My Ref-T amp

[IndyAudi]

Originally posted by BWRX
She's alive! Regardless, it sounds nice with a wallwart for the unregulated input to the Vdd regulator and a 9V battery before the 5V regulator,I can tell channel separation is already a bit better as the imaging is very good.

Brian,
The more you tell us the more intriguing this sounds,yes I'm interested - you betcha !

[BWRX]

A little bit of info about an incident I had involving that 9V battery...

As it turns out, the Tripath chip does not like it when the 5V supply drops to around 3V... I was listening and the amp just started reproducing white noise that gradually rose in volume until I jumped out of my seat and unplugged power. Moral of the story, the analog input section needs a nice steady 5V Since then I've hooked up another regulator to for the 5V section and there is definitely a bit of improvement in the sound. Not to mention the bass! Holy cow, I used 820uF 35V Nichicon HEs right next to the chip and there is absolutely no lack of bass whatsoever. My speakers are playing low notes I haven't heard them reproduce before! Almost sounds overpowering but I'm not used to much bass at all since my speakers are only 2 ways with a 1in tweet and 6.5in mid driver. Interestingly enough, the highs don't seem to have been reduced even though the bass is definitely more substantial.

I also found some .22uF mylar caps (I think they're mylar) that I put in the output filter as they're the appropriate values for my 8 ohm speakers.

She sounds quite good in current form but should improve a bit with a good low noise on-board 5V regulator and the DC offset nulling circuit and the regulators on board too! I've got new boards in the works with a slightly different layout and will build those up for a few people to test.

[BWRX]

Tonight I hooked up the amp using my beefy (for amps of this power level anyway) linear supply consisting of a 120VA Plitron toroid with 15VAC secondaries, an IR 25A bridge for each secondary, and a 22000uF Panasonic cap in parallel with a bleeder resistor and a 0.1uF Panasonic polypropylene film cap for each secondary. It puts out a rock soild 23.5V to my twin LT1085 regulators (set to 14V) for each output channel; the regulator for the one channel also powers the 14V analog driving section. A 12VDC 1.5A unregulated wall wart transfomer supplied DC voltage to my LMS1585 regulator (set to 5V) for the 5V analog section.

All I can say is that it works and sounds fabulous. There's a bit more detail and the soundstage sound a tad wider than before. I can't tell if it has more clean power because I don't want to play it too loud right now.

The power supply setup I'm using is most definitely overkill but that was all part of the experiment. Despite what Tripath said about using separate supplies, the Ref-T v1.0 board seems to work quite well but more testing is needed to see how it fares under differing conditions.

I have a rather large Digi-Key order coming in later this week and hope to have three boards fully built within two weeks so I can send a couple out to be tested/auditioned. I won't say who they'll be sent to so you'll just have to wait and see

Finally, here's a photo of the Ref-T with output caps next to my "beater" AMP3 for comparison.

[Pano]

Cool,
I especially like the heat sink attachment method.

Weren't you going to put regulation onboard? Or is it there and I don't see it?

[serengetiplains]

Looks great, Brian! Good to hear it sounds promising (as it should). Looks like a winner.

I second the praise for the heat sink attachment. Special audiophile elastomeric properties, I assume?

[BWRX]

Originally posted by panomaniac
Weren't you going to put regulation onboard? Or is it there and I don't see it?

All of the regulation will be onboard as soon as I receive my Digi-Key shipment

Originally posted by serengetiplains
Looks great, Brian! Good to hear it sounds promising (as it should). Looks like a winner.

I second the praise for the heat sink attachment. Special audiophile elastomeric properties, I assume?

Thanks! Yes, the audiophile elastomeric band was the heatsink attachment method of choice since I ran of thin double sided tape The chip doesn't even need a heatshink at my listening levels but it sure won't hurt anything.

[BWRX]

I've been probing the amp with my oscilloscope and found some interesting things. The waveforms look nice and clean and almost look overdamped. That may be because I'm using my actual speakers as the load instead of an 8ohm resistor. There is a 200mV (100mV from each output that are 180 degrees out of phase) 1MHz (exactly) residual signal that rides the output waveform regardless of the signal being amplified or the level of amplification. For this measurement I have two probes hooked up with the grounds at power ground and the probes at the output of one channel and the scope set to differential mode.

Michael or anyone else - have you seen this when probing around your amps?

I'll try to take some pics of the scope so you all can see what's going on.

Here's the switching waveform while the amp is idling. 5V/div vertical, 0.2us/div horizontal.

[BWRX]

Here's the residual waveform (best shot I could get). 100mV/div vertical, 0.2us/div horizontal.

[Pano]

That residual just looks like the switching sqaure wave after filtration to me. Isn't it?

Your full blast square wave above looks mighty clean, no ringing. Never seen that before. But I always scoped on resistors, not drivers. Will try the with drivers and report back.

You have both the L and C in the filters now, right?

[BWRX]

Hmm, I posted the horizontal scale wrong for that first photo. I thought I measured it at 0.5us/div. But yes, I thought it was just the switching frequency after the filter but was confused about the different frequency. I took my measurements at the output diodes before the inductors.