Thanks very much for the explanation. These amps sound perfect for my application, especially with low distortion at lower volumes.
I am in the process of converting my old Klipschorn L & R to 2 way with active crossovers. Same with my LaScala clone Center. I need a bunch of amps for this, but not a lot of power since these speakers are extremely efficient. The majority of listening is done at very low power levels, with occasional peaks.
Do you know where your distortion figures came from? I'm trying to reconcile them with the figures from post #712 which I guess only show near max output.
Thanks again.
Rod
Thanks very much for the link to the distortion chart. I guess that these figures are near or at max output as they seem very high. I sure wish we could get some actual curves of THD/Power for the T1-T4 series of amps. They would most likely show the amps in a much better light than this chart. That I think would help sales.
Rod
Answer depends on how loud you listen and on the size of the room. I am using a T2 into an 84db 4 ohm speaker in a small room and it is plenty loud for me. Sounds great too.
about T2 vs T1 into a 6 ohm load
Not having actually heard the hifimediy boards, I can only comment on similar amps from other vendors, but since they use similar chipsets their performance is likely very close...
So, in that light... with anything less than a true 8 ohm load, the higher current amps like the T2 seem to sound better, especially if you like the bass to have impact and control and still sound musical. And their midrange and hf qualities seem to be at least as good as the low-current versions.
So for the 86dB 6 ohm speakers my recommendation would probably be the T2
Note:
many of the Tripath amps have a lowish voltage gain 15X vs the more common 20X.
With less efficient speakers many of them will benefit from being driven by a buffer or pre-amp with +6 to +10 dB of extra gain, enough to drive the Tripath to full output. Otherwise you may find it won't play loud enough because the amp isn't being driven to full power. It won't be clipping, just on some music it won't get loud enough to be satisfying.
Some people change the gain setting on the amp to be more efficient, but this increases the noise level... this can be annoying if your speakers are 94+ dB efficient, much less so if your speakers are 84 or 86dB efficient!
Not having actually heard the hifimediy boards, I can only comment on similar amps from other vendors, but since they use similar chipsets their performance is likely very close...
So, in that light... with anything less than a true 8 ohm load, the higher current amps like the T2 seem to sound better, especially if you like the bass to have impact and control and still sound musical. And their midrange and hf qualities seem to be at least as good as the low-current versions.
So for the 86dB 6 ohm speakers my recommendation would probably be the T2
Note:
many of the Tripath amps have a lowish voltage gain 15X vs the more common 20X.
With less efficient speakers many of them will benefit from being driven by a buffer or pre-amp with +6 to +10 dB of extra gain, enough to drive the Tripath to full output. Otherwise you may find it won't play loud enough because the amp isn't being driven to full power. It won't be clipping, just on some music it won't get loud enough to be satisfying.
Some people change the gain setting on the amp to be more efficient, but this increases the noise level... this can be annoying if your speakers are 94+ dB efficient, much less so if your speakers are 84 or 86dB efficient!
With highly efficient speakers like Klipsch you could easily get by with the smaller Tripath sets using the 2020 or 2024 and a simple 12V battery supply.. but I like the sound of the 2050 boards better anyway.
For active systems, consider getting a Behringer DCX2496 and then get it modded for direct DAC output. The AKM 4393 DACs on that unit are outstanding when you strip out the cruddy drive circuits which are used for driving long cables in the pro audio world.
Unbelievably clean sonics, great flexibility, loads of parametric eq and crossover functions, and ability to correctly align the drivers in time... wow.. what a tool.
I can do the direct out mods or if you want to also change the DACS to the AKM4395 which are even better contact member Sendler on this site.
The distortion figures I referenced came from a T-amp shootout done on TNT-audio, they compared a bunch of them some time ago. The figures from hifimediy are a little but not significantly higher.
Another consideration about THD has to do with the prevalence of even order vs odd order harmonics and their audibility. The Tripath boards discussed here have a very transparent and neutral sound, like a combination of very good to excellent tube + solid state qualities, with very few of the drawbacks of either. True audiophile performance for peanuts.
Only, be prepared to experiment with power supplies, these amps will readily show the relative qualities of each. After extensive research and A/B comparisons I ended up with a very expensive LiFePo4 battery supply from RedWine Audio. The difference is not at all subtle.
Kind of funny to be driving $50 to $450 amps with an $800 supply, but then again, I'm getting sound that is much better than most $10K amps, so I'm not about to complain too much.
