On my QA450 (pcb RevI) I’ve blown and needed to replace one of the power resistors on a couple of occasions. I’ll quite happily take some advice on best FFT / sample rate to avoid this when running the THD v.s. Output Power script.
Anyway during removal I ham-fistedly overheated and damaged U4 so could someone please advise what U3/U4 are these as they’re not marked?
The temp sensors are Microchip TC77-3.3MCTTR in SOT23-5 package. The resistors are PWR163 series, 2 ohms each. These are special “pulse withstanding” resistors that can tolerate extremely high power for short duration, so don’t replace them with a substitute.
The QA450 manual has a detailed thermal discussion starting on page 12, and it also walks through building a thermal test budget. In short, if you don’t exceed 200W, the QA450 should be able to self protect if it gets too warm by opening the relays (60C is the cutoff–beyond that the loads are open). If you hit with 800W, though, the resistors will overheat and open before the temp sensor sees what is happening.
Finally, a PCB IR pre-heater can make removing the resistors quick. Use the pre-heater on the bottom to warm the area under the resistors to to about 100 to 150C, and then used localized heat on the top to bring the load resistor to 250C, and then pluck damaged R off with tweezers.
Thanks Matt, especially the preheat tip. Using hot air only was painfully slow hence damaging the TC77-3.3MCTTR. I’m from the through hole era not this new fangled smt one 
When running the thd vs power script for say a <80W amp, where would you recommend I set fft, sample size and db increment? Is it ok to regularly thermally trip the QA450?
It’s the second time I’ve burned out a load resistor (yes I’m using like for like replacements) so must be doing something wrong 
Hi @SABristol, yes, as long as your power is under 200W, you can trip the QA450 as much as you like it should fully self protect no matter how long the FFTs are.
Are you saying that at 80W you experienced a burned up load resistor?
The link below looks at FFT size versus THD. Take a look at the video at the end of the post to see the fast testing–55 points in 6 seconds.
Yep - I burned the first resistor when testing an Luxman L85v rated at 80W (well actually achieves nearer 100W), and the second and different resistor when testing a Cambridge Audio A5 which is only rated at 60W pc.
I wasn’t doing anything silly on either occasion and have never tested anything beyond the Luxman’s 100W. I usually keep the QA450 always on top to keep an eye on the temps too.
Thanks for the link to FFT sizing.
Hi @SABristol, below is a video showing the QA450 getting hit with 175W. Keep in mind this is just about worst case (the amp used is at its limit). Ideally, you’d want to pick an FFT size that takes expected power handling into account per the manual.
In the video, at t=6 seconds, the Class D amp is ready to go, but no loads are connected. the QA450 resistors are at ~50C each. This is a small FFT that is hammering away as fast as it can.
At t=9, the QA450 load gets engaged and thus the QA450 is seeing 175W into 8 ohms.
At t=11 seconds, note the power out drops suddenly. That is because the loads exceeded 60C and the QA450 disconnected the loads for protection.
Then there’s more cycles of waiting for it to reach below 60C (the cutoff) and as soon as both channels are below 60C, the load is re-engaged. And each time it’s uneventful and predictable. This test runs for 5 cycles in the video, but it can do this for for enough cycles to consider it indefinite.
Were you routinely hitting the thermal limit? These resistors are designed to operate for 1000 hours at 155C. But the QA450 should ensure they don’t get anywhere near that.
Let me know if you were seeing something different in your QA450. I’m genuinely curious how a 100W amp could cause a resistor to open.
Hi @matt
I’ve never used my QA450 for anything as brutal in your video. I worked on the basis that tripping the dummy load is a bad thing and something to be avoided. To that end I usually do longer term FR and THD tweaking with no more the 1-2W in to 8R. Usually I only trip it when I’ve done something like the thd v.s. power using too long a FFT.
I’ve wrote to support last week on this topic as I wasn’t sure the forum was the best place for this conversation.