| Author |
 Topic  |
|
|
noStuff
12 Posts |
 Posted - 12/20/2019 : 05:33:29
|
I'm trying to read an soldered W25Q16BV from a board ( so far it seems happening with all W25X16 series I'm trying )
The default line in devices.txt partially work already but the content in the "BUFFER" view is mostly incorrect by a good percentage.
I added the following line in devices.txt
Name="W25Q16BV",ID="EF4015",Class="W25X10",Category="SERIAL_EPROM",MFG="Winbond",Vcc="3.6V",CodeSize="2097152",Platform="GQ-4x4";
I'm pinning the pomona clip cable directly on the ZIF GQ-4x4 socket like in the attached photo but the problem still.
Sometime by reloading the "device" more or less data can be read.
Is there a particular setting to read and write the 1:1 content out the EF4015 with the GQ-4x4 hardware programmer by SPI?
Image Insert:
144494 bytes |
|
| Reply #1
anniel
2572 Posts |
Posted - 12/20/2019 : 06:34:20
|
quote:
Originally posted by noStuff
I'm trying to read an soldered W25Q16BV from a board ( so far it seems happening with all W25X16 series I'm trying )
The default line in devices.txt partially work already but the content in the "BUFFER" view is mostly incorrect by a good percentage.
I added the following line in devices.txt
Name="W25Q16BV",ID="EF4015",Class="W25X10",Category="SERIAL_EPROM",MFG="Winbond",Vcc="3.6V",CodeSize="2097152",Platform="GQ-4x4";
I'm pinning the pomona clip cable directly on the ZIF GQ-4x4 socket like in the attached photo but the problem still.
Sometime by reloading the "device" more or less data can be read.
Is there a particular setting to read and write the 1:1 content out the EF4015 with the GQ-4x4 hardware programmer by SPI?
Image Insert:
144494 bytes
EF4015? |
 |
|
| Reply #2
noStuff
12 Posts |
Posted - 12/20/2019 : 08:56:22
|
quote:
Originally posted by anniel
EF4015?
Yes, ID is EF4015.
W25Q16BV
Download Attachment: 
|
|
|
| Reply #3
anniel
2572 Posts |
Posted - 12/20/2019 : 09:53:43
|
quote:
Originally posted by noStuff
quote:
Originally posted by anniel
EF4015?
Yes, ID is EF4015.
W25Q16BV
Download Attachment:
Ok, I didn't follow you there for a moment. Your PCB is powered correctly? VCC is isolated? |
|
|
| Reply #4
noStuff
12 Posts |
Posted - 12/20/2019 : 10:42:42
|
quote:
Originally posted by anniel
Ok, I didn't follow you there for a moment. Your PCB is powered correctly? VCC is isolated?
Board is without power cable, battery or anything.
The GQ-4x4 is not powering on other 3v circuit ( no LED from button and stuff ) so, it's seems VCC is on it own and that's a good sing of possible GQ-4x4 capability.
Dunno how the GQ-4x4 is actually reading the content but it seems like the pulse is too long or too wide and by the time the data is actually aggregated on the buffer some already vanished, or something like that.
Anyway, it doesn't correctly read the full content, more the data more the losses.
I've also tried changing USB cable or adding external power to the programmer - nothing changed.
Pretty much 20% of the segment is always wrong. |
Edited by - noStuff on 12/20/2019 10:43:39 |
|
|
| Reply #5
anniel
2572 Posts |
Posted - 12/20/2019 : 11:11:51
|
quote:
Originally posted by noStuff
quote:
Originally posted by anniel
Ok, I didn't follow you there for a moment. Your PCB is powered correctly? VCC is isolated?
Board is without power cable, battery or anything.
The GQ-4x4 is not powering on other 3v circuit ( no LED from button and stuff ) so, it's seems VCC is on it own and that's a good sing of possible GQ-4x4 capability.
