Sonic Impact 5066 Parts List & Modifications

[planet10]

Originally posted by panomaniac
Here is the input schenatic for the Sonic. A bit different from the Tripath app notes.

This should help in bypassing any or all of it.

Good work... L1/C1 form a 2nd order low pass filter, C3 a 1st order high pass. Interesting that they use an L instead of another RC section -- guess that gives them better control over the Q. Gets one thinking about the buffered SI...

dave

[motherone]

I have a new multimeter on the way that can measure capacitence to within +/- 2.5%. It should be here within a week. At that point, I can probably finish mapping out the SMD caps on the board and publish a complete parts list.

I almost went for the DMM with Inductance testing too, but I couldn't justify spending 2.5x as much.

This new one (http://www.circuitspecialists.com/prod.itml/icOid/4631) should be better than the cheap ratshack one I have now. I skipped the one with True RMS function for one that could measure parts more accurately. Anyone have an opinion on this? I figure I'm going to be measuring parts values and such more often than taking true RMS measurements.

[Panelhead]

Originally posted by motherone
I have a new multimeter on the way that can measure capacitence to within +/- 2.5%. It should be here within a week. At that point, I can probably finish mapping out the SMD caps on the board and publish a complete parts list.

I almost went for the DMM with Inductance testing too, but I couldn't justify spending 2.5x as much.

This new one (http://www.circuitspecialists.com/prod.itml/icOid/4631) should be better than the cheap ratshack one I have now. I skipped the one with True RMS function for one that could measure parts more accurately. Anyone have an opinion on this? I figure I'm going to be measuring parts values and such more often than taking true RMS measurements.

EPO, a local parts supplier has a elcheapo brand for 59.95. It measures capacitance, inductance, and has a temperature probe. It seems as accurate as my more expensive meters, and the cap reading are about the same as my expensive standalone capacitance meter.
These are perfect, but I damaged the temp probe recently cramming it in my parts box. I am considering buying another just to get the temp function back.

George

[Pano]

FYI:
C3/C4 are 0.33uF C1/C2 are 100pF.
Thanks to Gary P over on Audio Circle for those values. He came up with a schematic at the same time I did.

The rest are yet to be mapped.

[theAnonymous1]

Soooooo, let me see if i have this correct. C3/C4 are first order high pass filters, correct? And they are responsible for the low frequency roll off, correct? So replacing them with a larger value should fix this "problem"?

If C3/C4 is really .33mfd and its before the 20k resistor wouldn't that give a roll off way lower than the 200hz mentioned before? Assuming I'm totally wrong, and theres a pretty big chance I am. What value would be good for the lowest possible rolloff before DC becomes a problem?

Thank you for accepting my humble ignorance.

[Panelhead]

Originally posted by theAnonymous1
Soooooo, let me see if i have this correct. C3/C4 are first order high pass filters, correct? And they are responsible for the low frequency roll off, correct? So replacing them with a larger value should fix this "problem"?

If C3/C4 is really .33mfd and its before the 20k resistor wouldn't that give a roll off way lower than the 200hz mentioned before? Assuming I'm totally wrong, and theres a pretty big chance I am. What value would be good for the lowest possible rolloff before DC becomes a problem?

Thank you for accepting my humble ignorance.

The way I see it, there is a 10K parallel to the 20K, so the result is 6.67K. This gives a -3dB point of 72 Hz. This is close to the measurements shown.
The frequency response of the Tripath evaluation board is shown as -0.5 dB at 20 Hz. This is mainly due to the evaluation board having a 2.2 ufd cap, compared to the 0.33 ufd.
My speakers are only flat to a lttle below 40 Hz. I hear a real softening on the lowend, but the upper bass seems okay. This is the area where the SI is only a couple dB down.

George

[theAnonymous1]

The way I see it, there is a 10K parallel to the 20K, so the result is 6.67K. This gives a -3dB point of 72 Hz.

I wasn't even thinking about how the 10k resistor would effect the total resistance because its before C3/C4. That makes a lot more sense though, thank you for enlightening me.

I think I'll replace them with some 3mfd caps. Another dumb question, will it effect the XO point if I put the cap before everything else on the input and remove C3/C4? I only ask because I measure my input impedance at 14k, 10k+ a 4k resistor I added in series.

[bwbass]

Originally posted by Panelhead
The way I see it, there is a 10K parallel to the 20K, so the result is 6.67K. This gives a -3dB point of 72 Hz. This is close to the measurements shown.

I read the graph posted here as -3dB around 45 hz or so. Are there other factors at work to explain this difference?

[Richard U]

Panelhead,

Who makes that mutimeter you mentioned and where did you purchase it?

Richard

[motherone]

Originally posted by panomaniac
FYI:
C3/C4 are 0.33uF C1/C2 are 100pF.
Thanks to Gary P over on Audio Circle for those values. He came up with a schematic at the same time I did.

The rest are yet to be mapped.

Great work Pano. More helpful info. I think this poor little sonic impact is going to be nude to the modders of the world soon. It sounds like you guys may finish mapping this thing before my multimeter even arrives

I'm still waiting on the good ol' postal service to bring me my goodies from Digikey. I'll probably start working on the boards tomorrow, but I don't know how far I'll get if the remainder of my parts don't arrive.

Does anyone out there have an inductance meter to measure the DCR and inductance of the coils? I'm wondering if they're the same as whats on the datasheet or not. It'd also probably be handy to know what the values of the SMD inductors (L1/2) on the board are as well.