[Vinnie R.]

So my:
Freq. Range -6dB at 52Hz and 30kHz  
Sensitivity 88dB spl (2.83V 1m)  
Normal Impedance 8 ohms (minimum 4.3 ohms)  
Power Handling 25W - 100W into 8 ohms on unclipped programme  
B&W DM303 aren't ideel :/

No, definitely not ideal for this amp.  It really thrives on higher efficiency speakers, as 6moons mentions using in their review.  This is similar to the low power SET tube amps I guess.

While my SI sounds really nice at "normal" listening volumes on my 90dB speakers, when I try to make it loud, the sound quality drops out.  This is demonstrated in the datasheet for the 2024 chip used.  It does really well with measurements until it pushes out more than 6 watts into 8 ohm loads, or over 10 watts into 4 ohm loads.  

It is a great sounding amp, but needs to be mated with higher efficiency speakers to sound good at louder volumes.  

IMO, it seems to sound best with a 12V SLA battery.  I tried it using a 12V Linear Regulated ac/dc power supply, and with the 8 AA batteries.  Using the 12V SLA sounds the best to me (best bass, best clarity and lowest noise).

[grkn]

If I use a 14V battery, could that help?

[KCHANG]

Even when the stock volume is at its max, the signal still passes through it (degradation), and it is adding 50k of load to the source,

Hi Vinnie,

  I feel that your statement needs a bit clarification.  Even though the signal does pass through the stock volume control even when it is at its max, that 50K resistance is in parallel with the signal input of the Tripath chip.  In other words, the volume pot in the maximum position does not impose a seriesl resistance before the chip input.  You get the 50K of the volume pot in series with the chip input only when the volume setting is at 0.

 Also, since the 50K is also in parallel with the second leg of the shunt-type volume control, it does not add 50K of load to the source.  In fact, for a given setting of the external volume cotnrol, the source sees a reduced impedance as the overall load, because the combined resistance of the 50K and the second leg of the shunt-type volume control in parallel is less than that of the second leg of the shunt-type alone.

Kurt

[Vinnie R.]

Hi Vinnie,

  I feel that your statement needs a bit clarification.  Even though the signal does pass through the stock volume control even when it is at its max, that 50K resistance is in parallel with the signal input of the Tripath chip.  In other words, the volume pot in the maximum position does not impose a seriesl resistance before the chip input.  You get the 50K of the volume pot in series with the chip input only when the volume setting is at 0.

 

Hi KCHANG,

Yes, this is true.  But if you use an external preamp with the SI, the signal is going through 2 potentiometers (1 in the preamp, and the 50k pot in the SI).  I guess I have a hard time understanding how the sound can really get better by using another passive preamp in front of the SI.  You are putting more "junk" in the signal path.  I suppose and active preamp would be a different story, as the sonic signature of the active preamp is sent to the SI (ex. tube preamp to get the "tube sound"), but the standard 2Vrms source (ex. CDP) has more than enough voltage to drive the SI into clipping when the volume it cranked all the way up, so is an active preamp needed?  I say keep the signal path as sure and pure as possible.  The SI already is a passive preamp (passive volume control), and being that it is 50k, I would say any source can drive it successfully.

And yes I agree/understand that when the volume is maxed out on the SI, the 50k is in parallel, not in series with the input.  But if you are using a passive preamp external to the SI, this is just adding more load (ie, an additional 50k in parallel with the passive preamp), plus, the signal is now going through the 50k pot's wiper and the extra connectors and wiring.  If you are going to use an external preamp, it is best to hard wire the input of the SI directly to the board, and disconnect the 50k pot.  Bypass the extra wiring, connectors, pot, etc and keep it simple and pure  

Good discussion,

Vinnie

[grkn]

Coulnd't you just put a better pot in there??

[Vinnie R.]

Coulnd't you just put a better pot in there??

You sure can!  If you really wanted to, use a stepped attenuator like a DACT or Goldpoint

But if you convert the stock pot into a shunt-mode pot, then you'll be taking the stock pot out of the signal path and it will only be used the shunt the unwanted amount of signal to GND (for attenuation).  This is probably the cheapest cost and biggest performance increase for the money.  Two resistor don't cost too much, unless you go crazy and get those Caddocks at $6 a resistor  

The only thing to be careful about in implementing a shunt mode pot is your choice of series resistor before the pot.  It will effect your attenuation curve, and also is parts of the input impedance, which will change based on volume position.  The lowest input impedance will be the series resistor (when the volume is at a min), and the highest input impedance will be the series resistor + the 50k of the pot (when the volume is at a max).  If you are not using an active preamp, I wouldn't go below 10k unless you know your source is up to snuff to handle the load.  

