[amb]

Heady, have you checked to see if your DMM is actually reading a reasonable current? I don't see anything obvious in the photos, but I assume you checked all nets with your DMM in ohms mode? Does it actually sound ok?

nysulli, 62mA is still within range. The AD8397 datasheet says that at high supply voltages (24V specified, but two 9V batteries come close), the quiescent current is typically 11mA but can be as high as 15mA per amp. There are two amps per chip, and there are two chips. Considering that Rled is 4.7K you probably have close to 4mA flowing through the LED. The ground opamp will run warmer than the L/R opamp after playing music because it has to source/sink the return current from both channels. I wouldn't worry too much about it.

__redrum, in a Mini³fied PINT, L1 is no longer within the feedback loop. You can substitute 10Ω resistors for L1 for the L and R channels but not for the ground channel, or you will kill stereo channel separation. For those who still have problems with cable-induced oscillations it might be a worthwhile experiment to add an additional 10Ω resistor in series with each of the L and R channel outputs to see if it makes any difference. Since the ferrite has much more impedance than 10Ω at high frequencies I wouldn't advocate removing them in favor of resistors only.

For those with too much quiescent current even without any cables connected, then the problem won't be solved by adding resistors or changing ferrites (because that's not where the problem is).

 

[Filburt]

Damnit...

I tried resoldering the cap on C3L...and after taking it off and putting it back on, my power draw shot up to 160mA. I ended up just replacing both the cap and the resistor thereafter, and it's still at 160mA. What happened?

 

[tomb]

Quote:

Originally Posted by Heady Yes, it is according to amb's instructions for mini3-fying the PINT.

Doh! my bad!

 

[amb]

Quote:

Originally Posted by Filburt I tried resoldering the cap on C3L...and after taking it off and putting it back on, my power draw shot up to 160mA. I ended up just replacing both the cap and the resistor thereafter, and it's still at 160mA. What happened?

Geez, I wish I could tell you, but I feel your pain...

 

[Filburt]

Quote:

Originally Posted by amb Geez, I wish I could tell you, but I feel your pain...

Maybe the board is bad. I had to use some CA to fix a manufacturing flaw on the side where the copper for the ground plate was exposed. It could just be some manufacturing defects are at work here, and my soldering simply aggrivated some other defect. Maybe I should try to get another board from tangent.

Gosh, I wish I could tell what's wrong with this thing. I don't think I've blown a chip up yet, and it seems like this is due to oscillation of some sort. I just can't figure out how to fix it!

 

[Heady]

Quote:

Originally Posted by amb Heady, have you checked to see if your DMM is actually reading a reasonable current? I don't see anything obvious in the photos, but I assume you checked all nets with your DMM in ohms mode? Does it actually sound ok?

amb - I haven't got the time to check my DMM yet but will soon. On another note, a sharp-eyed forumer in sgheadphone.net commented my C3 caps looked odd. I may have used the wrong values and will be checking soon. Sigh, I need a microscope.

The strange thing, is the amp sounds good to me now. Of course, I am old so my high frequency sensitivity is probably gone, so I won't hear any oscillations in the form of high pitch whistles or screeching.

But thanks, you have been a great help to all of us struggling to get this amp working.

 

[Heady]

amb - I have checked my DMM and it is reasonably accurate with the current draw. I used two AA batteries with 3.024V thru a 219.5ohm resistor and got a reading of 13.54mA which seems to be about right.

I have also checked the capacitors and I have all the correct ones placed. The supposed error was probably due to the color of my pic being off.

Can you elaborate on "checking the nets" as you have mentioned in one of your earlier post?

Thanks.

 

[cire]

just finished my PINT last night. mini'fied with a gain of 6. this thing is incredibly quiet, while listening through ksc75 with the volume turned all the way up, there was almost zero noise. i guess i'm lucky in that my pint had no complications.

 

[amb]

Quote:

Originally Posted by Heady Can you elaborate on "checking the nets" as you have mentioned in one of your earlier post?

Looking at the schematic, you'll see a number of components connected to each other. Each node of these connections is a "net". For example, the non-inverting input to the L/R opamp is connected to one leg of R2 and one leg of C1. That is one net, albeit a simple one with only three connections. Some nets have many more. Be sure to use the correct schematic. If you had Mini³fied your PINT, then the alterations should be reflected on the one you're working with.

With the amp power off (and all capacitors discharged), you check for continuity between each of those components on the same net with your DMM in ohms mode to make sure they are connected (0 ohms), and check that net against other nets to make sure they are not connected. Between nets that have resistors across them, you should measure the value of the resistor. For example, at the inverting input of the L/R opamps, if your R3 value is 1.2KΩ, then you should read 1.2KΩ between that net and the virtual ground net. However the meter should read about 101.2KΩ between the inverting input of the opamp and the non-inverting input (the resistance of R2 + R3) if your C1 is really a capacitor. If you Mini³fied your PINT, then the reading is more like 2.2KΩ at minimum volume position (R2 || R1) + R3, note that the R1 here refers to the 1K resistor installed in place of C1, not the actual R1 in the schematic. At maximum volume position on a Mini³fied PINT, the measurement should be (R2 || (R1 + Rpot)) + R3, which is about 11.1KΩ.

As you can see, you need to account for all the resistances between nets which can be a series or parallel combination of various resistors.

This is complicated slightly more by nets that are bridged by large electrolytic capacitors which will cause the reading to briefly show something between nets due to capacitor charging by the voltage from the DMM, so you'll want to let the reading settle (to "inifinite"). The opamps and diodes will also affect readings between certain nets, so measure the continuity with your meter both ways (swap the leads). For the opamp, since it contains circuitry inside, may skew the ohms reading between certain nets.

Nevertheless, going through this exercise will help you determine if there are problems with parts values used, solder bridges, poor soldering, possible defective parts, and force you to think about all the interconnections of the circuit which may yield a new understanding about any malfunctions.

 

[Heady]

Thanks amb for your very informative post. Unfortunately, the current built I have is beyond help. I will use your info for my next built. I am now thinking of soldering the opamp last, so I can check and measure all the connections prior to soldering them in. Do you think this is feasible?

 

[amb]

Quote:

Originally Posted by Heady I am now thinking of soldering the opamp last, so I can check and measure all the connections prior to soldering them in. Do you think this is feasible?

Yes, you could do that, but it might be a litle harder to do. With all the other parts in the way, you'll have to be careful with the soldering iron tip not to melt anything else.

 

[Filburt]

Alright, tangent sent me a new board. This one is fantastic...better build quality than any board I've seen thus far.

Anyhow, I did this thing on autopilot as I was up all night writing a paper, and ended up using ferrites instead of resistors. Everything turned out fine, though. Power draw is ~58mA, and DC offset is less than a mV.

So I now have a working 8397/6172 hybrid, and standard dual 8397. Yay!

 

[dhp]

Quote:

Originally Posted by Filburt Alright, tangent sent me a new board. This one is fantastic...better build quality than any board I've seen thus far. Anyhow, I did this thing on autopilot as I was up all night writing a paper, and ended up using ferrites instead of resistors. Everything turned out fine, though. Power draw is ~58mA, and DC offset is less than a mV. So I now have a working 8397/6172 hybrid, and standard dual 8397. Yay!

what's the difference in the boards? A pad came off on my old one though :|

 

[Filburt]

Quote:

Originally Posted by DieInAFire what's the difference in the boards? A pad came off on my old one though :|

The dimensions are exactly as they should be, ergo no parts of the board are shaved off where they aren't supposed to be (as was the case on the defective board). Overall build quality seemed better in general, including durability of pads.

Who/What is that in your avatar?

 

[dhp]

Quote:

Originally Posted by Filburt The dimensions are exactly as they should be, ergo no parts of the board are shaved off where they aren't supposed to be (as was the case on the defective board). Overall build quality seemed better in general, including durability of pads. Who/What is that in your avatar?

not me