DAC8820EVM: DAC8820EVM: U2 overheating oscillations in the REF

Ivan,

Just to confirm, because sometimes unexpected things happen with the EVMs in manufacturing etc., could you send a photo of the jumper configurations for a double-check?

How much current is being drawn from the +14V, -14V, and +5V power supplies?

The +5V DVDD connection I assume is being provided via J5.10 with W3 in the 1-2 position? Is anything supplied on J5.9? Though maybe unused by the DAC depending on the configuration of W3, the 3.3VD rail is still used by the supporting components which drive the digital interface to the DAC8820.

It would also appear that the EVM was designed with a larger capacitive load than the INA105 is rated for, so one suggestion would be to uninstall C7 to see if that removes the oscillation issue you have observed.

U2 getting hot would not add up to me with the default jumper position having W1 shorted in the 1-2 position, therefore effectively leaving U2 with no load in a simple voltage-follower configuration.

Kevin Duke
DAC Applications Manager

In reply to Duke of DACs:

Hi Kevin,

I appreciate your fast and detailed response. To answer your questions.

Just to confirm, because sometimes unexpected things happen with the EVMs in manufacturing etc., could you send a photo of the jumper configurations for a double-check?

Sorry for the bad quality picture.

How much current is being drawn from the +14V, -14V, and +5V power supplies?

(+14V,~120mA)  (-14V,<10mA)  (+5V,<10mA)

The +5V DVDD connection I assume is being provided via J5.10 with W3 in the 1-2 position? Is anything supplied on J5.9? Though maybe unused by the DAC depending on the configuration of W3, the 3.3VD rail is still used by the supporting components which drive the digital interface to the DAC8820.

It's correct, the +5V DVDD is provided via J5.10 with the W3 in the 1-2 position as is showed in the picture. I notice now that 3.3V is the power supply for ICs U6, U7, U8, and U9, the 5V is not re-direct to them, is need it? If I want to drive the DAC, this components should be power up, no?.

It would also appear that the EVM was designed with a larger capacitive load than the INA105 is rated for, so one suggestion would be to uninstall C7 to see if that removes the oscillation issue you have observed.

Removed C7, no improvements. The magnitude of the oscillations, (-VA, 71mV@PkPk)(110mV@PkPk)(REF&Output, 61mV@PkPk) In the REF and output the oscillations have a well define sinusoidal oscillation of 200 kHz, the rest is very random.

U2 getting hot would not add up to me with the default jumper position having W1 shorted in the 1-2 position, therefore effectively leaving U2 with no load in a simple voltage-follower configuration.

I have the same thought so I check for shortcuts and everything appears correctly wire up, as is in the schematics. 

I have to add a couple of things:

1. If I touch with my finger the IC U2 the oscillations decrease dramatically.
2. If I connect an external reference source the oscillations disappear, I did not drive the DAC yet, so I don't know if the EVM is fully functional, but the reference source for the DAC is not oscillating, the power lines are not oscillating and U2 is not reaching high temperatures.

Cheers

Ivan 

 

 

In reply to Ivan Benzaquen:

Hi Ivan,

Thanks for following up.

Ivan Benzaquen

It's correct, the +5V DVDD is provided via J5.10 with the W3 in the 1-2 position as is showed in the picture. I notice now that 3.3V is the power supply for ICs U6, U7, U8, and U9, the 5V is not re-direct to them, is need it? If I want to drive the DAC, this components should be power up, no?.

Correct - without providing the 3.3V supply the parallel interface will effectively be inaccessible, therefore you will not be able to drive the DAC from your host controller.

Ivan Benzaquen

• If I touch with my finger the IC U2 the oscillations decrease dramatically.
• If I connect an external reference source the oscillations disappear, I did not drive the DAC yet, so I don't know if the EVM is fully functional, but the reference source for the DAC is not oscillating, the power lines are not oscillating and U2 is not reaching high temperatures

Maybe one worthwhile test would be simply removing U2 - though I do not currently see any other reason for instability in the design.

When you do the second test with an external reference, I assume J4 is just moved to the 1-2 position with no other modifications to the board?

Kevin Duke
DAC Applications Manager

In reply to Duke of DACs:

Hi Kevin,

I'll remove U2 to do a further test.
When I was applying and external reference the jumper in J4 is in 1-2 position, no other modifications was done to the board.
But in this last test U2 was fully working since my external reference was applied to it. Could be the combiantion of REF102, U1, plus the OPA227, U2?

Cheers,

Ivan

In reply to Ivan Benzaquen:

Hi,

Removed U2 and the reference start to misbehave. U1, REF102, is only giving 4.17V through pin 6, no oscillations but an incorrect voltage reference for the DAC. Jumpers are set in the default position.

If I resolder U2, the U1 still giving an incorrect voltage in pin 6, the U2 IC is stable, no heating, no oscillations in the supply lines, references or output.

Since bypassing the reference, U1, by applying an external reference kill oscillations and temperature in IC U2 is normal, my conclusion is that U1, REF102 was/is the cause of the problems.

Cheers,

Ivan

In reply to Ivan Benzaquen:

Hi,

I've solder a new REF102, the reference part is working as expecteted. No oscillations and U2 with a normal temperature.

This problem was resolved, but I have anothe problem with this board, don't know if open an new discussion or continue here:
The output of the DAC is the higest level, 10V. I put RST pin low, that this should put the output of the DAC to zero voltage but is still high.
Any clue?

Cheers,

Ivan

In reply to Ivan Benzaquen:

Ivan,

Thanks for the update. It is interesting that replacing the component resolved the issue. I've certainly see strange things caused by flux contamination or cold solder joints in the past but it is not something I have seen very often from our EVMs in production as they are functionally tested before being stocked and ultimately delivered. In any case I appreciate you closing the loop on that side.

Concerning the RST pin, how exactly are you issuing the RST? Can you confirm that at the device pin the RST pin is low?

Kevin Duke
DAC Applications Manager

In reply to Duke of DACs:

Hi Kevin,

As well I do not understand what was the problem. I resolder the older REF102 and everything works fine again. You must be right cold solder or flux contamination.

I'm addressing the RST through the W6 connector. I've removed the jumper and connect one of the pins to GND with a cable, the pin goes directly to the RST pin of the DAC, and I've checked that the RST is low.

Cheers,

Ivan

In reply to Ivan Benzaquen:

Hi Ivan,

Sorry about the late reply - I was out of office this past week.

Does the device otherwise respond as expected to input codes? I.e. what is the output at both full-scale and zero-scale codes?

Kevin Duke
DAC Applications Manager

In reply to Duke of DACs:

Hi Kevin,

I didn't try but I'll do soon.

Since the ICs that are power by the 3.3V line are 5V tolerant, could I just power with 5V and connect the 3.3V line to the 5V?

Cheers,

Ivan