries.
Last edited:
Like this (Xray pict) !!
RC's across the secondaries ?
Better way, I found (but I do not remember where), 3,3 nF X2 type like Wima MP3 ?
Thank's.
Phil.
Power Filtering
AC filtering:
(practically applies to upper treble)
Snubbing the secondaries (the lower voltage side of your transformer) is a really good spot for an RC.
The audible effectiveness of this is fairly mild, so I'd consider it an "accessory" potentially useful for refining the treble qualities.
Range of C is 2u2 to 220n
Range of R is 0R to 18R.
Exact value depends on specific combination of specific transformer with specific bridge rectifier in specific applications; so, unfortunately, values copied from other applications may not work. However, it is possible to use a 2u cap and a variable resistor for R and fine tune for yourself. If center tap, it is super easy with just one variable RC, but if the effort is doubled for dual secondaries build, begin with the V- rail, tune to suit and then make a copy for the V+ rail.
I would like to assure you that AC filtering won't be getting any easier than this; however, there are two matters that make such an assurance impossible: 1) Fairchild Stealth (tested and works--slightly but well) and 2) clip on ferrite for the transformer secondaries cables (Caution: possible, but not tested at all--that may or may not work and/or explode something--no idea--not tested. But it is both plausible and easier).
DC filtering:
(practically applies to midrange and upper-midrange tone)
http://www.diyaudio.com/forums/chip...e-no-lossy-emitter-resistors.html#post3495575 See Post#30. Perhaps ignore most of the chat, but be sure to see the photographs at the bottom of post 30. Two diodes were added as series elements, at, locale, amplifier board v+ and v-. There's not much easier than that, and it is highly effective. Nice laid back powerful clean sound. It has a fairly big effect. NO edgy/forwards sound there! However, need of such is in direct proportion to suitability, so judge for yourself.
If used with a ferrite to increase filtering without increasing contour, the order is: Power supply, cable. . . ., ferrite, diode, amplifier board. The FB43 ferrite fits onto the diode pin, the diode fits onto the amplifier board. The dc umbilical cable for v+ and v- is series to the diode. For v+, the diode stripe faces the amplifier board. For v-, the diode stripe faces away from the amplifier board. Even then, still very easy!
The actual concept is most extremely simple; however, it is a little confusing why some applications receive so much benefit from something so simple and easy to do.
I would like to assure you that effective DC filtering won't get any easier than this, and I can do that, because there's nothing easier that works well.
Summary:
Power filtering for analog linear audio amplifier doesn't have to be complicated. There was two of the most practically, applicably, effective filters that I know of and they are the really easy to build.
P.S.
It may be premature to accessorize. Perhaps you'd rather hear the audio amp play, and then come up with a list of needs before further modifying it?
AC filtering:
(practically applies to upper treble)
Snubbing the secondaries (the lower voltage side of your transformer) is a really good spot for an RC.
The audible effectiveness of this is fairly mild, so I'd consider it an "accessory" potentially useful for refining the treble qualities.
Range of C is 2u2 to 220n
Range of R is 0R to 18R.
Exact value depends on specific combination of specific transformer with specific bridge rectifier in specific applications; so, unfortunately, values copied from other applications may not work. However, it is possible to use a 2u cap and a variable resistor for R and fine tune for yourself. If center tap, it is super easy with just one variable RC, but if the effort is doubled for dual secondaries build, begin with the V- rail, tune to suit and then make a copy for the V+ rail.
I would like to assure you that AC filtering won't be getting any easier than this; however, there are two matters that make such an assurance impossible: 1) Fairchild Stealth (tested and works--slightly but well) and 2) clip on ferrite for the transformer secondaries cables (Caution: possible, but not tested at all--that may or may not work and/or explode something--no idea--not tested. But it is both plausible and easier).
DC filtering:
(practically applies to midrange and upper-midrange tone)
http://www.diyaudio.com/forums/chip...e-no-lossy-emitter-resistors.html#post3495575 See Post#30. Perhaps ignore most of the chat, but be sure to see the photographs at the bottom of post 30. Two diodes were added as series elements, at, locale, amplifier board v+ and v-. There's not much easier than that, and it is highly effective. Nice laid back powerful clean sound. It has a fairly big effect. NO edgy/forwards sound there! However, need of such is in direct proportion to suitability, so judge for yourself.
