So, I could do for the help and advice of some of the wise wizards I see lurking around this site. In brief, I'm new to the diy audio scene, very good knowledge of sound engineering etc etc, but the electronics side, not so much :S
I've tried making a unbalanced to balanced board, with the 1646 chip, following an online diagram that pretty much follows the datasheet provided anyway. 15V+/- regulated on both rails.
At first, all was well, signal out was fine, but then I noticed the chip running very hot so removed power, checked voltages etc, all fine. Anyway, I temporarily lifted the ground on the power (not actually tied to ground, but just my 0V reference point) and tried again. Chip runs cold.... reinstate ground, chip runs hot.....?? Either way its died of death now anyway.
Any help would be greatly appreciated
I've tried making a unbalanced to balanced board, with the 1646 chip, following an online diagram that pretty much follows the datasheet provided anyway. 15V+/- regulated on both rails.
At first, all was well, signal out was fine, but then I noticed the chip running very hot so removed power, checked voltages etc, all fine. Anyway, I temporarily lifted the ground on the power (not actually tied to ground, but just my 0V reference point) and tried again. Chip runs cold.... reinstate ground, chip runs hot.....?? Either way its died of death now anyway.
Any help would be greatly appreciated
OK, I've just looked at the data sheet and see that the inductors are ferrite beads... hmmm... I'd guess that if you have the recommended inductance there then you could be looking at problems with physical layout. High speed high current output stages like this are going to demand the utmost care in implementation.
I would suggest adding some series resistance directly at the output pins as a test to see if that fixes the issue.
I would suggest adding some series resistance directly at the output pins as a test to see if that fixes the issue.
Posted together
They are bit expensive to experiment with. Try replacing the beads (is it a link and bead ?) with some low value resistors, say 33 ohms to begin with and see if the circuit runs cool.
Hard to say but layout really is everything on high speed circuitry like this. Ideally you should check with a scope to see if anything untoward is happening (oscillation).
Try the resistors, it will at least help determine if there is a stability issue. Also make sure the chip supplies are correctly decoupled.
Try the resistors, it will at least help determine if there is a stability issue. Also make sure the chip supplies are correctly decoupled.
So I think I've figured out where I went wrong. I connected the output ground to supply ground, which also has the input ground on it. Input ground is fine being on the supply, but as you said, I think I was getting a hell of a lot of positive feedback coming through and causing the chip to oscillate.
Nevermind, lesson learnt! Thanks again
Nevermind, lesson learnt! Thanks again
I'm playing with these and they work fine connected directly to mixer RCA out or Laptop. Here the schematic.
https://cdn.sparkfun.com/datasheets/BreakoutBoards/THAT_1646_Breakout.pdf
I think they are somewhat noisy without a buffer. Now I will adding a 2x or 4x buffer to increase gain.
https://cdn.sparkfun.com/datasheets/BreakoutBoards/THAT_1646_Breakout.pdf
I think they are somewhat noisy without a buffer. Now I will adding a 2x or 4x buffer to increase gain.
Pleased you seem to have figured it out
When experimenting with something like this it might help to include a couple of 47 or 68 ohm resistors in each supply. The chip only draws a around 6ma quiescent and so it can be a good guide that things are OK if you see the expected volt drop (around 0.5v) across each resistor. Anything significantly more indicates a problem but the bonus is the resistors limit current and save the chip.
When experimenting with something like this it might help to include a couple of 47 or 68 ohm resistors in each supply. The chip only draws a around 6ma quiescent and so it can be a good guide that things are OK if you see the expected volt drop (around 0.5v) across each resistor. Anything significantly more indicates a problem but the bonus is the resistors limit current and save the chip.
The that 1646 datasheet shows the equivalent circuit in Figure 2.
It shows that the input resistors are already inside the chip package.
It also shows that the output resistors are inside the chip package.
Figures 4 & 5 show with and without common mode reduction.
Figure 8 shows the output network that gives RF attenuation. NOTE !! that the italic "E" symbol is the chassis. Pin1, C3 and C8 all go direct to the enclosure, not to audio, nor power ground.
It shows that the input resistors are already inside the chip package.
It also shows that the output resistors are inside the chip package.
Figures 4 & 5 show with and without common mode reduction.
Figure 8 shows the output network that gives RF attenuation. NOTE !! that the italic "E" symbol is the chassis. Pin1, C3 and C8 all go direct to the enclosure, not to audio, nor power ground.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.