This is part one of the review and detailed measurements of Lyngdorf TDAI-3400 Integrated amplifier, streamer, DAC and Room equalizer. It is on kind loan from a member and costs USD $6,499 (£4,995.00). In this review I will be focusing on the overall features of the unit, and performance of its DAC and amplifier subsystems. In part 2 I will be hooking it up to my main system to test the RoomPerfect room equalization feature.
Note: My company, Madrona Digital does business with one of the companies in the Lyngdorf family (Steinway Lyngdorf). I consider them a good partner of our company so feel free to read as much or as little bias as you feel is right into my review.
The TDAI-3400 is of course quite expensive relative to typical desktop systems we test but one look at the outside and you can take comfort that you purchased a luxury audio product:
If you are not familiar with Lyngdorf, it is one of the earliest pioneers of room equalization. Indeed when Harman decided to develop its own (ARCOS) it used the Lyngdorf as a benchmark to beat. The Lyngdorf did very well in their double blind tests.
If you were to ask me what an ideal home audio product should be, I would be listing just about everything that is included in TDAI-3400. It is the all-in-one device that fulfills every need from Roon endpoint to aforementioned room equalization. There are tons of analog and digital inputs and of course networking to stream content to it. So I jumped at the chance to review the unit even though I know my wallet will hurt when I ship this heavy unit back to its owner!
There is a massive volume rotary volume control on the right. It is a bit light for my taste but still, it gives the manly control most of us desire. The volume control works in 0.1 dB increments which is quite accurate and nice.
Likewise the large dot matrix display on the left gives you a feeling that cost control was not an issue here. It is informative and makes navigating the unit quite a bit easier.
The one thing that caught me at first was that the Lyngdorf logo is always on when you throw the unit on using the rear rocker power switch. But that doesn't mean the unit is on -- it is actually in standby. Hitting the power button on the remote seemingly does nothing causing you to keep bashing it.
Turns out it takes a second for the logo to become brighter indicating it is powering on.
I reset the unit to factory to make sure no equalization was in there. That by default shuts off the pre-amp out there so I had to go through routing menus to enable them. If you have used Lexicon or pro products you would be familiar with this.
I installed the Lyngdorf USB drivers but I could not find the Asio interface as stated to be there. So for my USB DAC testing I used my usual Asio4all wrapper. The other input I tested was the XLR analog for the amplifier section.
I also briefly tested Roon endpoint which worked excellently and resulted in identical performance as if you used the USB input.
Let's get into measurements and see how she does.
DAC Measurements
Whenever we have a USB input, that is my choice for DAC testing and this is no exception:. Here are the results for RCA outputs:
At first the results were worse than this when the volume was set to 0 dBFS. That pushed out a lot more voltage though than the nominal 2 volt we need so I dialed that back down to -6.3 dB and performance much improved. That turned out to be exactly right for balanced outputs just the same:
As I have noted, at 0 dBFS performance drops by 6 dB or so. Normally I would complain about this but in the case of a Room equalization, it is very useful to have lots of headroom in a DAC to avoid clipping. Here, the volume actually goes to +5 dB so we have quite a bit of margin there.
SINAD (signal over distortion and noise) is no miracle at 94 dB. It would almost pass for transparency for CD's 16 bit format. I would have wanted this to be at least 10 dB better. As its, the TDAI-3400 lands in our third tier of performance:
Dynamic range story is a bit better but still fails to match state-of-the-art DACs:
And that is with 0 dB. With -6.3 dB it falls a few dBs.
Linearity shows competence but again, not exceptional performance:
Jitter performance is good:
Similarly multi-tone test shows little distortion:
Increasing the bandwidth of measurements to measure THD+N (distortion+noise) relative to frequency reduces performance fair bit indicating out of band (ultrasonic) noise:
Intermodulation distortion follows the story told by SINAD:
The curves in purple show the saturation with 0 dB volume control. As you see, distortion sets in earlier and rises fair bit at 0 dBFS.
Overall the DAC performance is "OK" but well below my expectations for a device this expensive.
Amplifier Measurements
My dashboard for amplifiers is 5 watts of output so let's see how TDAI-3400 does there using USB input:
Ah, just 61 dB of SINAD in channel 2? You can see its elevated noise in red graph in FFT.
