STM32F730 USB OTG_HS description is unclear in the datasheet.

[bas86]

Datasheet DS12536 Rev 1, page 45 says:

-----------------------------------------------------------------

The STM32F730x8 devices feature a UTMI low-pin interface (ULPI) for high-speed operation (480 Mbit/s). When using the USB OTG HS in HS mode, an external PHY device connected to the ULPI is required.

The STM32F730x8 devices feature an integrated PHY HS.

--------------------------------------------------------------------

As you can see, these records do not provide information about which specific devices require external PHY.

Additionally, I've found little bit more clear descritption in the very different part of mentioned document: page 13.

-------------------------------------------------

An USB OTG full-speed and a USB OTG high-speed with full-speed capability (with the ULPI only for the LQFP64 and LQFP100 packages and with the integrated HS PHY for the LQFP144 and UFBGA176 packages) 

-------------------------------------------------

That sounds much more clear. So it looks like records mentioned before shoud be clarified by the following way:

-----------------------------------------------------------------

The STM32F730R8 (LQFP-64) and STM32F730V8 (LQFP-100) devices feature a UTMI low-pin interface (ULPI) for high-speed operation (480 Mbit/s). When using the USB OTG HS in HS mode, an external PHY device connected to the ULPI is required.

The STM32F730Z8 (LQFP-144) and STM32F730I8 (UFBGA176) devices feature an integrated PHY HS.

--------------------------------------------------------------------

Let me know if I do not understand it correctly.

Best regards,

Aleksandr.

[Lubos KOUDELKA · Accepted solution]

Hello Aleksandr,

I can confirm your understanding is correct.

Probably best visibility of this fact is in datasheet, Table 2. STM32F730x8 features and peripheral counts.

Best regards,

Lubos

[waclawek.jan]

Generally, in STM32, the "value line" parts (with part number ending with 0) are "downgraded" parts from "normal" lines - they mostly see less testing, some portions of them are not tested i.e. for practical purposes those portions are "not existing", testing is at limited temperature range, some testing-related constants (e.g. unique serial numbers) are not available, etc.

This is nice as it yields cheaper parts for less demanding applications.

The problem with this approach is, that ST for some weird reason likes to "downgrade" similar but distinct parts but marks them with the same number, distinguishing them only by the suffixes. Here, the smaller packages are "downgraded" 'F722, the larger packages 'F723, hence the difference in whether they contain the internal HS PHY or not.

(A similarly confusing case is the 'F030, which in fact is one of 3 distinct chips, with respectively confusing suffix-dependent differences scattered throughout its datasheet and RM).

JW