Rick,
The one I used for the 3118D2 was the Bourns shielded SMD 10 uH 6A manuf # SRR1280-100M, Digikey # SRR1280-100MCT-ND. It feels warm during normal use at moderate volumes.
http://www.bourns.com/data/global/pdfs/SRR1280.pdf
For the 3116D2, I am using the unshielded Bourns thru-hole radial 10 uH 3A, manuf # RL622-100K-RC, Digikey # M9981-ND. Surprisingly, it does not get warm at all during normal use.
http://www.bourns.com/data/global/pdfs/rl622_series.pdf
The one I used for the 3118D2 was the Bourns shielded SMD 10 uH 6A manuf # SRR1280-100M, Digikey # SRR1280-100MCT-ND. It feels warm during normal use at moderate volumes.
http://www.bourns.com/data/global/pdfs/SRR1280.pdf
For the 3116D2, I am using the unshielded Bourns thru-hole radial 10 uH 3A, manuf # RL622-100K-RC, Digikey # M9981-ND. Surprisingly, it does not get warm at all during normal use.
http://www.bourns.com/data/global/pdfs/rl622_series.pdf
I have a question.
I'm using the parts-express sure board with the tpa-3110.
2x8W @ 4 Ohm TPA3110 Class-D Audio Amplifier Board Only 320-329
I use 8 aa batteries to power it.
On some music I hear a static like noise (like fm static) like a kssch noise.
It happens on the leading edge of piano notes on the first 30 seconds of the last track on Diana Krall's "live in paris". It is a cover of Billy Joel's "Just the way you are."
Every time there is a percussive piano note, I get a kscht noise.
I run an active crossover to it. 4ohm 93db woofer, 8ohm 110db compression driver tweeter. Problem goes away when I disconnect the woofer (use other amp to drive woofer).
Do I need more capacitance in my supply ?
Am I clipping ?
Is it TIM distortion ?
What's going on ?
I'll disconnect the tweet later to make sure has something do do with the woof.
Otherwise the resolution is astounding, probably the 2x100uF caps help.
Oh, I also get full range audio sounding like digitized noise pumped through my subwoofer.
Norman
I'm using the parts-express sure board with the tpa-3110.
2x8W @ 4 Ohm TPA3110 Class-D Audio Amplifier Board Only 320-329
I use 8 aa batteries to power it.
On some music I hear a static like noise (like fm static) like a kssch noise.
It happens on the leading edge of piano notes on the first 30 seconds of the last track on Diana Krall's "live in paris". It is a cover of Billy Joel's "Just the way you are."
Every time there is a percussive piano note, I get a kscht noise.
I run an active crossover to it. 4ohm 93db woofer, 8ohm 110db compression driver tweeter. Problem goes away when I disconnect the woofer (use other amp to drive woofer).
Do I need more capacitance in my supply ?
Am I clipping ?
Is it TIM distortion ?
What's going on ?
I'll disconnect the tweet later to make sure has something do do with the woof.
Otherwise the resolution is astounding, probably the 2x100uF caps help.
Oh, I also get full range audio sounding like digitized noise pumped through my subwoofer.
Norman
My tpa3110 started recently to add some noise over music, and I think chip is damaged .I was runing it directly (without a pot) with ESI Maya44 usb card (and Maya44 is not 'compatible' with sleep-mode of my PC, so I needed to replug it several times, sending a nasty 'thump' to the speakers) . Manuel of the Sure tpa3110 clearly states NOT to overload input of the amp with signals over 0.8V -it can damage the chip. So, I'm ordering a new one, it worked beautifully with my BetsyK mltl...
That noise could also be flux residue from the solder paste.
I'm jut sayin' cause we used to build the EVM, and I know
that sound. Wash and bake it to be sure, wouldn't hurt...
I have nothing to do with the Sure board at PE, its just
an assumption that maybe they had the same difficulty?
It can be especially bad if some moron re-worker mixes
water soluble with no-clean, and traps some junk that
needs washing under gunk that can't easily be rinsed.
Or sometimes clean only what they can see on top, like
if we can't see the flux, it doesn't exist. Yeah, right...
On the other end of the same, sometimes the center pad
might not melt, and the paste would get washed out. Or it
would melt, but all be sucked down into the vias and not
wet the chip. Hard to say without using an x-ray to inspect.
Flip it over and melt plenty of extra solder through the
vias with a heat gun, and maybe you can salvage the
connection without removing the chip. Again assumption
the Sure board even has such vias in the pad? Might not...
I'm jut sayin' cause we used to build the EVM, and I know
that sound. Wash and bake it to be sure, wouldn't hurt...
