Head-Fi.org › Forums › Equipment Forums › Headphone Amps (full-size) › Four-way integrated amp shootout: T-amp vs tube amps up to $1k
New Posts  All Forums:Forum Nav:

Four-way integrated amp shootout: T-amp vs tube amps up to $1k

post #1 of 27
Thread Starter 
Lately I happened to have four budget (<$1000), low powered (<30W) integrated amplifiers at hand, and had the chance to compare them while driving my speakers and K1000. It was quite an educational experience. Here are the four amps being compared:

Sonic Impact 5066 class T amplifier (~$30)
Response Audio Modified Jolida JD102B (~950)
Jolida JD301A (~$350) Hybrid amplifier
Aria MiniP1 tube amplifier, a.k.a. EL 84 version of Sophia Electric Baby (~$280)

Brief introduction to each amp:

1. SI T-amp
This is a switching amplifier using Tripath audio IC, and delivers 5.5W into 8 ohms. People interested in T-amp should read the collection of user reviews on TNT audio. There is no need for me to reiterate all the buzz generated by this little wonder. I am using the amp in its stock form while using big car booster battery (19Ah, SLA type) via a 5A-rated DC cable. I have previously posted some more details about this amplifier.

2. Modified JD102B
This amp is kindly loaned from a friendly head-fier, Randytsuch. It is heavily modified by Response Audio. The stock unit retails for $580 and the modified unit for $950. It uses 12AX7*1, 12AT7*2 and EL84*4. The power tubes are wired in ultralinear class AB1 configuration to provide 20W. I used JJ EL84, NOS Raytheon 12AT7 during the review. 12AX7 was either Chinese (selected and branded by TNT) or Yugo, and the difference was small. Chinese 12AX7 is slightly more resolving and the Yugo a tad warmer. The manual suggests 30 mV bias but Response audio suggests 17-22. Randytsuch suggested 21 mV and I concur. I tried 17, 19, 21 and 30 mV.

3. Jolida JD301A
This 30W integrated amplifier uses 12AX7*2 and LM1875 audio IC. LM1875 is a MOSFET op amp made famous by the 47 Lab Gaincard. I used Yugo 12AX7 during the review.

4. Aria MiniP1
It uses 2C51/5670*2 and EL84*4. The circuit is class-A push-pull, and the power is 10W, which is lower than the class-AB JD102B (20W). This is a trade-off between crossover distortion and output power. This amp is an alternative version of Sophia Electric Baby ($899). The main difference is that Sophia Baby uses Russian 6P1T (super-exclusive tube) and Aria uses 6P1P (=EL84/6BQ5). See my recent post about the Aria amp for more details. I bought mine off Ebay for $280 and use it with JJ EL84, which is a tad warmer than stock Chinese 6P1. For 2C51, I used stock Chinese 6N3 for review. I also have NOS WE 396A, the highly regarded 2C51, but found it intolerably bright.

Review System

Transport: Sony DVP-NS900V SACD/DVD player (SACD is occasionally used)
DAC: Benchmark DAC1
Speakers: Athena S2/P2
AKG K1000
Power treatment: Furman Power Factor Pro ($229) or Belkin Surgemaster II ($40)

Shootout Round 1: Driving Speakers
Amp configurations:
(1)T-amp + SLA battery
(2)JD102B, 21 mV bias, Yugo 12AX7, A/C Furman
(3)JD301A, Yugo 12AX7, A/C Furman
(4)Aria, Chinese 6N3, A/C Furman
(5)JD102B, 30 mV, Chinese 12AX7, A/C Furman
(6)JD301A, Yugo 12AX7, A/C Belkin

Performance: 4=2=1>3>5>6

First of all, my speaker is 88.5 dB anechoic, my room is 9’ by 13’ and I have two active subs to take care of <50dB bass. In such conditions, T-amp’s 5.5 W is loud enough, and therefore my review is about the sound quality in the first few watts and not related to the total output power of an amp.
There is a three-way tie between T-amp, modified JD102B and Aria. For resolution, speed, dynamics, bass impact, imaging and transparency T-amp is the winner. I would not hesitate to say that it has the highest fidelity of the four. However, its sound is a bit bright. For a good recording T-amp has pleasant treble sparkle, but for mediocre recordings the treble becomes lean. JD102B is the opposite of T-amp. It is warm, smooth and laid-back, which help keep mediocre recordings enjoyable. On good recordings I feel its sound is too distant and a bit veiled, and it does not take me close enough to the performers. Aria amp is sort of between T-amp and JD102B. Its presentation is forward enough to keep things exciting, but it does not bring much euphony to poor recordings. Aria does lead all other amps in terms of harmonic richness and ambience.

