Other Parts Discussed in Thread: TLVH431
Hi,
I would like to ask for your support regarding the TLVH431AQDBZR behavior in one of our customer's systems.
Our customer is using TLVH431AQDBZR in the feedback circuit of a 3.3 V isolated flyback converter. On some units, an output oscillation is observed.
We evaluated several devices from different lots and measured their small-signal gain and phase characteristics. From this analysis (please see the attached measurement plots), we observed the following:
-
Devices taken from units showing output oscillation ("NG devices") have lower loop gain compared to devices taken from units without oscillation ("OK devices").
-
The NG devices show less phase lag above roughly 100 kHz compared with the OK devices.
The device lots we have confirmed so far are:
-
NG devices: (31T) LOT: 5262916NFM
-
OK devices: (31T) LOT: 3404019HNA
In the actual application (3.3 V isolated flyback), the feedback network around TLVH431 has approximately the following conditions, under which we confirmed that the NG devices show smaller phase margin and gain margin, resulting in oscillation:
-
Cathode current is relatively low, around 0.4 mA.
-
A series RC network of 0.47 uF and 470 ohm is connected across the anode-cathode as a compensation network.
Based on the above, could you please help us with the following questions?
-
From the attached gain and phase measurements, can both the NG and OK devices still be considered within normal specification, or do the NG devices show a deviation large enough to suspect abnormal behavior?
-
What are the most likely reasons for the observed differences in gain and phase (especially the lower gain and different phase behavior above 100 kHz)? For example, could this be explained by normal process variation, internal compensation tolerance, or sensitivity to low cathode current?
If needed, we can also share the full application schematic by private message.
Thanks,
Conor