If used with a ferrite to increase filtering without increasing contour, the order is: Power supply, cable. . . ., ferrite, diode, amplifier board. The FB43 ferrite fits onto the diode pin, the diode fits onto the amplifier board. The dc umbilical cable for v+ and v- is series to the diode. For v+, the diode stripe faces the amplifier board. For v-, the diode stripe faces away from the amplifier board. Even then, still very easy!
The actual concept is most extremely simple; however, it is a little confusing why some applications receive so much benefit from something so simple and easy to do.
I would like to assure you that effective DC filtering won't get any easier than this, and I can do that, because there's nothing easier that works well.
Summary:
Power filtering for analog linear audio amplifier doesn't have to be complicated. There was two of the most practically, applicably, effective filters that I know of and they are the really easy to build.
P.S.
It may be premature to accessorize. Perhaps you'd rather hear the audio amp play, and then come up with a list of needs before further modifying it?
Hi Daniel,
>> Perhaps you'd rather hear the audio amp play, and then come up with a list of needs before further modifying it?
Yes, I think to do as, finish building and see..
But thank's for great explanation, as always !!
Work progresses :
I did not drill the sides (bottom and rear) for cooling, at this time..
Phil.
>> Perhaps you'd rather hear the audio amp play, and then come up with a list of needs before further modifying it?
Yes, I think to do as, finish building and see..
But thank's for great explanation, as always !!
Work progresses :
An externally hosted image should be here but it was not working when we last tested it.
I did not drill the sides (bottom and rear) for cooling, at this time..
Phil.
Arctic Ceramique and GC Type 44 are both good types of thermal paste to use between the LM3886TF and heatsink.
Thermal compound can also be used between the heatsinks and other bolt-on metal objects (such as support brackets).
Thermal compound can also be used between the heatsinks and other bolt-on metal objects (such as support brackets).
Hi,
Thank's for advices..
Thermal paste I use :
HC-131 une bonne conductivité thermique Thermal Grease (Blanc)
LM's are already mounted, and I put thermal paste under the heatsinks.
In my previous LM3886 amp from Chipamp (same schematic), temperature is only near 30 °, voltage 28 volts, DC offset 1,42 mV and 1,50 mV
Speakers are about 95/97 db, so I don't need so much power, maybe 5/10 watts max..
After reading :
http://www.diyaudio.com/forums/chip-amps/265771-lm3886-thermal-experiment-data.html
what is the correct or best operating temperature at sounding ?
I ask that, cause I have several amps (classA, mosfet..), and they work better when they are hot, maybe after half an hour or an hour..
Phil.
Thank's for advices..
Thermal paste I use :
HC-131 une bonne conductivité thermique Thermal Grease (Blanc)
LM's are already mounted, and I put thermal paste under the heatsinks.
In my previous LM3886 amp from Chipamp (same schematic), temperature is only near 30 °, voltage 28 volts, DC offset 1,42 mV and 1,50 mV
Speakers are about 95/97 db, so I don't need so much power, maybe 5/10 watts max..
After reading :
http://www.diyaudio.com/forums/chip-amps/265771-lm3886-thermal-experiment-data.html
what is the correct or best operating temperature at sounding ?
I ask that, cause I have several amps (classA, mosfet..), and they work better when they are hot, maybe after half an hour or an hour..
Phil.
Finished rear side :
Phil.
An externally hosted image should be here but it was not working when we last tested it.
Phil.
Yes, that's the right stuff. That is white silicon paste with aluminum oxide gap filler. It looks similar to GC Type 44.
Missing data: What was the room temp? And, the nature of the heatsink? And circulation/vents?
It is basically the difference between room temp and chip temp that we'd want to look at.
To warm a goodly sized aerospace aluminum support beam to 30c despite freezing temperature would make the LM3886 fairly unhappy; however, there are other conditions where 30c might be good news instead. 🙂 More information needed. 🙂
Due to the built in hard-clipper (spike system) that you never want to hear, the LM3886 will work better Cool (very close to room temp), lightly loaded, and not pushed over 45W, not even for peaks. The warmer it gets, the smaller the clean output capacity will go. So, keep it cool.
In the other part of your question, the heat was not desirable but the stable temperature is desirable, so that's why some amplifiers work better after the temp quit changing. Perhaps those other amplifiers need refit to functional thermal tracking (which was probably muntzed to simply and reduce transistor count).
That is very smart! The chip probably won't clip/distort with that usage.
I'm curious: What kind of speakers (details)?
So, I was wondering, is the additional amplifier for a second location or did the first amp leave some need unmet?
Last edited:
Hi Daniel,
It's for raising !!
First build, not so good, very noisy :
http://www.diyaudio.com/forums/chip-amps/79303-chip-amp-photo-gallery-283.html#post4126123
second, kit from Analogmetric, little better :
http://www.diyaudio.com/forums/chip-amps/247535-lm3886-kit-am-5.html
third, Chipamp, very good, works at this time :
http://www.diyaudio.com/forums/chip-amps/79303-chip-amp-photo-gallery-280.html#post4107696
And now I hope the best of alls !
Speakers are :
H frame with two Alpha15A per side, two Fostex FE126eN per side (back to back) for dual acoustic dipole..
About temperature in room, it's near 20/21°
Values that I have given are for the last amp (Chipamp dual mono)..
Phil.
It's for raising !!
First build, not so good, very noisy :
http://www.diyaudio.com/forums/chip-amps/79303-chip-amp-photo-gallery-283.html#post4126123
second, kit from Analogmetric, little better :
http://www.diyaudio.com/forums/chip-amps/247535-lm3886-kit-am-5.html
third, Chipamp, very good, works at this time :
http://www.diyaudio.com/forums/chip-amps/79303-chip-amp-photo-gallery-280.html#post4107696
And now I hope the best of alls !
Speakers are :
An externally hosted image should be here but it was not working when we last tested it.
H frame with two Alpha15A per side, two Fostex FE126eN per side (back to back) for dual acoustic dipole..
About temperature in room, it's near 20/21°
Values that I have given are for the last amp (Chipamp dual mono)..
Phil.
Neat! Those Fostex are big enough to cross over at woofer baffle step frequency. Good job!
Question:
Are the pairs of Eminence Alpha 15A woofers wired series or parallel? If parallel, that is a 4 ohm load, and in that case, we might want to discuss amplifiers suitable for 4 ohms.
Question:
Are the pairs of Eminence Alpha 15A woofers wired series or parallel? If parallel, that is a 4 ohm load, and in that case, we might want to discuss amplifiers suitable for 4 ohms.
Below impedance curve with filter, Alpha15A in // :
It seems suitable with LM3886, but if you have some advices, thank's in advance.
Just for fun, response curve, in red predict with MJ King sheet, black response in room at listening point :
Phil.
An externally hosted image should be here but it was not working when we last tested it.
It seems suitable with LM3886, but if you have some advices, thank's in advance.
Just for fun, response curve, in red predict with MJ King sheet, black response in room at listening point :
An externally hosted image should be here but it was not working when we last tested it.
Phil.
Work in progress :
Phil
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
Phil
Holes in the floor under the heatsinks and holes in the roof over the heatsinks to allow the cold air to enter, warmed air to rise and hot air to escape, as in "chimney effect".
Hi Andrew,
I agree, some detail about "chimney effect" :
Phil.
I agree, some detail about "chimney effect" :
An externally hosted image should be here but it was not working when we last tested it.
Phil.
Like Jimi Hendrix said : "rainy day, dream away"
So the wiring is finished, it remains to attach wires much "clean" and test amp.
PSU works, leds too !
Phil.
So the wiring is finished, it remains to attach wires much "clean" and test amp.
PSU works, leds too !
An externally hosted image should be here but it was not working when we last tested it.
Phil.
Hi,
Thank's Andrew, I forgot this..
It's works !
Dc offset without load 0,40 mV and 0,45 mV
With about 8 Ohm load 2 mV and 2,7 mV
In room temp 21°, and after 20 minutes maybe at 2/3 watts, temp at center heatsink is 27°, without top cover.
First listening : well balanced, very good soundstage, nice low bass, smoother sounding..
More details after several hours of music..
One thing about toroid mounting, I had to loosen the screws, because the more you shake the more we hear like a "cat sleeping"..
A hum that comes and goes, like with huge toroid.
Phil.
Thank's Andrew, I forgot this..
It's works !
Dc offset without load 0,40 mV and 0,45 mV
With about 8 Ohm load 2 mV and 2,7 mV
In room temp 21°, and after 20 minutes maybe at 2/3 watts, temp at center heatsink is 27°, without top cover.
First listening : well balanced, very good soundstage, nice low bass, smoother sounding..
More details after several hours of music..
One thing about toroid mounting, I had to loosen the screws, because the more you shake the more we hear like a "cat sleeping"..
A hum that comes and goes, like with huge toroid.
Phil.
- Status
- Not open for further replies.