Jolida JD301 is a tube/MOSFET hybrid amplifier. Ideally, it would have the warmth of tubes and the bass of solid state. Indeed on some recordings it shows such strengths. But its overall sound is a bit thin (like SS) and a bit blurred (like tube). On some recordings it can sound really sweet but on others lack airiness. It combines the strengths and the weaknesses of tube and solid state, depending on the music. Its lack of a consistent sonic signature puts it a bit behind the three other amps.

In configurations 5 and 6, I evaluated the effects of tweaking. I first experimented with the tube bias of JD102B. Using 30 mV (suggested by the manual), the sound becomes a bit brittle and harsh compared to 21mV. This instantly makes JD102B sounding worse than other amps under comparison, which indicates that the differences between the four amps are pretty small. Instead of using the Furman conditioner, I tried JD301 with common Belkin surge protectors. While I think every component in my system benefits from using the Furman conditioner, JD301 seems to improve the most with better power. Using the Belkin surge protector basically sounds the same as directly using the outlet. But using Furman conditioner provides across-the-board improvements for JD301 as everything sounds clearer. Ambience is better defined, details increased, and treble a bit richer—all quite subtle, though. Furman claims that Power Factor Pro is the first conditioner designed specifically for amplifiers. After hearing three amps sounding better when connected to it, I am sold.

Shootout Round 2: Driving K1000

Amp configurations (A/C from Furman):
(1) T-amp + SLA battery
(2) JD102B, 21 mV bias, Chinese 12AX7
(3) JD301A, Yugo 12AX7
(4) Aria, Chinese 6N3

Performance: 4=2>3>1

These four amps basically retain their sonic signatures when connected to K1000, compared to speakers. T-amp turns out to be a poor match for K1000. There is too much treble emphasis and the high shrill sometimes becomes annoying. {EDIT} After adding 10 ohm resistors in parallel, the treble emphasis is large tamed, but the sound is still on the bright side. Using a slightly bright source such as DAC1 certainly does not help, but playing SACD on my Sony as the source also does not alleviate the problem. However, T-amp+K1000 still has better transparency and imaging than any other combination. JD301 sounds particularly good with K1000, almost as good as Aria or modified JD102. Maybe it is because JD301 likes high impedance load. However K1000 still sounds a bit lean on JD301, while all-tube amplifiers alleviate this issue. JD102 provides a very smooth and forgiving sound, but is a bit recessed and not all that transparent. Weaker digital recordings benefit greatly with JD102. In terms of resolution as well as bass, I rank T-amp>Aria=JD301>JD102. But Aria is a head of the pack in ambience, airiness and rich harmonics. On my best recordings, Aria takes me closer to the concert than other amps.

Conclusion

Sonic Impact T-amp is truly a miraculous product. Its construction is flimsy and its appearance unimpressive. Its mini plug input and spring clip speaker terminals are disdained by audiophiles, but there is no fundamental reason why these would degrade the sound. Its volume pot is probably the cheapest kind but so far it turns smoothly without noise. I run mine off a big SLA battery, which should be some of the best DC available. It reminds me of what really affects sound quality: power and amplification signal path. Connectors, chassis and cables—do they affect sonics or aesthetics? Unfortunately my brain is very bad at differentiating aesthetics from sonics. In its stock form, the T-amp+battery ($70) can perform at the same level as modified JD102B ($950) and Aria MiniP1, a Sophia Electric Baby ($899) in disguise, plugged into Furman Power Factor Pro ($229). As a K1000 amplifier, I don’t enjoy T-amp’s brightness but some may cherish its ultra-resolution. Of course, T-amp is greatly limited by its low power (5.5W into 8 ohms), but most people also don’t realize they never use the full power of their bigger amplifiers, and that power should be compared in a logarithmic scale.

