Low noise symmetrical PSU TPS7A4701 and TPS7A3301

Status
Not open for further replies.
Jump to Latest
#44
Hi,

I received PCBs and currently I am testing them. Voltage configuration and under load circuit is working ok, though I still need to make noise measurments to see how it performs.

Along with the PCB I also ordered stencil, so I am thinking to offer PCB with SMD components soldered on, so the end user would only need to add through hole components (filtering capacitor, connectors, diodes, rectifying bridge....). This would also ensure that all PCBs would acuqire almost the same performance. But if there will be many requests for "chip only", then I will also offer them.

Best Regards,
Ales
 
#46

Attachments

  • 20140828_164358.jpg
    20140828_164358.jpg
    234.6 KB · Views: 651
  • 20140828_164439.jpg
    20140828_164439.jpg
    952.5 KB · Views: 633
  • 20140828_164451.jpg
    20140828_164451.jpg
    229.1 KB · Views: 583
#50
I tried today to make some measurments with an old Hammeg 20MHz scope, but the only thing I could get on the screen was flat line, which is good indicator for further measurments, if I'll find someone who has some proper gear. Any volunteers maybe? 🙂

Best Regards,
Ales
 
#56
mravlca said:
With or without heatsink maximum output current is 1A. Heatsink on some other designs is, in my opinion, just cosmetics.

BR,
Ales
Click to expand...

Yes, the specs don´t change. But depending on the used voltage drop and current drawn, maybe a lot of heat has to dissipate. I assume your PCB´s ground plane is fairly large, so this is not a critical point.
But nevertheless, a heatsink always helps to keep temperatures lower in case of large dissipation...

Florian
 
#58
gugi100 said:
The max. current is dependent on the difference between Vin and Vout.

http://www.ti.com/lit/ds/symlink/tps7a4700.pdf

"Dropout Voltage: 307 mV at 1 A"
Click to expand...

That figure gives the minimum headroom necessary for currents of 1A. Assuming a high enough headroom, maximum current is always 1A.

The tps7a4700 has an internal thermal protection, which is triggered at 170°C. Basically, it is just necessary to stay under this limit. This depends on the the voltage drop, the current and the heatsinking. As I said, on Ales´boards, there is a large groundplane doing most of the dissipation work.
You can actually estimate the temperature for a given current by taking into account the heat sink area of the PCB with the formulas given in the datasheet.
 
#59
Bottom side of PCB is completely covered with ground plane. As you can also see I haven't placed any components on the bottom, so heatsink can be applied. Option for heat bridge between PCB and casing is also possible using thermal conductive pads.
 
#60
mravlca said:
Bottom side of PCB is completely covered with ground plane. As you can also see I haven't placed any components on the bottom, so heatsink can be applied. Option for heat bridge between PCB and casing is also possible using thermal conductive pads.
Click to expand...

I did see that, and it certainly wasn´t meant as criticism. I guess for moderate loads the ground plane will be ok. But if you have heavier loads and large dropout voltage, I assume it would be a nice option to glue on a heatsink, to keep the temperature down. Usually, I use (electrically insulating) thermal-epoxy for making permanent heatsink connections. You can make some really wild constructions with this, so I never thougth this would be a problem...

Ales, could you give me the (estimated) ground pad area in cm^2? If I have the time, maybe I can calculate some estimated temperatures for certain load-situations... Just a thought.

Thanks
Florian
 
Status
Not open for further replies.
Top Bottom