Teac Tripath - thoughts to date

lcrim · « **Reply #120 on:** 19 Nov 2004, 04:28 am »

I spent most of this evening removing the volume pots and soldering jumpers between the RCA inputs and the input traces on the amp boards based on the digital photos from the site Mark posted a few pages back. I was feeling a bit stupid about the "monoblock" thing but as it turns out there may be something to it. One of the amps only worked on a single channel when I finished because of clumsy soldering. No matter what I did, I couldn't get it working. But that single channel sounded cleaner than the other amp playing on both channels (at least to my ears.) I ended up going back and unsoldering the jumpers for one channel and I have "monoblocks" again. I thought I would just keep this to myself but then I got back online and saw Vinnie's post and now I don't feel so dumb. There are sonic benefits to running a single channel out of a single Tripath board.
How do you "bridge" a board? Is it a complex task?
TIA

toobwacky · « **Reply #121 on:** 20 Nov 2004, 04:51 am »

Quote from: albee

Has anybody thought to try 4-conductor speaker wire to the L/R binding posts used with a Y-connector to the inputs for 60 watts per channel?

Is strapping the 2 channels of an amp, as suggested by albee, do-able? Anyone?

TW

lcrim · « **Reply #122 on:** 20 Nov 2004, 02:09 pm »

"bridging" a board internally to double the output would be a more elegant solution to the issue. Unfortunately, I don't know how to accomplish this. But there are probably people out there who do.

Occam · « **Reply #123 on:** 20 Nov 2004, 02:25 pm »

The tripath chip in the TEAC is already operating in bridged mode (Bridge tied load, BTL)

lcrim · « **Reply #124 on:** 20 Nov 2004, 04:45 pm »

Occam:
Is there a way to access this BTL output as presently I am only hearing the output from a single channel on the Tripath board. If you read about a page or so back in this thread VinnieR mentions that it is possible to "bridge " the two channels on a single board and enhance the power output. I don't know if this would also increase the noise from the power supply as a trade off but it seems likely.

mcgsxr · « **Reply #125 on:** 20 Nov 2004, 05:27 pm »

Larry, if you go to Tripath.com, and search around for products, and then audio, you will find a section dedicated to their differing chipsets, and pdf files outlining specs etc.

In one of these files, specific to the 2050 chipset, you will find a portion about bridging this chip into 75w into 4ohms - not sure how to accomplish this with the Teac, without major surgery to both the amp and main boards, but I am sure it could be done.

I too, would be worried about increased noise, as the trade off.

I think that people here are developing the DIY mod of simplifying the circuit, and removing the extra board - this is a significant upgrade, according to what people are reporting hearing.

The next level up, is to upgrade the PS caps etc, and for me, that meant turning to an established modder - Wayne. There are people who can do these things themselves, but not I...

dwk · « **Reply #126 on:** 20 Nov 2004, 06:38 pm »

You can't 'bridge' the outputs, because as others have said each channel is already a fulll H-bridge output stage - most digital amps do. The only thing you can do is run the two channels in parallel. This will not offer any voltage output improvement, but could in theory offer improved current delivery. You'd have to check the tripath data sheets to determine whether any special considerations have to be paid when paralleling channels.

Occam · « **Reply #127 on:** 20 Nov 2004, 06:41 pm »

The 2050 chip is a BTL amplifier whether it is configured as stereo (page 6) of the 2050 PDF or as a paralell BTL (page 7).
http://www.tripath.com/data.htm

As a stereo amp, whether you use both channels or not, the chip is not capable of driving a 4ohm load without significant distortion. This is not to say that one might not find that distortion utterly beguiling.

Configuration of the 2050 as a paralell BTL, capable of driving a 4ohm load with minimal distortion as well as increased power would require a substantial change in the TEAC's board topology; substantially beyond some wire changes.....

