RH850/F1L

Hello,

Do you get a E4000003 error ?

If so, please check for possible causes on this page:

www.renesas.com/.../rfp-error-guide

Is this Renesas hardware (e.g. starter-kit or piggy-board) or custom built hardware?

RFP uses the serial programming mode of the device.  Have you tried connecting with a debugger (e.g. CS+)?

A screen-shot of the RFP error would be helpful.

exactly.

tks for reply, i checked that guide already for that error

hi, it's automotive production board, i can't disclose lot of inside info.

but as you see in the waveform, for me it looks fine. any of the wires except vn or gnd removed, the clock question doesn't appear. 

flmdo, tdi, tdo seems ok regarding levels. mcu starts in serial programming mode at release fron reset after powering on according to this table, see my scope pls

so for me, it's in programming mode, asks for freq so some handshaking on tdi/tdo is happening, what pin is not connected or what else can interfere? on hw side, i'm experienced, as i said before wires are double checked.

i hope somebody knows why the timeout. finally, if i disconnect tdi and/or tdo, no frequency input form appears anymore, so mcu is running. btw, i can save eep/flash for the 2 boards using some factory tool with mcu started in app mode, i have validated data, i can play with the eep and values stored in it and it behaves accordingly.

my big question was if some security newer approach can do this or i'm just skipping something.

imho, mcu is started in serial prog mode, i expect to refuse reflash/dump or to ask a code or something, not timeout. my previous work was on older security approach on another boards, i never had issues to connect to renesas mcu, same E1 programmers (as i said, i got 2 E1 and 2 cables for this timeout-board)

to answer you, i was busy and tried only 2-wire serial prog, not anything more, my job is to reflash quickly in case of problem and test, i don't do debug (job description :( )

the screenshots are lost, but it's the standard 

E4000003 A timeout error occurred.

br, ovidiu

Maybe this is a dumb question but are you using F1L for a new design?  F1L has been "obsoleted" by F1Kx devices, and is not recommended for new designs..

The reason I asked about using CS+ was to find out if the device would boot into normal operating mode, and if a debugger could connect and gain control of it.  I believe this would be a worthwhile exercise for you to try, since it only involves installing the CS+ software and giving it a try in debug-only mode.

Looking closer at your scope data, there seems to be a very periodic reset being asserted, and while I might guess that is RFP retrying the connection, I only see a couple of instances where there is data on the xmt/rcv lines.  Is there an external reset circuit in this design?

Are you handling all unused pins according to the recommendations?  Are all power supply connections properly made?  Does the design conform the hardware design guide ()?

The frequency setting is critical for the UART communication as this is synchronous.  If you used the CSI / async communication this would not be required.  RFP only supports UART for the F1x devices unfortunately, but the PG-FP6 does support CSI.

hi, if i remember right they started in 2016 and newer version is like you say the F1K. it's automotive, we work in some projects with 9628 modems if you get what i mean. and it's fine imho, if old stuff make the job and it's stable, you don't change it after negotiating at least 5-year prod plans with renesas qualcomm etcaetera.

to put it shortly, in another needle adapter setup 1800 km away it works, mine is not. the guy operating that is not at work, that goes to my issue here, normally he should share schematic and insides, i see a switch on his needle adapter schematic, and it it's tied to supply, it is a signal not some power rail. the photo is somehow low resolution in a pdf, so i can only guess number of signals. I counted 11 wires, a switch and 5vDC supply connection. Me, I got 9 and no switch, starting from Renesas datasheets, all say the same thing more or less on 2-wire connection. BTW, the switch i see in his setup apparently goes to E1 !!!

that periodic reset, for me is E1 retry to get answer from mcu.

reset is classic 4k7/1nF being Y with some another IC, maybe you hit the spot.

I did not shared all this before cause I need to put photos and I won't because they are internal company property, schematic also.

So, that needle setup they have either has also debug (i need only reflash in disaster-recovery scenario) or something is escaping me.

Finally, I have wiring of production thing and they use TCK FLMD0 TDO RESET TDI (like me), power lines AND CAN line :) but Unicom flasher. So flashing should be sufficient with this minimal wiring, cause they do it for million pieces. CAN should be for 2 purposes, security unlock and various commands in production flow. but i don't expect jtag to discuss security on CAN line while in 2-wire serial prog mode, right???

i really appreciate the help, but if it isn't related to wiring scenario i use and the mistake is not clear, i don't want to waste time for anybody and close the topic. i will clarify with my prod guys (i avoid this for multiple reasons) or with the dev team (just broke in multiple pieces, it's called capitalism).

my conclusion: so i really appreciate your time, if you can clarify

  -  if i have some reset/wiring error (i doubt i messed up 2 wirings i have 3 decades on this game, but not renesas, only from 2 years i got this on desk) 

  -  OR security can generate this timeout error?

for me, rfp is not verbose and not clear like some other sw. no engineer working with renesas in my lab (i asked at least 4 of them working with different renesas mcu's) was able to respond this simple question. one of them told me only if jtag is locked, he gets a window asking for jtag pass. i repeat, in theory my jtag is unlocked with some successful end-of-line tool, i can't post snips of this. anyway, locked or not, why timeout message from rfp? shouldn't be something like device locked bla bla bla?

like in any IT branch, you didn't put the code or it's wrong, you got a message regarding this, not timeout tales...what do you expect, i'm mostly hw dev, for me sw gets madder day by day, i had to debug some crowdstrike rf testing station essential for what one colleague does, i'm tired of sw devs lack of competence, quality goes to bin those days

maybe i'll start to generate timeouts at job mee too :)

I do not believe the E4000003 error msg would be displayed for a secured device with the ID code.  Looking at the document Renesas Flash Programmer V3.16 Flash memory programming software User's Manual I would expect error E0004002:

It might be possible that the serial programming interface was disabled by a security setting.  Debugging would still be possible if the option byte enabled the function, and the ID code was known if set.  Are these not virgin devices as received from the factory?  If not, it would be difficult to guess what someone may have done.

You might try isolating the IC on the reset line, lift a component to break that connection.

This is a long-shot suggestion, but you could also try the last v2 RFP as this supported the F1L device.

thanks, i missed that one.

devices are programmed and secured (jtag locked with static code), then unlocked by a script hidden inside some end-of-line tool. the script says successful when i unlock, it may be correct...

so finally my answer is there, i should get E0004002. one day i'll come back with 

E4000003 snip and follow-up (too occupied now on another project), br ovidiu

i tried 2.05.03 you suggested:

now error became E1017023, i'm lost. powered from E1 or from the board 5V regulator, same message.

Different error code but same error msg (timeout), from the common U/M in the installation's "doc" folder: