Elekit TU-8900 2A3\300B Single Ended DHT Tube Amp - Build, Impressions, and Review

Discussion in 'Headphone Amplifiers and Combo (DAC/Amp) Units' started by Maven86, Oct 19, 2021.

  1. Maven86

    Maven86 Almost "Made"

    Contributor
    Joined:
    Jan 19, 2016
    Likes Received:
    395
    Trophy Points:
    53
    Location:
    DMV
    I was surprised to see quite a bit of interest in this amp so I thought it might be helpful to share my thoughts and experiences on the build, as well as sound impressions. A review is also in the works but I think it’s best to give things some more settling time while also allowing the “new toy syndrome” to wear off.


    IMG_3716.jpg

    The TU-8900 is Elekit’s latest model and uses directly heated triodes (DHTs) instead of pentodes like its predecessor. While most amps of this design tend to be built around one type of DHT, the 8900 has a unique auto detect circuitry that can recognize either 2A3 or 300B tubes. These power tubes are driven by either two 12AU7 or 12BH7 (my preference) and doesn’t require any tubes for rectification as this is accomplished via diodes.


    It’s sold as a kit that can be purchased through contacting Victor of VKaudio. Purchasing, build, and contact information can be found here:

    https://www.diyaudio.com/forums/elekit/374425-elekit-tu-8900-2a3-300b.html



    Aside from a brief intro, I’ll pass on covering the Elekit brand and their other products as others have done an excellent job of that in previous threads:

    TU-8600
    https://www.superbestaudiofriends.org/index.php?threads/elekit-tu8600.5132/


    TU-8800
    https://www.superbestaudiofriends.o...ded-pentode-tube-amp-build-impressions.10100/


    Intro\Background (Part 1)

    My first Elekit build was the TU-8800 over a year back. Before this, I hadn’t been much of a DIY guy and always hated soldering back in engineering school. However, cabin fever has a way of making you reconsider options and I soon found myself staring at a kit with a soldering iron in hand.


    To be completely honest, the 10 hours spent building the amp were a mixed bag. There were some very frustrating moments where I had to backtrack because I overlooked a step, and there were more than a few instances of re-soldering cold joints\solder bridges. Needless to say there was a lot swearing involved as well a new found appreciation for burn cream… Despite those small hiccups, the feeling of pushing the power button and watching everything light up as it should was extremely fulfilling and totally worth the headache.


    Over the next few months I would end building a few more kits (TU-8200, 8600, and 8800) for friends and family. The process became a lot easier and I was more comfortable making my own minor tweaks in these builds. When I heard rumors earlier this year about a new model using 2A3s, I immediately contacted Victor for details. Sure enough, the gears were turning behind the scenes and the rest is history.



    Build impressions to come but for now, sleep...
     
    • Like Like x 35
    • Epic Epic x 3
    • List
    Last edited: Oct 22, 2021
  2. Maven86

    Maven86 Almost "Made"

    Contributor
    Joined:
    Jan 19, 2016
    Likes Received:
    395
    Trophy Points:
    53
    Location:
    DMV
    From Box to Bench My Experience with the TU-8900 (Part 2)


    As usual, dealing with Victor was clockwork. I shot him an e-mail, chatted for a bit, and placed my order for the amp along with a TKD pot upgrade, Lundahl amorphous output transformers (OPTs), and a pair of Silvania 12BH7s. A quick note about the Lundahl upgrade. For the 8900 you can choose between the amorphous and standard steel cores of previous models. Personally, I'm happy with my choice but many will prefer the tone of the original.

    Victor also offers upgraded coupling caps in two flavors, the CuTF and ODAM from VHaudio. They’re excellent caps but I decided to do my own thing. Lastly, there’s no Takman resistor upgrade this time so the more meticulous builders might want to source their own stuff…or do what I did and just go stock.

    Ordered on Sunday, it shipped on Monday, and arrived on Tuesday. The goods were triple boxed and well packed. All components were accounted for with nothing missing or broken. Once again, clockwork and a well deserved shout out to Victor for consistently going the extra mile.

    IMG_1720.jpg


    I let things sit for a few days to plan logistics, all signs pointed to Friday. Partly because it was sunny and mostly because my wife was flying south for a business trip over the weekend. This meant being able to use the bottom half of the dining table to set up shop. “It is easier to beg forgiveness than to seek permission.” She was always the brains and taught me well:

    IMG_1664 2.jpg
    Those big blocks are the upgraded Lundahl OPTs. This year Victor tried something new by going with amorphous cores. A risky move considering the polarizing feedback these tend to have. I can definitely say these sound different but more on that later. As far as the rest of the parts go, there isn't much else that's remarkable. Also, I don't have any other pics of the the bagged components as this review was more of an afterthought (sorry). However, if you look really hard and use your imagination you'll see several packets of assorted resistors and capacitors.

    The stock resistors this time around are made by KOA and are actually pretty good. Likewise, the electrolytic caps are a mix of Nichicon and Rubycon (of black gate fame) for the power supply. I was originally planning on subbing these with some Audio Note Kaisei caps I had sitting around, but ultimately didn't feel the need to. For those who've sworn an oath against electrolytics, there doesn't seem to be much space for fancy oil caps. Polypropylene film?...Maybe if you're creative... Regardless, I'd love to see it happen. Finally, those big bulky copper kegs on the far right are my coupling caps. More on those later.

    In terms of tools, here are the basics: A good soldering iron\ station, desoldering pump, flux pen, wire cutters, tweezers, pliers, multimeter, screwdriver, and solder. For the case and volume pot, you'll also need two small allen wrenches. I used 2.0mm and 1.27mm respectively. I’d also recommend getting a head mount magnifier\light, similar to what jewelers use (shown below). It makes it a lot easier to work through the denser areas of the board where the soldering points are small and\or tightly packed together. You’re also less likely to miss these points or overlook mistakes.
    image3.jpg
    More of the chop shop along with my grumpy little helper.



    I was all good to go but the door bell rings and in comes an unexpected guest. Apparently my birthday was coming up and an old friend decided to drive down for the week. The next couple of days were a blast till I remembered I had an amp to build and roughly 8 hours to hide the evidence. I grab a couple Red Bulls and get to work.

    Compared to last year, things go much better this time. Here’s what I learned and did differently:

    1. Went with Cardas Eutectic solder instead of the the more expensive WBT silver. The lower melting point made it much easier to get a good flow. This made forming good joints a breeze while also preventing solder bridges caused by partial melting.

    2. As mentioned above, that head mount magnifier thing worked wonders.

    3. Red pens are helpful. Granted, it’s a small thing. But when you’re fighting through a caffeine crash at 11PM scrambling to get things done, it’s not hard to miss a resistor or put in a diode the wrong way. Small mistakes can cause big failures.

    4. Used a smaller more compact soldering iron. For me this meant better precision when soldering at weird angles or between small points closely packed together. There were quite a few of these.



    From here on out things were straight forward... I started with the resistors and worked my way up.
    image2.jpg

    IMG_1719.jpg
    Another helpful tip. As far as resistors go, when in doubt use the multimeter. In this case it measures a clean 3.3 ohms, the same as its rating. You'll have some variance but it shouldn't be more than 2-3% off spec.

    As for coupling caps, I ended up using Duelund CAST pure silver. For those not familiar with these, it's a not a big deal, life goes on. For those who are, yes I am crazy, and no I did not pay full price for these.
    IMG_1690.jpg

    image4.jpg
    They're a little big for the designated cap slots...but I had already made up my mind that these were going in. And in they went.

    After 5 hours and 54 minutes of tweaked out DIY and live guitar, I hit the power and the amp started to look like this.
    IMG_3652.jpg

    With my remaining energy, I reach for the Utopias and take a quick listen. 10 minutes later, I'm left very surprised and a bit troubled... (To be continued in sound impressions)


    Things are getting a little lengthy so I'll split these build impressions into two parts. In part two, I'll go over the guts of the 8900, some of its more interesting features, and my final thoughts and feelings about the experience. Stay tuned.
     

    Attached Files:

    • Like Like x 19
    • Epic Epic x 9
    • List
    Last edited: Oct 22, 2021
  3. Maven86

    Maven86 Almost "Made"

    Contributor
    Joined:
    Jan 19, 2016
    Likes Received:
    395
    Trophy Points:
    53
    Location:
    DMV
    Part 3: Build Impressions Cont.


    A couple quick shots under the hood, power supply and output transformers on the top bracket. In terms of inputs, the amp is pretty barebones with only one pair of RCAs. The ability to toggle output impedances for speakers is handy though, flip the switch up for 8-16 ohms and down for 4-6.
    image.jpg


    Output impedance can also be adjusted for headphones via the yellow jumpers. For the 8900, the jacks are now located under the front panel. We'll talk about this later, it's a pretty big deal.
    IMG_3731.jpg




    Finally, the two biggest features. On the bottom right are the fabled auto detect circuits, one for 2A3 and the for the 300B. You can basically plug either one and the amp will automatically detect the type and set the bias accordingly. The LED lights up green for 2A3s and Blue for 300Bs. If it's red... You're in for a long night.

    On the left is my personal favorite. Prior to the 8900, all Elekits to my knowledge have used some type of negative feedback (NFB). You can now toggle between feedback (NFB mode) or no feedback (Non-NFB mode) by moving a simple jumper on the circuit. Without going too far down the rabbit hole, amps that use feedback take a fraction of the signal from a later portion of the circuit (usually a tap on the output transformer) and "feed" it back to an earlier part (phase inverter).

    Supposedly this lowers distortion, improves frequency extension\response, and lowers the output impedance (basically a better measuring amp). However, some people don't like it for reasons google would be happy to provide. Jargon aside, the important thing here is that it sounds different and changes the volume output. Some people will prefer one mode over the other.
    image3.jpg

    And that's pretty much it for the build. So what are my thoughts and sentiments about this whole experience? Well here are some semi sober but completely honest rantings I scribbled down from a previous night...



    Props to Elekit For…

    -Addressing some of the ergonomic complaints of previous models. I loved my 8800 but boy was it a pain in the ass to get some basic things done. If I wanted to roll caps, the board needed to come off. Changing the output Z for headphones required the faceplate, tubes, and chassis cover to be removed. With the 8900, the Z-out jumpers are in the front, separated only by the faceplate and four screws. The chassis cover is now split into two parts that can be separately removed, with caps now accessible (within limits) by removing a small plate on top. Much better.

    -Doing things a little different with the DHT platform. The auto detect circuit for both 2A3 and 300B tubes is pretty interesting and the option to toggle between feedback and no feedback makes for some nice diversity with tuning options while avoiding the analysis paralysis of having too many options. Also, little things like improved tube sockets, a sturdier faceplate, and a relatively small footprint, sweeten the deal.

    -Going the extra mile with managing tube hum. The previous models were pretty good with hum but sometimes it took extra measures to make sure the wire from the OPTs wouldn’t touch the power transformer (avoids crosstalk). With the 8900, the PT now has discreet windings for left and right channels that minimize interference, the filament regulators and rectifier diodes have been improved, and the OPTs now have wire management notches built on their PCBs (props to Victor on this one). I’ll talk about it more in my sound impressions but it’s worked out very very well…
    image2.jpg
    Electrical tape. It's not pretty but it does the job... Which is something you should never say about your girl.






    Damnit Elekit, I wish…

    -There were labels on the resistor packets. I never ran into this problem with previous builds as I’ve always opted for the Takman upgrade offered by Victor. He used separate packets for each resistor value and clearly labeled them. A pretty basic feature, or so I thought…. There’s no labeling whatsoever on the stock packets and Elekit expects you to match values by reading the color coded bands on the resistors. Cool, but then the manual happens to be in black and white and some of the colors are pretty similar (gold, yellow, black, brown ect). Sure, you can easily mitigate this with a multimeter (shown earlier) but why force our hand? Oh yea, the stock resistors are also smaller than the Takman and some bags mix resistors of different values. God forbid you ever mix one of these up on the board……


    -There was a detachable plate at the bottom of the amp. I’ve acknowledged the ergonomic improvements made over previous models, much appreciated but it’s not enough. Let me explain by going back to the resistor scenario. So what happens if you dun fucked up? Best case scenario, the circuit protection kicks in and the LED lights up red. At which point you’ll have to whip out out that multimeter to isolate where the the voltage ratings are off spec. Then comes the fun job of disconnecting the transformers without cracking the board (extremely annoying), unscrewing and removing the board (pretty annoying), desoldering and swapping resistors (meh, not so bad), and then putting it all back together praying you didn’t miss anything else. This applies to anything that involves accessing the underside of the board such as tightening speaker posts, troubleshooting, rolling xyz ect… No bueno.


    In the Future it Would Be Cool to See…

    -Designs based on more esoteric DHTs. 2A3s are great and some like 300Bs but they’re like the apples and oranges of DHTs. Everyone’s tried them or at least knows what they are. How about an 10\801A or even a Type 50? I know the former can sound sublime and nothing glows quite as nice as thoriated tungsten.
    IMG_1434 2.jpg

    -Output transformers from different vendors, maybe even different core materials? The Lundahls have been great, and going with amorphous cores this time turned out to be a pleasant surprise. How about nanocrystaline or permalloy next? Last time we talked, Victor mentioned using Hashimoto opts in a different kit (probably Sun Valley). I thought why not go with something like that for the next model? There’s also, Tamura, ISO, and Monolith Magnetics among others I’ve liked.


    -An über Elekit using interstage transformers and DHT drivers. Something like a 26\50 with permalloy transformers all around. Considering the design philosophy of Elekit’s Yoshitsugu Fujita, this is most likely a pipe dream. Guess, I’ll have to figure this one out on my own (to be continued…)



    Up next are sound impressions. Here I'll discuss how the TU-8900 stacked up against some other contenders I've had experience with. The EC Aficionado, Elekit's own TU- 8800, DNA Stratus, and my current reference SET...
     

    Attached Files:

    • Like Like x 9
    • Epic Epic x 9
    • List
    Last edited: Oct 23, 2021
  4. Maven86

    Maven86 Almost "Made"

    Contributor
    Joined:
    Jan 19, 2016
    Likes Received:
    395
    Trophy Points:
    53
    Location:
    DMV
    The impressions I recorded from various test periods were very different. I’m currently at the 90(ish?) hour mark and things seemed to have stabilized. I do perceive some changes but they’re so subtle that I’ll willingly concede to placebo. Here are some thoughts…


    Test conditions\Rig info


    Qobuz, Tidal, or HQPlayer (native)--> iMac 27” usb——> Singxer SU-2 + 10M SRS rubidium clock with LPS ——> Schiit Yggdrasil A2 XLR out ——> Jensen 1:1 input transformers RCA out——> Amplifier

    Impressions were done with the 8900 in Non-NFB mode unless otherwise noted. Tubes used were NOS RCA 2A3 and Sylvania 12BH7 gray plate.

    Most of the test tracks used for AB or critical listening can be found here:

    https://forum.headphones.com/t/headphone-evaluation-test-tracks/348/174

    Headphones and speakers:
    Focal Utopia*
    ZMF Verite
    ZMF Aeolus

    Omega Super 3 HO monitors
    Coherent Audio GR-8 (coaxial radian drivers)

    As far as musical tastes go, I listen to just about every type. I sometimes go through extended spells of spamming one or two genres depending on mood... Lately it's been classical and metal.

    In terms of sound preferences... I generally like things slightly warmer than neutral but ultimately care more about tone and timbre...which I understand is extremely subjective. That being said, I tend to evaluate things more holistically instead of picking apart individual traits, unless something really stands out.

    Regarding my impressions: Preferences and tastes are very subjective. Assume phrases like "In my experience\opinion\system", and don't take anything I say as an absolute. It saves me the frustration of using those mitigating phrases all the damn time and helps free up my writing. I don't always write with nuance but I tend to (mostly) think along those lines.






    General Impressions.

    -Right off the bat, this amp initially sounded like trash. For the first 10 hours or so, the timbre was completely off. Steinways sounded like crappy keyboards and cymbals sounded like someone was was banging on an empty tin can with a spoon. It was so bad that I thought I must have screwed something up. I re-measured the test points and found nothing remarkable…

    -Things started sounding more “normal” after 10 or so hours but the overall sound signature still sounded much different than where things are currently. Timber was still somewhat artificial but not terrible, the upper midrange and above also sounded a bit peaky and unnatural.

    -As far as pairings go, it did well with everything I had around but had very good synergy with the ZMFs and Omega speakers. The Aeolus pairing was especially notable and sounded very fun with almost everything except maybe classical music. However, I ended up using the Utopias for most of the critical listening as they were the most transparent while still being well rounded.

    -As for tube rolling, I didn't have anything fancy like XXX monoplates. Just some garden variety NOS RCAs from GE, RCA, and Ken Rad. For the most part, the differences weren't huge when rolling 2A3s, just some slight changes in presentation. I felt that the 12BH7s had a much bigger impact. I tried 3 different pairs: Tung Sol black plates, Matsushita gold pins, and the stock Sylvania grey plates I got from Victor. All were decent but I preferred the more forward presentation of the Sylvanias. The Tung Sols had a huge soundstage but were way too relaxed and kept making me sleepy. The Matsushita's were somewhere in the middle in terms of presentation but with a sweeter tone than both. I haven't tried any 300Bs.

    -This is one of the quietest tube amps I’ve heard. After some adjustments, the previous Elekits were very good in this regard and I never really had any problems with noise. After using the 8900, the background of just about every amp I’ve owned seem almost gray in comparison. Whatever engineering improvements or unholy pagan rituals were done to make this happen, worked exceptionally well. Unless you have a faulty tube or something, the (perceived) noise floor is ridiculously low and the background is inky black.

    -Even after break in, this is still very much a modern sounding SET. This typically means less bloom, faster\sharper transients, less perceived roll off in the frequency extremes, and a somewhat more forward presentation. However, with NFB turned off, it’s a little more organic\natural sounding than previous Elekit’s I’ve tried. I've heard some feedback about OPTs with amorphous cores having washed out tones or sounding unnatural... I thought this might have been the case initially but after run in, my experience has been the opposite. It could potentially just be the caps balancing things out, but I actually prefer this tone over the other Elekits I've heard. I'd describe it as slightly dense and somewhat vivid without being romantic. The upper midrange and up is excellent on this amp. Very detailed but smooth. There isn't too much heft in the bass but there's definitely extension, texture, and punch. This is my preference, even for most modern music like EDM, but I can definitely see certain people wanting more.


    -I switched back and forth between feedback modes over the span of a few days. The differences were quite noticeable. With NFB turned on, things were cleaner, clearer, and more controlled. Things sounded “proper” and “polite”. Bass was not as impactful but seemed tighter and more extended. Treble was a little sibilant… When NFB was turned off, transients had more slam, edge, and impact. Soundstage, especially in terms of depth increased noticeably along with improved imaging. Treble was smoother… I personally preferred things with NFB turned off. To me, things sounded more engaging albeit a little exaggerated. Feedback made things a little too “behaved” sounding. Nevertheless, preferences will end up being very system dependent and I can definitely see people preferring things with NFB on. Finally, gain is higher with NFB off, this can annoying if your source has high output voltage.



    After the 70 hour mark, I figured it was fair to start making proper comparisons.



    Comparisons

    -The closest known references I’ve either owned or extensively used, which are comparable to the 8900, are few. The best are probably the DNA Stratus, EC Aficionado, and Elekit’s, 8600, and 8800.


    -Of the the four the Stratus is the most difficult to compare for a number of reasons. In terms of sound signature, it’s the least like the others in that it leans towards a more traditional SET sound. Furthermore, the system I had the Stratus in was very different than my current setup so I’m reluctant to compare technical performance. That being said, I am comfortable to say that I preferred the tone and signature of the Stratus relative to the other 3 amps, even when I cycled through different sources. Signature wise, the 8900 falls somewhere between the Stratus and the others but is closer to the modern side of things (say 30\70?).


    • The other three were all used in the exact same system. Direct AB comparisons were also made between the three, with either notes, or commentary made to other people for references. With the exception of the TU-8600, I’ve owned all the others for at least 6 months. The 8600 was in my system for ~3 weeks before being returned to its owner, but was IMO, enough time for me to get a fair impression of its qualities.

    • Of the three, I liked the EC Aficionado the least. While I didn’t quite jive with its overall presentation, I can certainly acknowledge its strengths. It was a little more resolving than the other two with better stage depth and imaging. The 8600 had a slightly wider stage but the AF went noticeably deeper. IMO, this is its biggest strength over the two. Both DHTs did micro dynamics better than the 8800 but AF still had a slight edge over the 8600 in that department as well. This was subtle and was mainly noticeable in good acoustic recordings with strings or woodwind instruments where the AF did the nuance thing a little better.

    • I liked the way both Elekits did macro-dynamics more. This was especially true with the 8600, which took things a step further. However, even in triode mode (on the 8800), everything was livelier, genres like rock and metal were just more visceral and tactile, orchestras …more grand. I’d take it a step further than simply saying they had better “macrodynamics”. I think they just did a better job of reproducing that “energy” you feel in a room during a live concert or performance.

    • Tuning aside, I felt like these three amps were more or less peers in terms of technical performance. Maybe the Aficionado might have been that one kid in the class who was a little bigger than the others and occasionally ran your pockets in the hallway… However, tradeoffs considered, the net results were similar.

    IMG_1048 (1).jpeg
    Some ABing while the AF was still around.




    • In this respect, the 8900 was at least a tier above the other amps in most categories. To me, this was especially true when considering the net effects as opposed to individual “traits”. In terms of resolution and microdetail, the differences were very noticeable when moving from the Aficionado, even without adequate “burn in”. I wasn’t able to compare the two extensively as the AF shipped off a couple days later but it was one of those things that wasn’t subtle. Everything was sharper and more defined. Instruments and voices had more depth and texture even on simple tracks. For example, the resin grit and “roughness” from violins was something I thought the Aficionado reproduced well. However, on the Elekit, this characteristic wasn’t just “present” but had a more dynamic element to it. There was much more nuanced gradation where you could now hear more variations in texture, at even the slightest shifts from the musicians bow.

    • Live music widened the gap even more when compared to the other amps. Going into specific details about the differences seems almost redundant. Saying it just presented more “detail” would be an understatement. Yes, there was more “information” present but comparatively, the difference wasn’t that stark in the grand scheme of things. The biggest contrast was with how those details were articulated, which IMO, happens to be what the 8900 does best. When making the switch, there was a much more convincing rendering of space and ambience of the recording with individual sources of sound having their own discrete sense of presence.

    • For instance, when listening to a rendition of “Vivaldi’s: Winter” in the “Freivogel & Voices of Music” soundtrack on my old Elekit TU-8800, everything sounded pretty good. Using the Focal Utopia, the stage is fairly compact but more than enough to hear the separate lines of music between the first violin and the rest of the instruments in the chamber orchestra. There’s a good amount of detail in that the instruments sound convincing and palpable. If you listen closely, you’ll even hear the hallmarks of a live recording like the random cough or sniffle from the audience. As the piece reaches its climax, you can feel the energy of the build up and there’s a lot going on at once. Despite this, nothing feels jumbled or congested and when things finally end, you’re surrounded by a tangible applause of an excited crowd from both sides.

