I had that info about advanced Delta Sigma modulators when I was trying to help out at Virtue, so I had some doubts about using Tripath technology for new designs. That's why I designed that "Tripathogee" amp, where I could switch between modulators using the same power supply, same output devices, same LPF, etc. That was a large undertaking, with a new circuit board design, lots of software for the micro to control the switching logic and relays, and expensive high quality parts. But I wanted to be able to LISTEN to the difference in a fair, well controlled blind test. I didn't state the difference as being too noticeable in that diyAudio post because I knew the results could irritate people. However, the difference was really fairly large, and you could switch back and forth and easily guess which modulator was which. The Tripath sound wasn't offensive, and I could see why people would like it, but it also lacked the clarity and "neutralness" (if that is a word) that I like in good analog amps. It's worth noting that this wasn't a test between the Tripath modulator and newer technology, because the Apogee modulator was developed in 1996 and released in 1999!
I think the SSM3XXX amps have that same clarity and neutral sound of a good analog amp such as a properly done LM3886. I haven't done any blind testing for the sound, but in every project I've used them they seem clear and have been a good choice for high quality drivers. Plus, they have good specs, for those who like to compare specifications
.The SSM3XXX amps also have very good efficiency even at low volumes, and they use a spread-spectrum modulation that means you don't have to synchronize the chips to a common clock. It also means you can eliminate the output filter for some applications. For that 22-amp board that I made for the line array speaker, there were no issues with heat dissipation or chip interaction.
Last edited:
Neil, how many different listeners auditioned your modulator test rig way back when? I recall you saying your son preferred the Tripath sound. Aren't young ears generally better than old? HA!
I just went to the Tripath ap note AN1, as I recall them claiming their modulator "adapted" by "learning" the switching characteristics of the output transistors! In the absence of a feedback loop, I wonder how they can accomplish that? The block diagram does not show a FB loop. If indeed there IS a FB loop, I do not see an input for that on their chipsets that incorporate a separate modulator and discrete output stage, as your test rig was fashioned. Is it possible that their fully-integrated chips, such as the TA2020 - do in fact "learn" the output stage and adapt, whereas the separate modulator config does not? Or are we just reading marketing department gobbledygook?
Tripath also claimed their modulator switching frequency was "spread spectrum-like" and adaptive in nature, maxing at 1.5 MHz, but averaging 600 kHz.
This is a snip from their AN1:
"Instead of using PWM, Tripath Class-T amplifier processors use proprietary algorithms and techniques to create the modulation that drives the switching transistors. A Class-T amplifier’s processors learn the characteristics of the output transistors. Then, based on the analog input signal, they switch the output transistors with exactly the right timing to eliminate Class-D PWM problems: transistors not being perfect switches, ground bounce, output transistor mismatches, dead-time distortion and residual energy from the oscillator in the audio band. The result is a high power efficiency, audiophile-quality audio amplifier – a Class-T amplifier.
If one were to compare the waveform before the output filter of a Class-D PWM amplifier to a Tripath Class-T amplifier, some significant differences would be evident. The waveform for a Class-D PWM amplifier would be a pulse-width varying digital signal at the fixed, 100-200kHz, frequency of the triangle wave generator. The waveform for a Tripath Class-T amplifier would be a complex digital waveform of varying frequency. A Class-T amplifier switches the output transistors in a fashion similar to spread spectrum technology, at a varying rate up to 1.5 MHz and averages 600kHz to 700kHz."
I'm listening to my old TA2020 right now and I gotta say...it sounds damn good...really damn good! I've owned a LOT of excellent amplifiers to give me a fair frame of reference. Clarity and detail are on a par with what I can hear on a good HP amp and Sennheiser HD650s.
I'll say this...if there is a Class D topology out there right now that is significantly better sounding...I want to sign up...IMMEDIATELY! That is a sincere statement, and not a knock on anyone's opinion or conclusions. I refer back to the title of this post!
I just went to the Tripath ap note AN1, as I recall them claiming their modulator "adapted" by "learning" the switching characteristics of the output transistors! In the absence of a feedback loop, I wonder how they can accomplish that? The block diagram does not show a FB loop. If indeed there IS a FB loop, I do not see an input for that on their chipsets that incorporate a separate modulator and discrete output stage, as your test rig was fashioned. Is it possible that their fully-integrated chips, such as the TA2020 - do in fact "learn" the output stage and adapt, whereas the separate modulator config does not? Or are we just reading marketing department gobbledygook?
Tripath also claimed their modulator switching frequency was "spread spectrum-like" and adaptive in nature, maxing at 1.5 MHz, but averaging 600 kHz.
This is a snip from their AN1:
"Instead of using PWM, Tripath Class-T amplifier processors use proprietary algorithms and techniques to create the modulation that drives the switching transistors. A Class-T amplifier’s processors learn the characteristics of the output transistors. Then, based on the analog input signal, they switch the output transistors with exactly the right timing to eliminate Class-D PWM problems: transistors not being perfect switches, ground bounce, output transistor mismatches, dead-time distortion and residual energy from the oscillator in the audio band. The result is a high power efficiency, audiophile-quality audio amplifier – a Class-T amplifier.
If one were to compare the waveform before the output filter of a Class-D PWM amplifier to a Tripath Class-T amplifier, some significant differences would be evident. The waveform for a Class-D PWM amplifier would be a pulse-width varying digital signal at the fixed, 100-200kHz, frequency of the triangle wave generator. The waveform for a Tripath Class-T amplifier would be a complex digital waveform of varying frequency. A Class-T amplifier switches the output transistors in a fashion similar to spread spectrum technology, at a varying rate up to 1.5 MHz and averages 600kHz to 700kHz."
I'm listening to my old TA2020 right now and I gotta say...it sounds damn good...really damn good! I've owned a LOT of excellent amplifiers to give me a fair frame of reference. Clarity and detail are on a par with what I can hear on a good HP amp and Sennheiser HD650s.
I'll say this...if there is a Class D topology out there right now that is significantly better sounding...I want to sign up...IMMEDIATELY! That is a sincere statement, and not a knock on anyone's opinion or conclusions. I refer back to the title of this post!
Neil, how many different listeners auditioned your modulator test rig way back when? I recall you saying your son preferred the Tripath sound. Aren't young ears generally better than old? HA!
I was just my wife, my son and myself, and my wife and I did a lot of listening both with and without wine
. I offered the amp to people on the PE board and here, but I didn't get any takers. I ended up giving the amp to Pete Basel, who was interested in sharing it with members of an audio society in Connecticut. But the last time I check, he hadn't done that. If you want the amp for your own evaluation, check with Pete.My son preferred the Tripath because it had superior bass to the Apogee. This is probably due to the digital DC blocking filter in the Apogee. There are some serious limitations to the Apogee filters that show up in low frequencies. Yes, young ears like bass.
The analog feedback for the TC2000 is on pins 2 and 4 (chan 2) and 6 and 7 (chan 1). The data sheet describes the equations used to calculate the feedback to keep the feedback in an optimal range. I had to adjust the reference values to work with the 48V output devices that I used.
Probably a bit of that thrown in as well
That's true, but at idle (no input), frequency is relatively fixed. In fact, you had to adjust the capacitor CFB on the feedback pins to ensure that the idle was adequately offset: ...The value of CFB should be offset between channel 1 and channel 2 so that the idle switching difference is greater than 40kHz. This makes multi-channel implementations with the TC2000 difficult, as each channel has to be "de-tuned" to avoid interference.
Again, I don't want to claim that anyone's baby is ugly, but I think Michael is right in saying that the Tripath amps have a characteristic sound. Some people love that sound, and others don't. And it's not an overpowering coloration--those Tripath amps were and still are very good quality.
For me, the larger IRF/Infineon amps (IRS2092-based) are a good way to go if you need high power, as are the N-core and some of the newer TI "HD-series" amps. For lower powered active speaker applications, Analog Devices is probably the most "accurate". I haven't tried the Merus series from Infineon but they look promising, as does the new class D/ class H hybrid from TI, the TAS5825P. Fortunately, the technology has matured so that all of these amps should sound very good and you probably couldn't distinguish one from the other at the same power output It might be interesting to see some critical A/B testing of various new devices, but that's not something I'm interested in any longer.
