[Filburt]

If I get some new 8397s in, I *might* try to perform the modification amb mentions and will take pictures so people can see how it works.

Thanks for the explanation about LG, amb. That's what I figured was happening, but I didn't trust myself on it since my major is Philosophy, not electrical engineering

Tangent - Yeah, I thought I'd share on that one, in case someone else was having problems. It seems to be very stable in this configuration, and is very black on the background, even with the volume turned at full. Even in my quiet room, I can barely detect that it's on...in fact it's so quiet I really can't confirm that there is any noise. This particular aspect of it is amongst the best I've encountered in portable amps. It does seem to accentuate background noise on recordings more than my other amp, but it seems to be proportional to the greater degree of forwardness and slight increase in brightness, so I think it's probably okay. It's running fairly cool, too. With the lid open, it runs just above room temperature with my earphones. With the lid closed, it gets a bit warmer, but not too bad.

 

[amb]

Quote:

Originally Posted by tangent It'd probably be simpler to just jumper L1G and use a ferrite bead around the OG wire at the board end.

Yes, but clumsy to do with SMD parts (difficult to make a mechanically-sound wire connection to the end of a SMD ferrite. If one is to do this, then a through-hole "bead-on-lead" would be a better choice. The wire to the jack can then be twisted and spliced onto the leaded bead, and insulated with heat-shrink tubing.

 

[Heady]

Quote:

Originally Posted by amb In addition, we also put L1 and L2 outside of the feedback loop by populating R4 on the pads for C3 (leaving the original R4 pads unused), and then stacking C3 on top of R4. This is a little frustrating to do due to the small SMD part size, but it all worked out very well.

Hi amb, I may be wrong but I think tangent mentioned that C3 should be left out, if using the ferrite beads. In which case, there is not need for the fiddly stacking? What do you say?

Thanks.

 

[amb]

Quote:

Originally Posted by Heady Hi amb, I may be wrong but I think tangent mentioned that C3 should be left out, if using the ferrite beads. In which case, there is not need for the fiddly stacking? What do you say?

C3 can and probably should be omitted if you're using the high value feedback resistors as listed on tangent's schematic and parts list. However, when you Mini³-ize a PINT, you switch to lower resistor values, substitute a 1K resistor in place of the input coupling cap, and change the ferrites to be outside the feedback loop. With this scheme, C3 is no longer optional.

 

[Teerawit]

Quote:

Originally Posted by amb However, when you Mini3-ize a PINT, you switch to lower resistor values, substitute a 1K resistor in place of the input coupling cap, and change the ferrites to be outside the feedback loop.

The ferrite for just the ground channel or for all three channels?

 

[Heady]

Quote:

Originally Posted by amb C3 can and probably should be omitted if you're using the high value feedback resistors as listed on tangent's schematic and parts list. However, when you Mini³-ize a PINT, you switch to lower resistor values, substitute a 1K resistor in place of the input coupling cap, and change the ferrites to be outside the feedback loop. With this scheme, C3 is no longer optional.

Thanks for your comments. I am planning to try both arrangements. According to your earlier post, with the pictures, does it mean that:
1. make the cut as shown in the picture
2. make the jumper as shown
3. solder R4 and C3 in C3 position
4. solder L1(L), L1(R) and L1(G) in the positions as silk-screened on PCB
is what is required to mini3-ised the PINT?

Thanks again for making clear your comments.

 

[splaz]

Quote:

Originally Posted by amb C3 can and probably should be omitted if you're using the high value feedback resistors as listed on tangent's schematic and parts list. However, when you Mini³-ize a PINT, you switch to lower resistor values, substitute a 1K resistor in place of the input coupling cap, and change the ferrites to be outside the feedback loop. With this scheme, C3 is no longer optional.

Hmmm...

So if you replace the input caps and then use lower value resistors like said in the other topic, then you should also do the other fiddling around you just mentioned or is it fine just without input caps and with lower resistor values ?

 

[Filburt]

Quote:

Originally Posted by splaz Hmmm... So if you replace the input caps and then use lower value resistors like said in the other topic, then you should also do the other fiddling around you just mentioned or is it fine just without input caps and with lower resistor values ?

I would have thought my findings in this thread would make it abundantly clear that you _cannot_ simply replace the input caps and resistor values and expect everything to work correctly. Modifying the design in some manner is essentially mandatory if you go this route, and amb or tangent's suggestions will work. I will probably perform these modifications on my next build and post pictures if I can (probably next week).

 

[amb]

Quote:

Originally Posted by Teerawit (change the ferrites to be outside of the feedback loop) The ferrite for just the ground channel or for all three channels?

We did it for all three channels, but I think at minimum the ground channel is the one that would most benefit from this.

 

[amb]

Quote:

Originally Posted by Heady 1. make the cut as shown in the picture 2. make the jumper as shown 3. solder R4 and C3 in C3 position 4. solder L1(L), L1(R) and L1(G) in the positions as silk-screened on PCB

Yes.

Quote:

is what is required to mini3-ised the PINT?

The above, plus install 1KΩ resistors in place of the two input coupling caps, change R4 to 6.2KΩ, R3 to 1.2KΩ, and C3 to 10pF.

 

[dsavitsk]

Quote:

Originally Posted by Heady 3. solder R4 and C3 in C3 position

? What did I miss?

edit: never mind, I looked at the schematic

 

[splaz]

I was forgetting you were trying it Filburt.

It's just that I can't recall amb or morsel mentioning anything else apart from this new piece of information. So I thought what they said before about replacing the input caps without this new info on ferrite shifting and putting in C3, would work.

Now bear with me, still a bit rusty on the whole theory side. This is what the schematic would now look like ?

Same goes for the right side.

edit: I know it probably shouldn't be called R0 I just found it easier to copy and flip a nearby 0 than put in a 5. Lazy I know...

 

[amb]

Quote:

Originally Posted by splaz This is what the schematic would now look like ?

Yup.

Quote:

Same goes for the right side and similar for ground.

Except, of course, there is no R0, R2, R3, R4 or C3 at the ground channel.

Edit: I see that you added a ground channel schematic. It's right on.

 

[morsel]

I am so scatterbrained, thank goodness amb remembers these details. We Mini³ified our PINT weeks ago. I wasn't going to mention the pcb surgery, as the ferrites should also work inside the feedback loop, but perhaps not for the ground channel, since it has no feedback resistor to isolate the inverting input from the load.

 

[Filburt]

Well, after three boards and 3 dead 8397s, I certainly wouldn't take the chance again on a non-modified configuration. I did take the opportunity to experiment a bunch with other ideas while I was at it, so I ended up turning the failed ones into a self-teaching thing.