    • Switch over to the 8900 and things get eerie. The stage widens considerably and deepens even more. The first thing that sticks out is the reverb from an AC. I’m sure it was there before but this time it sticks out like a sore thumb because of how black the background is. As the violins start playing you can notice that you can now easily pinpoint their orientation relative to the mic, not just in width but in depth. Each one has their own space in the stage and you notice that the second one from the left is a little behind on her tempo. While this is going on, you notice a lot more going in the crowd. Some guy is being an ass and tapping his palm on his thigh. That random cough you heard during the first take? Now it’s almost as palpable as the lute playing a few feet behind and you know it’s coming from a guy sitting near the front right side of the room. Speaking of room, given the reverb and decay you’re hearing, you can probably get a good sense of its size and height. Again, the piece starts reaching its climax but this time you hear the gradual build up of each individual instrument in sequence, all at their own pace, until everything blends into one massive crescendo. The words “jumbled” and “congested” don’t even cross your mind as nothing feels strained or forced. Finally it ends and the crowd starts clapping. What sounded wide but flat on the older Elekit is now a convincing illusion of 3D space in which applause comes from every direction with each one has it’s own sense weight and cadence. Hyperbole? Sure. But I honestly don’t think I’m far too off.

    • The 8600, probably had better “heft” and body in terms of macrodynamics, but I felt that 8900 may have articulated things better even in this category. Unfortunately, I wasn’t able to compare the two directly and so I’ll leave it at that. However, in terms of transient performance I haven’t owned an amp that does things better.

    • The only other amp I have that’s comparable in performance is a custom SET that use 801A tubes. However, the sound signature is quite different and closer to that of a classic DHT amp. It’s a dual mono design with a discrete power transformer for each channel. All chokes and transformers are ISO-Tango.
    IMG_1404 2.JPG

    - From the upper midrange up, along with the sub-bass I prefer the presentation of the 8900. It sounds liver in those regions while the 801A takes a more relaxed approach. The 8900 also has more bass punch in the lowest registers.

    - The closer you get to the midrange, the more the 801A pulls ahead. If I were to describe the signature in terms of headphones. Think HD650 with the transient performance of the Utopia and the staging of the HD800. The midrange is colored but very detailed and rich. It's not the most accurate sounding amp but it is a beautiful sounding one.

    -In terms of raw resolution, the 8900 probably has a small edge. It's more "detailed" but the 801A sounds more "real". Soundstage wise, the 801A is noticeably better. It's the only amp I've owned that can give an out of head experience with the Utopia. The 8900 gets close but isn't quite there.

    -The Elekit is a lot more well rounded and sounds better with most things, while the 801A was tuned for the Utopias and sounds best with it. My ZMFs and HD650 don't pair nearly as well.


    Closing thoughts

    Overall, I'm pretty happy with my 8900 and think it's Elekit's best model thus far. If you can jive with it's sound signature, it more than holds its own against other amps around its price range. In fact, out of all the amps I've tried in this range, it's easily my first choice.

    A few people have asked whether I still think the 8800 is still a good value alternative. Considering my 8800 will shortly be up for sale, I feel like I'm shooting myself in the foot... But what the hell, this wont be my worst life choice. The simple answer is no, at least not for me. Even with the fancy caps, I actually spent LESS, on the 8900 by foregoing takman resistors, and considering one other overlooked fact. The 8800 uses four coupling caps while the 8900 uses two.

    When using pricey caps, which everyone seems to be doing these days...the two extra caps add up. In terms of my personal taste, even with my favorite tubes, I can't think of anything I actually prefer on the 8800 in terms of tone or presentation. As far as technical performance goes, I'll say it again, these aren't peers. If you want my more candid inner thoughts, I respectfully feel that the 8900 takes an explosive soupy s#&t on the 8800... But hey, players wanna play and tube rollers wanna roll. The 8800 will still have its fans and that's perfectly fine.


    Anyhow, that's pretty much it from me. Hopefully, in a couple of weeks we'll be getting some new impressions and maybe even some counterpoint from someone with more street cred in these parts. @yotacowboy the stage is yours.
     
    • Epic Epic x 12
    • Like Like x 9
    • List
    Last edited: Nov 9, 2021
  5. EagleWings

    EagleWings Friend

    Pyrate Contributor
    Joined:
    Jul 20, 2017
    Likes Received:
    1,709
    Trophy Points:
    93
    Location:
    India
    Good stuff Maven. Regarding the way you have mounted the caps, have you not soldered the leads to the turret pins? I was thinking of this exact idea of putting the caps on the bottom plate. But I realized, if I had to take my board out tomorrow to trouble shoot or mod, I’ll have to desolder the caps first, unless I have a way to reach down and release the caps, specifically the one on the left, as the headphone board is directly below the gap. Btw, I was intending on using zip ties instead of a tape, as I might be moving my amp every now and then and wasn’t sure if a tape can hold the caps down. But then I haven’t figured out yet, as to how to secure them using zip ties.
     
