Tripath Linguini

albee · « **on:** 21 Aug 2005, 02:28 pm »

Anybody checked out the Italian mods here?

http://www.autocostruire.com/catalog/index.php?cPath=42

tianguis · « **Reply #1 on:** 21 Aug 2005, 10:42 pm »

Albee:
Not exactly mods. Looks like a pretty reasoned and complete approach to most of the problems inherent in the T-Amp boards: inductors, coupling caps, adjustable gain, impedence matching, replacing SMD's, etc.. It seems as if they're aware of all the issues talked about in the fora. Having modded a few T-Amps and being in the process of building a 41hz amp, it looked good enough for me to order one. I'm currently using a modded SLA-powered T-Amp as my primary amp.

Regards,
Larry Welsh

albee · « **Reply #2 on:** 22 Aug 2005, 12:10 am »

I was mulling over the idea of getting one myself. Currently I'm using to two unmodded SI 5066's to run my biampable speakers. Using an amp per channel really improves the imaging as well as using one channel per driver increases dynamic impact.

tianguis · « **Reply #3 on:** 23 Aug 2005, 11:07 am »

Albee:
I've used four T-Amps to drive a pair of Hammer Dynamics Super 12's. I agree about the benefits of bi-amping. However, my current modded T-Amp has some goodies in it and is really good with Lowthers.
The folks in Italy are pretty quick, as they shipped within 12 hours. Should have it Thursday. Building it should take an hour or so.
Here's a thread on DIYAudio. The folks in Europe have some up and running and the results look encouraging. There are comparisons with 41hz amps and T-Amps.

http://www.diyaudio.com/forums/showthread.php?s=&threadid=60046&perpage=10&highlight=autocostruire&pagenumber=1

Regards,
Larry Welsh

ohenry · « **Reply #4 on:** 23 Aug 2005, 11:55 am »

I've used the air core inductors from that company replacing the toroids on several amp3's and I really liked the results. It was a pleasant surprise to find that the air cores imparted more heft to the presentation by enhancing the amp3's bass and mid-bass. At about $30/set, they are a bit pricey, but worth it to me and the trans-Atlantic shipping is quick.

Also, don't overlook trying some nice paper in oil input caps if you have some hiding in a forgotten corner. I found that Sprague 2.0uF's result in addictive midrange, although Auricaps have the edge on bass.

It seems crazy to throw high dollar parts into a $25 kit, but the results are equally insane.

tianguis · « **Reply #5 on:** 23 Aug 2005, 01:57 pm »

ohenry:
Yup, I intend to try a few input caps: I've got Jensen, Jantzen, Auricap, Hovland, Vit Q, AudioCap Theta and Rel Caps to mess around with.

I'll also be changing the caps on the rails to Panasonic FC's.

Regards,
Larry Welsh

KT · « **Reply #6 on:** 24 Aug 2005, 12:53 pm »

Larry,

Keep us updated as you try out these different input caps.

I've only tried a Hovland in this position so I cannot make any comparisons. I'd be interested, however, to hear your opinion of the Vit Q. I have a bunch of them and tried them once as coupling caps in my SE amp. While the sound was warm and fluid, I found that it was way too muddy, syrupy, and detail-obscuring to my ears. It also threw a heavy blanket over the pacing and propulsiveness of the sound and glossed over the dynamics and attack of the leading edge of notes. I didn't enjoy it.

I also had a Cary Audio SLP-94 preamp that used Audio 1 caps for coupling. These looked very similar to the Vit Q type vintage sealed metal-cased oilers. They deadened the sound in a very similar way. I couldn't stand to listen to music on it because it sounded too restrained and the sense of timing was off. I changed them out to VTV Ultratones and that transformed the pre into something I really enjoyed.

In some of the reviews I've read, there seems to be a slight lack of harmonic richness in the upper midrange on these T-amps. Perhaps the right input cap would help out a bit? How about the Audio Note coppers or Jupiter beeswax?

Also, I've seen a site that recommends the Blackgate C caps for the input cap. These must be a recent addition to the Blackgate line - I never recalled seeing a C cap before (the Cs are specified for coupling applications).

Best,
KT

tianguis · « **Reply #7 on:** 24 Aug 2005, 03:12 pm »

KT:
I, too, have usually found the paper/oil caps to sound a bit tubby and flat when used as coupling or X-O caps. I much prefer Auricaps. However, I'm going to try some caps which I've been testing in another application (Jantzens) which (in my system and to my ears) sound better than Auricaps at about 1/10th the price. The caps I mentioned above are ones of the right value I have kicking around. Gotta try some Jupiters at some point.
I ordered the Autocostruire kit Sunday night and it was just delivered! I'll build it later today.

Regards,
Larry Welsh

tianguis · « **Reply #8 on:** 24 Aug 2005, 07:46 pm »

I've had this running for about an hour so far. My first impression was the huge soundstage it throws. It images way better than my modded SI. It has more bass authority, also. Top end is still a bit dodgy, but that's typical until several hundred hours. Detail is as good or better than the SI. Better dynamics.
So far it's stock, unhoused. I'll be making changes after getting some hours on it.

Regards,
Larry Welsh

Barry · « **Reply #9 on:** 28 Aug 2005, 01:08 pm »

Are you guys powering the Audiodigit T-amp simply connecting a 12V SLA battery to it?

tianguis · « **Reply #10 on:** 28 Aug 2005, 02:31 pm »

Barry:
Yup, 12v SLA. Additionally, I put a bank of 3 1800 uf Panasonic FC's across the battery posts (observing proper polarity, of course). I've found it to have a positive effect on dynamics (especially DAC's) even though it shouldn't.
BTW, I changed the power rail caps to Panny 2200 uf. FC's. Noticeable improvement.

