Part Number: UCC28C43-Q1
Other Parts Discussed in Thread: TIDA-01505, , UCC28C43
Hello,
we are debugging a board based on the TIDA-01505 reference design, using only the flyback with optocoupler version.
The converter is expected to start from low input voltage, around 40 VDC, as described in the TIDA-01505 design guide. The UCC28C43-Q1 should start when VDD exceeds the UVLO threshold, around 8.4 V.
At the moment we observe the following behavior.
With the original configuration, applying VIN_HV = 40–60 VDC, the voltage on C54 / UCC28C43 VDD does not rise correctly. It remains around 1.8–2.7 V, so the controller does not start.
We verified that the UCC28C43-Q1 itself is functional:
- by externally supplying C54/VDD with 10 V, the controller starts correctly;
- VREF = 5 V is present;
- COMP ≈ 6.3 V;
- CS = 0 V;
- OUT pin switches correctly
Therefore, the UCC28C43 and its oscillator/control section appear to be working.
We then checked the HV startup circuit:
- Q9 was checked and appears correctly mounted and functional;
- Q11 does not appear to be forcing Q9 off;
- removing D35 did not change the issue;
- the HV startup resistor chain is not open. For example, with VIN_HV = 60 V we measured approximately:
- high side of R48: 21.5 V
- low side of R48 / high side of R49: 11.4–11.5 V
- low side of R49: 3 V
- Q9 pins around 3 V / 2.6 V / 1.85 V
- C54 around 1.85 V
The important observation is related to the level-shifter branch:
- If R59 is removed, C54/VDD rises to approximately 10 V with VIN_HV = 60 V.
- If Q19 is removed and R87 is changed to 10 kohm,
- C54/VDD can rise to about 9 V even from VIN_HV = 40 V.
- However, when Q19 is mounted, C54/VDD collapses again to approximately 2.7 V.
- Q19 was replaced with a new component, but the behavior did not change.
- Removing R96 allows C54/VDD to rise to about 9 V, but without R96 the SiC gate is no longer driven, so R96 appears necessary for Q19 / level-shifter operation.
- Changing R96 to 10 kohm still causes Q19 to become polarized and the startup remains blocked.
- R87 was also tested as a stronger pull-up, for example 3 kohm, but the issue remained.
- Measured VGS of Q19 is around −0.75 V in the problematic condition
So the current conclusion is that the Q19/R96/R87/R59 level-shifter branch loads or clamps the VDD/startup node during startup, preventing C54 from reaching the UCC28C43 UVLO turn-on threshold.
Could you please clarify the intended startup behavior of this branch in the optocoupler version of TIDA-01505?
In particular:
- Are R58/R59 intended to be mounted as 0 ohm links in the default optocoupler configuration, or are they only for a specific level-shifter supply option?
- Is the level shifter supply expected to be pre-biased through J14 before applying the main HV input?
- What are the correct default values / population options for R87, R96, R58, R59, Q19, and the related level-shifter startup components?
- Is there any known correction or modification to the TIDA-01505 schematic/BOM for reliable startup from 40 V using the optocoupler version?
- Based on the measurements above, which component or population option should be changed to prevent Q19 from pulling down the VDD/startup node during startup?
Thank you.
