Muse M20 EX2 TA2020 amp pictures and some info

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I just got my Muse M20 EX2 TA2020 amplifier and I wanted to share some info and maybe get some ideas.

First impression out of the box is that it seems like a quality product despite the price. Feels solid with nice finish. The volume knob is aluminum and so is the power push button. Gold plated connectors on the backside and a psu connector.

I quickly opened it and had a nice surprise. No parts are missing and everything seems to be laid out real nice and tight. The board has a white coating with parts notation almost everywhere. Most parts are hole mount aside from 4 Surface Mount Schottky Power Rectifier that are between the inductors. On the backside things are this nice thou. The board is dirty and solder points are not perfect with some pins not being perfectly covered with solder. But this can be fixed as I will desolder some parts anyway and it will be easy to remake the bad solder points.

Components wise this amp is really neat. For 35$ delivered (no PSU) you get an aluminum case with aluminum knobs. Nice connectors, the ta2020 chip of course and for power there are 2 paralleled Rubycon 470uF/24V caps (one is ZLH and one ZLJ thou). There's only one more electrolytic on the board, Nichicon Gold 100uF/16V FW(M). This is near the opamps and tripath chip. The coupling caps are Epcos 0.22uF 250V polyester (+ - 20%). The large inductors are Toko 10uH rated 5A (that's what I got while looking for them online, might not be accurate). Opamps are 2SC1815.
I am using a cisco router psu rated at 12V 5A :D This is temporary until I get my 13.5V smsp.
I managed to listen for 3-4 hours last night. I was very pleased from the start. I couldn't say more as I used some inefficient speakers that I have right now. And they are bi-amped so I had to listen to tweeter+mids only. I selected some vocal and jazz for that matter. I turned it all the way up and had no problems. On the back of my speakers they are rated at 120W(music) 8 ohm for highs+mids. Still this amp managed to make them loud. I had to back it up at half volume for relaxed listening. I also connected it to the bass driver on both speakers and played some bass tracks. It did manage to move the membrane quite a bit. At some frequencies my glass windows began to resonate. Also full volume with bass tracks the led from the volume knob didn't flicker at all. So until I manage to build enclosures for my fostex speaker I decided to mod this little amp so I squeeze some more quality out of it. I decided to replace the power cap with a BC (vishay) 4700uF/25V 150 RMI low ESR cap. Even thou it's large I will rest it on one side and it has plenty room. The decoupling caps I will change for some Wimas as that's what I have at hand. I might go some other route if it's worth it. Also I don't know the right spec. Now it has 0.33uF, and I have at hand 0.22uF MKS4, 0.47uF MKP10 and 1.0uF MKS4. I would't stretch for more than 10-15$ for 2 caps so shoot away if you have any ideas in this range.
For the inductors I was thinking of getting the amp6 inductor kit from 41hz. It is 10uH 5A rated and I like DIY-ing so it will be fun. I had some issues with 41hz in the past, I payed for a amp6 kit and didn't receive any update for about 1-2 months. I decided to get my money back from paypal and now I'm reluctant at dealing with them. As a back-up I decided to buy some ferrite cores and wind them myself but I got stuck as there is no info on the inductance of the material of the cores. I got exactly this item Ferrocore . If anyone knows how to wind this for 10uH 4-5A I would very much appreciate that. I also got some 0.5mm copper wire. I saw this gauge used in the 41hz kit but wondered on how do they rate it at 5A as this gauge would only support 3.5A max. So if needed I could get some 0.7 or 0.8mm. For the rest of the caps I have no idea on what to make of them as I have no schematic and the Lepai is different in layout as it has tone controls. Also is there a way in which I could tell if I have metal film resistors where I should? Attached some pictures with the internals.
If you have any ideas on improving this amp please share them.
 

Attachments

  • 1.jpg
    1.jpg
    522.7 KB · Views: 1,185
  • 2.jpg
    2.jpg
    484.8 KB · Views: 1,089
  • 3.jpg
    3.jpg
    432.7 KB · Views: 1,048
  • 4.jpg
    4.jpg
    627 KB · Views: 1,065
  • 5.jpg
    5.jpg
    534.7 KB · Views: 995
  • 6.jpg
    6.jpg
    251.1 KB · Views: 366
Last edited:
An update with the power caps replaced by the 4700uF one. There is still room for larger decoupling caps.
First picture represents the way the cap went in. In second and last one it is soldered.
 

Attachments

  • 9.jpg
    9.jpg
    652.6 KB · Views: 491
  • 8.jpg
    8.jpg
    568.1 KB · Views: 495
  • 7.jpg
    7.jpg
    453.5 KB · Views: 453
Disabled Account
Joined 2002
The new 4700 µF cap has too long lead wires which does not improve things. It is better to use 2 caps just as the amp had but the new ones should have the same footprint and larger value/better quality. Measure the voltage from the power supply please if you can use 16 V (better 25 V to be safe) 2200 µF low ESR caps like Pana FC/FM.

The coils could be better as those are not 5 A rated IMO and the output relays are 2 A rated which is too light in my book. I would use larger value input caps.

I think I have better 10 µH coils in my stock which you can have for the price of the envelope and the stamp.
 
Last edited:
Hmm... the Tripath chips have separate V+ and Gnd pins for each channel's output, so they can each have their own electrolytic and bypass. This is also mentioned in the datasheet.

But at the same time... many implementations use a single cap for everything.

Well just something to consider if you are touching that part further.
 
Hmm... the Tripath chips have separate V+ and Gnd pins for each channel's output, so they can each have their own electrolytic and bypass. This is also mentioned in the datasheet.

But at the same time... many implementations use a single cap for everything.

Well just something to consider if you are touching that part further.

What part?
There's only one electrolytic besides the 2 from psu.
 
Tripath chip has multiple power supply pins, each channel's output MOSFET stage has one set of pins of their own.

That two bigger electrolytics that you replaced, some boards may route the traces such that each of them is closer to one of the power supply pins but further from the other. i.e. each channel get their own psu capacitor. Some boards ignore this.
 
Ah, ok.. On the back the traces are identical in the area of the caps. There's large traces for + and - and the caps are centered on them . I hope that I can source tomorrow 2 Nichicon FW 2200uF/16-25V that will be a perfect fit in the case. Also going to swap the output caps, hope I can find some small mkp/fkp in the bin.

What are those transistors for and can I put better ones if they affect sound quality? I found the original ones for almost nothing - 30 cents or something.
 
As small digression I 'think' I damaged my Lepai's Ta2020 Chip By tripping over a speaker wire and yanking it out of it's Speaker spring clip... when it was operating.
Left channel no longer produces Sound.
But shorting the Chips' pins from underneath with a fingertip produces a hum equally loud on both channels' speakers.
Any basic tests of Ta chip functionality possible?

Seems odd to buy a 'superior' to Lepai Ta2020 implementation..then have to mod it as well. I wonder if an Amp 6 might require less 'improvement'?
 
I figured it out with the transistors and the other electrolytic cap. They are for the speaker protection relays.
I managed to replace some caps. Cdo got swapped for wima 0.01uF/100V FKP and Cz for wima MKS 0.22uF/100V.
I have some questions for whoever wishes to help me out.
First, I saw that Co is made up from 2 different caps. One 0.47uF/100V paralleled with another 0.001uF/100V. I looked on the backside and they share both connections. Is it supposed to be this way? On the ta2020 spreadsheet there's only one Co (actually 2 for each channel), but I have 4 caps in total for one channel. 2x0.47uF and 2x0.01uF. If the 0.01uF one is not needed then that's actually a very good thing as I will get both out and put a larger one instead :D I might fit a 10mm lead one in diagonal.
Then another thing that I don't get is that I saw 2x330pF/630V (?) connected between the pot and ground. Pot is Alps A503 50K. If I set the sound to minimum (mute) then the cap is shorted out to ground. If I increase the volume on the knob it isn't shorted anymore. Is this creating a low pass filter? And is it ok? Could I do without it and short it out?

All that's needed now is to find some caps that fit for Co and maybe some 0.1uF/100V ones to replace the many rest on the board. They are really small and packed together so I need to find a package small enough. I found wima mks4 0.1uF/100V would be a perfect fit for all 0.1uF/100V on board. And Wima mks4 0.47/100V would almost fit perfect for Co. Would it worth it to go for MKP (polypropylene) wimas for Co and try to make it work with wires? Or would the difference be inexistent between MKS and MKP in my setup?

In the pictures the rectangle represents all 4 Co and circle is all 4 grounding pins (one channel only). In the other picture those are all the 8 Co caps.
 

Attachments

  • 10.jpg
    10.jpg
    326 KB · Views: 528
  • 11.jpg
    11.jpg
    819.6 KB · Views: 539
  • 12.jpg
    12.jpg
    558.8 KB · Views: 492
As small digression I 'think' I damaged my Lepai's Ta2020 Chip By tripping over a speaker wire and yanking it out of it's Speaker spring clip... when it was operating.
Left channel no longer produces Sound.
But shorting the Chips' pins from underneath with a fingertip produces a hum equally loud on both channels' speakers.
Any basic tests of Ta chip functionality possible?

Seems odd to buy a 'superior' to Lepai Ta2020 implementation..then have to mod it as well. I wonder if an Amp 6 might require less 'improvement'?

Not odd at all. In fact if you think about it I really do wonder why everyone buys that ugly case Lepai with tone control and numerous failures for different reasons. You as well ;) Ok ok it was your fault but you get the point...
The case is much nicer, has an actual Alps pot, that works as it should. You do not have any op amps (to bypass or) to fiddle with. No popping noises on start-up (lepai's put speaker protection at some point but didn't work that good). Lepai actually had some kind of chip protection that was causing the chip itself to fail. Kinda ironic...
Then there's the challenge to find good parts that actually fit and look nice. Price wise it's the same thing. I payed 35$ shipping included. Lepai goes for what..19-20-23$? I mean usual price not black friday type price. 12$ I'd pay just for the case, let alone pot, nice led glow not that overpowering flashlight, nice laid out board, good electrolytic caps by default (rubycon and nichicon), vishay decoupling caps, beads instead of inductors, aluminum knobs for volume and power switch. And the actual inductors seem pretty sturdy as well :) So, why not start better then continue to mod? Plus it is a great amp for anyone who doesn't want to replace anything, just listen to it out of the box without that Lepai type headache, crackling, failing, burnt parts, reversed connected ones etc. I do like the lepai thread thou, I learnt a lot from there! But I'd like to see as a challenge how nice and neat could one get this Muse amp inside :)
Amp 6 kit is something that you buy to put into a case that you buy and you just use it. No modding there, at least that's not it's scope anyway. You have fun assembling everything for the 41hz amp kit. It has it's own market. I for one was inclined to get one, even tried once last year but for some reason got no reply from them for 1-2 months until I decided to get my money back from paypal. But I wouldn't have tried again anyway to get another amp6 kit, too much work to find a case that looks nice, fits and is affordable.


Forgot to ask, would my amp get louder if I used a preamp?
 
Neither questioning Nor caring about your Consumer choices :) just wondering.

Yes.. it Does get louder with a Preamp... careful though, it's too easy to overload the
erm.. Fragile.. TA chip.
A Cheapy Stepped Attenuator is a Whole world 'better' than an Alps (have both) same or even less price too.
Try one, you won't go back to a goofy wiper /carbon track gizmo :)

No arguement, the Cheap Lepai.. it's typical Chinee crap .
ONLY claim to fame is it's $15/18 delivered to your door pricings.
Although once heavily modded ie: ALL parts replaced except the SMD ones, Does help enormously.
FYI; I did follow the Parts Express 'recomended upgrades' They claim replacing the 047 uf Output caps with Decent quality Film ones makes the Most noticeable improvement.. Yesss it does. They seemingly know what they are saying.
Look it up on their Forums.
Bypassing caps is an easy 'instant' improvement.. but it creates just as many issues as it masks. Do Not bypass caps in the signal path ..thats' a Mickey Mouse trick in my experiences.
Casting genuine doubts re the competence and or integrity of that Amps' builders.
Seemingly you got a decent Case for your Extra price.. likely that's about all imo.
Critical ingredient is the Ta 2020 chip. A circuit built exactly to the chip maker's specs/schematic is likely THE very best implementation possible. Copy it whenever, wherever possible
I have a spare Ta chip ( Hey $5.. delivered to my door ;-) from which I will copy/build that exact Tripath circuit.
Hopefully I won't wreck my fragile DIY Amp.. this time.
 
The value of Co affects the cutoff frequency of the class-D's output filter, better to not change the value. Same value larger physical size is fine as long as the legs fit in the hole. Don't use wires.

The 0.01uF is probably there due to the "I need to parallel big caps with faster small ones" disease. Or maybe it is really needed there because the 0.47uF is really ineffective at the higher frequencies. Can't trust those China components. I'd follow the datasheet and put one decent capacitor.

330pF capacitors are most likely to filter out RF noise. Cutoff frequency is usually high enough to not affect audio a bit.
 
As small digression I 'think' I damaged my Lepai's Ta2020 Chip By tripping over a speaker wire and yanking it out of it's Speaker spring clip... when it was operating.
Left channel no longer produces Sound.
But shorting the Chips' pins from underneath with a fingertip produces a hum equally loud on both channels' speakers.
Any basic tests of Ta chip functionality possible?

You may have stressed the board and could damage the traces. If possible test your traces for continuity. If some trace does not have continuity anymore you could always put solid wire instead. Check the left channel relay as well, unsolder it and do some tests. I don't know if the chip has short protection on outputs. Also check the inductors, I saw a fried one in some post on Lepai thread.
I have two transistors for the speaker relays, check those as well if you have them.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.