Dunno how the GQ-4x4 is actually reading the content but it seems like the pulse is too long or too wide and by the time the data is actually aggregated on the buffer some already vanished, or something like that.
Anyway, it doesn't correctly read the full content, more the data more the losses.
I've also tried changing USB cable or adding external power to the programmer - nothing changed.
Pretty much 20% of the segment is always wrong.
Power up the target. |
|
|
| Reply #6
noStuff
12 Posts |
Posted - 12/20/2019 : 11:20:41
|
quote:
Originally posted by anniel
Power up the target.
Powering up the target board (power on system off) with or without programmer VCC make the W25Q16BV unrecognized.
A boot and a forced power off is required to get it recognized again as EF4015. |
Edited by - noStuff on 12/20/2019 11:34:09 |
|
|
| Reply #7
anniel
2572 Posts |
Posted - 12/20/2019 : 14:32:13
|
quote:
Originally posted by noStuff
quote:
Originally posted by anniel
Power up the target.
Powering up the target board (power on system off) with or without programmer VCC make the W25Q16BV unrecognized.
A boot and a forced power off is required to get it recognized again as EF4015.
How long are your wires to the chip? |
|
|
| Reply #8
noStuff
12 Posts |
Posted - 12/20/2019 : 14:46:42
|
quote:
Originally posted by anniel
How long are your wires to the chip?
I'm using standard jumper wire 2point cable ( 20cm / 8inch ) on the Pomona 5250 pin |
Edited by - noStuff on 12/20/2019 14:49:46 |
|
|
| Reply #9
noStuff
12 Posts |
Posted - 12/21/2019 : 00:12:42
|
Supposedly all of these ID might affected ( taken from the bios file descriptor table ) on the board.
C22015 (Macronix MX25L16)
1F4600 (Atmel AT26DF161)
EF3015 (Winbond W25X16)
898911 (Intel 25F160S33B8)
BF2541 (Microchip SST25VF016B)
EF4015 (Winbond W25Q16)
EF4015 (Winbond W25Q16) and EF3015 (Winbond W25X16) surely are.
It seems SPI operation are not fully supported while in circuit.
To recap the situation is as follow;
ATX PSU power connector attached without the power cable NO battery > ID and data with percentage error
ATX PSU power connector attached with power cable ( power ON system OFF ) NO battery > no ID
NO ATX PSU power connector NO battery > no ID
NO ATX PSU power connector with battery > ID clock switch EF4015 <> FFFFFF |
Edited by - noStuff on 12/21/2019 06:54:39 |
|
|
| Reply #10
anniel
2572 Posts |
Posted - 12/21/2019 : 06:51:29
|
quote:
Originally posted by noStuff
quote:
Originally posted by anniel
How long are your wires to the chip?
I'm using standard jumper wire 2point cable ( 20cm / 8inch ) on the Pomona 5250 pin
Try with ADP-056. |
|
|
| Reply #11
anniel
2572 Posts |
Posted - 12/21/2019 : 06:54:41
|
quote:
Originally posted by noStuff
Supposedly all of these ID might affected ( taken from the bios file descriptor table ) on the board.
C22015 (Macronix MX25L16)
1F4600 (Atmel AT26DF161)
EF3015 (Winbond W25X16)
898911 (Intel 25F160S33B8)
BF2541 (Microchip SST25VF016B)
EF4015 (Winbond W25Q16)
EF4015 (Winbond W25Q16) and EF3015 (Winbond W25X16) surely are.
It seems SPI operation are not fully supported while in circuit.
To recap the situation is as follow;
ATX PSU power connector attached without the power cable > ID and data with percentage error
ATX PSU power connector attached with power cable ( power ON system OFF ) > no ID
NO ATX PSU power connector > no ID
Look in the devices list for compatibility.
Did you isolate VCC to the programmer for your tests? |
|
|
| Reply #12
noStuff
12 Posts |
Posted - 12/21/2019 : 09:26:29
|
quote:
Originally posted by anniel
Did you isolate VCC to the programmer for your tests?
Can you elaborate the VCC isolation on the programmer?
or you mean if I tried to add a small diode in between the GQ-4x4 VCC and the in-circuit W25Q16 PIN 8?
Using the following line doesn't change anything
Name="W25Q16BVSSIG",ID="EF4015",Class="W25X10",Category="SERIAL_EPROM",MFG="Winbond",Vcc="3.3V",RVPP="3.6V",CodeSize="2097152",Platform="GQ-4x4";
First read;
mostly bad or missing data.
Second read;
less corrupted or missing data.
Third read;
mostly bad or missing data.
Fourth read;
less corrupted or missing data.
ETC ETC
Interaction with speed might did something once and I got a good chuck of good content with less shifted data ( 0.302 difference )
Board is already stuffed, I'm using a SO TEST CLIP on a soldered W25Q16. I'm not trying to evaluate the embedding of an ISP connector, the hardware is 10 year old and today with the latest GQ-4x4 software available ( GQUSBprg 7.23 Mar-4-2018 ) the GQ-4x4 is not working correctly, so, chance are it won't unless action is taken directly by the software developer.
Anyway, the GQ-4x4 is partially working it's just not always reading the right data.
Perhaps the board is tricky, perhaps the programmer is, I don't know, I sort of have (can use) already a know working in system hardware programmer. It's just not mine and I'm trying to find an alternative on budget before going for it.
If the GQ-4x4 doesn't 1:1 work while in-circuit and\or work only in-adapter (out of circuit) I'm going to refund till I'm in time.
Thanks anniel for all the support! |
Edited by - noStuff on 12/21/2019 15:03:56 |
|
|
| Reply #13
anniel
2572 Posts |
Posted - 12/21/2019 : 18:18:49
|
quote:
Originally posted by noStuff
quote:
Originally posted by anniel
Did you isolate VCC to the programmer for your tests?
Can you elaborate the VCC isolation on the programmer?
or you mean if I tried to add a small diode in between the GQ-4x4 VCC and the in-circuit W25Q16 PIN 8?
Using the following line doesn't change anything
Name="W25Q16BVSSIG",ID="EF4015",Class="W25X10",Category="SERIAL_EPROM",MFG="Winbond",Vcc="3.3V",RVPP="3.6V",CodeSize="2097152",Platform="GQ-4x4";
First read;
mostly bad or missing data.
Second read;
less corrupted or missing data.
Third read;
mostly bad or missing data.
Fourth read;
less corrupted or missing data.
ETC ETC
Interaction with speed might did something once and I got a good chuck of good content with less shifted data ( 0.302 difference )
Board is already stuffed, I'm using a SO TEST CLIP on a soldered W25Q16. I'm not trying to evaluate the embedding of an ISP connector, the hardware is 10 year old and today with the latest GQ-4x4 software available ( GQUSBprg 7.23 Mar-4-2018 ) the GQ-4x4 is not working correctly, so, chance are it won't unless action is taken directly by the software developer.
Anyway, the GQ-4x4 is partially working it's just not always reading the right data.
Perhaps the board is tricky, perhaps the programmer is, I don't know, I sort of have (can use) already a know working in system hardware programmer. It's just not mine and I'm trying to find an alternative on budget before going for it.
If the GQ-4x4 doesn't 1:1 work while in-circuit and\or work only in-adapter (out of circuit) I'm going to refund till I'm in time.
Thanks anniel for all the support!
Simple piece of paper folded in V shape inserted in the ZIF.
Try with ADP-056. |
|
|
| Reply #14
noStuff
12 Posts |
Posted - 12/22/2019 : 02:25:22
|
quote:
Originally posted by anniel
Simple piece of paper folded in V shape inserted in the ZIF.
Try with ADP-056.
By description it seems the ADP-056 JTAG/SPI could work but the shipping cost in mine area (IT) would be over 60$ and it's not available in mine amazon store region for a reasonable price.
