LM3886 In parallel - overheating issues - Behringer Truth B2031

Thread Starter

acme

Joined Jan 3, 2018
21
Hi all. I'm hoping someone can help shed some light on an issue I have with one of my active Behringer Truth B2031 studio monitors.
These monitors use 3 x LM3886 amp ics. One 3886 for the high frequencies & two in parallel for the mids/low frequencies.
As the title of this thread says, one of the monitors overheat, it gets extremely hot, then starts "ticking" as the thermal protection kicks in & out.
Even when it is sitting idle, it overheats. I have read up on this & it "seems" to be related to the input dc offset.
Unfortunately, I'm not too sure exactly how & where to test this dc offset voltage. What I have done so far is: I removed one of the LM3886 ic's.
Then the speaker works nicely & doesn't heat up at all, but it's obviously not as loud as the other one. I have measured all the resistors related to the output of both ic's & they all measure perfect according to the schematic. I've measured for any dc on the outputs - none. I've tried to measure dc voltage on the input pins of the ic (thinking this is where I would find the offset voltage) - none. Impedance of the driver measures at 6.3 ohms.
I have attached the schematic diagram in the hopes that someone with more knowledge can take a look & possibly point me in the right direction here.
Any help will be appreciated. Thanks in advance.
 

Attachments

Thread Starter

acme

Joined Jan 3, 2018
21
So after removing the resistors R1 & R2, I measure 0.050v on ic1 & 0.125v on ic2. Could this could be where the problem lies & how would I go about balancing these voltages? I checked the feedback resistors all measure perfect, so Im not too sure where to tackle this.
 

dendad

Joined Feb 20, 2016
4,451
I 'd think 100mV difference would be ok, but maybe not.
Do you have an oscilloscope? If so, leaving R1 and R2 out, then feeding a signal in, do both Pins 3 look to be the same signals?
 

Thread Starter

acme

Joined Jan 3, 2018
21
I 'd think 100mV difference would be ok, but maybe not.
Do you have an oscilloscope? If so, leaving R1 and R2 out, then feeding a signal in, do both Pins 3 look to be the same signals?
I do have a scope, I'll check quickly
 

dendad

Joined Feb 20, 2016
4,451
If you have 2 input, differential input capability, the resultant signal should be close to zero. That is, both the same.
 

Thread Starter

acme

Joined Jan 3, 2018
21
The output of both ic's are clean & pretty much identical. Using a low level signal of 500hz on the input, I measure 1.12v rms on ic1 & 1.13v rms on ic2.
 

Thread Starter

acme

Joined Jan 3, 2018
21
Just out of interest, can yo do the same to the working channel? It could be interesting to see if the other one is within 100mV or closer.
Maybe the ICs need to be selected for a good match.
Do you mean test on the HF amp ic?
 

dendad

Joined Feb 20, 2016
4,451
Not too sure I understand, I have a basic function generator with only one output.
You can feed one amp into each scope input, invert one and add them together (if your scope allows this). The result is the difference between the signals.

That should be close to zero.
I'm off to bed now. Hopefully you can get more help if you need it.
 
Last edited:

Thread Starter

acme

Joined Jan 3, 2018
21
You can feed one amp into each scope input, invert one and add them together (if your scope allows this). The result is the difference between the signals.

That should be close to zero.
I'm off to bed now. Hopefully you can get more help if you need it.
I presume using the math function on a digital scope would be the same as using add/invert. Seems to be about 150mv difference between the 2 amp ic's.

Thank you for your help so far, it really is appreciated.
 

Attachments

Top