Sonic Impact 5066 Parts List & Modifications

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
[QUOHow about those caps across the output terminals?? Is it recommended to replace those?? If yes, then with what?? Again, my apologies if I missed a post answering this question...TE]

I replaced them with 0.1uf M Cap poly's and removed the on board C19 & C20. It seems to really open up the detail. It is early days with this last mod and things may change, it sounds good at the moment.
PS make sure you twist the speaker cables from the board to the terminals..


Barry..
 

Attachments

  • t amp output mod.jpg
    t amp output mod.jpg
    85.6 KB · Views: 1,598
Administrator
Joined 2004
Paid Member
C19/C20 are the Differential Output Capacitors. They are meant to drain off some of the RF.
Tripath calls for 0.01uF. That should have an impedance of about 800 ohms at 20kHz and 16 ohms at 1MHz. Some Tripath based boards use much larger caps. I don't know what the Sonic uses.

The terminal output caps could make a difference, as they are larger. A lot of people have reported better results without the caps across the speaker terminals, or the use of better caps.

The output filters of Class-D amps is wide open to research!
 
Thanks again!

Thanks all for the tips! I have a real good idea of my final mods list now, and will start the ordering process in the next week-ish. I even have an idea of what I want to do for a re-box, but I'll keep that card close to my chest, in case it doesnt work out... it'll be neat if it does.

Pano - Thanks for the tip on the power supply. I have a BK Precision 1686 bench supply that should do the trick for now. I checked out wluk and will be ordering a half dozen from him.

I also added the .15 uF output caps to the parts list (Digikey number P3876-ND).
 
panomaniac said:
R01/R02 set the input impedance. They simply tie the input to ground via 10K ohms. This is a bit low for normal purposes. 50-100K would be better for most sources. 10K was proabably choosen to give a good load for the headphone output of portable CD/MP3 players.

So you can remove them if you are using a 100k or 50K volume control, or replace them with those values if you are not. You can place them anywhere upstream of the input cap, and after your input connectors.
I think for the time being I'll be using the stock volume control still (50k pot I believe) so I'll just take R01 and R02 out. Thanks.
 
Hello to everyone, i am new to the forum (but have been reading quite a bit lately). What a great forum and also a great community! Keep up the good work!
I also have a little (easy) question regarding my t-amp. I just want to replace the input caps to foil types with 2.2uF. Since they are *big* i was wondering if i could just unsolder the input cap of each channel, bypass it with wire, and then wire the new cap between the pot and the board. Basically this would leave the input circut intact, but the decoupling cap would be right "after" the pot in the schematics. (A bit like the "stealth-mod" from Panomaniac but easier)
Thanks a lot!
 
Heat dissipation

Hi,
I'm new to the game of T-amps, but I have an idea to put two t-amps in a box (http://www.overstock.com/?page=proframe&prod_id=1662439 with external power supply in the smaller, matching box) and bi-amp my speakers, which are relatively inefficient at 86db/W. I'm wondering if the heat generated by the chips will cause the temperature to rise substantially within the case. What kind of power do they dissipate during music? Should I get a little chip heatsink for the chips?

Also, I am planning on using a 5A switching power supply with a large smoothing capacitor on the output terminals of the power supply and another on the rail close to the amps. Is this a good idea or does someone object? I think it's almost like a CRC filter, due to the cord from the power supply to the amp box, but only with a very small resistance. Suggestions? Comments? I hope this bi-amping works.

I'm also wondering about the grounding. Should I worry about doing a proper star ground? probably not, right? But I just want to make sure I don't get any ground loopeys.
 
Re: Heat dissipation

philibuster said:
Hi,
I'm new to the game of T-amps, but I have an idea to put two t-amps in a box (http://www.overstock.com/?page=proframe&prod_id=1662439 with external power supply in the smaller, matching box) and bi-amp my speakers, which are relatively inefficient at 86db/W. I'm wondering if the heat generated by the chips will cause the temperature to rise substantially within the case. What kind of power do they dissipate during music? Should I get a little chip heatsink for the chips?


A heatsink won't do any good if there's nowhere for the heat to go. Based on the Tripath Datasheet, the amps are 88% efficient into 8 ohms, and 81% efficient into 4 ohms. That means for 15 watts of power into each speaker, the chip will need to dissipate roughly 5.5 watts, and for 10 watts into 8 ohms, it will need to dissipate roughly
2.5 watts.

Some small airholes in the top and bottom probably wouldn't hurt. I guess it will depend on how hard you plan on driving them.

philibuster said:

Also, I am planning on using a 5A switching power supply with a large smoothing capacitor on the output terminals of the power supply and another on the rail close to the amps. Is this a good idea or does someone object? I think it's almost like a CRC filter, due to the cord from the power supply to the amp box, but only with a very small resistance. Suggestions? Comments? I hope this bi-amping works.

I'm also wondering about the grounding. Should I worry about doing a proper star ground? probably not, right? But I just want to make sure I don't get any ground loopeys.