Shouldn't the GQ-4x4 work right away?
About the VCC isolation you mean insulation on the VCC PIN?
like making sure it's not making short on other contact?
We are not talking about separate no loop VCC\IO\GND on top? right?
I'll evaluate mine option for science. |
Edited by - noStuff on 12/22/2019 05:06:26 |
|
|
| Reply #15
anniel
2572 Posts |
Posted - 12/22/2019 : 06:54:22
|
quote:
Originally posted by noStuff
quote:
Originally posted by anniel
Simple piece of paper folded in V shape inserted in the ZIF.
Try with ADP-056.
By description it seems the ADP-056 JTAG/SPI could work but the shipping cost in mine area (IT) would be over 60$ and it's not available in mine amazon store region for a reasonable price.
Shouldn't the GQ-4x4 work right away?
About the VCC isolation you mean insulation on the VCC PIN?
like making sure it's not making short on other contact?
We are not talking about separate no loop VCC\IO\GND on top? right?
I'll evaluate mine option for science.
Maybe you can find one on eBay?
For VCC this picture will be better than any explanation. As the adage goes, a picture is worth a thousand words.
Image Insert:
|
|
|
| Reply #16
noStuff
12 Posts |
Posted - 12/22/2019 : 08:33:03
|
quote:
Originally posted by anniel
Maybe you can find one on eBay?
For VCC this picture will be better than any explanation. As the adage goes, a picture is worth a thousand words.
Image Insert:
If that's the DO(io1) pin I can just leave it out by not jumper wiring the PIN 2 on the CLIP.
Without the DO(io1) in position the W25Q16 get not recognized.
Or you proposing to not using the VCC programmer?
NO VCC programmer > no ID
About the ADP-056 I'm not sure it will do, probably the signal is too weak or the noises too great on the starting point or by the PSU GND routing.
So, mine best option is to send the GQ-4x4 back because it's not working, order a ADP-056 from Britain (@loose), wait 2 month's for the delivery and try again to report if it work or not. |
|
|
| Reply #17
anniel
2572 Posts |
Posted - 12/22/2019 : 14:45:26
|
quote:
Originally posted by noStuff
quote:
Originally posted by anniel
Maybe you can find one on eBay?
For VCC this picture will be better than any explanation. As the adage goes, a picture is worth a thousand words.
Image Insert:
If that's the DO(io1) pin I can just leave it out by not jumper wiring the PIN 2 on the CLIP.
Without the DO(io1) in position the W25Q16 get not recognized.
Or you proposing to not using the VCC programmer?
NO VCC programmer > no ID
About the ADP-056 I'm not sure it will do, probably the signal is too weak or the noises too great on the starting point or by the PSU GND routing.
So, mine best option is to send the GQ-4x4 back because it's not working, order a ADP-056 from Britain (@loose), wait 2 month's for the delivery and try again to report if it work or not.
If you don't get the ID you definitively have power issues on the target. |
|
|
| Reply #18
noStuff
12 Posts |
Posted - 12/22/2019 : 17:22:34
|
quote:
Originally posted by anniel
If you don't get the ID you definitively have power issues on the target.
On standard SPI instruction DO (Data Output) and DI (Data Input) are unidirectional.
By the technical PDF and definition DO is used to read and DI to write.
Image Insert:
In absence of other active power supply source and without the supply voltage from the VCC programmer I find a good thing to have the whole board off W25Q16 included.
Other cheapo stuff may don't do that and the 3v3 circuit would be like the Christmas tree behind me, so, that's why I didn't yet sent it straight back.
With the on-board battery inserted perhaps I could get an ID for the timed clock check. Didn't checked, and wouldn't be useful.
|
Edited by - noStuff on 12/22/2019 17:53:38 |
|
|
| |
Topic |
|
GQ 4x4 W25Q16BV checksum mismatch W25X16
