Sure Electronics New Tripath Board tc2000+tp2050

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Sendler,
thanks for the explanation about the extra fan. I did fire it up last night and it worked fine, sounded very good across the midrange, and I'm not aware of any problems with the fan on this version. I'm only using a 12 vote smps right now while I make a decision about power supplies. I did some comparison with my Hlly t-90 and Arjen's t-2020 bd. There are some differences, but there's also a strong familial resemblance between the 3 boards sonically. The 2022 remains distinctly darker though. The Sure could use more attack, but I'm pretty sure that has to do with my running it so far below its recommended voltage.
 
The importance of the output filter chokes is often underestimated, I have noticed several improvements of using really good ones. One advantage is lower current draw, another is less saturation at higher frequencies and higher power. All those cheapo ebay boards use output filters of the category "pure crap".

The input caps, even a 1uF cap will allow bottom deep bass frequencies to pass through. You can calculate the cross over frequency, basically the cap forms a highpass filter with the input impedance of the amplifier input. It's a simple PLLXO. I have the experience that under 200nF there is noticeably less lower bass extension. Above that it is often hardly noticeable, depending much on the rest of the system.
 
Hi Scott, thanks for the info on the input caps. So perhaps I should buy a couple more 0.47uf then at least all four caps will be of the same value, I can pick up a couple of cheapish ones tomorrow.

Regarding the coils, will these do the trick do you think ?

WUERTH ELEKTRONIK|744824622|CHOKE, COMMON MODE, 2.2MH 6A | Farnell United Kingdom

Thanks

Nope, TOTALLY unsuitable. What you need is something that responds well with high current signal of more than 1Mhz. Common mode chokes filter fairly low current high frequency noise. You need a choke with a low perm core of sintered iron powder or MPP. Ferrite is quite unsuitable too. Beware that on saturation the cores can become excessively hot and even mess up the current sensing of the amplifier. Stuff blowing up due to this is not unlikely.

Apart from the risk of more emission I would rather recommend to use no core (aircore) than the wrong core material.

I have had TA3020 amps that went down from 90mA per rail of idle current (already quite good) to a spectacular 60mA per rail with a different inductor. Always beware that the output filter Q determines much of the in band linearity of a class T amplifier. This means you can have a rise of higher frequencies in the audio band. Some may experience this as pleasant and write raves about such an inductor, while in fact a treble boost would have the exact same result.

It's OK to fool yourself but best to try to stay objective in public. Also for this reason do be careful with what you read on forums.
 
This is not easy, I don't think they have any suitable ones at Farnell. There are suitable ones from Panasonic, but the cheapest ones and probably the best ones are what you can wind yourself. Core manufacturers often are willing to send you samples. MPP cores are by far the best and I would recommend low perm (something like 20 to 60) Iron powder cores are also very suitable, this is what Tripath used themselves.

Over at 41hz you can possibly buy separate cores and wire. Or these:
http://www.xfmrs.com/pdf/XF100910DL-xxxx.pdf
 
This is not easy, I don't think they have any suitable ones at Farnell. There are suitable ones from Panasonic, but the cheapest ones and probably the best ones are what you can wind yourself. Core manufacturers often are willing to send you samples. MPP cores are by far the best and I would recommend low perm (something like 20 to 60) Iron powder cores are also very suitable, this is what Tripath used themselves.

Over at 41hz you can possibly buy separate cores and wire. Or these:
http://www.xfmrs.com/pdf/XF100910DL-xxxx.pdf

Ok, yes will wind them myself*, thanks for the 41hz link, they have various ones to choose from:

41Hz Audio:Sounds Good to Me!

The difference between these escapes me as they give little information...
 
Coil Pile

Ok, yes will wind them myself*, thanks for the 41hz link, they have various ones to choose from:

41Hz Audio:Sounds Good to Me!

The difference between these escapes me as they give little information...

I've listened to quite a few different coils including the ever popular T106 type 2 toroids from 41Hz and the Ferroxcube toroids as found in the ClassD audio amps (much better sounding but leak like an air core) with three different types of wire. 14ga Litz wrapped and unwrapped, 18ga solid, and Belden hard insulated 14 guage solid. And other recommended favorites from Toko, Sagami, CoilCraft, ect. The Wurth XXL were tried on a tip from another member or I wouldn't even have glanced at them. They look just like what you are taking out. Luckily I am open minded. I like them by far the best. They even sound better than my Belden air toroids in many ways, have a 13 amp saturation rating and they fit right in the stock location. I will use them exclusively. There are tens of very similar looking coils and tens more of larger types but the Wurth XXL have the best sonics of the ones I have tried.
http://katalog.we-online.de/kataloge/eisos/media/pdf/7447709004.pdf
 

Attachments

  • Coil Pile comp.jpg
    Coil Pile comp.jpg
    203.7 KB · Views: 1,336
have a 13 amp saturation rating

Be careful with this to check the saturation frequency it was tested on!! A 13A saturation current at 100Khz can become a whole different story at 1Mhz!

Just connecting a million chokes doesn't necessarily make you an expert either... Have you done any measurements with all these chokes? It is not that hard, just connect up a tone generator and the amp on a lab supply (or PS with Ampère meter(s) in line) , make a sweep up to 20K and check the current consumption of the amplifier. When the cores hit their saturation point the current will go up rapidly.

Not to say that I'm personally an expert, but not a noob either. I have understood from application notes and personally from an expert I know that ferrite is an inherently a bad material for this purpose because it has very poor high frequency/high current capability/HF saturation. What you may find providing a "nice sound" can perhaps be mere distortion....

EDIT: Gapped ferrite is suitable by the way, the air gap improves the HF stability in a large degree:
http://www.ferroxcube.com/appl/info/gaptoroids.pdf
 
Last edited:
Sagami


I have some Sagami 7G17B. They sound very nice but 10uH is the lowest value and I have found that going lower than the text book predicts is very beneficial to the sound.
If you look at the Wurth XXL docs you will see that the 13 amp rating is conservative, the current vs inductance charts showing a very gentle saturation characteristic similar to the carbonal toroids and they have a high resonant frequency. The Wurths XXL are very safe to use with the TK2050 chip set and have a much higher current capacity than the hair fine coils that come in the Sure 2X100 amps.
 
Last edited:
the current vs inductance charts showing a very gentle saturation characteristic

Okay, maybe I missed it, but I don't see the meased frequency?

They sound very nice...

Ok, but have you measured anything? IR thermal, current etc...?

I would be most interested in this, afterall you have tried more than a dozen different ones...:)
 
Last edited:
IR Thermal?

Okay, maybe I missed it, but I don't see the meased frequency?



Ok, but have you measured anything? IR thermal, current etc...?

I would be most interested in this, afterall you have tried more than a dozen different ones...:)

IR thermal? Yeah, I measured IR thermal. With my finger. The filter caps get warmer than the coils.
 
I'd mentioned that I had my Sure board up and running with 12 volts. I know the difference has been talked about here, but I found a 19.5 volt SMPS in the garage and gave that a shot. Big difference! I expect to hear more of the same with 24 volts (I have a pair of batteries). Better extension, better dynamics, etc. After that, I guess I need to try 30 volts.

12 volts was like driving a Ferrari in second gear all day. I'm obviously not done yet, but I think there's a reason so few companies have messed with the t-2022 vs. the t-2050. I'm pretty sure that the t-2000 t-2050 combination sounds better and when you take the simplicity of the power supply into account, it's a lot cheaper to implement and much less finicky.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.