Sonic Impact Mod Question...

[mourip]

I have been working on my SI in a "Bud Box" experimenter's chassis while gathering parts for my "all out" version. The nice part about this phase is that I have lots of space for trying out the various mods that are being discussed. My goal is to have an excellent amp for driving my AKG K-1000 headphones.

One suggestion that I read was to put a 5-10 ohm resistor across the outputs in order to give this amp a load that it will be happier with instead of the 120-300 ohm load that many headphones offer. I am trying this right now and the results are superb with a 5 ohm 10 watt Mills non-inductive wire-wound. For me it sounds much better with the resistor than without it.

My question concerns the implications for battery life? My amp runs from a 3.4AH SLA. It seems to me that putting a resistor across the outputs will drain the battery substantially faster than having just a 4 ohm speaker. Is that correct? I am very good on building but a little short on theory and so would enjoy discussion about this.

Thanks a lot,

Paul

[protos]

Well, yes.A 4ohm load is going to eat up more current than a 300ohm load when you have a signal of course.But since you will only need a couple of volts in headphone use you should be able to manage at least 5-7 hrs with your battery before a recharge.

[motherone]

Probably even more. Even with a 10 ohm resistor across it, you're still presenting an approximate 9 ohm load to the amp. You may be better off going with the 5 ohm resistor, since the output networks in the SI are designed around a 4 ohm load rather than an 8 ohm load.

Your battery life will depend on your playback levels. I would guess that with a 3.7Ah SLA, you'll probably get far more than 8 hours of play out of it.

I guess worst case scenario. you can always get a bigger battery

[Pano]

Be carefull usign this amp with headphones, there is a lot of ultrasound left in the output signal. Mostly around 1Mhz and at about 30-50mV P-P.

Putting a hi quality cap on the power supply rail on the PCB will drop ultrasonic hash by about 6dB. I've found that a Panasonic FM series electrolytic works as well as any film cap (Solen or other) and they are smaller and cheaper. It should help your transients too.

The SI uses about 70mA at idle. It can use as much as an amp or more running hard into 8ohms. Double that for 4 ohms. The 1st day I used mine the AA batteries lasted about 5 hours driving very hard into 4 ohms, music uses less power than full volume test tones, natch. YMMV.

[mourip]

Originally posted by panomaniac
Be carefull usign this amp with headphones, there is a lot of ultrasound left in the output signal. Mostly around 1Mhz and at about 30-50mV P-P.

Putting a hi quality cap on the power supply rail on the PCB will drop ultrasonic hash by about 6dB. I've found that a Panasonic FM series electrolytic works as well as any film cap (Solen or other) and they are smaller and cheaper. It should help your transients too.

The SI uses about 70mA at idle. It can use as much as an amp or more running hard into 8ohms. Double that for 4 ohms. The 1st day I used mine the AA batteries lasted about 5 hours driving very hard into 4 ohms, music uses less power than full volume test tones, natch. YMMV.

Thanks for the warning. I have seen a bunch of posts regarding the possible ultrasonics. That is one reason I put the resistor on the output. It is a 5 ohm so should be about right. I also replaced the stock supply cap with a much larger Panasonic and have a couple Blackgates in the bull pen. Glad to hear about battery life. I seldom listen for more than a couple of hours so I should be good to go unless I forget to switch back to charge mode.

Best,

Paul

[mourip]

Has anyone seen a schematic of the SI amp? I vaguely remember see an attempt by someone to reverse engineer it. I already have the Tripath pdf but am hoping to see something closer to the SI which I know has different parts values...

Thanks,

Paul

[motherone]

The SI is very similar to the Tripath datasheet. I think the major differences are:

1) There is an inductor at the input along with a 10k resistor to ground, prior to the DC blocking capacitor

2) The DC blocking capacitor is 0.33ufd instead of 2.2uf

3) The feedback resistor is 36k instead of 20k

Panomaniac has documented what he's found in this thread:

Sonic Tripath Investigations


And has put up a schematic of the input here:

http://www.michael.mardis.com/sonic/.../input-new.gif

I believe panomaniac is still working on documenting the output section. You may find more information at audiocircle.com.

Anyohw, you might want to search the forums, since there's a couple of threads documenting this information

Good luck!

[jmateus]

I hope this posting is not off topic...
My Sonic Impact 20 something dollars BLEW up yesterday...
Why? Because I connected it to my electrostatic speakers, it
did play for a second or two and then gave up...Took it to
a pair of dynamic speakers (the ones I had it connected to
previously)...and nothing, not a note of music...Caput!...

Now I would like somebody to help me to unsderstand why
didn't this amplifier like my ESL's, they have an aproximate
capacity of 1,200 pf, but in between there is an impedance
transformer. Another words the amplifiers always see 8 ohms
as the impedance load...

Please, can anyone explain to me why do electrostatics are
so peculiar about the amplifiers they are connected to?
I'd really apreciate some light...

[protos]

Probably some kind of oscillation into hi capacitance.Don´t forget the transformer you are using will also increase the capacitance seen by the amp output.

[jmateus]

You've got a good point there, I didn't think of that...
But on the other hand I've heard stories about regular class AB amps
with bipolar transistors at the output that dont react very well either
to the ESL. I've never found a credible (or not so) explanation why
push pull amplifiers get destroyed as soon as you plug the ESL's.
To me is a technical mistery, however any tube or solid state class A
work really well with them. I have both and I enjoy my speakers, but
yesterday was a day to forget and I'm thinking, would the ESL destroy
the Class D module I'm working on now? That's a risk that I don't to
take, this module is expensive, not the rinky dink SI cheap one.

Morou? Obrigado seu brasuca....