Regards,
Larry Welsh

toxteth ogrady · « **Reply #11 on:** 28 Aug 2005, 05:20 pm »

Larry

Good stuff... thanks for sharing your experience. To what do you attribute the improvements over your modified SI? I'll be building one in the next couple of weeks but haven't yet made a decision on which one. I've seen the boards from 41hz, Autocostruire, and Diyparadise... all of them look very promising.

tianguis · « **Reply #12 on:** 28 Aug 2005, 06:04 pm »

O:
The Autocostruire has addressed most of the problems with the SI and corrected them: good air-core inductors, proper-sized input caps, impedence matching, location of the caps on the rails, quality of the parts, 2020 chip instead of the 2024. The sound is a big improvement over my modded T-Amp. I had the T-Amp back in my system for an hour while I changed some things on the 2020 and the experience was painful.
I'm waiting for one of Yeo's Diyparadise 2020's. Maybe I'll try one of the 41hz through-hole 2020 kits. I need a couple of amps for bi-amping anyway.

Regards,
Larry Welsh

tianguis · « **Reply #13 on:** 29 Aug 2005, 12:34 am »

All:
I feel like I'm hogging this thread, but if your're bothering to read my blather, it may be because you'd like to build one.
I discovered how critical matching speaker impedence is. I've listened with two values before today. It figures that the third is the charm in my system and the difference is not subtle. It's HUGE. If I were to build another, I'd substitute some posts in the four cap positions and experiment with .22 and .33 on the posts. It would then be possible to parallel both on the same posts. This would obviate the need to desolder and resolder the board, which can lift solder pads and lead to repairs.
Whst I've arrived at (with hybrid Lowther/Fostex ML TL's) was only a happy accident, but it's shown me the value of something I was never concerned about.

Regards,
Larry Welsh

zeta4 · « **Reply #14 on:** 31 Aug 2005, 05:47 pm »

Hi
Just registered so if Ive repeated anything please forgive.

Ive been a user of the the bigger Tripaths for years and just love them. Ive moded them just like the smaller versions. One extra thing I found was critical, the +5V supply had to be very very clean.
A friend gave me an SI to mod for him and the previous posts have been most useful, thanks but the +5V area doesnt seem to have been mentioned. The TA202x chips look interesting in that they have separate +5v digital and analogue supply pins which are supplied on the SI chip at least from an on board regulator in turn supplied from the main power rails. I want to try to separately regulate the +5V analogue supply with something like an LT1763 supplied from its own battery. Anyone heard of this?
Geoff

Occam · « **Reply #15 on:** 31 Aug 2005, 09:02 pm »

Geoff,

A most excellent suggestion! Another poster here has mentioned the importance of the 5v supply in the Tripath amps -
http://www.audiocircle.com/index.php?topic=18443.msg179339#179339&highlight=tripath+felicia#179339

I guess that while the 2020 and 2024 chips provide their own +5v regulators, the fact that it comes out via a pin for external routing to the analog and digital +5v pins is indicative that an implementer might have good reason to provide alternative sources for that voltage.

I like your choice of that LT1763 LDO, extremely low noise regulator. The whole family is very impressive -
http://www.edn.com/article/CA46885.html
I've not been able to suss out the +5v current requirements for the 20/24 chips, but from the 2050 specs it seems that 200ma might be more than enough, making the LT1962 feasable. But for these tired old eyes, if I have to work with smt, I'd much prefer your SO8 over a MS8 package

.

Have you tried separately regulating the analog and digital +5v lines? or is the major benefit to be had with a single regulator?
As the audiokit board under discussion here already has decoupling on the 2020 input pins, what do you reccomend for the regulator's input and bypass cap? Are you using a ceramic .01uf or a more bespoke cap for bypass duties?
Sorry for the multiple questions (I actually have scads more), but your post was very intriuging......

Welcome to AudioCircle!

Regards,
Paul

zeta4 · « **Reply #16 on:** 1 Sep 2005, 05:18 pm »

Hi Paul

Thanks for the encouraging reply. I guess the hard part comes now. I think I will have to lift the +5V analogue pin off the board as I think the connecting track goes under the board as far as I can see on the SI.

Initially I was just going to try separately regulating the analogue supply
with the intent of keeping any noise from the +5V digital out. If it is succesful I will probably try a separate reg for the digital +5V

I have some little boards for the LT1763 left over from another project
which will do nicely. I used a 1uf/16V oscon on the reg i/p and 10uf/16V oscon on the o/p which seems to work well. Ill let you know how it turns out.

Occam · « **Reply #17 on:** 1 Sep 2005, 05:53 pm »

Quote

Ill let you know how it turns out.

Hot damn!

doug s. · « **Reply #18 on:** 11 Sep 2005, 06:39 pm »

Quote from: tianguis

All:
I feel like I'm hogging this thread, but if your're bothering to read my blather, it may be because you'd like to build one.
I discovered how critical matching speaker impedence is. I've listened with two values before today. It figures that the third is the charm in my system and the difference is not subtle. It's HUGE. If I were to build another, I'd substitute some posts in the four cap positions and experiment with .22 and .33 on the posts. It would then be possible to parallel both on ...

hi larry, the autocostruire website mentions the ability to match speaker impedence for 4-6-8 ohms. do you know if it's possible to match for higher ohms, like 14 or 16?

thanks,

doug s.

tianguis · « **Reply #19 on:** 12 Sep 2005, 06:01 pm »

Doug:
Without doing the math or investigating availability (I'd try .1 polypropylenes), smaller than .22 uf matching caps MAY be better. However, given the impedence curve of your speaks the .22's are probably fine. My experience showed that 3 times the recommended capacitance didn't work too well!

Regards,
Larry Welsh