Just finished replacing the vol pot on the T1 with an Alps yesterday, a big improvement over the individual guitar quality pots I had in there. Notice steel plug where one of the pots used to be.
What do you recommend for low volume detail other than the Alps and chip types? Could use some ideas.
Thanks.
Attachments
Last edited:
So, how to find out the right source parameters for a Tripath - especially for a T2 ? For example: with a CD Player (output regulated) directly connected you cannot change much, or can do ?
Or how can I read it out from specifications of a source like Pre-Out of an integrated amp or Pre-amp if it is the right one for the Tripath ? Are the +6 to +10 dB extra gain a variable or fixed parameter of a pre-amp ?
Can it be that these 4 small white plastic switch combinations of the sure board are the gain settings to match the amp with the source correctly ?
I would go for a good stepped attenuator. I nearly was going to invest in chip types but there is no need to add more and more regulation between source and powersection.
Maybe this one should be interesting:
Valab Linear type Stepped Attenuator Potentiometer 10K - eBay (item 270723691712 end time May-20-11 07:36:57 PDT)
Gain settings
On the Sure: yes, the switches set the gain. There are 2 for each channel. When all 4 switch combinations are in the UP position on the Sure it has a lot more gain than the Helder. When it's (1 & 3) up or (2 & 4) up the gain seems to be about the same as the Helder. The Sure is pretty quiet at max gain, the Helder also. I have to get my 'scope running again to be able to measure the actual amount of gain...
Will be trying the Hifimediy T2, I hope it has more gain.
To change gain on the Helder one would have to pry up the heatsink and change R4... not what I call user friendly...and R4 is a SMD... a real PIA to change.
I would really like to encourage Helder to give his amps about 6 dB more gain in stock form and make R4 a thru-hole mount instead of SMD. Make it easier, not more difficult to mod!
About the pre-amp: I would want to to have that extra 6 or 10dB gain with an attenuator, not fixed. If the music is recorded really loud it's probably not necessary at all. But on quieter music, mastered at very low levels there are times when I really would like to have that extra gain.
It's only about 15% of the time that I feel more gain is needed, but it's enough to where I'm seriously thinking of getting a Burson Buffer (with attenuator) which has that extra 6 dB if needed.
On the Sure: yes, the switches set the gain. There are 2 for each channel. When all 4 switch combinations are in the UP position on the Sure it has a lot more gain than the Helder. When it's (1 & 3) up or (2 & 4) up the gain seems to be about the same as the Helder. The Sure is pretty quiet at max gain, the Helder also. I have to get my 'scope running again to be able to measure the actual amount of gain...
Will be trying the Hifimediy T2, I hope it has more gain.
To change gain on the Helder one would have to pry up the heatsink and change R4... not what I call user friendly...and R4 is a SMD... a real PIA to change.
I would really like to encourage Helder to give his amps about 6 dB more gain in stock form and make R4 a thru-hole mount instead of SMD. Make it easier, not more difficult to mod!
About the pre-amp: I would want to to have that extra 6 or 10dB gain with an attenuator, not fixed. If the music is recorded really loud it's probably not necessary at all. But on quieter music, mastered at very low levels there are times when I really would like to have that extra gain.
It's only about 15% of the time that I feel more gain is needed, but it's enough to where I'm seriously thinking of getting a Burson Buffer (with attenuator) which has that extra 6 dB if needed.
That looks like a good pot! Only, 25 USD with Vishay SMD resistors? This could be good! I'm getting one!
Not linear please
You need Log scale I have their log 25k stepped pot with T2. It's excellent quality.
Last edited:
about gain:
T1 and T2, powered up to 30V, have a 15v/v gain.
T1 510A, powered up to 39V, has gain of 20.
T3 and T4 I am not sure, but guess it is higher.
Gain setting has to do with amp power supply voltage. The tripath tk2050 datasheet has details.
A typical consumer market CD player, DAC or preamp has a 2Vrms output. Gain of 15V/V for an amp powered at 30V is good fit. Amp will even clip at very high level.
If you use a source like netbook motherboard output, mp3 player or ipod with output limited to 1Vrms, you can not reach max amplifier output power. If you want to, you need a preamp -or better- change for a better source !
On their board, Sure has switches to set gain. The price to pay is more noise. I'd rather change for a better amp than the Sure and better source than an mp3player.
