DAC8734: EVM - DAC A-GROUP output negatively biased. User ZERO Calibration won't help

Stefano Alfredo La Spina

Part Number: DAC8734

Hi forum. For the prototype of my project I am using the DAC8734-EVM board due to electronic part shortage. Apart from that, the board is awesome since it is useful for hands-on approach. I have rewritten some lines of code to steer the board from an 8-bit MCU. Everything seemed to work flawless since after a while I noticed some strange behavior:

With all power-on sequence correct (LOGIC-> NEGATIVE ANALOG RAIL -> POSITIVE ANALOG RAIL ) and power supply in the specs (+-15V from a low ripple PSU) and for the bipolar use (GAIN 4x ON-BOARD 5V REFERENCE) in order to have -10V to +10V output, I have noticed that GROUP-A DAC outputs are negatively biased.

Sending DAC0 0x0000 command shows me -230mV on output 0 and -220mV on output1 of Group-A Here the screenshot from a digital acquisition multimeter:

Same things won't happen on Group-B both outputs:

with a user zero calibration register 2LSB I have some uV span:

Now the fact is that whole range is shifted. Maximum GROUP-A positive value is 9.3V (and not 9.9V) while negative is more closer to minimum specs (9.8V)

The strange thing is also that I am getting near 780LSB the zero of the group but even with that I do not get full zero as for group-B:

sending DAC0 783 gives me:

while sending DAC0 782 gives me still negative:

as from the datasheet 1LSB should have a value of: 305uV and here gives me more than 600uV span.

Is there a way to reset to factory the trimmed value since USER CALIBRATION of maximum 32LSB is useless here.

Best regards

over 3 years ago

0 Paul_Frost over 3 years ago

TI__Guru 53773 points

Hi Stefano,

Can you share your Jumper configuration for the EVM? I suspect you have some kind of misconfiguration or some parasitic resistance that is skewing the output Gain. What are us for JP6/7?

0 Stefano Alfredo La Spina over 3 years ago in reply to Paul_Frost

Prodigy 240 points

Hi Paul. Thank You so much. Here is the photo of the board.

Best regards

0 Paul_Frost over 3 years ago in reply to Stefano Alfredo La Spina

TI__Guru 53773 points

Hi Stefano,

Your jumpers seem correct. Can you measure the reference pins? What is the uncalibrated DAC output voltage at negative full-scale and positive full-scale?

0 Stefano Alfredo La Spina over 3 years ago in reply to Paul_Frost

Prodigy 240 points

Hi Paul. Reference Voltage A and B are both +4.9997 V

Maximum negative full-scale on output A : +9.334V

Minimum negative full-scale on output A : -9.8368V

Maximum negative full-scale on output B : +9.987V

Minimum negative full-scale on output B : -9.969V

0 Paul_Frost over 3 years ago in reply to Stefano Alfredo La Spina

TI__Guru 53773 points

Do both channels in group A have the same values?

Can you inspect near the VOUTx and RFB pins for any debris, like solder or flux? If you have re-worked the device on this board, I suggest a thorough cleaning of the DAC area to ensure there is no parasitic current between RFB1 and RFB2.

0 Stefano Alfredo La Spina over 3 years ago in reply to Paul_Frost

Prodigy 240 points

No rework, no debris. Rather than thinking about it mabye at the very beginning I have messed a little bit with the analog power supply. No inversion, no voltage overload. Could it be damaged inside? It is a very strange behaviour. Seems it is negatively biased.

+1 Paul_Frost over 3 years ago in reply to Stefano Alfredo La Spina

TI__Guru 53773 points

What concerns me is that their is gain values are wrong as well. That usually indicates extreme reference error.

Do both channels in group A have the same values?

Do you have a narrow probe that you can use to measure the VREFA at the pin itself?

Please shoot me an email at frost@ti.com. Maybe we can just send a replacement.

0 Stefano Alfredo La Spina over 3 years ago in reply to Paul_Frost

Prodigy 240 points

Thank You Paul.

Here the values measured:

Dac0 0x0000 = -0.2285 V
Dac0 0x7FFF = +9.393 V
Dac0 0x8000 = -9.851 V
Dac0 0xFFFF = -0.2288 V

Dac1 0x0000 = -0.2286 V
Dac1 0x7FFF = +9.394 V
Dac1 0x8000 = -9.851 V
Dac1 0xFFFF = -0.2290 V

Dac2 0x0000 = -0.01 V
Dac2 0x7FFF = +9.963 V
Dac2 0x8000 = -9.963 V
Dac2 0xFFFF = -0.28 V

Dac3 0x0000 = -0.48 V
Dac3 0x7FFF = +9.967 V
Dac3 0x8000 = -9.965 V
Dac3 0xFFFF = +0.23 V

When You say :

"Do you have a narrow probe that you can use to measure the VREFA at the pin itself?"

You mean PIN of the DAC IC?

If yes, it shows 4,9995V at VREFA as at the TP2

Thank You.

0 Paul_Frost over 3 years ago in reply to Stefano Alfredo La Spina

TI__Guru 53773 points

Can you try disabling the Reference by removing the jumpers on JP7 and JP6, then power up the board, then re-apply the reference voltage by inserting JP6&7?

0 Stefano Alfredo La Spina over 3 years ago in reply to Paul_Frost

Prodigy 240 points

Paul. Nothing happens. Have tried it several times.

0 Stefano Alfredo La Spina over 3 years ago in reply to Paul_Frost

Prodigy 240 points

Hi Paul. Free replacement board has arrived today. I have not yet measured it but I would like to thank officially You and all TI for the kindness and the great customer support. Thanks again. Have a nice day. All my best. Stefano

Data converters

Data converters forum

DAC8734: EVM - DAC A-GROUP output negatively biased. User ZERO Calibration won't help