JD301A is the most powerful amp by specs among the four although its sound is a bit less refined. Considering its low price and its tubeness to tame the treble, it would be a great buy for a budget system.

Modded JD102B is really smooth and tubey, which is a great virtue in an age filled with overly bright components and recordings. I hate to send it back to Randy, but I can’t afford to buy another speaker amp.

Aria amp from Ebay is a risky purchase but my unit has proven itself as a good performer and turns out to be a great bargain. Its class A, push-pull tube sound is clean, warm and rich in harmonics, qualities that have special place in my heart. The differences between these four amps are pretty subtle and some of them might disappear with tweaking, especially with tubes. Using a “highly regarded” WE NOS 2C51 ($30 ea) would instantly make Aria the worst speaker amp in the pack. In the end I decided that Aria is a keeper and I am ordering different tubes to tune its sound. An Adcom speaker selector enables me to power K1000 and speaker with one amp. I just got some JAN GE ($7 ea) tubes from 80’s, supposedly the worst NOS 2C51 you can find, and everything sounds really good.
post #2 of 27
Nice work, man.
post #3 of 27
Yup, SI has changed the way I look at other amps, any amps.

I don't find SI to be overly bright with my gear, cables, and linear PS, but then again, my K1000 has a warmer bass due to mods.

Using 6-8 Ohm resistor across speaker outputs of SI does work for K1000 to adjust the treble down accordingly. Have you tried this?
post #4 of 27
Thanks for the thoughts. This is very helpful. I also find that the T-Amp (even upgraded, as mine is) and a slightly bright source (HeadRoom Overture) makes my K 1000 too metallic. But it drives the K 1000 much much better than my much more powerful hybrid amplifiers, which causes distortion with the K 1000.

Best regards,

-Jason
post #5 of 27
Thread Starter 
Quote:
Originally Posted by Jon L
Using 6-8 Ohm resistor across speaker outputs of SI does work for K1000 to adjust the treble down accordingly. Have you tried this?
Interesting idea. I will give it a try.
post #6 of 27
Re: JD301A and the LM1875 amp chip
The datasheet of the LM1875 chip says nothing about MOSFETs, and the published internal schematic diagram shows that it has all bipolar junction transistors. Where did you get the idea that it contains MOSFETs?
post #7 of 27
Thread Starter 
Quote:
Originally Posted by amb
Re: JD301A and the LM1875 amp chip
The datasheet of the LM1875 chip says nothing about MOSFETs, and the published internal schematic diagram shows that it has all bipolar junction transistors. Where did you get the idea that it contains MOSFETs?
Jolida describes its hybrid amp as tube/MOSFET:http://www.jolida.com/catalogue/hybridamp.shtml

Opening up my unit, the only solid state device I see is LM1875. So I thought it is a MOSFET IC. But LM1875's datasheet says nothing about MOSFET. What does MOSFET look like? I wonder if they are there but I failed to see them.
post #8 of 27
I looked at the Jolida page and it says "Mosfet output" for all their hybrid models. I wonder if this is an error. I have not seen an integrated chipamp that uses MOSFETs internally as its output stage. Typically amps employing MOSFETs output stage have discrete MOSFETs.

It doesn't make sense to use an IC chipamp that already has the appropriate output power capability and then add discrete MOSFETs, so perhaps only some of the Jolida models have MOSFETs? Without inspecting one in person or seeing a schematic diagram I won't know.

Btw MOSFETs come in various packages but the most common for power output is the TO-220 case style (like this) mounted to a heat sink. However bipolar transistors, voltage regulators, and other devices also come in the same style package, so you'd have to read the markings on them to find out what they are.
post #9 of 27
Thread Starter 
Inside JD301A I can only see two packaged ICs, two LM1875. Even the power supply section uses diodes instead of voltage regulating ICs. Close examination of LM1875's schematic diagram I see one FET in the lower left corner. I am not sure what role it plays in the circuit, which has dozens of BJT.
post #10 of 27
Quote:
Originally Posted by Ferbose
Close examination of LM1875's schematic diagram I see one FET in the lower left corner. I am not sure what role it plays in the circuit, which has dozens of BJT.
That is a JFET wired as a constant current source.
post #11 of 27
A note to thank Ferbose for a great contribution. This is precisely the kind of thoughtful evaluation that others should try to emulate in their Head-Fi reviews. And just what folks like me who haven't a lot of experience, technical knowledge, or sophisticated listening skills should seek out in learning about audio.