The amplifier in TDAI-3400 is calle "equibit" and is a custom PWM (pulsewidth modulated) design using switching technology. So I expect some out of band noise. Wanting to isolate that and remove the DAC from equation, I hooked up the unit using XLR input and my AES-17 filter and remeasured. At first, that created very odd results. Audio Precision software decided to just show a DC measurement in the scope graph and would refuse to give me a reliable power measurement. I was going crazy until I realized there was fair bit of DC leakage out of the unit. So I told the AP to use AC measurement and that all of a sudden cured the problem:
Note that the wiring is reversed in this test from the last one. As such, it is the blue channel now that is bad and gotten actually worse than using USB input without the AES-17 filter! Analog inputs are likely digitized using an ADC and used. Still, this is very odd behavior so I set out to investigate what is going on.
Before we go there, I cranked the input to 4 volts (XLR) and ran the best case signal to noise ratio test:
For a home processor I like to see 20 bit response would put us at 120 dB and we are failing that here.
Keeping the output at 5 watt, running THD+N test against frequency shows stunningly high noise (and distortion?) results:
And that is by cheating and using 45 kHz bandwidth instead of the usual 90 kHz I use for this test. Clearly there is junk above audio band here. Let's run a broadband (spectrum) test using more power to reduce effect of background noise using a 1 kHz test tone:
I expect to see switching noise in these types of amplifiers but not one that lands so close to the audio band at such elevated levels.
Edit: after consulting with the company, the switching frequency is the second peak (around 384 kHz). What is prior, below 100 kHz is due to noise-shaping.
Let's zoom in to see that better:
Noise starts to increase immediately after 20 kHz and rises up to just -20 dB or so in one channel (red). Hopefully your tweeter stops way short of 55 kHz or it will be cooking good.
The problem is apparent even in time domain scope:
Noice all the spikes on the red channel and thick waveform for blue. Normally my AES-17 filter gets rid of switching noise but the frequency is so low here that it does nothing to the components above 20 kHz but below 100 kHz.
All of this noise results in variable and pretty poor THD+N versus output power:
Edit: The external filter I use has wider bandwidth than the typical "20 kHz" one that the industry uses (mine is flat to 40 kHz). As such, it doesn't filter the noise shaped output of the TDAI-3400, resulting in much worse measurements than otherwise would be there.
The good news is that there is copious amount of power here to the tune of nearly 350 watts, beating out even the Hypex Ncore 400 we tested (with a single power supply). Outside of that, the switching noise is either confusing the analyzer input or is really elevated.
Let's see if we can get some reprieve in a simple frequency response test. Once again I had to set the input to AC coupled to get proper measurements:
The blue channel response starts to rise at or below 20 kHz, peaking at around 50 kHz. I suspect that is the corner frequency of the passive filter in the switching amplifier. Given that my load is a simply resistive dummy load, I expect much more pronounced and hence load dependent frequency response.
Conclusions
The pedigree of the company and very nice industrial design got me in a very positive mood for this review. The initial DAC measurements brought my excitement level down a notch but I still held on hope for a positive outcome.
Then came the power amplifier and all of that went out the window. Tons of ultrasonic noise makes a mockery of trying to play any high-resolution content with levels of noise that would dwarf anything real music would have. Frequency response variations as early as 20 Hz show very high speaker load dependency. There is a reason every engineer designing a switching amplifier tries to get the frequency up as much as they can. We have designs from a few hundred kilohertz to even a megahertz or two.Yet it appears that Lyngdorf has opted for something close to 50 kHz. That is just too low.
EDIT: as noted in the review now, the extra noise below 100 kHz is from noise shaping. The actually switching frequency is indeed much higher and competitive with other designs.
Yes, from pure audibility point of view, the huge amount of power will be a more audible benefit than ultrasonic noise takes away. Here though, we like to see excellence in design and I don't see that in Equibit amplifier technology. It may have been fine decades back when Lyngdorf first designed it but it simply is not competitive today. You can't have an amplifier that has so much noise that it can't properly be even measured with industry standard analyzer!
What a missed opportunity to impress with one-of-a-kind device with ideal feature list.
I will withhold ultimate judgement until I finish testing the room equalization.
EDIT: Part 2 testing of RoomPerfect is published here: https://www.audiosciencereview.com/...measurements-of-lyngdorf-roomperfect-eq.6799/
------------
As always, questions, comments, recommendations, etc. are welcome.
