Electronics > Projects, Designs, and Technical Stuff
The TLV760xx regulators are a nice piece [s]of crap[/s]
bloguetronica:
--- Quote from: floobydust on February 10, 2019, 04:59:17 am ---250mA and 500msec/div :palm:
--- End quote ---
Unfortunately, they haven't made a plot representing the capacity of this regulator to output the input voltage when failing. This regulator family is expunged from my current and future designs. I ven deleted the Eagle libraries that I had created previously. It would be a crime not to do so.
Kind regards, Samuel Lourenço
tszaboo:
--- Quote from: bloguetronica on February 10, 2019, 02:54:09 am ---
--- Quote from: NANDBlog on February 09, 2019, 11:13:34 pm ---Just drop in a 470 Ohm resistor on series with the input pin.
But why would you not drive the LED directly from the 24V in the first place?
--- End quote ---
If you need to do that, then you know you have a crappy regulator. Anyway, a resistor set as described won't save your regulator in a no-load condition. It is best to use a better, more robust regulator.
Kind regards, Samuel Lourenço
--- End quote ---
A) You are clearly overloading the regulator. "Continuously running the TLV760 into thermal
protection degrades device reliability."
B) Yes, it will save it, because it will dissipate the heat instead the TINY LDO that you are using with higher loads, provide short circuit protection, and reduce the turn on transient.
C) If shorting the regulator output blows it up... Then dont short it?! Why is this even discussed? It is not standard operation to short the output, a SOT23 LDO is not a CC CV power supply.
iMo:
The TIs datasheet calculates (8.1.3) the max power with RThJA=275degC/W, however.
With 24V input, and 3.3V/16mA output it gives you 91degC above ambient.
When the output is shorted the current could be 60mA (see the Fig.12 for 10V input), thus close to 400degC above ambient..
bloguetronica:
--- Quote from: NANDBlog on February 10, 2019, 09:23:46 am ---
--- Quote from: bloguetronica on February 10, 2019, 02:54:09 am ---
--- Quote from: NANDBlog on February 09, 2019, 11:13:34 pm ---Just drop in a 470 Ohm resistor on series with the input pin.
But why would you not drive the LED directly from the 24V in the first place?
--- End quote ---
If you need to do that, then you know you have a crappy regulator. Anyway, a resistor set as described won't save your regulator in a no-load condition. It is best to use a better, more robust regulator.
Kind regards, Samuel Lourenço
--- End quote ---
A) You are clearly overloading the regulator. "Continuously running the TLV760 into thermal
protection degrades device reliability."
B) Yes, it will save it, because it will dissipate the heat instead the TINY LDO that you are using with higher loads, provide short circuit protection, and reduce the turn on transient.
C) If shorting the regulator output blows it up... Then dont short it?! Why is this even discussed? It is not standard operation to short the output, a SOT23 LDO is not a CC CV power supply.
--- End quote ---
A - It is possible.
B - No it doesn't protect the LDO from surges which, in any case, didn't happened when the AD5551s got fried.
C - This was done to assess the reliability of the regulator.
--- Quote from: imo on February 10, 2019, 12:12:39 pm ---The TIs datasheet calculates (8.1.3) the max power with RThJA=275degC/W, however.
With 24V input, and 3.3V/16mA output it gives you 91degC above ambient.
When the output is shorted the current could be 60mA (see the Fig.12 for 10V input), thus close to 400degC above ambient..
--- End quote ---
Since the regulator is not floating in mid-air, you should use the junction to board figure of 56.8ºC. You then have 0.3312W of dissipated head, giving you a rise of 18.84ºC above ambient.
Anyway, the schematic is attached (the TLV76050 was replaced by the TPS70955, and the respective capacitors as well). You can see that there is no reason why this board failed on power-up.
Kind regards, Samuel Lourenço
bloguetronica:
Well, it seems that the TPS70933 is even worse. It requires a 1K resistor in series with the input. Direct to 24V it blows instantly. Ok, back to the TLV760. I wonder if a choke could absorb those spikes. Or perhaps a 220 \$\Omega\$ resistor.
Kind regards, Samuel Lourenço
Navigation
[0] Message Index
[#] Next page
[*] Previous page
Go to full version