Post by Graham on Aug 26, 2019 15:23:38 GMT 12
A few week ago my little Polk Audio subwoofer that I use only for home theatre died for the second time. Just over 2 years ago I had to replace the plate amp as it started making a heartbeat type sound and then started buzzing. It has done it again. Searching on line I find this is a common failure as several caps on the board are barely up to the task, but that replacing or upgrading these is near to impossible because of restricted space on the tiny board. So I figured I would try a different approach as most of the features and adjustments on the original plate amp are not necessary for my application as I feed the amp from the LFE output from the Denon AVR which takes care of crossover duties internally. I bought off ebay a little 100 amp class D mono amp ( a whole $30 NZ !!) . The input to this is fed from the LFE output and I have powered it with an ex laptop 19v DC 3.5 amp switchmode ecu. The amp specs say 12 - 24 volts DC. After drilling a hole through the original plate amp and fitting a grommet the speaker wires are connected to the 2 speaker posts. I find excellent volume and performance are achieved using only about 1/3 volume on the little amp and it sounds equally as good as the original plate amp for movies and even the few music dvd discs that I have. Sorted  
 
|
Post by colinf on Aug 26, 2019 19:08:20 GMT 12
Nice one! 😀 Yes the caps on some subwoofer amps seem to go prematurely IME. Especially the ones that control the mute relay circuit and power supply for the amp section. Some electrolytic caps in the 90s and 2000s suffered from an incorrect formula for the electrolyte (remember computer motherboard failures?) so they dried out and sometimes even shorted out internally and caused loads of problems.
AMR-iFi R&D
|
Post by Owen Y on Aug 26, 2019 21:06:48 GMT 12
A few week ago my little Polk Audio subwoofer that I use only for home theatre died for the second time. Just over 2 years ago I had to replace the plate amp as it started making a heartbeat type sound and then started buzzing. The 'heartbeat sound' could probably be what's called 'motorboating' - an oscillation due to deteriorated and/or insufficient PSU capacitance, say.
|
Post by Graham on Aug 26, 2019 22:29:23 GMT 12
Yep the capacitors that are known to fail are on the PSU circuit board.
I investigated replacing them but it looked like a major mission just to get at them, so opted for a simpler and hopefully more reliable fix.
You must know by now my motto is KISS (keep it super simple) !!
|
Post by colinf on Aug 27, 2019 3:07:25 GMT 12
Replacing the original caps could be a longer term goal perhaps?
AMR-iFi R&D
|
Post by Graham on Aug 28, 2019 19:41:28 GMT 12
Yeah Naah. I probably should but I doubt if I'll bother.
The little class D amp performs just a good as if not better than the original, and as I said I don't need the speaker in and out terminals, the crossover, or the phase switch.
The little amp uses the well regarded TPA3116D2 chip with an NE5532 opamp buffer.
|
Post by colinf on Aug 30, 2019 18:59:42 GMT 12
It’s quite a powerful chip for its size, needing only a small heat sink to pump out almost 100w. Class D amps have come a long way in the last 5 years or so. They now have things like AM interference avoidance by continually changing the switching frequency, anywhere form 400kHz to over 1MHz! I use a class D amp in my 301 turntable power supply. Is your woofer 4 ohms?
AMR-iFi R&D
|
Post by Graham on Aug 30, 2019 21:52:42 GMT 12
Yes Colin, it is a 4 ohm 8 inch speaker.
|
Post by colinf on Aug 31, 2019 7:00:54 GMT 12
Ok, at 19.5v power supply voltage you could expect about 40 or so watts at less than 0.1% thd (according to the data sheet). That’s most likely ample at LFE frequencies and as you say, works fine!
AMR-iFi R&D
|
Post by Graham on Aug 31, 2019 11:54:53 GMT 12
In this amp the TPA3116D2 is configured in bridged mode so 50+50 on 24 volts so maybe with only 19 volts it could produce around 80 watts.
However I find plenty of volume with 1/3 to 1/2 on the volume knob.
I couldn't resist having a peek inside and the IC has a huge heatsink that hardly gets warm after running for hours.
|
Post by colinf on Aug 31, 2019 18:40:34 GMT 12
The most a 4 ohm load can draw at 19v in bridged mode is about 45w. At low distortion and allowing for amp inefficiency I’d expect around 40w. Remember the class D amp has its output rails sitting at half supply voltage (9.5v in this case). That’s an RMS output voltage of 13.44v in bridged mode, and into 4 ohms that’s 45w. Unless your amp generates a negative power supply inside as well, to complement the +19v from the laptop power supply, in which case the power supply might be + and - 19v. Is that the case?
AMR-iFi R&D
|
Post by colinf on Aug 31, 2019 19:53:55 GMT 12
The first line of the 3116 data sheet reads: “The TPA31xxD2 series are stereo efficient, digital amplifier power stage for driving speakers up to 100 W / 2 Ω in mono.” The 100w rating is bridged at 2 ohms, not 4. The maximum supply voltage is 26v, so a + and - power supply can’t be used.
AMR-iFi R&D
|
Post by Graham on Aug 31, 2019 21:05:07 GMT 12
Wow, if its only producing 40 watts that makes the performance even more amazing.
I guess it shows how little of the available power we use for normal listening levels in all types of speakers.
|
Post by colinf on Aug 31, 2019 21:52:19 GMT 12
Yes that’s what Nelson Pass was on about when he said most of the power needed to reproduce instrumental harmonics at normal volume is only around 1 watt, depending on speaker efficiency. Hence his company is called First Watt. He has an oscilloscope connected to his speakers while listening and can see how much power is being used at normal volume. The peaks are much higher than 1 watt of course. I guess that’s why there’s an explosion of high quality low powered amps these days. I’ve put an oscilloscope on my speakers as well before and the Focals draw about 60w on drum peaks at high-ish volume.
AMR-iFi R&D
|
Post by Owen Y on Sept 1, 2019 17:16:20 GMT 12
The amplifier should run AT LEAST 20dB attenuated, ie. no more than 1/10 power, in order to have enouigh power available for dynamic swings, dynamic peaks. Rock & Jazz might contain only 10-15dB dynamic peaks (3-6x) above average SPL, but Classical crescendoes are frequently 20-25dB (10-18x) above average SPL. Especially the LF end, which is where music reproduction is most demanding of watts. Would that be about right, colinf?
|
Post by colinf on Sept 2, 2019 4:43:10 GMT 12
You could get away with a low power amp above 200Hz in the 5 to 10 watt range assuming an about 90dB/w/m speaker, and you’re not really caning it. The bass can use much more power, say 20 to 100w. But pianos and drums take lots of peak power in the mids to reproduce without clipping if you’re playing at highish volume, so can use much more than 10w, and even as much power as a woofer. The tweeter is the one where you can use a low power amp most of the time. Cymbals have a lot of upper midrange, and only a relatively small amount of power is needed above 10kHz. But the more efficient your speaker the less overall power you need. I remember hearing the Tannoy Westminsters (99dB/w/m) driven by an Audionote Otto, only 8w single ended EL84 amp. They played John Cougar Mellencamp. It was awesomely dynamic and alive, and almost at rock concert volume! I couldn’t believe it was only 8 watts.
AMR-iFi R&D
|