It is snowing here for the first time this year. I need some money to buy snowshoes. Yes, it is only one inch of snow but you don't want me to fall and get hurt, do you? So please consider donating money using:
Patreon: https://www.patreon.com/audiosciencereview), or
upgrading your membership here though Paypal (https://audiosciencereview.com/foru...eview-and-measurements.2164/page-3#post-59054).
Note: My company, Madrona Digital does business with one of the companies in the Lyngdorf family (Steinway Lyngdorf). I consider them a good partner of our company so feel free to read as much or as little bias as you feel is right into my review.
The TDAI-3400 is of course quite expensive relative to typical desktop systems we test but one look at the outside and you can take comfort that you purchased a luxury audio product:
If you are not familiar with Lyngdorf, it is one of the earliest pioneers of room equalization. Indeed when Harman decided to develop its own (ARCOS) it used the Lyngdorf as a benchmark to beat. The Lyngdorf did very well in their double blind tests.
If you were to ask me what an ideal home audio product should be, I would be listing just about everything that is included in TDAI-3400. It is the all-in-one device that fulfills every need from Roon endpoint to aforementioned room equalization. There are tons of analog and digital inputs and of course networking to stream content to it. So I jumped at the chance to review the unit even though I know my wallet will hurt when I ship this heavy unit back to its owner!
There is a massive volume rotary volume control on the right. It is a bit light for my taste but still, it gives the manly control most of us desire. The volume control works in 0.1 dB increments which is quite accurate and nice.
Likewise the large dot matrix display on the left gives you a feeling that cost control was not an issue here. It is informative and makes navigating the unit quite a bit easier.
The one thing that caught me at first was that the Lyngdorf logo is always on when you throw the unit on using the rear rocker power switch. But that doesn't mean the unit is on -- it is actually in standby. Hitting the power button on the remote seemingly does nothing causing you to keep bashing it.
Turns out it takes a second for the logo to become brighter indicating it is powering on.I reset the unit to factory to make sure no equalization was in there. That by default shuts off the pre-amp out there so I had to go through routing menus to enable them. If you have used Lexicon or pro products you would be familiar with this.
I installed the Lyngdorf USB drivers but I could not find the Asio interface as stated to be there. So for my USB DAC testing I used my usual Asio4all wrapper. The other input I tested was the XLR analog for the amplifier section.
I also briefly tested Roon endpoint which worked excellently and resulted in identical performance as if you used the USB input.
Let's get into measurements and see how she does.
DAC Measurements
Whenever we have a USB input, that is my choice for DAC testing and this is no exception:. Here are the results for RCA outputs:
At first the results were worse than this when the volume was set to 0 dBFS. That pushed out a lot more voltage though than the nominal 2 volt we need so I dialed that back down to -6.3 dB and performance much improved. That turned out to be exactly right for balanced outputs just the same:
As I have noted, at 0 dBFS performance drops by 6 dB or so. Normally I would complain about this
but in the case of a Room equalization, it is very useful to have lots of headroom in a DAC to avoid clipping. Here, the volume actually goes to +5 dB so we have quite a bit of margin there.SINAD (signal over distortion and noise) is no miracle at 94 dB. It would almost pass for transparency for CD's 16 bit format. I would have wanted this to be at least 10 dB better. As its, the TDAI-3400 lands in our third tier of performance:
Dynamic range story is a bit better but still fails to match state-of-the-art DACs:
And that is with 0 dB. With -6.3 dB it falls a few dBs.
Linearity shows competence but again, not exceptional performance:
Jitter performance is good:
Similarly multi-tone test shows little distortion:
Increasing the bandwidth of measurements to measure THD+N (distortion+noise) relative to frequency reduces performance fair bit indicating out of band (ultrasonic) noise:
Intermodulation distortion follows the story told by SINAD:
The curves in purple show the saturation with 0 dB volume control. As you see, distortion sets in earlier and rises fair bit at 0 dBFS.
Overall the DAC performance is "OK" but well below my expectations for a device this expensive.
Amplifier Measurements
My dashboard for amplifiers is 5 watts of output so let's see how TDAI-3400 does there using USB input:
Ah, just 61 dB of SINAD in channel 2? You can see its elevated noise in red graph in FFT.
The amplifier in TDAI-3400 is calle "equibit" and is a custom PWM (pulsewidth modulated) design using switching technology. So I expect some out of band noise. Wanting to isolate that and remove the DAC from equation, I hooked up the unit using XLR input and my AES-17 filter and remeasured. At first, that created very odd results. Audio Precision software decided to just show a DC measurement in the scope graph and would refuse to give me a reliable power measurement. I was going crazy until I realized there was fair bit of DC leakage out of the unit. So I told the AP to use AC measurement and that all of a sudden cured the problem:
Note that the wiring is reversed in this test from the last one. As such, it is the blue channel now that is bad and gotten actually worse than using USB input without the AES-17 filter! Analog inputs are likely digitized using an ADC and used. Still, this is very odd behavior so I set out to investigate what is going on.
Before we go there, I cranked the input to 4 volts (XLR) and ran the best case signal to noise ratio test:
For a home processor I like to see 20 bit response would put us at 120 dB and we are failing that here.
Keeping the output at 5 watt, running THD+N test against frequency shows stunningly high noise (and distortion?) results:
And that is by cheating and using 45 kHz bandwidth instead of the usual 90 kHz I use for this test. Clearly there is junk above audio band here. Let's run a broadband (spectrum) test using more power to reduce effect of background noise using a 1 kHz test tone:
Edit: after consulting with the company, the switching frequency is the second peak (around 384 kHz). What is prior, below 100 kHz is due to noise-shaping.
Let's zoom in to see that better:
Noise starts to increase immediately after 20 kHz and rises up to just -20 dB or so in one channel (red). Hopefully your tweeter stops way short of 55 kHz or it will be cooking good.
The problem is apparent even in time domain scope:
Noice all the spikes on the red channel and thick waveform for blue. Normally my AES-17 filter gets rid of switching noise but the frequency is so low here that it does nothing to the components above 20 kHz but below 100 kHz.
All of this noise results in variable and pretty poor THD+N versus output power:
Edit: The external filter I use has wider bandwidth than the typical "20 kHz" one that the industry uses (mine is flat to 40 kHz). As such, it doesn't filter the noise shaped output of the TDAI-3400, resulting in much worse measurements than otherwise would be there.
The good news is that there is copious amount of power here to the tune of nearly 350 watts, beating out even the Hypex Ncore 400 we tested (with a single power supply). Outside of that, the switching noise is either confusing the analyzer input or is really elevated.
Let's see if we can get some reprieve in a simple frequency response test. Once again I had to set the input to AC coupled to get proper measurements:
The blue channel response starts to rise at or below 20 kHz, peaking at around 50 kHz. I suspect that is the corner frequency of the passive filter in the switching amplifier. Given that my load is a simply resistive dummy load, I expect much more pronounced and hence load dependent frequency response.
Conclusions
The pedigree of the company and very nice industrial design got me in a very positive mood for this review. The initial DAC measurements brought my excitement level down a notch but I still held on hope for a positive outcome.
Then came the power amplifier and all of that went out the window. Tons of ultrasonic noise makes a mockery of trying to play any high-resolution content with levels of noise that would dwarf anything real music would have. Frequency response variations as early as 20 Hz show very high speaker load dependency. There is a reason every engineer designing a switching amplifier tries to get the frequency up as much as they can. We have designs from a few hundred kilohertz to even a megahertz or two.
EDIT: as noted in the review now, the extra noise below 100 kHz is from noise shaping. The actually switching frequency is indeed much higher and competitive with other designs.
Yes, from pure audibility point of view, the huge amount of power will be a more audible benefit than ultrasonic noise takes away. Here though, we like to see excellence in design and I don't see that in Equibit amplifier technology. It may have been fine decades back when Lyngdorf first designed it but it simply is not competitive today. You can't have an amplifier that has so much noise that it can't properly be even measured with industry standard analyzer!
What a missed opportunity to impress with one-of-a-kind device with ideal feature list.
I will withhold ultimate judgement until I finish testing the room equalization.
EDIT: Part 2 testing of RoomPerfect is published here: https://www.audiosciencereview.com/...measurements-of-lyngdorf-roomperfect-eq.6799/
------------
As always, questions, comments, recommendations, etc. are welcome.
It is snowing here for the first time this year. I need some money to buy snowshoes. Yes, it is only one inch of snow but you don't want me to fall and get hurt, do you? So please consider donating money using:
Patreon: https://www.patreon.com/audiosciencereview), or
upgrading your membership here though Paypal (https://audiosciencereview.com/foru...eview-and-measurements.2164/page-3#post-59054).
Last edited:
