MB1440 circuit diagram

[PHolt.1]

Hi All,

Does anyone have the circuit diagram (schematic) for this board?

Basically (and I have posted about this before, without any conclusion) I have several boards which work perfectly (sw download and SWD debug) via the big 20-way CN2, but are unreliable via the little 10-way CN1 (sw download works but SWD always fails within some tens of seconds or minutes).

Buzzing the thing through with a meter, I can verify CN1 and CN2 are interconnected directly, on all the relevant signals, so this difference is a mystery. But maybe CN1 has some funny stuff on it. I see no useful explanation in the PDF

https://www.st.com/resource/en/user_manual/dm00526767-stlinkv3set-debuggerprogrammer-for-stm8-and-stm32-stmicroelectronics.pdf

It may be that JP4, JP5 and others around there made CN1 different to CN2.

The issue is not affected by whether Cube IDE debug (STLINKV3) is set to 24MHz or 8MHz. SWD is set to 2MHz and reducing that makes no difference to reliability either.

Many thanks in advance.

[Tesla DeLorean]

@Amel NASRI​  @Nawres GHARBI​  @Imen DAHMEN​  can we get some eyes on this

[Tesla DeLorean]

Cross linking for reach

https://community.st.com/s/question/0D53W000005rD0VSAU/stlinkv3mini-mods-schematic

[Imen.D]

Hi @PHolt.1​ , @Community member​ ,

Thanks for pointing out to me this post.

I am checking your request internally and I will come back to you with update.

Imen

[Imen.D]

Hello,

For schematics related question, I've heard back from the team that it has been decided internally not to publish any schematics around V3 family to prevent as much as possible any attempts of illegal copy.

And all information required for the expected use cases of the product should reside in the user manual.

So, if I understand: there is an issue with MB1440 but also another using B-STLINK-ISOL that connected. 

If only focus on the use of MB1440 (no V3 ISOL connected), CN1 (STDC14) and CN2 (JTAG20) provides the same signals, all duplicated as a same wire (for JTAG connection). Then there is no reason observing a different behavior whatever using CN1 or CN2 JTAG connections. Then check attentively the used and connected signals in both CN1 or CN2 configurations.

  

If now consider the addition of the MB1599 (B-STLINK-ISOL) into MB1440 (STLINK-V3 ADAPTER), then the connection cannot be done on the CN2 connector of MB1440 since the supposed purpose is to have the isolation function added.

The connection of the target now should be done on the CN1 connector of MB1599. For this configuration, the jumpers settings must be modified according to section 13.2.2 of UM2448.

This consists in removing the JP1 jumper from the MB1441 and to place it on JP2 header of MB1599. I maybe this is the rootcause.

Imen

[Tesla DeLorean]

This does however significantly diminish the potential for the Maker community to take and expand on the board set/stack

It also really doesn't build confidence in ST's ability to protect its own IP within the STM32

[LChis.1]

That ship has already sailed. Somebody has decompiled the Java programmer, found the f/w AES encryption keys, downloaded his own snooper app to the chip and extracted the bootloader, patched it to not burn the read-lock bits... and there are already V3 clones on blogs or Github.

[PHolt.1]

"then the connection cannot be done on the CN2 connector of MB1440 since the supposed purpose is to have the isolation function added.

The connection of the target now should be done on the CN1 connector of MB1599."

Yes; there is no way around this since the ISOL board covers up all the connectors of the base board :)

I did the JP1 -> JP2 jumper move.

Something strange is still going on. I don't buy the counterfeiting issue because the MB1440 has almost nothing on it and anybody wanting to copy it will work it out in a few hours. I just didn't want to spend a few hours...

I suspect there is something affecting the clock signal. When I get some time I will examine it; I have a 400MHz scope and a 1GHz low-cap probe.

[Tesla DeLorean]

I'm not buying it either.

ST has had plenty of opportunities to just sell pre-programmed STM32F103 or STM32F723 IC's that people could integrate on to boards, like ATMEL did, and end up with clear pin definitions and a reference design. The ability to massively expand the eco-system and reach would seem to outweigh what industrial counterfeiters might do with some superficial hurdle erected in their path.

#MisreadTheTeaLeaves

[PHolt.1]

Yes; also with modern isolators (the RF based ones, not old-style optoisolators) there is no reason for the low speed when the ISOL board is installed. It reduces the speed of STLINK V3 from 24MHz to 8MHz. I did tests and it sort of runs up to about 16MHz but not reliably.

I am now testing the Segger J-Link Edu which, while not isolated, is cheap at some GBP 60, and it runs the debug console at up to 4.5MHz, which is relevant because debugs come out 2x faster than with STLINK V3.

Yes; also with contemporary isolators (the RF primarily based ones, now not vintage-fashion optoisolators) there's no cause for the low velocity when the ISOL board is established. It reduces the rate of STLINK V3 from 24MHz to 8MHz. I did checks and it sort of runs as much as approximately 16MHz but no longer reliably.