T1 and T2, powered up to 30V, have a 15v/v gain.
T1 510A, powered up to 39V, has gain of 20.
T3 and T4 I am not sure, but guess it is higher.
Gain setting has to do with amp power supply voltage. The tripath tk2050 datasheet has details.
A typical consumer market CD player, DAC or preamp has a 2Vrms output. Gain of 15V/V for an amp powered at 30V is good fit. Amp will even clip at very high level.
If you use a source like netbook motherboard output, mp3 player or ipod with output limited to 1Vrms, you can not reach max amplifier output power. If you want to, you need a preamp -or better- change for a better source !
On their board, Sure has switches to set gain. The price to pay is more noise. I'd rather change for a better amp than the Sure and better source than an mp3player.
Thanks for your tips. First I wanted to write : "if you want louder than make louder"
After reading some opinions and reviews it makes not only sense regarding vol. gain to settle on a buffer. There could be heavier impedance mismatches in every system ...
Last edited:
Not linear please
I have their log 25k stepped pot with T2. It's excellent quality.
Thanks for the hint ...
oops, I forgot to mention in my previous post that tc2000 (the controller part of tk2050) input sensitivity is 1.5Vrms
see table page 3 of datasheet http://www.datasheetcatalog.org/datasheet/Tripath/mXyzwqtx.pdf
gain setting and input stage details page 12 & 13.
see table page 3 of datasheet http://www.datasheetcatalog.org/datasheet/Tripath/mXyzwqtx.pdf
gain setting and input stage details page 12 & 13.
Gain may be different
The gain referenced on the Tripath data sheet is the "standard", and a 1.5V input would get 20X multiplication, i.e. 1.5V would produce 30V at the output, so definitely it gets full voltage swing. But Helder and Hifimediy appear to have reduced the gain on their amps, and a 1.5V input to them would produce only 22.5V, less than full voltage swing.
If as you posted earlier on the HifimeDIY T2 the gain is 15X, then you need a full 2V to get full drive, i.e. the T2 is 2.5 dB less sensitive than the "stock" Tripath.
Like Helder, Hifimediy have probably reduced the gain to get less self-noise at the output. So for these implementations, extra gain may be required, especially if you have inefficient speakers, or as in my case are running with significant eq that reduces the midband level (I prefer subtractive EQ rather than boost, especially at HF).
Also, Alkasar, the assumption that only poor quality devices are being used is not entirely accurate !
I am using a Behringer DEQ2496 which has been modified to bypass the output section and get direct DAC output, and I can assure you, this produces VERY high quality sound, albeit with a somewhat lower drive level.
The direct out DAC on these units is as good or better than many of the top DACs available (clearly much better than the Benchmark for example)... but it may need more drive when used with low gain amps.
Something like that Burson Buffer would work well with its extra 6dB of gain. And, yes, the buffer really does reduce the issues arising from insufficient drive and drive vs load impedance mismatches.
BTW, while the Sure runs overly hot and does not appear have the same quality parts as the Helder, in stock form I like its sound better. I had higher hopes for the Helder.
Let's see how they compare when modded !
The gain referenced on the Tripath data sheet is the "standard", and a 1.5V input would get 20X multiplication, i.e. 1.5V would produce 30V at the output, so definitely it gets full voltage swing. But Helder and Hifimediy appear to have reduced the gain on their amps, and a 1.5V input to them would produce only 22.5V, less than full voltage swing.
If as you posted earlier on the HifimeDIY T2 the gain is 15X, then you need a full 2V to get full drive, i.e. the T2 is 2.5 dB less sensitive than the "stock" Tripath.
Like Helder, Hifimediy have probably reduced the gain to get less self-noise at the output. So for these implementations, extra gain may be required, especially if you have inefficient speakers, or as in my case are running with significant eq that reduces the midband level (I prefer subtractive EQ rather than boost, especially at HF).
Also, Alkasar, the assumption that only poor quality devices are being used is not entirely accurate !
I am using a Behringer DEQ2496 which has been modified to bypass the output section and get direct DAC output, and I can assure you, this produces VERY high quality sound, albeit with a somewhat lower drive level.
The direct out DAC on these units is as good or better than many of the top DACs available (clearly much better than the Benchmark for example)... but it may need more drive when used with low gain amps.