I see no techincal reason why a paralell BTL configured 2050 chip would be noisier than a stereo configured BTL 2050. This is true whether one of those channels is 'disabled' in that stero configuration, or not. This is due to the fact that in paralell BTL configuration, the swithing frequecy is synchronized (and strapped) on both sides, whereas in a stereo configuration, they are not. Nevertheless, implementing the Teac board as a paralell BTL is far beyond a mod/tweak; it is a significant topological circuit change.

mcgsxr · « **Reply #128 on:** 20 Nov 2004, 10:27 pm »

Occam, thanks for adding to that, had read through that info on Tripath's site, but it is well clear of my understanding, so I appreciate your extra info, and thoughts.

lcrim · « **Reply #129 on:** 22 Nov 2004, 02:42 pm »

Well, I think I'm going to listen to these for a while and stop opening these things up. I have a single board in thr L,R position with a jumper soldered between the RCA input and one of the channels. I tried doing this with the Center channel board a number of times and for some reason found the L,R board w/ a single channel implemented has better "grunt."
I got some copper sheeting from Home depot that is fairly thin and paper backed. It is sold in 8" x 20" rolls. I cut two pieces and cupped one over the power supply board and one over the amp board. So there are two thicknesses between the power supply and the amp board.
Earlier I had soldered in a better power cable with a shield and attached the shield to the chassis. The round, ferrite thingy was kept in line also the connector to power supply board was kept.
So , with all the mods, there is less than $250 spent for these Teac "monoblocks", and the bang for buck ratio is through the roof. With some experimenting re: speaker matching, the quality is extraordinary and can only improve w/ further "burn-in."
Thanks to all.

mcgsxr · « **Reply #130 on:** 22 Nov 2004, 02:56 pm »

Larry, sorry to ask you to open them up again.... could you post some pics of the copper install - sounds like a relatively cheap, and simple procedure that others might benefit from...

geofstro · « **Reply #131 on:** 22 Nov 2004, 03:45 pm »

I'm just back from a trip where I managed to hear a lot of live music. The fabulous Lila Downs in Nantes, followed by the London Jazzfest.

Haven't had any time to check on things here until now and I'm so happy to find so much quality input and great ideas on modding the Teac.

My second Teac turned up last week and I shall definitely be pulling the L/R boards on both amps and running them in "monoblock" mode.

Vinnie has confirmed this puts less strain on the PS and I'm not surprised that reduction and simplification will always result in better sound.

With my ears re-sharpened I'm looking forward to hearing the improvements.

I also brought some female IEC outlets at Maplins electronics in London. I'm hoping to fit these to each amps PS and use sheilded power cable from the PS into these outlets. I'm hoping this will allow me to experiment with different power cables without having to cut a bigger hole in the case.

http://www.maplin.co.uk/Module.aspx?ModuleNo=1360&TabID=1&source=15&WorldID=2&doy=22m11

Any tips on which shielded cable I should buy? I'm looking at all the Volex power cables at mouser.com and am overwhelmed by the choices.

Thanks

Geoff

lcrim · « **Reply #132 on:** 22 Nov 2004, 06:54 pm »

geoff:
The Volex 17604 (mouser part # 686-17604) is the low end of what should be tolerated as a power cord. It is available from most large electronics suppliers and there is a story that Jon Risch noticed that things just sounded better when it was used and then he improved the connectors --- the Asylum power cord! Don't know how true that is but as a starting point its a huge improvement over most stock cords.
Chris venHaus offers much better raw materials for this kind of thing @http://venhaus1.com/VH_Audio_Test.html
He's one of the good guys and has great service. If his stuff seems pricey, check what the same things cost elsewhere.
Chris has a 12AWG double shielded wire (cryoed) for $4.00 a foot that w/ a Marinco male connector and an IEC plug of similar quality, makes for some very nice power cables.
I personally couldn't justify spending that much on power cords for a $100 amp. But, hey guys spend $700 modding $200 receivers so what do I know?
Mark: I'll try to post some pictures tonight , if I can figure out how to upload to the gallery here.

ludavico · « **Reply #133 on:** 22 Nov 2004, 07:23 pm »

"...and the bang for buck ratio is through the roof." -Larry (lcrim)

Ahh...that is music to my ears.

Teac #2 just arrived 2 minutes ago. Teac #1 was "modded" yesterday by soldering a jumper from the RCA of the center channel to the input of the amp directly. (Pulling the input pots out of the A-L700P was a breeze).

Thanks for the tommytube URL that shows pics of this Mark!!!

