+5V on USBi ( EVAL-ADUSB2Z ) is only 4.3V
When using the USBi board to program the DSP (ADAU1701) I have noticed that the +5V voltage is only 4.31V when connected to a DSP, and 4.47V when left unconnected. My circuit draws approx 60mA.
This gives me a hard time, as my voltage monitoring circuit kicks in at 4.36V and resets the entire circuit.
This also gives me another issue - the +3.3V i create from the 5V, is only +3V, so I don't like to lower the voltage monitoring bar - while I would like to be able to power the system with the USBi as the only connection.
I have looked at the schematic for the USBi and the Q1 transistor has 5.00V on Emitter, 3.66V on the Base and 4.31V on the Collector which goes out to the board.
I tried two different USBi's and they measure the same. Is there a failure in the USBi or why is the +5V not 5V?
Best regards
Thomas
Hello Thomas,
Thanks for bringing this up, I never noticed it. Looking at the USBi schematic, I see that the low output voltage isn't a component failure, it's a design choice. Q1's Darlington configuration pretty much guarantees a 0.7 volt drop at the target board. Apparently this was not judged a problem for several reasons:
USBi-powered evaluation boards have low-dropout 3.3V regulators.
Customers' target boards would not be USBi-powered in actual use.
So what can we do to work around this? Depending upon what matters in your application, you could:
Temporarily set a lower voltage threshold in your battery detector, reverting to the design value in production.
Replace your target board's 3.3V regulator with a low-dropout version.
Replace Q1, the FZT705 Darlington transistor on your USBi with a FZT795A High Gain PNP transistor:
https://www.diodes.com/assets/Datasheets/FZT795A.pdf
This transistor manages a Vce drop of about 0.1 volt at 100 mA with a forced gain of 40. I tried looking for logic-level P-channel MOSFETs which could be an even better choice, but could not find any suitable as a drop-in replacement. At the same time reduce R11 to 470 ohms, to give this non-Darlington transistor enough drive to power your external board. Apparently your project has some ADAU1701 sections powered down. The last time I measured it, a running ADAU1701 uses about 100-110 mA.
Best regards,
Bob
Please log in to reply.