    Last edited: Oct 19, 2021
  6. Maven86

    Maven86 Almost "Made"

    Contributor
    Joined:
    Jan 19, 2016
    Likes Received:
    395
    Trophy Points:
    53
    Location:
    DMV
    The leads actually are soldered onto the turret pins but as you pointed out, I soldered it in such a way where it would be easy to desolder if I needed to remove the board (which I have). I initially planned on using zip ties but didn't have any on hand. I opted for that temporary solution just to get things fired up. With the zip ties, since they'd be parallel with the vent slits at the bottom, I'd fasten two on each cap and run a long piece of electrical tape through them and wrap it around the vent slits. Still a spartan solution for sure, optimally it would be best to just get something like these:

    https://www.amazon.com/ASC-Capacitors-RS752152-Capacitor-Hardware/dp/B00DWI1RKO
     
  7. EagleWings

    EagleWings Friend

    Pyrate Contributor
    Joined:
    Jul 20, 2017
    Likes Received:
    1,709
    Trophy Points:
    93
    Location:
    India
    That’s a nice idea and I might take that route if I pick up Duelunds in the future. Given the Audyns have axial leads, I might be able to use zip ties through the slits in the bottom plate. But first, I plan on trying to squeeze the caps on the board itself. They are just a mm longer than the provided slot. So if I can file away a little of the PCB on either ends of the slot, I might be able to fit them. The problem would the turret would be too close to the place where the lead connects to the cap.

    I look forward to your sound impressions.
     
  8. Jh4db536

    Jh4db536 Friend

    Pyrate
    Joined:
    May 25, 2016
    Likes Received:
    926
    Trophy Points:
    93
    Location:
    Los Angeles
    Home Page:
    i used dsub pins to mount the CuTF couplers on the 8900 that i built for someone. It's a very tight fit but allowed to rapid rolling or playing with directionality.

    [​IMG]

    They also sell the amp with the Lundahl 2785B non amorph OPT. That's a 3rd option as well.
     
    Last edited: Oct 20, 2021
  9. Maven86

    Maven86 Almost "Made"

    Contributor
    Joined:
    Jan 19, 2016
    Likes Received:
    395
    Trophy Points:
    53
    Location:
    DMV
    That's an interesting solution, it's definitely a tight fit. Did you purchase your V-caps straight from Chris? I noticed that the ones Victor offers are rated for 400Vdc unlike the regular 600V I've typically gotten. I wonder if the 400V ones are smaller.

    Yup, the standard core Lundahls are also an option. Would love to hear the difference between the two in this circuit. Speaking of options, supposedly pure silver windings are on the table but when I asked for a quote....:eek:

    All of a sudden the song "Window Shopper" started playing in my head...
     
    Last edited: Oct 20, 2021
  10. Jh4db536

    Jh4db536 Friend

    Pyrate
    Joined:
    May 25, 2016
    Likes Received:
    926
    Trophy Points:
    93
    Location:
    Los Angeles
    Home Page:
    the owner of the amp bought these 600vdc Vcap from Victor and dropshipped it to me. I ended up buying my own pair to try in another amp that i built and have on hand for rolling purposes.

    I am not opposed to buying silver trafos from Lundahl even at that price. The issue for me is that i'm not sure if i prefer Lundahl.
     
    Last edited: Oct 20, 2021
  11. Maven86

    Maven86 Almost "Made"

    Contributor
    Joined:
    Jan 19, 2016
    Likes Received:
    395
    Trophy Points:
    53
    Location:
    DMV
    Oh I completely agree. If you're looking at things relatively, the Lundahls can even be considered a bargain (as crazy as it sounds). I've also heard about amorphous and silver going well together, hence my inquiry with the Elekit but it's way to expensive of a gamble for me right now.

    That being said, I was actually very close to going silver with a pair of custom interstage transformers I got recently. These were permalloy core so I was very curious about the net effect. In the end, it was still a little over my budget...maybe one day.
     
  12. Jh4db536

    Jh4db536 Friend

    Pyrate
    Joined:
    May 25, 2016
    Likes Received:
    926
    Trophy Points:
    93
    Location:
    Los Angeles
    Home Page:
    I'd have to build an entirely new amp to use those lundahls. I like the sound of low dampening factor primaries. I personally wouldn't go higher than 2.75k for 2a3. So they would need to be customized. Tough choices that limit what applications i could use it in or roll into.

    my current experimentation is using amorphous materials in low level signal and low DC application. So that would mean input, line level, passive gain stage, and interstages. also, THAT solid state buffer stages combined with AM lundahls caught my interest as implemented in the new ECP dac, have had some really interesting observations with the THAT buffer combo even in this Nutube thing i built.