Modding of Teac #2 will commence this evening.

Larry: did you ground the copper shields that you put between the amp and PS? Thanks, BTW, for the info on power cords. I been thinking of trying the DIYcable JR filter and the "asylum" cord. Dang, so the center channel amp board was not as good as a "mono'd" L,R amp?

Hmmm...

I should be cocked, locked and ready to rock tonight (with just the level 0 mods).

John

geofstro · « **Reply #134 on:** 22 Nov 2004, 07:36 pm »

Larry,

Thanks from me also. I'm sure I'll find the cables I'm looking for from those options.

Also to Mark. Those pictures are so clear I might even try to bypass the pots myself.

Cheers

geoff

sluggo · « **Reply #135 on:** 24 Nov 2004, 04:30 am »

I have some experience with the Tripath parts and can offer some advice to those interested in modding theTeac amp. I don't own one the amps myself, so I'm going based on the digital photos I've seen.

Each 2-channel amp is implemented in a two-package product - the TP2000 (the visible part) is the "controller" for the H-bridge power die, the TP2050 (mounted on the "back" of the PCB (under the large aluminum heatsink). There are two 2-channel amps in the Teac, one on each board. On the L/R amp board, the outputs are connected to the L/R speaker connectors. For the center channel board it APPEARS that they have left unconnected one of the outputs, driving the center channel with one channel of a stereo pair. This seems very strange, as there are output inductors for both channels on each board, and if you are not driving one channel you do not need these parts. I've never known CE manufacturers to give away free parts, so I'd have to assume that the "center" channel is actually a paralleled output. The way to tell would be to look at the top of the main board and see if there's traces present on the center channel's 7-pin output connector that parallel the outputs.

IF the center channel output is a paralleled stereo pair, then you have a higher output current capability on the center channel. You could drive (for instance) a 4 Ohm load to 90W (at .5% THD), assuming you had enough mojo in the power supply.

If you are only going to use this as a 2-channel design, then you might want to consider paralleling the L/R channels as well. Driving 4 Ohm loads, this will give you a 2x70W amp (.005% THD) if the power supply can provide 7 Amps at 30V (210W). Looking at the photos of the Teac supply, my guess is that it's closer to 120W than 210W, so maybe it's only good for 2x40W. Who knows? After all, it's only spec'ed for 6 Ohm loads.

Anyway, there's nothing to paralleling the L/R channels. Just run a heavy wire from the left red post to the right red, and another from left black to right black. Connect the load to either output. On the input side MAKE SURE that you have the same mono source connected to BOTH channels, and NEVER run in a stereo source into the L/R inputs. If the L/R outputs are paralleled and the L/R inputs are not the same, bad things will happen.

If you're driving 8 Ohm loads, there's really no point in paralleling the outputs. You won't get any increased power output, as the voltage swing is the same either way, and at 8 Ohms you're not going to current limit the device. Paralleling on the TP2050 is really only useful for 4 Ohm loads.

If you have a 4 Ohm sub and 8 Ohm satellites that you like, this amp is set up really well as it is.

If I were hot rodding it, I'd first check the power supply. The TP2050's sweet spot is in the 28-30V range. If you can find a nice quiet supply that provides 30V (regulated) at 7A, you are set. Switching supplies are not a disaster with a switching amp like this, so you can get away cheap if you need to. If you do change the power supply, remember that the TP2000 requires +5V as well (it looks like there's a 7805 on the main board). And if you change the supply voltage, you'll need to adjust the output offset. Fortuntely, they have the pots for this (VR80 and 81).

Second thing I would do would be to bulk up the local storage cap. There's not a lot of room, unfortunately. Probably have to lay it across the top of the chip. Bypass it with a high quality film.

With the increased current capability I'd look at the four output inductors L309-L311. These may be perfectly capable for higher current outputs, but there's no harm in replacing them with bigger cores. Maybe swap them in one channel only and compare them with an unmodified board to see if you like the change. I'd stay with the default filter value of 15uH.

