Help - Revelprog never worked properly for me [solved]

[michael chiklis]

Hello,
my revelprog programmer never worked as it should.
Since i bought it, i always had to read ROMs dozen of times until i can get identical output file because programmer often doesn't read well (i always used Hex Workshop to compare output file).
If you get some difference on few bytes when you read twice a rom, then you know the reading was not good.
If you get same identical read, then very likely reading is good!
Or simply, just watch the CRC32 code on the bottom right of revelprog software, if you get the same code on every reading attempt on a chip, the read is good, otherwise is bad.

Until now i thought the issue was due to dirty or damaged contacts of SOP8 to DIP adapter, but i found out that it isn't the problem because i tried many new of those and i still get error in reading (and of course when writing too).
After many test (and after few years), i understood that the problem was inside the programmer, because in the last days is getting worse
I find out that my Revelprog is faulty too late, i think my warranty is over, so i decided to open it and try to fix it by myself.

wp_20180802_19_14_55_pro.jpg (657.83 KiB) Viewed 8792 times

wp_20180802_19_18_57_pro.jpg (629.61 KiB) Viewed 8792 times

hardware test.png (112.88 KiB) Viewed 8787 times

I did resoldered MCU (IC STM32F103RBT6), the four bus chips and the voltage regulator with my hot air soldering station and the help of Amtech flux, but this didn't help me. I also changed the 8Mhz quartz oscillator, my programmer still behavies the same (few bytes still not readden properly, testing on SPI 25X40AL chip at the moment).

Finally, i discovered that when i use refrigerant spray on MCU, i get nice good reading for a while, this means that MCU is faulty.

I'm considering to change MCU chip, i can do by myself, but i need to program MCU chip via JTAG.
@ArT, please can you help me by sending the firmware file for programming new STM32F103RBT6 mcu ?
Thank you.

[ArT]

Hi Michael,

It seems to me that something is with MCU - have you checked it with lower speed? (settings -> hardware -> flash spi SCK = 10%)
HW test at 5V gives 0.04V difference on buffer - it seems that voltage drop is to high due to high current consumption (in idle).
It may be also something with filtering capacitors - I see that you have one of the earliest hw version and after few years they may have different parameters. You can try to remove capacitor. I marked on photo with yellow (it is few pF and is very crucial for high speed SPI transmission, but in latest versions there is software filtering and it can be removed).
We can check it, and if MCU is faulty we will replace it free of charge.
Is it possible to send programmer to us? It can be with cheapest registered mail.

[michael chiklis]

I did many test.
Yes, i did try all of the 3 speed option, always behavies the same.

Can you explain me which are the filtering capacitor?
I might check them with my capacitor and ESR tester.

We can check it, and if MCU is faulty we will replace it free of charge.
Is it possible to send programmer to us? It can be with cheapest registered mail.

Even if i already did soldering work and with outdated warranty?

[ArT]

I did many test.
Can you explain me which are the filtering capacitor?
I might check them with my capacitor and ESR tester.

The crucial for FLASH SPI is capacitor I marked at the photo above (it is on SCK line), it is few pF (4,7pF - 10pF as I remember) but you can remove it and it should work 100% fine with latest software without this capacitor.

We can check it, and if MCU is faulty we will replace it free of charge.
Is it possible to send programmer to us? It can be with cheapest registered mail.

Even if i already did soldering work and with outdated warranty?

Yes, your satisfaction is important to me
It has to work and every read should be the same.

[michael chiklis]

You can try to remove capacitor I marked on photo with yellow (it is few pF and is very crucial for high speed SPI transmission, but in latest versions there is software filtering and it can be removed).

Ohh sorry, i just see it.
Hmm, i don't think so it's that capacitor the problem, because it reads good when MCU is very cold (when i spray refrigirator can on mcu)

[michael chiklis]

Please, can you explain me where and how to ship it?
Can i send it to you without soldering again the 10 pin strip that connects the 2 boards together?
(i can send you new strip pin if you want).

[ArT]

You can try to remove capacitor I marked on photo with yellow (it is few pF and is very crucial for high speed SPI transmission, but in latest versions there is software filtering and it can be removed).

Ohh sorry, i just see it.
Hmm, i don't think so it's that capacitor the problem, because it reads good when MCU is very cold (when i spray refrigirator can on mcu)

When capacity changed (aging) then SCK high speed switching signal states may be not recognized properly with MCU (when you changing MCU temperature then lo/hi levels threshold are changing too) - I had similar issue before with old unit. You can check it before shipping, but it may be also MCU faulty.

Please, can you explain me where and how to ship it?
Can i send it to you without soldering again the 10 pin strip that connects the 2 boards together?
(i can send you new strip pin if you want).

Yes it can be without 10pin strip, just send 2 boards and enclosure. I'm sending you shipping details on PM.

[michael chiklis]

ehi, you're great.
I removed that capacitor, now it is reading always good
I will do more test tomorrow, if i will get further reading issue then i will consider to send it to you.
Thanks a lot

[ArT]

great! Take your time and check it carefully - you can warm up and freeze MCU for tests. If you will not downgrade to older version then it should work without this capacitor. If you will have any more issues please send it to us.

PS. you do not have to use winhex/hex workshop for file comparison anymore - since v1.6.x there is DiffChecker tool built in REVELPROG. Please check it.