ground symbol confusion

[madtecchy]

Hi can anyone please clear up a question about the ground symbols on this schematic. I have highlighted the offending symbols in yellow and put my guess next to them. Also you will notice two back to back electrolytics on the input. would there be any benefit in sound quality if i replace these two capacitors with one polypropolene LCR 8mf audio grade cap.

regards Ian

[firechief]

The two caps is series whould result in 5 ufd with the idea that the combination would be non polarized (that is why the polarity is back to back). A replacement could be 5 ufd non polorized in you favorite type. The ground symbols are confusing, there shoud be some indication on the drawing. Look for some note.

[madtecchy]

Thanks Fire chief.
I have had a look at the psu schematic and have found the symbol i have marked as power ground is the 0 volt junction between the two large res caps in the psu. so i guess power ground. as for the input caps i mentioned i did understand why the caps are back to back and that the sum of both equals 5uf. to replace the two 10uf input caps will mean i have to remove them and replace with links as there is no room to fit the larger film caps in that location .i was planning on locating them near to the input connectors.

After the input caps there is a 1k resistor that conects to the inverting input of the opamp wich i understand from my extensive reading i have done on this forum to be virtual ground. Now the two caps and the resistor to ground create a high pass filter with a corner frequency of 31.83Hz. the spec sheet of this amplifier says it has a frequency response starting at 7 Hz. The confusion grows

[sofaspud]

Your guesses for the different grounds seems correct to me, just from looking at the schematic. But I think it is R102 not R101 that's used for the fc formula, with R101 being a small enough value that it can be ignored.

[madtecchy]

Originally Posted by sofaspud

Your guesses for the different grounds seems correct to me, just from looking at the schematic. But I think it is R102 not R101 that's used for the fc formula, with R101 being a small enough value that it can be ignored.

Thankyou Sofaspud.

I recalculated the cut off frequency using the value of R102 47k5 and 5uf cap and i get the cut off frequency of 0.67 Hz and a time constant of 0.2375 second.

If i calculate with 8uf (as these are what i have) I get cut off frequency at 0.42 Hz and a time constant at 0.38 second. Do you think it would be ok to use the 8 uf film cap instead of the two back to back electrolytics.

After reading many posts on this forum i am led to belive that so called audio grade polyprop caps are better in this position. Size and cost left out of the equation.?

and finallly just to clear things up a little is the inverting input of the opamp not seen as virtual ground as it is in the lm3886.?

Best Regards Ian

[sofaspud]

Electrolytics are best left out of the signal path - that's pretty much universally agreed upon (See? It does happen!). Larger value film caps are more of a premium item and nowhere near as common as electrolytics. The 8uF film would be the caps to use.
I don't really have an answer for your op amp question. It depends on how it is connected. In the circuit above, the inverting input is connected to DC ground thru R102; the noninverting input is connected to AC ground thru C101. At least I think that's correct.
I had thought an op amp amplifies the difference on its inputs, then I find a circuit with the inputs tied together. What's going on? Is that a unity gain configuration? Oooo my head hurts

[jitter]

Originally Posted by sofaspud

I had thought an op amp amplifies the difference on its inputs, then I find a circuit with the inputs tied together. What's going on? Is that a unity gain configuration? Oooo my head hurts

An opamp with inverting and noninverting input tied together obviously isn't going to be useful. A differential voltage of 0 amplified by any gain will have 0 as result.

Did you mean output tied to inverting input? If so, that's a voltage follower and that has a gain of 1.

[cliffforrest]

Originally Posted by jitter

An opamp with inverting and noninverting input tied together obviously isn't going to be useful. A differential voltage of 0 amplified by any gain will have 0 as result.

Did you mean output tied to inverting input? If so, that's a voltage follower and that has a gain of 1.

To be pedantic that is not quite true

Tying the inputs together means that the input Vos (offset error), almost no matter how small, will get amplified by the open-loop gain causing the ouput to whack against one of the power rails.

Cliff (pedant)

[madtecchy]

If You look at the schematic i posted . the the inv and non inv inputs are not tied together. I have another question to ask about the schematic i posted . is it using what is known as nested feedback? to me it looks simler to the myref chipamps that can be found on this forum.
Regards ian

[sofaspud]

madtecchy, I don't think yours is a nested op amp circuit. "Nested" is, I believe, another term for having 2 op amps with one inside the feedback loop of the other (as seen in this composite amplifiers thread). Your circuit reminds me more of the headphone amp circuits that include a discrete output buffer. I haven't checked out the MyRef circuits yet, but it did come up in the aforementioned composite amp thread.
(see Linear Technology AN-21 Composite Amplifiers by Jim Williams, Figure 2, for the OT stuff)