I have nothing to do with the Sure board at PE, its just
an assumption that maybe they had the same difficulty?
It can be especially bad if some moron re-worker mixes
water soluble with no-clean, and traps some junk that
needs washing under gunk that can't easily be rinsed.
Or sometimes clean only what they can see on top, like
if we can't see the flux, it doesn't exist. Yeah, right...
On the other end of the same, sometimes the center pad
might not melt, and the paste would get washed out. Or it
would melt, but all be sucked down into the vias and not
wet the chip. Hard to say without using an x-ray to inspect.
Flip it over and melt plenty of extra solder through the
vias with a heat gun, and maybe you can salvage the
connection without removing the chip. Again assumption
the Sure board even has such vias in the pad? Might not...
The fact that it goes away when not driving woofer is interesting. I think the woofer connected causes amp to have to drive more current which overshoots giving the sibilance. This my be one of the problems of running 6 ft of wire without a proper inductor and capacitor filter setup but only a ferrite bead. Try connecting the inductor and cap filter that the TI circuit recommends at the amp output before going to the speaker wire to see if it goes away. IIRCC, for 4 ohm speaker use a 15 uH inductor (on each wire to speaker) in series and a 2.2 uF (50 volt) to ground after the inductor. See if this will clean up the excessive hash produced by the PWM to drive the woofer, which then gets onto the tweeter.
Last edited:
Monitor your supply with an oscilloscope during high current passages.
Even a DMM may show supply drop.
http://data.energizer.com/PDFs/e91.pdf
Run the amp on a fixed power supply good for >5A and see if there is a difference.
IMHO AA size may be a small.
Even a DMM may show supply drop.
http://data.energizer.com/PDFs/e91.pdf
Run the amp on a fixed power supply good for >5A and see if there is a difference.
IMHO AA size may be a small.
But Sure tpa3110 worked perfectly well, an sounded reeeealy good for the first 20 days,I was listening music using program-attenuation only (WMPlauer), so it's not bad solder I think. Stupid as I am, I did replugin' Maya's usb without cutting the power to amp, an after 10 or 15 such 'thumps' (plus some blackouts - diferent kind of noise thru the speakers) , litle businesscardsize amp started to put some 'krrrr...' over music. I 've bought Alps 50K log pot, but it was too late. So, Im waiting for a new one, at 9$ it's no brainer
Water soluble flux: You can bake it dry, test it,
put it in a sealed bag, and pretend it works.
Till starts drawing water right back out the air.
Weeks or months later, depending humidity...
Even "dry", it seems to me viscous and oily.
Can't neutralize or get rid WS flux by baking.
It likes to hide under the edge, between the
pins and the ground pad. And the more dry,
the harder it is to get out with a short blast.
Takes time to re-absorb water, but a minute
is plenty long enough...
One of my pet peeves was a re-worker who didn't
understand every time you touch-up with a WS flux,
you have to thoroughly wash the board all over again.
Even if its been washed before.
Not just dab at the top of the rework with an alcohol
soaked q-tip and throw it in the oven. Like I won't
know the difference when it comes to test...
The q-tip is fine for cleaning enough to inspect the
progress of your work. But you really got to drown it,
else you are just smearing that flux around. It goes
into the vias and hides up under everything...
Again, I know exactly that sound you just described.
Nothing technical, just experience. If I didn't see or
measure anything else obviously wrong, send it back
to wash before I would request replacing anything.
----
I did have fun accident at the bed of nails testing house.
Needed to tombstone a part (lift up one end) to isolate
where a short was, there was nobody to do that for me.
Tried to use what looked like our WS flux, at their bench.
But it was actually a yellow inkjet refill bottle! I did not
spot the slight difference in color, and touched the drop
with an iron. What a mess one tiny ink drop can make!
Splattered everywhere...
put it in a sealed bag, and pretend it works.
Till starts drawing water right back out the air.
Weeks or months later, depending humidity...
Even "dry", it seems to me viscous and oily.
Can't neutralize or get rid WS flux by baking.
It likes to hide under the edge, between the
pins and the ground pad. And the more dry,
the harder it is to get out with a short blast.
Takes time to re-absorb water, but a minute
is plenty long enough...
One of my pet peeves was a re-worker who didn't
understand every time you touch-up with a WS flux,
you have to thoroughly wash the board all over again.
Even if its been washed before.
Not just dab at the top of the rework with an alcohol
soaked q-tip and throw it in the oven. Like I won't
know the difference when it comes to test...
The q-tip is fine for cleaning enough to inspect the
progress of your work. But you really got to drown it,
else you are just smearing that flux around. It goes
into the vias and hides up under everything...
Again, I know exactly that sound you just described.
Nothing technical, just experience. If I didn't see or
measure anything else obviously wrong, send it back
to wash before I would request replacing anything.
----
I did have fun accident at the bed of nails testing house.
Needed to tombstone a part (lift up one end) to isolate
where a short was, there was nobody to do that for me.
Tried to use what looked like our WS flux, at their bench.
But it was actually a yellow inkjet refill bottle! I did not
spot the slight difference in color, and touched the drop
with an iron. What a mess one tiny ink drop can make!
Splattered everywhere...
Last edited:
So, I can try baking and if problem temporarely goes away, its a bad flux problem? What temperature and time of baking do you recomand?
-On the other end of the same, sometimes the center pad
might not melt, and the paste would get washed out. Or it
would melt, but all be sucked down into the vias and not
wet the chip. Hard to say without using an x-ray to inspect.
Flip it over and melt plenty of extra solder through the
vias with a heat gun, and maybe you can salvage the
connection without removing the chip. Again assumption
the Sure board even has such vias in the pad? Might not...-
Unfortunately, I'm not experienced enough to play with solder here...
might not melt, and the paste would get washed out. Or it
would melt, but all be sucked down into the vias and not
wet the chip. Hard to say without using an x-ray to inspect.
Flip it over and melt plenty of extra solder through the
vias with a heat gun, and maybe you can salvage the
connection without removing the chip. Again assumption
the Sure board even has such vias in the pad? Might not...-
Unfortunately, I'm not experienced enough to play with solder here...
Attachments
65C for an hour will temporarily get rid of absorbed moisture,
but won't get rid of flux and any corrosion that has formed.
You need hot water and a brush.
I recommend going at with running water as hot as your hands
can tolerate, a soft toothbrush, maybe a final dip in distilled, or
de-ionized. On assumption you haven't a sink that runs clean
de-ionized to begin with.
Then blow that water off with dry compressed air (not the kind
that spits more dirty water and oil) from every possible angle.
Blow the hardware too, not just the circuitry. Then put in 65C
oven for at least an hour.
I've seen where a board comes out of the oven nice and dry,
but still had water up in the legs of the board standoffs. That
would be a real problem if you were fixin' to seal it into a bag.
Why I was saying you need to blow out the hardware too, or
else it takes a LOT longer in the oven to dry.
-----
About cleaning with alcohol or other solvents...
I can't remember if it was Coilcraft or their competitor I was
talking to at one of the TI tradeshows. The rep explained that
about alcohol being the reason why so many of their parts
were "no-clean". It had nothing to do with water...
If using a bottled WS flux that includes alcohol as a solvent,
you have maybe damaged the enamel before it ever gets to
wash. They don't want you using that sort of flux on any of
the inductors. Not that you can't or shouldn't rinse it if you
got some on by mistake.
The WS flux in solder paste doesn't have alcohol, so you are
not seeing "no-clean" warning with surface mount inductors
so often as with through-hole.
but won't get rid of flux and any corrosion that has formed.
You need hot water and a brush.
I recommend going at with running water as hot as your hands
can tolerate, a soft toothbrush, maybe a final dip in distilled, or
de-ionized. On assumption you haven't a sink that runs clean
de-ionized to begin with.
Then blow that water off with dry compressed air (not the kind
that spits more dirty water and oil) from every possible angle.
Blow the hardware too, not just the circuitry. Then put in 65C
oven for at least an hour.
I've seen where a board comes out of the oven nice and dry,
but still had water up in the legs of the board standoffs. That
would be a real problem if you were fixin' to seal it into a bag.
Why I was saying you need to blow out the hardware too, or
else it takes a LOT longer in the oven to dry.
-----
About cleaning with alcohol or other solvents...
I can't remember if it was Coilcraft or their competitor I was
talking to at one of the TI tradeshows. The rep explained that
about alcohol being the reason why so many of their parts
were "no-clean". It had nothing to do with water...
If using a bottled WS flux that includes alcohol as a solvent,
you have maybe damaged the enamel before it ever gets to
wash. They don't want you using that sort of flux on any of
the inductors. Not that you can't or shouldn't rinse it if you
got some on by mistake.
The WS flux in solder paste doesn't have alcohol, so you are
not seeing "no-clean" warning with surface mount inductors
so often as with through-hole.
Yeah, you have vias in the center pad. So there is a real
possibility that the solder paste and flux could be wicked
away before it wet the back of the chip. You need an x-ray
machine to tell. There's nothing you can see from looking
through the holes in the back, except that dry empty holes
are a better sign of success here than shiny clogged ones.