Best,
Beau
post #12 of 27
Thread Starter 
Updated the original post a bit. With T-amp and K1000, adding a 10 ohm resistor in parallel cures the excessive brightness. Thanks, Jon L, for the advice. However, DAC1, T-amp, and K1000 are still on the bright side on neutrality, respectively and as a whole. I found that Grove Tube reissued Mullard 12AX7 works very well with Jolida JD301. Now JD301 has SS amp's bass, details, transparency and the pleasant harmonics and ambience of tubes. The sound is a bit bright but energetic. I now use it for jazz with K1000, as well as SACD. For classical CDs, whic are the mainstay of my collection, I use Cayin HA-1A in single ended triode mode with K1000. Although SET topology is not the last word in accuracy or resolution, its rich harmonics, smoothness, airiness and sense of space cures the last trace of digititis on my DAC1. I am a fan of SET now, at least for classical music.
post #13 of 27
Quote:
Originally Posted by Ferbose
Updated the original post a bit. With T-amp and K1000, adding a 10 ohm resistor in parallel cures the excessive brightness. Thanks, Jon L, for the advice. However, DAC1, T-amp, and K1000 are still on the bright side on neutrality, respectively and as a whole. I found that Grove Tube reissued Mullard 12AX7 works very well with Jolida JD301. Now JD301 has SS amp's bass, details, transparency and the pleasant harmonics and ambience of tubes. The sound is a bit bright but energetic. I now use it for jazz with K1000, as well as SACD. For classical CDs, whic are the mainstay of my collection, I use Cayin HA-1A in single ended triode mode with K1000. Although SET topology is not the last word in accuracy or resolution, its rich harmonics, smoothness, airiness and sense of space cures the last trace of digititis on my DAC1. I am a fan of SET now, at least for classical music.
Sounds like you're having fun. Too bad I broke my T-amp and can't find another one cheap. I've been looking at the Red Wine Audio Clari-T, but I just received my Dynahi yesterday, so maybe I'm not looking any more

You got a listen to my modded AKG K340 via your amp at the meet, but you won't believe how K340 sounds from the Dynahi. It's quite scary...
post #14 of 27
Quote:
Originally Posted by Ferbose
Updated the original post a bit. With T-amp and K1000, adding a 10 ohm resistor in parallel cures the excessive brightness. Thanks, Jon L, for the advice. However, DAC1, T-amp, and K1000 are still on the bright side on neutrality, respectively and as a whole. I found that Grove Tube reissued Mullard 12AX7 works very well with Jolida JD301. Now JD301 has SS amp's bass, details, transparency and the pleasant harmonics and ambience of tubes. The sound is a bit bright but energetic. I now use it for jazz with K1000, as well as SACD. For classical CDs, whic are the mainstay of my collection, I use Cayin HA-1A in single ended triode mode with K1000. Although SET topology is not the last word in accuracy or resolution, its rich harmonics, smoothness, airiness and sense of space cures the last trace of digititis on my DAC1. I am a fan of SET now, at least for classical music.
If you get the opportunity, try a true triode for an honest SET experiece. A tetrode or pentode wired in triode is no where near as good. A good SET amp will blow you away if you're anything like me. They give nothing up in way of detail, accuracy, transparency, or resolution. I find SETs are most incredible at retrieving subtle nuisances which PP, OTL, or SS never seem to get right.

I like the T amp. There is no doubt it is an outstanding value. While I like it better then some amps costing several times its price, it's really not the giant killer I was led to believe. This is strictly coming from a amp->speaker perspsective. I never got the chance to try it with my K1K, though I'd imagine the sound would be tipped up as you experienced. Just one man's opinion.
post #15 of 27
T-amp rocks my house every day.
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Headphone Amps (full-size)
Head-Fi.org › Forums › Equipment Forums › Headphone Amps (full-size) › Four-way integrated amp shootout: T-amp vs tube amps up to $1k