How we burned four ST1S10

[Євген Хуторний]

Creative Director - myself

Executive Producer - a friend of mine

Episode 1.

At first we had issues with staring preamp due to invalid powering sequence. However, after we fixed the sequence, preamp was not operational. Despite the enable voltage has been applied to U101 (S1S10) it was not providing 8V on its output.

So we decided to experiment with EN, pulling it to VIN, as the data sheet suggests. It has burned almost immediately with lots of smoke.

Episode 2.

Suspecting that we may have soldering issues, on another board we unmounted R303, U303 and applied 12V on PS and 3.3 on envpa. For some short time U101 has provided approx 3V on output and burned with flame and lots of smoke.

Episode 3.

We purchased another pair of ST1S10 and recreated schematics on a break board. This time we got success and S1S10 gave us 8V on output. So we continued experiment, unmounted parts from the preamp board (U101, C104-C109, R102-R103) and connected the break board output and ground to the preamp board. After powering and enabling ST1S10, it has burned with flame and lots of smoke.

Episode 4.

We cleaned up the break board, disconnected it from preamp and mounted fourth S1S10 to the break board. ST1S10, has burned immediately after powering, again with flame and lots of smoke.

[Samuel Lourenço]

Hi,

I've had a breakthrough. It seems that the ST1S10 has low margin, and it is not fit for this project. It works mostly well on 12 V (better yet, so far I had no mishaps), but on 13,2 V it starts to fail badly. Probably, you will have to lower the voltage to 12 V or a bit less (although at 11.5 V TX starts to misbehave, clicking the relay erratically).

Probably, the culprit of what happened to one of the ST1S10 I had installed was some inductive spikes due to a long supply cable, and that happened when 13.2 V were being supplied to the board. The ST1S10 has a maximum rating of 20 V, so I find hard to believe, but I can guess that it was the case. The other case of a dead ST1S10 was due to a short circuit in one of the MLCC capacitors (the ST1S10 should handle direct shorts, but it does not). This makes me believe that the ST1S10, fake or not, is not fit for the job, because if is not robust. That DC-DC converter should definitely be replaced. I got the ones I used from Mouser.

Really, products from ST Microelectronics are a joke.

Kind regards, Samuel Lourenço

[Євген Хуторний]

Hi Samuel,

Thank you for looking at this.

Our colleague has suggested that we exceeded ST1S10 capabilities on capacitative load.

Unfortunately, we ran out of spare ST1S10 to verify this hypothesis.

Kind Regards,

Євген

субота, 22 червня 2024 р. о 20:32:58 UTC+3 Samuel Lourenço пише:

[Samuel Lourenço]

Hi,

I don't think that's the case. The ST1S10 is seemingly implemented on the Radioberry PreAmp with an extra capacitor on the voltage divider to handle loads, and it seems that the datasheet was taken into account in that regard. I think that the ST1S10 might not be tolerant to inductive spikes. Certainly, an electrolytic capacitor and a more resistive PPTC would mitigate this tendency. MLCCs in general can aggravate such event, because with the supply cable they will create an under-dampened LC response to transients.

Kind regards, Samuel Lourenço