Elekit TU-8200 DX Headphone/Speaker Amp Review

[Dimu]

I think I'll take it apart next weekend and give it all the once over to see if there is something obvious such as a bad solder point or something?

Bad solder joints aren't visible, often not even under microscope. That's why welders test their work using xrays. The only way to deal with a suspected cold joint is to reheat properly every joint- this will reflow the solder. Sometimes the problem is with surface prep in which case even reheating won't help- need to clean up and remove whatever is preventing solder from joining.

 

[ctrlm]

Bad solder joints aren't visible, often not even under microscope. That's why welders test their work using xrays. The only way to deal with a suspected cold joint is to reheat properly every joint- this will reflow the solder. Sometimes the problem is with surface prep in which case even reheating won't help- need to clean up and remove whatever is preventing solder from joining.

Yeah - I was afraid of that

 

[dawktah2]

Bad solder joints aren't visible, often not even under microscope. That's why welders test their work using xrays. The only way to deal with a suspected cold joint is to reheat properly every joint- this will reflow the solder. Sometimes the problem is with surface prep in which case even reheating won't help- need to clean up and remove whatever is preventing solder from joining.

I did this and still have crackle I posted about earlier. I am looking for troubleshooting help, now.

 

[Dimu]

First experiment- JFET input with OPA1642. Sounds much better, finally I like the sound I get out of vinyl. Really noticeable improvement. The next step will be OPA1641/LME49990 hybrid where OPA1641 handles initial pickup from the cartridge and LME49990 drives the tube in the preamp. Might even play with gain between two stages to shift more gain towards LME49990.

I like TU8500 a lot after the updates that I did, but there is a bit of topology trouble as left channel picks up more hum than the right channel. Not audible, but visible with oscilloscope.

 

[Dimu]

Had a musician friend visit me yesterday. Played his studio-recorded tracks via Gungnir Multibit and TU8500/TU8200 on LS50. He liked it a lot, said reproduction was excellent including both voice and drums. He was surprised by base actually. The tubes I currently have are Mullard 12AT7 in TU8500 and Dutch E80CC with Telefunken EL12 spez in TU8200.

 

[Dimu]

And now I am breaking in this beauty, arrived today from Japan. Tube rolling, op amp rolling, cartridge rolling... I guess speaker rolling is next

 

[Dimu]

I think I am finally happy. Built a JFET/bipolar hybrid today- input stage OPA1641 at half gain and driver stage LME49990 at full gain. Hum is much more controlled than even with OPA1642 at half gain. Sounds great with Nagaoka MP-200, I like it better than OPA1642, which was in turn much better than anything else I tried before, with Burson being the worst offender due to horrible hum. I might as well stop here. The only other idea is to try dual LME49990 but that might get more noisy on the input end.

I think this would be my recommendation for anybody looking to spin vinyl into TU8200- build a TU8500 with Mundorf caps, TDK pot, improved bypass for opamps, halved gain for the input stage and these babies:

 

[Dimu]

Tried dual LME49990. Sound is awesome. As expected noise is slightly higher- at full amplification of TU-8200 I measured 60 dB at the speaker for dual LME49990 and 59 dB for hybrid OPA1641/LME49990 version (first stage gain is still suppressed to 50%). I think I will stick these into spectrum analyzer to measure actual difference. I now have five different opamps that I can compare- the original opamp, the upgrade recommended by Victor, OPA1642, OPA1641/LME49990 and LME49990/LME49990. With two channels in the amp that's 10 measurements- sounds like a fun project...

 

[Effusion]

@Dimu very cool! I'm a big vinyl fan as well and after following everything you recently did with your rig, I may have to do something similar to mine in the future too. I've been down the crazy opamp road before in other equipment and I would roll quite a few in the past, so I know how much of a difference they can make to the sound in the end and how satisfying it is to finally find that magic one.

Also, that MP-200h looks amazing! I'm sure it sounds just as good as it looks too. There is just so much to getting great sound out of vinyl, from the cartridge, to the table, to the phono preamp, and beyond. I always get a bit of a chuckle when vinyl newbies initially think it is as simple as their digital rigs and then find out how complex it truly can be. In the end though, vinyl offers something that digital audio really just can't, not that it is superior in anyway, just different. I actually prefer certain recordings on vinyl versus digital and even with a few, when I hear them digitally now, I can't help it, I just think about how they sound different on vinyl while listening digitally.