    I like the use of silver transformers no matter what the application (input, put, but it is theorized that the high price performance ratio is usage in secondary windings. I like the sound of iron in the output i guess. Maybe i need to try monolith summit plain ole iron OPT, i have a bunch of stuff in my room that i need to try including vintage telefunk and i want to try hashimoto as well. More $ is not always better.
     
    Last edited: Oct 20, 2021
  13. yotacowboy

    yotacowboy McRibs Kind of Guy

    Pyrate Contributor
    Joined:
    Feb 23, 2016
    Likes Received:
    10,695
    Trophy Points:
    113
    Location:
    NOVA
    Home Page:
    Aside: Searched but couldn't find anything using Google-fu. Is/are there any discussions here about the construction methods and materials as they pertain to audible traits in different use cases for transformers (or even other electrical components)? I.e., as interstage/small signal versus output? It's one thing to read "amorphous core transformer" and know what that exactly means per the materials and construction method, but for me at least, beyond hand-waving or just stating things like, "there's less magnetic saturation than a hammond" or something, could we have a thread dedicated to these types of things? Or how different transformer manufacturers have certain house sounds? I'm game to call up Victor to see if I can get a hold of a pair of amorphous core transformers to compare between the stock Atene's, versus the LL2785s, versus the LL amorphous cores. This is a lot like rolling coupling caps where materials and construction methods (including secretive, pagan-worshiping processes of setting gaps using certain materials that compress over time?) have audible affects, i assume? Or am I an idiot?
     
  14. Maven86

    Maven86 Almost "Made"

    Contributor
    Joined:
    Jan 19, 2016
    Likes Received:
    395
    Trophy Points:
    53
    Location:
    DMV
    To answer your last question, not at all, these are excellent points. The problem is that in terms of materials, manufacturers don't agree on what's "best". I think this partly due to the subjective nature of audio but mostly due to different design philosophies that each transformer winder has. As far as amorphous cores go, some winders don't like using them (Electraprint, Hashimoto ect.), others will swear by them and by extension, nanocrystaline cores (Monolith, Tribute ect.) It's definitely worth a discussion, perhaps via PM or in another thread.
     
  15. Jh4db536

    Jh4db536 Friend

    Pyrate
    Joined:
    May 25, 2016
    Likes Received:
    926
    Trophy Points:
    93
    Location:
    Los Angeles
    Home Page:
    Transformers are like voodoo audio art more than a science. Implementation and application matters a lot as well. "best is subjective". More expensive doesn't guarantee anything.

    There's a lot of information on the remnants of a collapsed star that lead to this universe that we call the milky way Sbaf

    For example, there are many amp reviews where the magnetics are known (vs "custom wind") or you might have purchased or heard these amps at a meet. I have said many things myself in many threads. Mr Marvey has provided plenty of insights if you know what to look for.

    I have accumulated a lot of datapoints on magnets/capacitorss through loaners, meets, tinkering, rolling opt, building amps for people, modifying amps, etc.
    [​IMG] [​IMG]
     
    • Like Like x 6
    • Epic Epic x 1
    • List
  16. Maven86

    Maven86 Almost "Made"

    Contributor
    Joined:
    Jan 19, 2016
    Likes Received:
    395
    Trophy Points:
    53
    Location:
    DMV
  17. RestoredSparda

    RestoredSparda Friend

    Pyrate
    Joined:
    Jan 7, 2021
    Likes Received:
    3,638
    Trophy Points:
    93
    Location:
    WI
  18. Maven86

    Maven86 Almost "Made"

    Contributor
    Joined:
    Jan 19, 2016
    Likes Received:
    395
    Trophy Points:
    53
    Location:
    DMV
    Appreciate it. Build Impressions Part 2 = Part 3. Sorry, I should probably change the wording...

    Edit: Fixed
     
  19. Whitigir

    Whitigir New

    Joined:
    Nov 30, 2015
    Likes Received:
    14
    Trophy Points:
    3
    Man, I agree with you that the resistors has no detailed informations, and that Elekit actually mixed them here and there in different bags. That is pretty “harsh”!!! Like tricking you into a magical world with smokes or so .

    congratulation on such a sweet built. I also had mine built and took some notices down as well. If anyone want to take a look, I have got it here
     
  20. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,778
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    What @Jh4db536 said. It's all over the place. Put aside stuff like core material, interleaving different laminations, interleaving windings aside. Trying to get a grasp of that stuff gets in the way. Personally, I take an initial measurements approach to check bandwidth (no, we don't want -3db at 30Hz or 18kHz) and make sure no peaks at 50kHz. It's basically a simple "is this not shit" test before going to the next step: does this sound good? This assume that you have the transformer sized and spec'd to your needs: primary inductance, current, expected loads, etc.

    I'm not even sure that there is even a sound to amorphous, iron, nickel, nano-crystalline. There's probably more of a house sound to each manufacturer IMO.

    I suspect the Monolith plain iron core to be fantastic.
     

Share This Page