I'm slowly talking myself into going out and buying one of these things. It looks like a nice design for the money, with lots of room for mods. If I do, I'll post here and let you know what works for me.

regards,

Sluggo

albee · « **Reply #136 on:** 24 Nov 2004, 04:57 am »

sluggo, you are a genius! You need to be in the Tripath mod business. ;0)

Well, gang, my first brave move was to yank my "center" channel board this evening. Like others, I heard an immediate lifting of a veil or two. Sounded cleaner and a tad more defined! Amazing.

I'm not a modder of anykind so instead of trying to remove the center binding posts from the chassis I just grabbed some pliers and "unscrewed" (with great effort) the knobs. Helps to get them out of the way.

Well, I have the bug now. I would like to remove the volume pots and go direct but I don't trust me soldering skills (what skills?). Any suggestions on gauge and type of wire?

Occam · « **Reply #137 on:** 24 Nov 2004, 06:19 am »

Quote from: sluggo

I have some experience with the Tripath parts and can offer some advice to those interested in modding theTeac amp. I don't own one the amps myself, so I'm going based on the digital photos I've seen.

..... so I'd have to assume that the "center" channel is actually a paralleled output. The way to tell would be to look at the top of the main board and see if there's traces present on the center channel's 7-pin output connector that parallel the outputs.

If it were, the center channel would be rated at higher than those 30 watts...

Quote from: sluggo

Anyway, there's nothing to paralleling the L/R channels. Just run a heavy wire from the left red post to the right red, and another from left black to right black. Connect the load to either output.

Given the modulator feedback scheme of this intrinsically BTL, bridge tied load, topology [whether stereo or paralelled] this is not the appropriate architecture for paralelling the 2050 outputs. Please download the PDF for the 2050 and look at the implementation of the stereo BTL (Page 6) as well as the proper implementation of of the paralelled BTL configuration (Page 7). Indeed, properly modding a dual channel Teac board for paralell operation, is not topologically difficult.; its a rerouting of a few traces and a rationalization (removal) of some redundant components on the output filter, and the change of some smt feedback capacitors. The difficulty is that the output inductors will not support the increased current and will saturate. Approriate purposemade inductors would be available from Pulsus or Cambion, but mounting such might be difficult.

sluggo · « **Reply #138 on:** 24 Nov 2004, 08:27 am »

Quote from: Occam

If it were, the center channel would be rated at higher than those 30 watts...

Not necessarily. If the amplifier is voltage limited into 6 Ohms, then paralleling the outputs will make no difference in the available power output.

Quote from: Occam

Given the modulator feedback scheme of this intrinsically BTL, bridge tied load, topology [whether stereo or paralelled] this is not the appropriate architecture for paralelling the 2050 outputs. Please download the ...

The circuit you refer to is appropriate when designing a board for paralleled outputs. It eliminates duplicated parts and saves the manufacturer money. For whatever reason, Teac has populated the outputs of both channels on the center channel board even though only one channel is used. Since my comments are directed toward people who want to mod their existing Teac (as opposed to designing their own amplifier), it seems more appropriate to describe what can be done with the existing circuit.

Feedback is taken prior to the inductors. Paralleling the outputs after the inductors will have a negligible impact on the operation of the loop.

Paralleling two fully populated outputs doesn't change the current through any particular inductor.

ludavico · « **Reply #139 on:** 24 Nov 2004, 03:16 pm »

"Any suggestions on gauge and type of wire?" -Albee

For what it is worth (I have no significant experience in elctronics) I used 18 gauge solid core hookup wire and Cardas eutectic solder. I bought a cheapie 30W soldering iron with a pencil tip, and I practised with one of the Elenco Electronics soldering kits (Vellemen has some cool ones too) before I attempted anything with the TEAC boards. The practise helped for sure! Still, I mucked up TEAC #1 by inadvertently creating a solder bridge between the C-out pin and G.

Damn! I ended up having to jumper directly between the RCA center pin and the input pin of the Tripath chip.

Albee, you have to be real careful if you are a piker like me...

Dude, if you get a Elenco kit, make sure to practise what you are going to do on the TEAC board.

Teac #2 went much better.

sluggo - good on ya... I just wish I had the background and skills to play around with the A-L700P design. Definitely sounds like TEAC is leaving a lot of what the Tripaths can do in the box.

Of course VRossi Audio is spinning up now

Happy Thanksgiving!

Cheers,
John