Neil, thanks for the response. I don't have the data sheet handy for the separate Tripath modulator, as I'm a low-power guy who runs flea-powered single ended triode amps and efficient speakers to match. My Tripath amp is a low-powered one. The data sheets for the fully-integrated Tripath chips do not show any feedback loop...IMHO a rather large oversight when you tout being able to "learn" the output transistors.
Does Analog Devices offer a higher power chip using the same technology as the very interesting SSM3302? I searched but couldn't find one. I guess I could always bridge one to get 20 Wpc, the same as my Tripath 2020. Even though my primary amp is only 2.5 wpc (2A3 triode) it clips so gently that it never sounds harsh. When I run solid state amps, I like to have at least 20 Wpc - just to make sure I'm not listening to square wave crescendos. Again back to the Tripath...they touted their "soft clipping" characteristic as a sonic benefit, similar to a vacuum tube. Clipping characteristic is something people don't always consider when evaluating the "sound" of a particular amplifier. I have some rather inefficient mini monitors lurking around here for these kinds of evaluations.
Last question, as you have been very generous with your time: do you know of any commercially-manufactured PCBs for the SSM33XX that I can get to use as the basis for eval? I don't want to drop $200 on the AD eval board...I'm retired now and don't benefit from a corporate R+D sample budget anymore!
Regards,
Does Analog Devices offer a higher power chip using the same technology as the very interesting SSM3302? I searched but couldn't find one. I guess I could always bridge one to get 20 Wpc, the same as my Tripath 2020. Even though my primary amp is only 2.5 wpc (2A3 triode) it clips so gently that it never sounds harsh. When I run solid state amps, I like to have at least 20 Wpc - just to make sure I'm not listening to square wave crescendos. Again back to the Tripath...they touted their "soft clipping" characteristic as a sonic benefit, similar to a vacuum tube. Clipping characteristic is something people don't always consider when evaluating the "sound" of a particular amplifier. I have some rather inefficient mini monitors lurking around here for these kinds of evaluations.
Last question, as you have been very generous with your time: do you know of any commercially-manufactured PCBs for the SSM33XX that I can get to use as the basis for eval? I don't want to drop $200 on the AD eval board...I'm retired now and don't benefit from a corporate R+D sample budget anymore!
Regards,
I don't understand why ADI rates the SSM3302 as a 10W/ch chip. You should be able to get more output from the device than the TA2020 because of the somewhat higher voltage. However, ADI specs this chip at 1% distortion rather than the more common 10% level used for most class D amps. Also, ADI shows 24W/channel at 15V at 10% distortion in the spec sheet, but they have two notes that say there are package limitations and that higher outputs require special care in the PCB design. Most vendors would simply specify the higher wattage output. Instead, ADI specifies the chip at 12V and no additional heat sinking. The SSM3582, which uses the same 40-pad package, is rated at 31.5W, and there are no notes about additional heat sinking. I don't understand why they don't spec the chips the same way. But with a decent board with a heavy ground plane to remove the heat you should get 20 Wpc.
The SSM3582 is the chip I'm using for my new line array amp. It's got a built-in limiter for soft clipping, but you need software to set it up. I don't know of anyone who mass produces boards with these chips and sells them at low cost. However, I can send you some of the next round of boards I make if you want to try it. The boards are a very small (1-3/4" by 1-3/8") piggy-back assembly that plugs onto a "motherboard". The piggyback board has two SSM3582 chips and the "motherboard" has an ADAU1701 DSP module. It's Revision 4a of the stereo 3-way DSP board described in this article. I've got a prototype of the motherboard working and I partially built up two of the SSM3582 boards. Right now I'm still debugging the software. I'll post more info about these boards on that same page as the design gets completed.
The SSM3582 is the chip I'm using for my new line array amp. It's got a built-in limiter for soft clipping, but you need software to set it up. I don't know of anyone who mass produces boards with these chips and sells them at low cost. However, I can send you some of the next round of boards I make if you want to try it. The boards are a very small (1-3/4" by 1-3/8") piggy-back assembly that plugs onto a "motherboard". The piggyback board has two SSM3582 chips and the "motherboard" has an ADAU1701 DSP module. It's Revision 4a of the stereo 3-way DSP board described in this article. I've got a prototype of the motherboard working and I partially built up two of the SSM3582 boards. Right now I'm still debugging the software. I'll post more info about these boards on that same page as the design gets completed.
