So i decided to dig into a KRK V8-II set that i bought almost 10 years ago. One of the monitors broke after a few months only, but KRK didn't even answer my emails back then.
Turns out there is a triple LM3886 setup, with what appears to be a parallel config for the woofer. All fair enough, but the output resistors are <1 and ~660k ohm respectively. Why would you use such a setup? Shouldn't they generally be well matched and small?
Turns out there is a triple LM3886 setup, with what appears to be a parallel config for the woofer. All fair enough, but the output resistors are <1 and ~660k ohm respectively. Why would you use such a setup? Shouldn't they generally be well matched and small?
I opened the other speaker, turns out both resistors should be .22, so it appears the issue has been found.
~660 kΩ is not a standard value. 680 kΩ is. Is the resistor marked as 680 kΩ or did you measure ~660 kΩ? If the latter, could it be that the resistor used to be <1Ω (whatever that means) and is now ~660 kΩ because it's dead?
Tom
Tom
Burned resistors usually go open circuit rather than high value unless your meter considers 680k as open circuit ?
Yeah, but who knows what else is in parallel with that resistor that the ohmmeter might pick up. If the resistor was de-soldered prior to measurement, I'd agree with you.
Tom
Tom
I see your thoughts here, but measuring output res of a parallel coupled amp at 660 k gives an indication, reason to wonder and also turned out to be the issue in the amp. I thought maybe there would be a reason to set up the circuit in such a way that i couldn't think of, and of course i'd like to look into that before ordering spare parts =)
<1Ω means less than 1 ohm.
<1Ω means less than 1 ohm.
I understand what "less than" means. <1Ω still does not convey much information. 0.99Ω? 1mΩ?
The 660 kΩ you measure across the burnt out resistor could be the resistance of everything that's in parallel with that resistor at that point. Your meter will be sending current through the 0.22Ω resistors back through the ESD protection within the LM3886es, etc. If you want a true measurement, you'll need to de-solder it.
However, as you've already identified the part as burnt out, it would be pointless to de-solder it just to measure it.
Before you just replace the burnt out resistor, I think it'd be good to figure out why it fried. Did the LM3886 feeding it die?
Tom
The 660 kΩ you measure across the burnt out resistor could be the resistance of everything that's in parallel with that resistor at that point. Your meter will be sending current through the 0.22Ω resistors back through the ESD protection within the LM3886es, etc. If you want a true measurement, you'll need to de-solder it.
However, as you've already identified the part as burnt out, it would be pointless to de-solder it just to measure it.
Before you just replace the burnt out resistor, I think it'd be good to figure out why it fried. Did the LM3886 feeding it die?
Tom
It needs powering up with a variac or light bulb in series with the mains.
And strictly no speaker connected.
Measure output for DC offset.
If no DC offset it might be OK.
OP never said the 660K "burned".
It seems more likely the 3-amp board failed because of a resistor error throwing all the load on 2 amplifiers.
Not really. The LM3886 is pretty rugged. It also has features such as over-current detection, SOA limiter, and over-temperature shutdown.
Populating 680 kΩ instead of a 0.22 Ω power resistor does not happen in production. Having the board pass with 680 kΩ populated rather than 0.22 Ω, is equally unlikely. I think you can throw that theory out.
OP has not answered any questions about the markings on the resistor. If it's actually marked 680 kΩ, I'll eat my hat. I bet it's a 0.22 Ω that's gone open circuit.
Tom
Populating 680 kΩ instead of a 0.22 Ω power resistor does not happen in production. Having the board pass with 680 kΩ populated rather than 0.22 Ω, is equally unlikely. I think you can throw that theory out.
OP has not answered any questions about the markings on the resistor. If it's actually marked 680 kΩ, I'll eat my hat. I bet it's a 0.22 Ω that's gone open circuit.
Tom
The other res is .22 Ω, same as both in the functioning amp. Given the context, the discrepancy was the point, but yes, i see what you are getting at. How do you read a resistor starting with black black black, just 0 Ω ? It measures 1.3 M out of the board..
Not gonna spend much time on the project, as it turns out the woofer is completely fried, but would be fun to figure out where the error was.
- Status
- Not open for further replies.