I haven't seen them in assembled form elsewhere and SMT of that small size is beyond me.
But, but ... in one of your layouts you had SMD there already.
The TPS7A4700 is much larger than the TPS7A49. The 'DGN' package (HVSSOP with exposed DAP) version of the TPS7A49 is hand-solderable. When you make the footprint for it, just put a reasonably large via in the middle of the thermal pad. Something like 10-15 mil diameter hole if you can fit it without violating the DRCs.
You then solder the IC from the top first. Whet one pad with solder and place the IC such that it lines up properly. Then solder another pin to hold it in place. Once you're satisfied with the alignment, solder the remaining pins. You're guaranteed to have solder bridges. Clean those up with solder wick. Then flip the board over and solder the thermal pad through the via.
Or... Get a solder paste stencil made. Many of the overseas outfits will now make a stencil for something silly like $10 when they make your boards. Solder paste + stencil + toaster oven = DIY heaven. If you go this route you can use the manufacturer recommended footprint for the small parts.
Tom
Demodulated RF shows up as noise behind desired signal. Anecdote: I demoed some shielded speaker cables at a friend's place downtown with lots of stray RF. Compared with his set the things sounded less highs, but putting the default ones back in showed the "missing" high frequencies were all garbage. I suppose that's what audiophiles would call a "black background".
As a point of interest, when Wendy Carlos put her entire New York City studio in a Faraday cage some previously noisy gear quieted right down. So maybe you won't directly hear Earl And Duhg's Zany! Morning Zoo! amidst the Mozart, but it doesn't mean their noxious influence wouldn't barf on the bassoons anyway.
How important is it to resolve the two halves of the charging pulses through decoupling caps?
I've seen discussions of this around.
I can see how to do it. Simple connect the ground pins of opposite decoupling caps together before connecting to anything else.
What I can't get my hear around is how this can be done with a ground plane.
Easy enough for me to connect the opposing caps via a trace. Then connect the centre point of thus trace to ground with a via.However if I were to run a ground plane then the ground pins would be on the same net as the gnd plane and be connected to gnd plane independent of my trace.
The best I have come up with so far is to use signal ground as the gnd plane. Power ground as a pour as wide as possible down the centre line of the opamps. That way it doesn't touch the gnd pins and then I can use the single via connection from the resolved gnd pairs.
I would post a diagram but baby is asleep on me and she won't let me sit down while she does so. Cellphone it is.
I've seen discussions of this around.
I can see how to do it. Simple connect the ground pins of opposite decoupling caps together before connecting to anything else.
What I can't get my hear around is how this can be done with a ground plane.
Easy enough for me to connect the opposing caps via a trace. Then connect the centre point of thus trace to ground with a via.However if I were to run a ground plane then the ground pins would be on the same net as the gnd plane and be connected to gnd plane independent of my trace.
The best I have come up with so far is to use signal ground as the gnd plane. Power ground as a pour as wide as possible down the centre line of the opamps. That way it doesn't touch the gnd pins and then I can use the single via connection from the resolved gnd pairs.
I would post a diagram but baby is asleep on me and she won't let me sit down while she does so. Cellphone it is.
How important is it to resolve the two halves of the charging pulses through decoupling caps. That us between a pair of opposite polarity decoupling caps.
I've seen discussions of this around.
I can see how to do it. Simply connect the ground pins of opposite decoupling caps together before connecting to anything else.
What I can't get my hear around is how this can be done with a ground plane.
Easy enough for me to connect the opposing caps via a trace. Then connect the centre point of this trace to ground with a via. However if I were to run a ground plane then the ground pins would be on the same net as the gnd plane and be connected to gnd plane independent of my trace.
The best I have come up with so far is to use signal ground as the gnd plane. Power ground as a pour as wide as possible down the centre line of the opamps. That way it doesn't touch the gnd pins and then I can use the single via connection from the resolved gnd pairs.
I would post a diagram but baby is asleep on me and she won't let me sit down while she does so. Cellphone it is. Later I will post
I've seen discussions of this around.
I can see how to do it. Simply connect the ground pins of opposite decoupling caps together before connecting to anything else.
What I can't get my hear around is how this can be done with a ground plane.
Easy enough for me to connect the opposing caps via a trace. Then connect the centre point of this trace to ground with a via. However if I were to run a ground plane then the ground pins would be on the same net as the gnd plane and be connected to gnd plane independent of my trace.
The best I have come up with so far is to use signal ground as the gnd plane. Power ground as a pour as wide as possible down the centre line of the opamps. That way it doesn't touch the gnd pins and then I can use the single via connection from the resolved gnd pairs.
I would post a diagram but baby is asleep on me and she won't let me sit down while she does so. Cellphone it is. Later I will post
Or TINA-TI: https://www.ti.com/tool/TINA-TI. I can't stand the user interface in LT Spice. It's a powerful simulator but I prefer something that's also usable.
KiCAD can simulate circuits now too. I haven't played much with it, but I know it's one area of the tool that has seen much improvement lately.
Tom
You won't have charging pulses through the decoupling caps. You'll have charging pulses between the main capacitors in the power supply, before the regulators. Look at the layout on the last page of Bruno P's "The G Word" article linked to some posts back. You can see how the main supply caps have their own ground traces while everybody else is on a ground plane.
Tom
Regarding the tradeoffs between different filter topologies, one of things that has been written about is the effect on noise and distortion.
Microcap 12 is now free to download (and has been since 2020). Much bigger library than Tina-Ti, and includes tubes/valves (although I use Tina-Ti too). My Microcap version is 12.2.0.5
I've just been using it to simulate a RACAL tubed power supply, looking at ripple voltages and currents from the rectifiers, smoothing caps, and currents through the series regulator tubes/valves.
I wanted to avoid charging pulses so I chose a choke-input filter for my WHAMMY, resulting in capacitor currents being 120 Hz sinusoids at much lower amplitude than a capacitor-input filter would have. The PSU section on the PCB wanted some mods, of course, but the thing is quieter than an embarrassed church mouse.
Are files compatible between the different software?
Seems to me that ltspice is the 'industry standard' here on DIYA
Seems to me that ltspice is the 'industry standard' here on DIYA
The answer is no. All simulator packages save schematics and analysis files in their own format. Zero compatibility
Craig.
Craig.
for folks that are new/beginning to doing PCB layout, tomchr gave some good tips here. i'll be a little stronger and say although you might be anxious to get it done, it is always worth it to spend quite a bit of time up front spinning parts around and playing in the rats nest a while with placements before laying any tracks down!