Neil, thank you...I'd be very interested in trying the 3302, as I'm only looking for analog inputs. Your PCB layout only accommodates the 3582 (data input only?) correct?
All my DSP is handled via Foobar, which I use as a media player in conjunction with some streaming hardware. Just want to compare some different Class D amplifiers using analog inputs.
All my DSP is handled via Foobar, which I use as a media player in conjunction with some streaming hardware. Just want to compare some different Class D amplifiers using analog inputs.
The TA2020 was always my favorite Tripath chip. For me it was the sweet spot.
Now that we are in the year AD2020 I'm sure there are Class-D chips that sound just as good, or better. Plus they have some of the great technical advantages that Neil mentions.
IMO, the Tripath chip proved to the Audiophile crowd that Class-D could hang with good Class-AB and not be embarrassed in front of Class-A. For many it was a turning point.
Michael, do you remember our email exchanges way back when the Sonic Impact amp first hit the scene and started all of this? I was living in San Diego, you in Hawaii. I recall that I designed some improved inductors for your early TA2020 modifications, which eventually led you to recommend I purchase the 41Hz kits out of Sweden. I did! That is the 2020 amp that I'm still using today, although it is not my primary amp! Time flies!
I've been out of the DIY game for quite a long time, as graduate school and managing my growing MIL communications business robbed all my free time. I'm retired now and living in beautiful, mountainous Northern AZ. And yes...my soldering iron is being warmed up again!
Regarding the sound of Tripath vs. more recent Class D offerings, the majority of responders here said there might be MINOR improvements in sound, but nothing that puts the old gal to shame. And indeed, if the Tripath sound is now considered euphonic by current standards - with that "SE tube warmth" - I'd probably still prefer it to more neutral Class D amps - as I'm still a tube guy and always will be. Everybody has different taste.
I'm still very intrigued with Class D, and will simply have to audition a couple of modern designs to satisfy my curiosity. Imagine, a single-ended tube guy who is also fascinated with Class D! That's the engineer in me! The music lover in me loves the warm sound!
Regards,
Dennis
Yes I remember. Was the 41Hz TA2020 called the AMP 3? I was a very nice board layout. I might still have an unbuilt kit.
Hi Michael-- This is off topic, but what do you do with your electronic creations and acquisitions when you move? I've been "cleaning house" in preparation for a move, but I've got drawers full of prototypes, nice parts and even completed amps that I don't know what to do with. I just keep looking at this stuff and putting off the decisions...
Same problem with test gear: yesterday I took a Tek 465 and a Sony/Tek 323 oscilloscope to the electronic recycling at the dump. It kind of hurt to do that, but they both needed some maintenance and shipping them to a worthy recipient would cost more than buying a new hand-held LCD scope. *sigh*
It is the AMP6 Michael, and I still have your notes for how to build it in a more logical order.
Do you want to get rid of the unbuilt kit? I know a young engineering student here in town who is short on cash, but has a real appetite to learn all about audio. She (yes, a female) aspires to become an audio engineer someday!
Drop me a PM if ya want...I sent friend request.
-Dennis
My gosh, Ken...is he still at it? I've blocked him so I don't have to read his arrogant rants anymore. His false airs of superiority and condescension over everyone who disagrees are rather offensive.
People whom cannot behave respectfully, and communicate on an intellectual rather than an emotional level - have absolutely no business causing drama here on my thread. I sincerely wish he would leave.
People whom cannot behave respectfully, and communicate on an intellectual rather than an emotional level - have absolutely no business causing drama here on my thread. I sincerely wish he would leave.
It was Amp 6. A fantastic amplifier.
I’ve built 6-7 of those and I’ve not kept one for me, all gone to friends (don’t make the same mistake)
I am left with an Amp10 (TA2022) and a Truepath (TA3020) from them.
George
YES...it IS a great little amplifier! George, are the other AMPs not as good as the AMP6 in your opinion? Or do they all sound the same?
No, they don’t sound the same to me.
IMO AMP10 is very good at LF and low mid but not at higher frequencies.
The Truepath shines everywhere !
(but none of those is a small amplifier like the Amp 6)
George