It's going to be difficult to create a star ground since you'll really have two sets of grounds -- the groundplane of the SI's and the other grounds in the chassis. That said, a star ground probably couldn't hurt.

As for the switching PSU, I think that should be sufficient. The large smoothing cap may or may not make a difference. If you did go CRC, maybe you should think about an actual resistor in there. Also, what's the voltage output of the PSU? If it's an unregulated switcher, you may want to think about using some 3 pin regulators to reduce the switching noise.
 
Thanks for the reply. I was thinking of cutting some holes in the box for ventilation, but I have to put them strategically, so it doesn't screw up the look of the amplifier that doesn't look like an amplifier.

The power supply is regulated. So you think the large caps might not make a difference? The supply has ripple and noise at 1%, so 120mV. Should I get some bypass caps for these too? Thanks again.
 
I don't think you'll need to go crazy with large caps to make a difference. Maybe some 1000uF low-esr's to start with. You could always add or remove depending on what sounds better to you, though I doubt you'll hear a huge difference.

As for cutting the holes, maybe if you did it at the top and bottom of the back of the case, it would keep it from ruining the looks. Just place the boards close to the back if you can.
 
Administrator
Joined 2004
Paid Member
Some cooling on these chips is needed, a solder slug is what Tripath recommend, but a lot of PCBs don't use it.

Those little ramdac heat sinks from "wluk" on eBay work very well. In a closed box you'll not get al the cooling you could, but it's better than nothing. Or you could figure out a way to press a copper strip to the top of the chip with some thermal paste in between. You could sort of spring load the strip against your box. More mass, more cooling. You don't need a lot, but some cooling is a very good idea.

120mV is a lot of ripple. I try to keep mine under 5mV at all frequencies. The chip has high ripple rejection, but why not start off clean? A few nice Panasonic FM series caps or other low ESR caps would really help, and a few film and ceramic types will help kill high freq. noise. Put some capacitance right where the power comes into your box and also some as close to the PCB as you can.

As you are using a wooden box, you'll have little RF shielding. Be careful when you get the amp near other audio gear. Usually not a problem, but al least you know to careful.

Well, that's what I think, anyway. ;)
 
yeah, 120mv ripple is the quoted figure from the data sheet. I am planning to put 10000uf at the output terminals of the power supply and another 10000uf on the +12V line. I am also upgrading the tank cap to a 680uf from nichicon (mouser doesn't carry the panasonics, and I was buying everything else from mouser anyway). Is this enough capacitance?

Thanks for the newegg tip. Going to have to get some kind of heatsink...

I might have to get out the jigsaw for the holes...

How much RF comes out of the amplifier? I'm going to be using an EMI filter on the power inlet for the mains, so there's not going to be any leakage back to the mains. What about airborne? That scares me a little, because I would like to listen to AM radio on this. Has anybody tried it? Will this mess with my wireless? it's 2.4ghz, but what's the highest switching frequency of this little guy? Should I line the box with some steel plate? Funny... on other equipment, you try to keep the RF out of the box. This is the complete opposite.

What's the charge pump for in this chip? Does it just do nothing? Is there any way to disable it?
 
Ex-Moderator
Joined 2005
philibuster said:
How much RF comes out of the amplifier? ... That scares me a little, because I would like to listen to AM radio on this. Has anybody tried it?
Can't say for sure but a poorly designed switching amp will interfere with AM reception. The closer the amp and tuner are the more interference you'll get.

Originally posted by philibuster
Will this mess with my wireless? it's 2.4ghz, but what's the highest switching frequency of this little guy? Should I line the box with some steel plate? Funny... on other equipment, you try to keep the RF out of the box. This is the complete opposite.
It doesn't seem to interfere much with wireless communications signals up in the GHz range. I have an unenclosed amp3 always on right behind my laptop (6in apart) on my desk and get great wireless signal strength. Tripath claims the 2020, 2021B, and 2024 can switch up to 1MHz. Their switching frequency varies based on the level of the input and is likely to be somewhere in the 600-700kHz range.

Originally posted by philibuster
What's the charge pump for in this chip? Does it just do nothing? Is there any way to disable it?
The charge pump is used to drive the high side mosfets. The output stage will not work properly without it.
 
Administrator
Joined 2004
Paid Member
As Brian stated, the Triapth amps switch in the 600kHz to 1 MHz range. In other words, right in the AM radio band.

There are some chips that are made for car radios that have an "AM mode". This turns the chip into a regular linear amp - not switching - to prevent killing the AM band when you select it on your car radio. Switching harmonics probably don't go much over 3MHz, and will be very weak there, so it won't hurt Wi-Fi, corless phones and that sort of thing. But the AM radio band won't like it.

You can run one amp per speaker, many people do. You can even use one channel for Hi one for Lo. But be sure your speaker and crossover do NOT have a common ground between Hi and Lo if you go this route.

In what way, and where, do you want to reduce IM distortion?
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.