likely not thermal shutdown then.
there is high/low voltage and short circuit protection
should be able to go down to 18 volts
there is a resistor value to select overcurrent threshold
could be conservative for whatever load its running
likely a board or component issue
resistor/inductor and numerous other guesses where
a short or open could be. which would cause low /high voltage error
unless the actual chip die was damaged as well
probably start with basic voltage measurements
power pad is grounded I believe so can check for any
input/heatsink/case grounds.
if bridged or floating load, output connectors, damaged twisted speakon jacks
just numerous guesses where to trouble shoot
if it works for awhile then shutsdown.
is it a difficult load like 2ohms
or a 4 ohm load with a tweeter crossover that might dip
down to 2 ohms.
there is high/low voltage and short circuit protection
should be able to go down to 18 volts
there is a resistor value to select overcurrent threshold
could be conservative for whatever load its running
likely a board or component issue
resistor/inductor and numerous other guesses where
a short or open could be. which would cause low /high voltage error
unless the actual chip die was damaged as well
probably start with basic voltage measurements
power pad is grounded I believe so can check for any
input/heatsink/case grounds.
if bridged or floating load, output connectors, damaged twisted speakon jacks
just numerous guesses where to trouble shoot
if it works for awhile then shutsdown.
is it a difficult load like 2ohms
or a 4 ohm load with a tweeter crossover that might dip
down to 2 ohms.
Last edited:
I took a look at the board and it looks like the TI recommended circuit. They have a 22k OCP resistor.
I noticed however that the 4 - 1uF caps near the output of the chip are not populated on the board. I don't know if that could cause this or not.
I've tracked down when the fault happens. It's only when there are sharp transients (i.e. slap bass on low notes) that the amp goes into limp mode.
I monitored the supply voltage and it does not visibly (Fluke meter) dip on the transients. It's never below 48v.
I noticed however that the 4 - 1uF caps near the output of the chip are not populated on the board. I don't know if that could cause this or not.
I've tracked down when the fault happens. It's only when there are sharp transients (i.e. slap bass on low notes) that the amp goes into limp mode.
I monitored the supply voltage and it does not visibly (Fluke meter) dip on the transients. It's never below 48v.
Attachments
This might be the PS being recharged by the load, causing the rail voltage to increase past the maximum (OVP) protection. This happens under these kind of conditions with some class-D amps and is called "bus pumping":
https://www.maximintegrated.com/en/design/technical-documents/tutorials/4/4260.htmlhttps://www.diyaudio.com/community/threads/bus-pumping.164953/
Bus pumping happens when charge is returned by the load to the amp board. The SMPS often cannot absorb any charge being returned by the amp board and so the rail voltage rises, and this triggers the OVP on the amp board or potentially the SMPS itself. One potential fix is to place capacitors between the PS and the amp board. The caps absorb charge more than the PS can and this tends to help. If you can just turn everything off and then on again, and it all works as though nothing happened, this may be the cause.
Another approach that would solve this problem would be to use a linear PS (transformer+rectifier+caps) instead of an SMPS since the caps are built into it already.
https://www.maximintegrated.com/en/design/technical-documents/tutorials/4/4260.htmlhttps://www.diyaudio.com/community/threads/bus-pumping.164953/
Bus pumping happens when charge is returned by the load to the amp board. The SMPS often cannot absorb any charge being returned by the amp board and so the rail voltage rises, and this triggers the OVP on the amp board or potentially the SMPS itself. One potential fix is to place capacitors between the PS and the amp board. The caps absorb charge more than the PS can and this tends to help. If you can just turn everything off and then on again, and it all works as though nothing happened, this may be the cause.
Another approach that would solve this problem would be to use a linear PS (transformer+rectifier+caps) instead of an SMPS since the caps are built into it already.
Also, see page 12 for a good explanation of bus pumping:
http://www.irf.com/technical-info/appnotes/an-1071.pdf
http://www.irf.com/technical-info/appnotes/an-1071.pdf
This sounds like it may be the problem. I have a Peavey CS-800 sitting close by with a couple of big 15000uF caps. I could probably just put one of these across the switching power supply and see what happens. The class D amp board itself has 2x 2200uF caps in parallel for 4400uF at 63V rating. Wouldn't this likely be enough?
Is everything definitely wired up correctly?
I'm not sure why but with my tpa3255, I had my RCA wires the wrong way round in the jumpers (I think this was the issue, somebody was helping me out) and my amp kept powering off too. Since the wiring was fixed, it's never cut off again and I've ran it at pretty much full volume with a 48v PSU
I'm not sure why but with my tpa3255, I had my RCA wires the wrong way round in the jumpers (I think this was the issue, somebody was helping me out) and my amp kept powering off too. Since the wiring was fixed, it's never cut off again and I've ran it at pretty much full volume with a 48v PSU
What amp board is this? Photos would be helpful along with a sketch of how you have it hooked up.
Is your SMPS going into hiccup mode? If you have bad bootstrap caps or if your amp has a linear regulator to drop the input voltage to the 12v needed by the amp for its analog section, and the linear regulator is not a special one, it will not work above 35v input and the regulator will shutdown.
Is your SMPS going into hiccup mode? If you have bad bootstrap caps or if your amp has a linear regulator to drop the input voltage to the 12v needed by the amp for its analog section, and the linear regulator is not a special one, it will not work above 35v input and the regulator will shutdown.
Sorry for not posting more details earlier. This is the unit:
https://www.aliexpress.com/snapshot...d=8145071997970759&productId=1005003635917956
and Ive watched the meanwell clone with a meter an it holds a steady 48v (10a) during all operation and shutdown conditions. Failure seems to occur when clipping (preamp).
https://www.aliexpress.com/snapshot...d=8145071997970759&productId=1005003635917956
and Ive watched the meanwell clone with a meter an it holds a steady 48v (10a) during all operation and shutdown conditions. Failure seems to occur when clipping (preamp).
I have no clip indicator on the preamp, but I'd assume it's clipping on transients. It wouldn't clip with professionally recorded music since it has the transients compressed. This is with an electric bass when slapped particularly on low notes that this happens.
You will need an oscilloscope to measure your transients for sure. On the TI reference EVM amp and many clones that copy the TI reference design, the clip indicator is standard issue for this reason. You want to know why the amp shutdown. It sounds lien there is nothing wrong with the amp and you are simply playing it beyond its rated capability. Try a 54v power supply that will buy you about another 10vpp and reduce occurrence of clipping.
No I didnt. I chalked it up to a cheap aliexpress board. I also ordered a half dozen tpa3118 boards on aliexpress and none of them worked at all. I have several that work from a previous ebay order. So lesson learned, dont trust ali sellers since its harder to return fake or defective products. Theres plenty of that on ebay also, but its easy to open a case and get a refund when this happens.