Lepai T-Amp with TA2020

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
col said:
have you attached a LED to the fault pin? Is the light on when the amp is unmuted?

edit: try touching the fault pin to the unmute. That should reset it.

col.

When I had the led and resistor wired up to the fault pin, it did not light up at all, which makes me thing the output section is good. I've checked the output diodes and they are ok along with the inductors and they appear to be ok.

I'll try touching the fault it to the mute pin and see if that does anything. :)
 
col said:
when you had the led on the fault pin did it light up when you muted the amp?

Maybe try inputing a signal directly to the IC pins.
:eek: Turns out the LED I was using wasn't working, I've now wired up a working red LED and it now lights up when unmuted when wired up to the fault pin which makes more sense.

When the amp is muted the LED is no longer lit... Basically the Fault LeD only comes on when it the amp is unmuted.

I'll try inputting the signal straight to pins 10 and 13, and see if that works at all.
 
I think the TA2020 must be dead, but at least I have learned a little about how the amp works.

It turn out the zener diode from the fault pin goes to the mute pin, so at stock this amp has overcurrent limiting disabled. I'm not really sure why they use a zener instead of a normal connection, but I found when I had my fault LED working I could short the mute pin to the fault pin and the all familiar clicking would return.

I think I will remove the zener between the mute and fault pins on my next lepai to arrive as I think overcurrent protection my have saved the tripath IC in this case...

As the fault LED was coming on all the time I was thinking it's probably one of the outputs shorting, and true enough it was!

I pulled out the output diodes and output inductors, and measured the resistance from the outputs from the Tripath IC to ground, and I found that 3 of the outputs had a resistance of around 40kR to ground, while the other input was a short to ground.

I thought it could be a bridged PCB track so I desoldered the pin and pulled it from the PCB, and yet there seems to be a short thru the Tripath IC! I guess something burned thru within the IC... :hot: :(

Still, my next Lepai should be here in a few days, I'll try and be a little more careful with my next one, and I've always got this one for spares... :smash:

Thanks for you help col, I didn't have a clue where to start before but I'm now pretty familiar with the TA2020. :)
 
It must have been too hard to implement the overshoot diodes on a one-sided PCB, so they left them out. I don't see the logic behind it being a cost cutting measure when it obviously puts the chip at risk operating at voltages above 13.5v especially, if they have disabled the over voltage feature :|

I think the LM338T 5A regulator circuit before the power input is the way to go. That way you can connect it to any DC voltage from 32v down. Just remember that it requires the voltage to be at least 1.5v above what you want to operate at for regulation to work.

here is the data sheet:

http://pdf1.alldatasheet.net/datasheet-pdf/view/8646/NSC/LM338T.html

I based my circuit on "FIGURE 3. Regulator with Protection Diodes". The caps I'm using are much bigger than the diagram and the one after the regulator is low ESR. The power is very clean.

http://pix.minirig.org.au/main.php?g2_itemId=1057

With the trimmer you can adjust the output voltage to what ever is desired. Iv'e got mine set on 13.3v.

col.
 
Re: Bulk capacitor

Calamaro said:
I think it would be very useful to replace bulk capacitor (180uF: TA2020 schematics have two x 180uF, someone goes up to 1000uF, I noticed that two x 470uF give better mids...). Any experience?


Hi Calamaro,


Do you mean that I should change the 180uf cap and metal cap in pair as attached photo?


Thank you.
 

Attachments

  • bulk cap1.jpg
    bulk cap1.jpg
    64.7 KB · Views: 1,105
I apologize for my delay in answering, but in Italy we had a long weekend due to the bank holiday of today, so I went out of town for a little break :car:

cheric:
if you did the bypass mods you must replace input resistor, 20/22k will works fine.
As explained by Gmorris and Buckapound, input capacitors form a high pass filter with input resistor so go over 1 uF. Datasheet suggests 2.2uf, on the Bay you can easily find some mkt small enough to fit in the amp case. Ero 1822 is a very good cap, MKC dielectric is polycarbonate that is regarded to be much better, as good as polypropylene.

aszu:
further improvement can be done replacing bulk capacitor and output inductors (see Col experience), and adding some capacitance to 5v side (I did it on another pcb so I have to study a little bit this board to see how to do it on Lepai. I'll be back soon ;) ).

panda360:
no, just replace capacitors pointed by the arrow (one electrolitic and one ceramic, please see pict.). As electrolitic replacement Tripath suggests Panasonic FC or HFQ, but I think anything will be better than the existant one! Anyway go for a low ESR and add some capacitance (470uf or even more). Replacement will be more confortable if you invert electrolitic and ceramic cap place (they are paralleled), as ceramic has more room on the board, but pay attention to electrolitic polarity! Plus on chip side, minus on pots side.
 

Attachments

  • attachment.jpg
    attachment.jpg
    76.9 KB · Views: 1,090
100nF monolithic bypass on the opamp rails

For those that have kept their opamp in the circuit a minor mod but well worth doing is bypassing the opamp power rails.

http://pix.minirig.org.au/main.php?g2_itemId=1063

Best caps for this purpose are 50V 100nF monolithic, these are designed for bypass applications. Polyester caps can be used for bypassing but they are not as effective at very high frequencies.

col.
 
Re: 100nF monolithic bypass on the opamp rails

I've upgraded the power cap to a 25v 6800uF cap, and I was wondering if I could bypass it with a 100nF Polyester cap as I don't have any film ones?
col said:
It must have been too hard to implement the overshoot diodes on a one-sided PCB, so they left them out. I don't see the logic behind it being a cost cutting measure when it obviously puts the chip at risk operating at voltages above 13.5v especially, if they have disabled the over voltage feature :|

I think the LM338T 5A regulator circuit before the power input is the way to go. That way you can connect it to any DC voltage from 32v down. Just remember that it requires the voltage to be at least 1.5v above what you want to operate at for regulation to work.

here is the data sheet:

http://pdf1.alldatasheet.net/datasheet-pdf/view/8646/NSC/LM338T.html

I based my circuit on "FIGURE 3. Regulator with Protection Diodes". The caps I'm using are much bigger than the diagram and the one after the regulator is low ESR. The power is very clean.

http://pix.minirig.org.au/main.php?g2_itemId=1057

With the trimmer you can adjust the output voltage to what ever is desired. Iv'e got mine set on 13.3v.

col.
Thanks for the info, it does look some sort of regulator would be the best solution. I just this minute managed to fix my lepai, I replaced the TA2020 and it's not making music again. :cool:

Right now I've got a diode in series (where the inductor was) which drops 0.5v and I've also removed the zener diode that joins the fault pin to the mute pin so the overcurrent protection is now enabled.

I'll see how it goes, I bought 2 TA2020 IC's so if I cook this one I've still got a spare...

Also, I think overshoot diodes could be added, but it would be a bit messy as the heatsink would be in the way...
col said:
For those that have kept their opamp in the circuit a minor mod but well worth doing is bypassing the opamp power rails.

http://pix.minirig.org.au/main.php?g2_itemId=1063

Best caps for this purpose are 50V 100nF monolithic, these are designed for bypass applications. Polyester caps can be used for bypassing but they are not as effective at very high frequencies.

col.
I'll give this a go, I've only got Polyester caps at the moment but I guess they will be better no caps at all?
 
Empty section of board

Does anyone have any idea to what the empty section of the board is?

If you look in between where the output filter inductors and the power input inductor is you can see "1k*4" and 4 empty positions for resistors, under that there is a "0.47u" and the place for a cap.

Next to the power stiffener cap there is the place for a diode marked "5v". Also, down by the power switch there is a place marked "A733" and a couple of wire links that are empty.

col.
 
Re: Empty section of board

I've been listening to it all day today now I've fixed it, it sounds lovely. :cool:
col said:


I think it might be in the "members only items for sale" section. You have to be logged in to access it. They are changing the site though and its a bit hard to navigate at the moment.

col.
I can't login right now, I went to get some build instructions for my amp9 yesterday and it says my password is wrong. I then asked them to e-mail me my password and it sent me a blank password! :(

I've posted this on their forum so with any luck they should get it sorted soon.

Would these work if they physically fit on the board?

http://uk.farnell.com/jsp/search/productdetail.jsp?sku=1612716

It's rated for 3 amps but I'm not sure if it would give off more EMI compared to a toroid?
col said:
I think the two most important mods that still need to be worked out are:

1. Implementing the overshoot diodes
2. Tone control detent switch

col.
For number 1, would there be enough room under the board for 4 more diodes of the same sort? They just need to be wired in a similar way to amp6...

For number 2, you could short the mute pin to ground, then replace the power/mute switch with a DPDT one that is the same shape so it would fit in the hole. With a dual gang switch you could then switch the audio input pins on the TA2020 between a direct feed from the RCA ins or a feed from the op-amp (with tone controls). Would this work?

If you don't mind drilling you could use a round toggle type switch and this way you could keep the mute/power switch as it is...

col said:
Does anyone have any idea to what the empty section of the board is?

If you look in between where the output filter inductors and the power input inductor is you can see "1k*4" and 4 empty positions for resistors, under that there is a "0.47u" and the place for a cap.

Next to the power stiffener cap there is the place for a diode marked "5v". Also, down by the power switch there is a place marked "A733" and a couple of wire links that are empty.

col.
I haven't a clue, but that extra piece of PCB could come in handy if you wanted to short it to the +ve power input, then you could have the overshoot diodes coming from that extra piece on PCB...
 
i think my lepai just broke. when its on there just comes a soft ticking sound and the tripath chip gets very hot. what might be wrong? some of the solderings of the chip are shorted? i just changed the pre amp to LM4562, but worked fine with it. after a night of sleep the amp was dead.
 
Would the volume pot still work OK if I used a 50k log POT?

I've managed to break the volume pot trying to get thr front panel off! :eek:
whelibob said:
i think my lepai just broke. when its on there just comes a soft ticking sound and the tripath chip gets very hot. what might be wrong? some of the solderings of the chip are shorted? i just changed the pre amp to LM4562, but worked fine with it. after a night of sleep the amp was dead.
Mine did exactly the same thing, constant ticking sound when unmuted?

Try removing the zener diode between pin 18 (fault pin) and pin 11 (I think), and see if you still get ticking.

With the zener gone, try connecting an LED and 2k resistor between pin 18 (fault pin) and GND and see if the LED lights up.

Your TA2020 could be dead, but it may just be a short in the outputs...
 
i think my lepai just broke. when its on there just comes a soft ticking sound and the tripath chip gets very hot. what might be wrong? some of the solderings of the chip are shorted? i just changed the pre amp to LM4562, but worked fine with it. after a night of sleep the amp was dead.

whelibob, sounds like one of your outputs has shorted. What voltage were you using? My suspicion is that the output inductors can't handle much current and fail, this then takes out the channel on the TA2020 chip. It happened on one of mine, luckily I had not modded it and only had it turned on for 20 mins. I was able to return it and get a replacement.

Mike has just had a similar failure, he was able to replace his chip.


I just put some more capacitance on the power cap:

http://pix.minirig.org.au/main.php?g2_itemId=1066

I might take out the zener too! Mike, did you replace it with a wire link or just remove it?

col.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.