ST1S10 not fit for the job

[Samuel Lourenço]

Hi,

I could handle any other quirks from the Radioberry PreAmp, but this one is so repeatable that it shall not pass without a rant, and after I've wasted my time building the Radioberry PreAmp and replacing many damaged components, I'm rightly disappointed. The components that I didn't replaced due to blown ST1S10 were, curiously, the final transistors (which were not the ones recommended in the BOM, but of type AFT05MS004NT1).

So, the board worked before I had to replace the final transformer, after finding out that I had not used the appropriate wire (used enameled wire after trying to strip, unsuccessfully, the recommended PFTE-coated stranded wire). However, after that replacement, when supplying 13.2 V to the board and activating the PTT, the ST1S10 immediately malfunctioned and locked in permanently, without regulation. This happened, as I later found out, due to a short circuited MLCC capacitor. The ST1S10 should have short circuit protection, but, in reality, it doesn't. The event also damaged the IC that is used to measure the bias current.

After replacing all the damaged components, second try with a new ST1S10 in place. Same 13.2 V in the supply. TXing on 20 meters went well. TXing on 10 meters completely damaged the ST1S10, again. No over-current, no nothing. Just unexplained smoke.

Replaced the damaged ST1S10 with a third one, but decided to lower the voltage. At 11.5 V the relay randomly clicks while on TX. At 12 V, so far, so good. However, I don't think the ST1S10 will not fail in the future, even if I limit the voltage to 12 V and not a mV more.

So, I see several things happening. Apart from the lack or robustness of the ST10S10, which is transversal to many other devices from ST Microelectronics, I see that the field is quite open for inductive spikes due to long supply cables. I guess that the PPTC doesn't offer enough resistance to destroy the under-dampened resonant voltage spike that happens when TXing.

So, I would replace the DC-DC from ST Microelectronics with one from Texas Instruments, as they are usually quite robust to shorts and other events. For example, the TPS54233, as 2 A are quite enough, and the 28V maximum rated voltage can accommodate any spikes. Also, a big electrolytic capacitor at the DC-DC input would go a long way on absorbing inductive spikes. The soft-start feature is also a plus.

Please, don't take my ranting as negative criticism. I see too many people with damaged ST1S10s on their projects. This really needs to change!

Kind regards, Samuel Lourenço

[László Ocskay]

Hello!

I don't have a preamp circuit yet, so I couldn't take measurements. But I don't think the C104 4.7nF in the wiring diagram is right there. If you soldered that out, wouldn't the situation be much better?

[Samuel Lourenço]

Hi,

It seems that C104 was put there so the ST1S10 would handle capacitive loads. I don't know why, but whatever. However, what I would discuss further are the positions of C102 and C103, because they should be near the respective pins they are intended for (i.e. VIN_SW and VIN_A). The power trace should fork before those capacitors, and each capacitor should be dedicated to each pin. Overall, chances are that the ST1S10 is inadequate. I see too many reports of failed chips in the wild. At least one person that toasted four chips, and one user reporting a defect on a video I've done. So, something is not right.

Kind regards, Samuel Lourenço