These new chips seem very impressive but I am skeptical that Topping, Sure, Dayton, 41Hz and the numerous other Chinese companies will use them to replace their Tripath amplifiers. Texas Instruments has been making superior chips for a few years now and they have mostly gone widely ignored. I would suggest that some new business firms launch new amplifiers and maybe knock some of these older firms out of business. There are lots of opportunities to make money and it is just a matter of going to the trouble to get it done.
41hz is Swedish and are made in Sweden exclusively.
But I disagree. TI has made some good chips in the past 5 years but they were still lacking behind the superior Tripath chips in most areas. That is the main reason they haven't been used as widely until now in the DIY market.
Remember, to TI, the DIY market is a niche market. Most of their class D chips go in well-known brand name TVs, recievers, and so on. For Tripath chips the DIY is the only market as the company went bust over 5 years ago.
But I disagree. TI has made some good chips in the past 5 years but they were still lacking behind the superior Tripath chips in most areas. That is the main reason they haven't been used as widely until now in the DIY market.
Remember, to TI, the DIY market is a niche market. Most of their class D chips go in well-known brand name TVs, recievers, and so on. For Tripath chips the DIY is the only market as the company went bust over 5 years ago.
I would not call the 10 year old Tripath chips market DIY only because they are still used pretty much exclusively in all Topping and similar small class D amplifiers on the market which are quickly becoming obsolete. The opportunity to replace these poor sounding amplifiers with much improved sounding lower cost units does exist and the first business firms to bring them to market will reap the benefits.
Topping, Sure and all the other small amp companies are peanuts for TI. They don't build and design chips with this market in mind at all.
I don't know yet if TI is finally going to match Tripath in sound quality, so far all their attempts have been inferior. I see great potential in the TPA3116D2 though. Might still not be up to Tripath in sound quality but on other parameters it's an improvement, most significantly in efficiency at real music levels.
Last edited:
Exactly. The only commercially available (in the US) TPA-based amp currently is the TBI Millenia for $500. It lingered in obscurity for the past few years, then an audiophile reviewer got a hold of it and now it is getting rave reviews from people with $$$$ amps, including Pass amps. Not that I'm saying the TPAs are that good, I don't own $$$$ gear - but the TPA chip has now put TBI on the map. And it's using the older TPA3100D2, not the newer 3116/8.
If audiophiles want to spend $500+ that's fine, but I'd rather have something much cheaper and tweakable. Enclosures are for wimps anyway
@ Saturnus: So far I have yet to read of someone listening to a TPA-based that did not think it equal if not better than a similar powered Tripath. Not that that means anything. Just my experience so far across different forums. I have heard enough Class D amps (except Hypex) to be genuinely excited about what I hear from the TPA chips.
Last edited:
Wushu,
It's interesting that when I read the review of the TBI Millenium it was after I had written my own impressions and noticed the main thing they noticed too: the resolution and micro details that you hear in your favorite CD's, the spatial imaging (or holographic), and the neutrality of the amp. I am not surprised that they compared it too really high end amps costing as you say, $$$$$ and for audiophiles, $500 seems like a steal. I think this amp will be in the $100 price point when packaged with a case, binding posts, etc. I think it can be made very compact - which is not always a good thing because people equate massive size with how good it sounds. After reading the review on the TBI Millenium I don't feel so crazy to think that this is one of the best sounding amps I have ever heard.
It's interesting that when I read the review of the TBI Millenium it was after I had written my own impressions and noticed the main thing they noticed too: the resolution and micro details that you hear in your favorite CD's, the spatial imaging (or holographic), and the neutrality of the amp. I am not surprised that they compared it too really high end amps costing as you say, $$$$$ and for audiophiles, $500 seems like a steal. I think this amp will be in the $100 price point when packaged with a case, binding posts, etc. I think it can be made very compact - which is not always a good thing because people equate massive size with how good it sounds. After reading the review on the TBI Millenium I don't feel so crazy to think that this is one of the best sounding amps I have ever heard.
I did a switch between the 3116 and 3118 listening to the same track with same 65 watt power supply, same CD player volume setting, and I am noticing some minor differences: obviously the power output is different for the same input setting (determined by voltage divider split on gain pins) the 3118 is lower in volume. It required me to move the volume knob from 10 o'clock to 12 o'clock to get about the same volume as I got on the 3116. There is a slight difference in the power caps between the two (560 uF vs 470 uF on the 3118), I think it might make a bit of difference because I like the bass on the 3116 better - a little more powerful. But in all other regards, once the power output was adjusted to be equivalent (by ear) both amps were very similar in their fantastic detail and ability to bring out micro dynamics and wide sound stage with pinpoint spatial imaging capability. Right now, I am going to listen to the 3118 for a few more days to see if I can hear any more differences. I just realized that I now have two very similar and well matched amps suitable for driving a bi-amped FAST. I am thinking of using TI's ultra low distortion audio opamp quad package (LME49740) to make a nice preamp for buffering, splitting signals, and active high pass and low pass filters prior to driving the 311x amps. Let's me skip that passive crossover business. I think a home built active crossover with quality components is a lot less expensive than components for a passive crossover.
Last edited:
I am wondering the same thing but at present I don't have anything better. I think some Mark audio Alpair 7's may be one of my next driver purchases but I will still build my own cabinets. I have an old pair of Boston Acoustics HD5's but the foam on the surround has long since disintegrated and need replacement. Not sure if that can beat my Vifa's in the FH3i enclosure. They can get louder no doubt but the imaging and detail of the Vifa's is superb.
I'd highly recommend high sensitivity full-range speakers like Fostex for these small single power class D amps. It really pulls out the very best of the speakers. Failing that, go 2-way or more but with active filtering. Class D does not seem to like passive filters very much and perform much better when coupled directly to the drivers.
With high sensitivity speakers also means you don't need a lot of output so you can use solar charged batteries for your whole set-up. Although not noise free, batteries, especially lifepo4 types, does seem to a be much better power supply than even the very best and most optimized grid connected power supply.
With high sensitivity speakers also means you don't need a lot of output so you can use solar charged batteries for your whole set-up. Although not noise free, batteries, especially lifepo4 types, does seem to a be much better power supply than even the very best and most optimized grid connected power supply.
Saturnus,
I agree on full range drivers although with 20 to 30 clean low THD watts on hand I don't see need for high sensitivity (90 dB and above) drivers. Not sure why class D is inherently not good with passive crossovers - although I don't like them myself and prefer DSP or active preamp crossovers with bi-amping instead. I think as long as a full range driver is above 85dB I will consider it. There are lots of choices in the 88 to 89 dB range. What drivers do you personally use with your amps? Do you use the 41Hz amp6?
I agree on full range drivers although with 20 to 30 clean low THD watts on hand I don't see need for high sensitivity (90 dB and above) drivers. Not sure why class D is inherently not good with passive crossovers - although I don't like them myself and prefer DSP or active preamp crossovers with bi-amping instead. I think as long as a full range driver is above 85dB I will consider it. There are lots of choices in the 88 to 89 dB range. What drivers do you personally use with your amps? Do you use the 41Hz amp6?
Fostex FE206 in back loaded horns. Yes, I use an amp6.
Ideally completely passive filtering would be preferable with class D amps. However, you'd most likely be restricted to 2nd order filters, and the amp has to be built for the filter. You "simply" replace the active filters with a pllxo input instead of the input cap for the HP, and replace the output filter with one optimized for the desired LP xo. That way you have a 2-way system to drive each speaker driver directly without need any additional stages beyond those that are already present in the amp.
Also note that using a pllxo for the HP section can always be used in combination with an active LP section. That limits the required amount of op-amp to 2 (if you're want 2nd order filters that is).
Ideally completely passive filtering would be preferable with class D amps. However, you'd most likely be restricted to 2nd order filters, and the amp has to be built for the filter. You "simply" replace the active filters with a pllxo input instead of the input cap for the HP, and replace the output filter with one optimized for the desired LP xo. That way you have a 2-way system to drive each speaker driver directly without need any additional stages beyond those that are already present in the amp.
Also note that using a pllxo for the HP section can always be used in combination with an active LP section. That limits the required amount of op-amp to 2 (if you're want 2nd order filters that is).
Last edited:
What is a pllxo? Passive crossover? My normal understanding of pll is phase locked loop and not sure how that applied here. What you describe is interesting and reducing opamp stages is nice so you have a link or circuit schematic that you can point me to? One advantage of active low pass and active high pass is ability to put a potentiometer and make it variable. The TI LME49740 seems very clean with THD in the 0.00005% range so I don't think having 3 or 4 stages of opamp will be too bad on sound quality. Although purists will say one should not have anything more than 3 or 5 discrete transistors in the signal path and no capacitors . Once I have embraced claasD I have no qualms of adding a few more stages of opamps. After all, most class D consists of an input buffer/gain opamp followed by a voltage comparator opamp that performs the PWM via comparison with a sawtooth waveform.
The FE206'a are nice drivers in a BLH. You don't need more than 3 watts to get very loud levels beyond comfort. I have heard many good things about Amp6 but wonder if anyone out there has heard both the Amp6 and TPA3116d2 to compare?
. Once I have embraced claasD I have no qualms of adding a few more stages of opamps. After all, most class D consists of an input buffer/gain opamp followed by a voltage comparator opamp that performs the PWM via comparison with a sawtooth waveform. The FE206'a are nice drivers in a BLH. You don't need more than 3 watts to get very loud levels beyond comfort. I have heard many good things about Amp6 but wonder if anyone out there has heard both the Amp6 and TPA3116d2 to compare?
This is much more subjective but I did find experimenting with the two power supply caps with different capacities and brands to have an audible impact, much more so than other amps. I settled on 2200uf Nichicon KG, but every time I swapped those caps (I tried out Elna Silmic, Panasonic FM, FC, KA) the presentation changed. It was quite audible and was done with both the 3100D2 and 3110 boards to same effect.
So you are saying that even at 2200 uF there was a difference? Wow, that is like 10x what they spec for their EVM boards. At that size capacitor the amplifier size is not as miniature as I would like. I am thinking of doing an SMD board with qnty 6 per side (16 total) SMD 100 uF caps per side to keep profile low. Wondering if I ought to try adding caps to see if I can hear a difference?
That's been my experience. It was not subtle. Caps are so cheap, it's worth trying out.
Maybe you could mix and match?
I was originally thinking this would be optimal:
2x Nichicon 1000uF (UVY1E102MPD1TD)
1x Rubycon 470uF (25YXF470MEFC10X16)
That's for each channel btw. So about 5000uF total capacity.
Arranged in a T-shape power plane going to the power pins on either side with the Nichicons on either side of the Rubycon.
I was originally thinking this would be optimal:
2x Nichicon 1000uF (UVY1E102MPD1TD)
1x Rubycon 470uF (25YXF470MEFC10X16)
That's for each channel btw. So about 5000uF total capacity.
Arranged in a T-shape power plane going to the power pins on either side with the Nichicons on either side of the Rubycon.