The crazy thing is how vinyl has been resurrected in recent years as CD sales have fallen off the map... while digital downloads/streaming is so much quicker/easier, there is a disconnect from the physical media that I believe can create a hidden void in one's psyche. There is just something to getting a new record, examining the artwork/cover, feeling it in your hands, and the physical action of putting a needle to it. Then there is always the spontaneous nature of it, the random pops/crackle, etc. add to the overall presence of the sound. The great thing is how many new albums offer digital downloads with the vinyl versions, so one can now justify the extra cost of the record. Also, some of my favorite genres on vinyl are ambient, classical, and jazz, as it just seems to add so much extra character and nuance to the sound, creating an environment where the music just seems more real or in the moment. Now, I do know a lot of guys who stay away from vinyl like it is the plague, but that is really just because they burnt-out on it back in the day and it is just not as simple as digital music or they just have never been exposed to it properly, but I think they are sometimes missing the point. It's not that we think vinyl is superior or even equal necessarily, although with some recordings it actually is, but different in such a way that we feel closer to the music in general when we experience it.

Keep letting us all know how it goes!

 

[Dimu]

Well here is a DUH moment- went poking into those opamps with oscilloscope and could not figure out why the output is different between the two channels until I noticed that the channels are wired upside down- right channel has second op amp as input stage while the left has the first op amp as the input stage. So I need to flip op amps in one of my hybrids and I have to watch which one I put into which channel... I kept looking just at the left channel on the schematic assuming that they are symmetrical. Turns out, for some funny reason, they are not.

 

[Dimu]

Regarding opamps- waiting for SOIC AD797 to throw into the mix. Will also flip one of the hybrid OPA1641/LME49990 at the same time.

I looked into RIAA equalization filter to see where it is at. Shifted balance point using the additional resistor network at the top of the board and tuned one of the caps with additional polysterene caps (bottom of the board). This aligned the filter component measurements to theoretical calculations. Then measured frequency response with LME49990 relative to RIAA curve. At the output of TU8500 top end is within 0.1 dB of ideal levels (the last column in the measurements table). Bottom end is suppressed though, will need to tweak that further and re-test. And I want to play with noise- I want the phono amp to be quieter.

 

[Dimu]

So I cleaned out phono section today:

then soldered back better metal film resistors and low pass filter caps with polypropylene dialectric. Dropped input impedance to 47k.

Back is now cleaner too- only bypass caps remain on the back side:

Running frequency response traces now on the whole unit at max volume with 3x gain switch and MM mode switch.

 

[Dimu]

Traces confirmed what I measured earlier. Here are left and right channels below 1 kHz (100 Hz per division)- there should have been a bit more drop- supposed to be 19.35 dB:

These are left and right from 1 kHz to 21 kHz- perfect 20 dB drop:

These were measured across whole TU-8500- from phono input to the output, at max volume setting. Channels track each other perfectly both below and above 1 kHz- that is because I used highly matched components on both sides.

 

[Dimu]

Results of noise measurement on op-amps below. Measured using SPL1000 at the left speaker with TU-8500 and TU-8200 volume turned up to the max setting.

original JRC 2068 DD - 54 dB
OPA 2227P - 58.5 dB
OPA1642 - 56 dB
AD797/AD797 - 55.5 dB
OPA1641/LME49990 - 55 dB
LME49990/OPA1641 - 59 dB (this is simply flipped installation of hybrid opamps between left and right channels)
LME49990/LME49990 - 59 dB

Here is what I think based on this:

1. the original op amp has the lowest noise- that is a surprise!
2. predictably LME49990 in the first stage leads to high noise due to high current noise of bipolar input.
3. AD797 has very respectable result.
4. OPA1642 is not great likely because of the second stage noise (voltage noise due to JFET architecture).
5. OPA1641/LME49990 has the lowest noise of high quality opamps. This is what I expected, except JRC 2068 DD is somehow quieter...

Based on these measurements I decided to choose between AD797 and OPA1641/LME49990 hybrid. So I did a listen test. And I liked the hybrid better. Sound is more dynamic, more punchy. This is subjective of course.

So I am leaving the hybrid in there. First JFET stage has half the gain of the second bipolar stage (I put 4.7K resistor instead of 10K in the feedback loop of the first stage). Basically first stage picks up MM cartridge signal and drives the RIAA filter. Actual amplification is done by LME49990 in the second stage. TU-8200 tubes add just 3x gain on top, followed by two driver stages and output stage of TU-8200.

Listening to this whole setup as I am typing it- sounds awesome!
If you want turntable in your future grab a TU-8500 while they still make these kits!

 

[Dimu]

This is to document how hybrid opamp is installed with mixed up channels. The picture is upside down because it is taken from the back of the case- left channel is on the right. The top op amp is #1 and the bottom is #2 in these adapters. In the left channel (right on the picture) there is a stripe- this is OPA1641. The right channel has top opamp with first pin marked with a dot- this is LME49990.