-Vinnie

[Vinnie R.]

If I use a 14V battery, could that help?

I'd stay away from 14V.  The max voltage for the TA2024 is 13.2.  
If you run it at 14V, it may not blow, but you'll probably be straining it and causing it to get hot.  This little chip does not have mucho heatsinking, so be careful.  

Either you enjoy the SI with your inefficient speakers at moderate volume levels, or you look into trying higher efficiency drivers.  For example, if you have 87dB speakers, then using 90dB speakers (3dB higher efficiency) should give you the output level as if your SI had double the wattage, right?  This becomes 4x for 93dB, etc.  I think that is how it works.  Remember, double the wattage does NOT equal double the volume.

[KCHANG]

Hi Vinnie,

 I see that we pretty much share the same view, and it's just that we apparently interpreted differently the terms "signal passing through the pot" and "adding the 50K load to the source."    Like you said, the 50K pot might not be that bad, and if one really wants to improve the sound by using an external "high-quality" passive pre it might be better to remove the 50K pot to simplify the connections.  I thought about doing that but decided to keep the pot because I liked its small size and the built-in power switch.  By the way, I've been powering the SI amps with a big car battery (how about 850CCA) connected to a trickle charger.  I thought the amp sounded more potent than when it was powered by a 2A 12VDC regulated power supply.

 The UPS guy just delivered my JVC RX-F10S receiver 20 minutes ago.  Thanks to DMason for drawing my attention to that amp.  It'd be interesting to compare the two digital amps driving my FX200's on open baffles.

Kurt

[Vinnie R.]

Kurt,

Keep us posted!  

Have Fun!

Vinnie

[grkn]

I guess this is the stuff to read for making a shunt

[tianguis]

All, and Vinnie, too!:
       I received a SLA battery today from BatteryMart, and the difference is impressive. I thought it might be. I've been running the SI with wall warts, battery packs I made up, or the internal AA's. The capacity of the SLA increases the dynamics, detail, bass and sound stage. I never considered a switching or regulated power supply as they're all too "dirty".
       Looking at this thread (I just found it, never posted here before), I find my experience seems to contradict several opinions. I'm driving the SI with a Transcendent Grounded Grid pre at the moment, feeding it from a Tosh 3960 with Vinnie's and John Swenson's mods (using Auricaps and Bybees). I've also used it being driven directly by a Hagerman Cornet phono pre with equal results. I find that the least-distorted settings result from a cranked GG and low-volume SI. I can dial digital glare and hash in or out depending on those relative settings.
        Before putting a Panny XR50 in my system, I'd been using all tubed stuff for years. Since getting the SI (bought as a sub for the Panny while I do some mods) I wonder which will sound better? I have another SI coming, which I intend to butcher (caps, damping, posts, RCA's, wire, chassis).
         I've used the SI to drive both Hammer Super 12's and noRh 4.0's (which do surprisingly well at ~87 dB efficiency). I think this T chip stuff is here to stay. My audio friends think I've lost it.

Regards,
Larry Welsh

[TheChairGuy]

Larry/tianguis,

Thanks for the tip.  I HAD the SI cranked to full, and volume controlled with my Dynaco PAS-4.....thought that was the logical limitation of this $39 (list) amp.

Well, I dialed down the SI to about 10 O'clock after reading your post, and tweeked the Dynaco up a notch or two - bingo!   Lots of grain, hash and plenty of noise now out of the picture...it drives my 87 db Linaeums quite well in the midrange.

Bass is flaccid, as I've found with this Dynaco in other set-ups, but the midrange is rather terrific now.  I am thoroughly enjoying it.  I am just stunned at how decent this presentation is now.

The sheer listenability of the SI (with the Dynaco) bests that which I remember with either my old Norh MultiAmps or AudioSource Amp3..using two half-respected budget amps for reference.

I'm sure upon more thorough review of everything in an extended listening session I'll find something to nit-pick, but for $39 it'll come off as pretty slight.  Go battery folks on the SI, you will be surprised at the fidelity it's capable of (in tandem with a good pre) on speakers as insensitive as 87db (extremely simple crossovers in the Linaeums, tho)

[tianguis]

Chair Guy:
         Great! I gave the SLA battery a full charge overnight (it was originally 12.1 v.) and notice even more of what I heard yesterday: better bass and dynamics. What prompted me to try a lower setting on the SI's volume control was the THD plot I saw somewhere, with very bad things happening above ~5 watts. Seems to work.

Regards,
Larry Welsh

[Vinnie R.]

Chair Guy:
         Great! I gave the SLA battery a full charge overnight (it was originally 12.1 v.) and notice even more of what I heard yesterday: better bass and dynamics. What prompted me to try a lower setting on the SI's volume control was the THD plot I saw somewhere, with very bad things happening above ~5 watts. Seems to work.

Regards,
Larry Welsh

Hi Larry,

In terms fo the THD plot, setting the SI's volume lower and using an active preamp at a higher setting, or setting the SI's volume higher and using an active preamp at a lower setting is all relative.  It all has to do with the level of the input signal reaching the Tripath chip.  However, the
improvement you are hearing probably has to do with changing the impedance ratio between your source and amp.  In terms of distortion, there is going to be a point when the signal going into the SI is too great, and then you will get distortion.  If fact, there is a OVERLOAD pin on the chip inside that you can use to wire to an indicator (like an LED) that tells you when the input signal is overloading the chip and causing distortion.  There is also a FAULT pin that you can use an indicator to tell you if a fault condition has occurred, such as a shorted output, over-heated chip, too low input voltage, etc.  Pretty sweet built in monitoring in these Tripath chips.  

If you are going to use an external preamp, the best thing you can do is just bypass the stock volume control completely and send the input straight to the amp board.  

I am actually staying away from a preamp, as I find that this modded little beast sounds the best without it.  IMO, any active preamp is just adding noise and coloration.  Maybe the stock SI benefits from such coloration (ex. tubed preamp adds even-ordered harmonics) if you system, but after mods, you may never want to use it again  

Have fun!

Vinnie

[TheChairGuy]

Have fun!

Vinnie

....for $39!  Or, less than $60 with battery (that you can use for other toys down the line). Serious cheap thrills here with the SI.

I think running it without a preamp, on further listening, at moderate levels (below 12 o'clock) is the best route to go.  To really make the best use of this little feller', 90db and higher seems warranted.  Tho it plays okay for background on 87 db'ers, it doesn't have enough guts to quite make it.  It's really got a pleasant tone to it, overall.

Still, what a nice little back-up it is to play with.  I'm not advising folks with Maggies or Martin-Logans to try the Sonic Impact  

[tianguis]

Vinnie, CG:
        I replied at length to Vinnie's post, but it appears to have been gobbled up by the Internot. The gist: Vinnie's points are all valid and probably closer to the actual situation than my ham-fisted approach. What I was trying to get across was I thought that perhaps the T chip was as sensitive to input signal strength re: THD. I've browsed the specs without seeing anything regarding this. Time for more homework. Of course, at any given listening volume,  output power is the same no matter the relative settings of the T-Amp and pre pots.

Regards,
Larry Welsh

[ToddSTS]

...I am actually staying away from a preamp, as I find that this modded little beast sounds the best without it. ..

Vinnie,

What mods have you done on the SI?  I reread the thread and only found the SLA battery mod and metal chassis mod.  Have you changed out any components on the board itself?

Thanks.

Todd

[Vinnie R.]

...I am actually staying away from a preamp, as I find that this modded little beast sounds the best without it. ..

Vinnie,

What mods have you done on the SI?  I reread the thread and only found the SLA battery mod and metal chassis mod.  Have you changed out any components on the board itself?

Thanks.

Todd

Hi Todd,

I have been working on a bunch of mods to the board and am getting exceptional results!    It is very tedious working with the surface mount components, but that's what it takes!  

I suppose I need to "let the cat out of the bag" and mention that I have been working hard on a battery-powered prototype that uses a highly modified version of this board in a nice new enclosure  

I am planning on starting small volume production shortly and getting some feedback.  Someone who is really into Class-T is getting one on Friday for evaluation.  

I'm really looking forward to this...

-Vinnie

[BillyM]

Keep us updated, as there are still a few of us quite interested in these little buggers.  *enjoying mine more and more*  ...and some little mods might be right up my alley...  Keep good notes, take pictures, etc during your mod excursion, some of us might like to follow your footsteps...

Another idea, bridge it down to one single RCA input and a pair of binding posts for mono-blocking.  Should allow those down into the mid-high 80's-db efficency use of this little bugger (and good interconnects too!)

I still have yet to hook the little guy up to my main Athena SF-2's  Being 92db's efficent, I'd say they will put out decent sound.

--billyM

[pboser]

Hi,
I just found this thread - Vinnie, how about an update?

Also, are you saying if you remove the 50k pot you just take the signal straight into the board?  Is there any need for a 50k resistor per channel to ground?

Thanks,
Pete