Something like that Burson Buffer would work well with its extra 6dB of gain. And, yes, the buffer really does reduce the issues arising from insufficient drive and drive vs load impedance mismatches.
BTW, while the Sure runs overly hot and does not appear have the same quality parts as the Helder, in stock form I like its sound better. I had higher hopes for the Helder.
Let's see how they compare when modded !
it's 1.5V rms, not peak. Also, the full 30V power supply voltage is not available at the output.
I don't get it. Input sensitivity is the same for all boards based on tc2000.
Gain setting depends on a few resistors around it. It's a designer choice.
Tripath standard design has 15V/V gain. Helder amp as well (according to schema). Same for hifimediy T1/T2. Makes sense. For the higher power chips, Hifimediy adjusted the gain accordingly.
They both set the gain according to specs for this voltage. Sure decided to provide gain setting. The main reason they gave is to get full power from the low level output mp3players around. Fair enough to adress a large market that I would not qualify as HiFi
misunderstanding here
dcx2496 is excellent and I of course agree for the role of a buffer. Just make sure that the extra 6dB will not result in clipping at full volume
Hi Jack,
Thanks very much for all the good information.
We actually have a Behringer DCX2496 that we've been using for EQ with our Danley DTS-10 sub. It has worked okay for that but can not do a HPF below 20 Hz. We'd like one around 11 Hz. My recent plan for actives is to get the Mini DSPs. The new 2 x8 looks really good but is a bit pricey. It will take two of these for our system, but that would allow us to have 3 way on all three fronts if we should want in the future. It will also allow the high pass filter for our sub. And it has both balanced and unbalanced inputs/outputs. Also has very user friendly PC programming.
Thanks you for your offer of doing the mods on our DCX. If the Mini DSPs work out as we hope I will probably sell our DCX.
Thanks also for the further info on distortion.
Very interesting on the power supply issue. I sure do hope that we will be happy with something way less expensive: especially considering that we will have several amps and I am figuring a separate supply for each, ie for each two channel amp. But thanks for commenting that the supply differences were not subtle. So often these kinds of changes are.
Cheers,
Rod
Alternate way of HPF on DCX
I presume you are doing this to limit subsonic excursions on the sub?
It might be worth mentioning that if you do the direct out mod on the DCX 2496, you can use smaller coupling caps on each leg of the DAC output, thereby creating a 1st order rolloff at 11Hz if you choose.
OR... use an even smaller value to create HP at 25Hz, then add in a little boost on the PEq section to bring it back up near flat at 20Hz, and voila, you have a pretty good HP filter...
No need to spend big$$ on those mini-dsp units when you've already got a diamond in the rough in the form of the DCX... just mod it, it's easy. Look up the pinouts for AKM4393, trace the DAC pinouts to where the connectors for the ribbon cables protrude from the bottom of the board, connect flying leads to the signal (+) line and signal (-) line, cut those lines on the ribbon to bypass the gazillion op-amps in the output circuitry....
Then connect coupling caps to each (both signal lines from the DAC have +2.5VDC offset) and now you have direct out lines that feed into the amps.... No prying up the smd legs on the DACs or cut traces on the board, just connect to protruding pins... see? It's easy ! no need to ship it or pay me to do it!
BTW the quality of the bass on the modded DCX is stunning.
I presume you are doing this to limit subsonic excursions on the sub?
It might be worth mentioning that if you do the direct out mod on the DCX 2496, you can use smaller coupling caps on each leg of the DAC output, thereby creating a 1st order rolloff at 11Hz if you choose.
OR... use an even smaller value to create HP at 25Hz, then add in a little boost on the PEq section to bring it back up near flat at 20Hz, and voila, you have a pretty good HP filter...
No need to spend big$$ on those mini-dsp units when you've already got a diamond in the rough in the form of the DCX... just mod it, it's easy. Look up the pinouts for AKM4393, trace the DAC pinouts to where the connectors for the ribbon cables protrude from the bottom of the board, connect flying leads to the signal (+) line and signal (-) line, cut those lines on the ribbon to bypass the gazillion op-amps in the output circuitry....
Then connect coupling caps to each (both signal lines from the DAC have +2.5VDC offset) and now you have direct out lines that feed into the amps.... No prying up the smd legs on the DACs or cut traces on the board, just connect to protruding pins... see? It's easy ! no need to ship it or pay me to do it!
BTW the quality of the bass on the modded DCX is stunning.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.