That revelation for me was totally counter-intuitive...
You could put you some low temp lead solder on the back
and melt it through to the front with an iron and heatgun.
In this case only, shiny clogged is what you are after...
Don't waste your time if you havn't a heatgun. The board
will sink away the heat of an iron by itself and the solder
will never go through.
---
If you have to heat the front, then you want aluminum
foil tape to mask off the caps and plastic parts. Never
use plain aluminum foil without a backing. And never
peel to expose the glue. Fold it such that molten glue
won't oooze from folded edges or corners that will be
touching your board. And it will ooze from all the other
unfolded edges, so plan ahead carefully for that.
Paper and glue are important parts of the barrier and
add mass. Ordinary foil conducts heat too well and the
heatgun can blow it around too easily. As messy as the
glue is when it drips, no other mask works better.
The professional stuff is a water soluble and peel away
mask that comes in a pink bottle. It takes too long to
set up, and offers less protection than foil. If you are in
a rush (which always happens) and the pink goo is still
runny, it gets blown from where its needed most.
Wouldn't waste my time with pink goo as a heat barrier,
but its very good for keeping tiny surface mount parts
from being blown around unintentionally.
You will be moving a lot of hot air, because the ground
plane will conduct it away. If you are trying too hard to
present all the heat to one spot, its just going to burn
before you get anything on that ground plane to melt...
possibility that the solder paste and flux could be wicked
away before it wet the back of the chip. You need an x-ray
machine to tell. There's nothing you can see from looking
through the holes in the back, except that dry empty holes
are a better sign of success here than shiny clogged ones.
That revelation for me was totally counter-intuitive...
You could put you some low temp lead solder on the back
and melt it through to the front with an iron and heatgun.
In this case only, shiny clogged is what you are after...
Don't waste your time if you havn't a heatgun. The board
will sink away the heat of an iron by itself and the solder
will never go through.
---
If you have to heat the front, then you want aluminum
foil tape to mask off the caps and plastic parts. Never
use plain aluminum foil without a backing. And never
peel to expose the glue. Fold it such that molten glue
won't oooze from folded edges or corners that will be
touching your board. And it will ooze from all the other
unfolded edges, so plan ahead carefully for that.
Paper and glue are important parts of the barrier and
add mass. Ordinary foil conducts heat too well and the
heatgun can blow it around too easily. As messy as the
glue is when it drips, no other mask works better.
The professional stuff is a water soluble and peel away
mask that comes in a pink bottle. It takes too long to
set up, and offers less protection than foil. If you are in
a rush (which always happens) and the pink goo is still
runny, it gets blown from where its needed most.
Wouldn't waste my time with pink goo as a heat barrier,
but its very good for keeping tiny surface mount parts
from being blown around unintentionally.
You will be moving a lot of hot air, because the ground
plane will conduct it away. If you are trying too hard to
present all the heat to one spot, its just going to burn
before you get anything on that ground plane to melt...
Last edited:
Sorry folks, but this thread is titled "TPA3100D2" I guess it has turned into a general TI Class "D" amp thread now.
As a note I hand solder my TPA3100D2 chips in. In the two years that I built my first one, I have had no issues, I used GlowCore no-clean (15mil) , it is great stuff. Water washable solder/flux is a waste of time. Hand soldering does not require any moisture baking operation.
I read that someone want to run a TPA3100D2 or equiv from USB power or a battery. Maybe a 12V car battery, but not small batteries, this can get expensive or be a pain to re-charge all the time. Totally forget USB power, I will only run a headphone amp off USB power.
As a note I hand solder my TPA3100D2 chips in. In the two years that I built my first one, I have had no issues, I used GlowCore no-clean (15mil) , it is great stuff. Water washable solder/flux is a waste of time. Hand soldering does not require any moisture baking operation.
I read that someone want to run a TPA3100D2 or equiv from USB power or a battery. Maybe a 12V car battery, but not small batteries, this can get expensive or be a pain to re-charge all the time. Totally forget USB power, I will only run a headphone amp off USB power.
'These boards are $10 a pop, assembled, correct?'
Yup! You can find it at PartsExpress or on the e- bay
2 X 8Watt Class D Audio Amplifier Board - TPA3110 2W Stereo Power Mini Amp | eBay
Yup! You can find it at PartsExpress or on the e- bay
2 X 8Watt Class D Audio Amplifier Board - TPA3110 2W Stereo Power Mini Amp | eBay
You are wright, sorry... TPA3110 deserves new thread, sounds good and is extremly cheap.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.