Hello folks, I’m new here, and hoping for help. I have a Dell 3531 giving a one beep error at boot with a black screen indicating bios failure. I ordered a CH341A and cable clip, and after a several hour fight, finally managed to pull a dump file out of the EN25Q32 bios chip using flashrom and the CH341A provided utility. Both dumps match in file size, and the 341 utility gave a message saying the file content didn’t match. I didn’t catch the error exactly, but I can probably reproduce it if necessary. Anyway, I was hoping for a clean bios file to load into the chip with the CH341 utility to get the laptop working again. I have several versions loaded already, and if any information is needed from the case for Ids, please let me know
I have some electronic background from years ago, but the newer platforms of pc and laptop are a bit over my head these days. I read in the specs for the EN25Q32 bios that it was write protected and I’m not sure reflashing it is possible, but I have it apart and the hardware still set up, so I’m hoping for some more experienced input. I tried loading the extracted file via USB repeatedly with no luck also. I’m not entirely sure the extracted file I have is correct though. Thanks in advance! Here is the latest bios version from dell 3531A03.exe
Attached: dump.bin is my dump file, and bk2.bin was another supposedly working backup from a different forum that was listed as the same A03 version of my Laptop model. I have other extracted bins from the dell exe file, but assumed they weren’t extracted correctly because of the file size, at least until I saw the bk2.bin. Again, thanks in advance for any help here!
@Zerowin - The BIOS image itself extracted from exe is 8MB (8192kb), but you only dumped one chip that’s why your dump is 4MB. You need to also dump other chip and send to me so I can compile, fix, then split and send back to you. Second chip will have 32 or 64 in ID, only needs to be 32 but they may have put 64 on there.
If they are both the same size and ID, you MUST be sure you remember which dump goes to what chip/location on the board So make a note of where the chip is on the second dump you send me, and name the file as such, and I will keep that name so you can keep things straight based off the names you use.
Thanks for the reply Lost_N_Bios. I didn’t know there was a second chip, but the first one was UH6 marked on the board, and the only other chip I found made by eon was marked UH2(MB chip marking, not the chip id number). The first posted dump file was from the 4/32 Mb chip, this second one, found on the other side of the MB detected as 2/16 Mbs. The bin file I read from it verifies with information in the chip. The first file did not. This is only a 2Mb file though, so if I need to keep looking, or use a different version, please let me know. Getting all 8 pins on the SOIC clip to connect was a pain because it’s physically right up against another component, so I’d rather not move the clip from where I have it if at all possible. Thanks again, you do amazing work here!
@Zerowin - Thanks, that is the correct chip, sorry I didn’t think to mention that it may be a 2MB chip too.
However, I just now noticed your original comment (second sentence) about dump not matching. This means bad dump, and you need to redo with other software version or other software etc. This is required because if you can’t read/verify then you can’t write properly (it will fail to write/verify, so failed write). Sorry I didn’t notice that before!! I would have had you redo all I’m going to mention below before you got to the other chip. What software version did you use for the first chip dump anyway? Sounds like it’s not compatible since you can’t get a verified dump/
Did the second chip dump verify OK after read? If yes, whatever software you used for that, you can use same again to write back to it and it should be OK, But you will need to use something else for the first chip otherwise it will just be a brick until you can get a proper write/verify (and really, read/verify needed first too so I have a proper dump from you.)
I do know what you mean about tight spots, sometimes you may need to file down some of the plastic on the clip, that will help. I\m sorry, but before I do anything here I need you to make a verified read from the first chip and send to me, and let me know if second chip you just did in post #3 verified properly before you saved it (If so, great, no need to redo it)
I would also try 1.30 and 1.34 normal CH341A software, as well as “CH341A v1.31(1.4) (CH341AFree)” from the package below (for this one, just choose correct type “25” and size "4MB/32" http://s000.tinyupload.com/index.php?fil…695330485827902
Thanks for the patience and prompt replies, Lost_N_Bios.
The second chip read out was with CH341A ver 1.29 and verified just fine, so you have a good copy on that one already.
After trying again with the first chip, CH341A ver 1.29, 1.30, 1.31, and 1.34 all failed to verify because the chip and main memory were in disagreement (or something very similar). Oddly enough, ver 1.18 read and verified correctly, but I assume this is a false positive?
Just to be sure, I downloaded the ASProgrammer 1.41 and selected the first of the three chips listed (EN25Q32A) and read, it failed verify. The error reported stated “Verification error on address 00291885”. I selected the second EN25Q32A chip and read it and it seemed to pass verify, no error reported. Last I tried the EN25Q32B chip selection, read it, and it appeared to verify reporting no error. So now you have three files, one named CH341Aver1.18chip1, and an EN25Q32Achip1 and EN25Q32Bchip1. I know exactly what version each was read with, but the first is from the lowest version of the 341Tool that worked and the second two were from the ASProgrammer 1.41 named respectively after the chip setting that produced the good dump. I assume three files to compare should show you if there are any differences. If you need anything else, please let me know, and thanks so much for the time you’ve spent on this already. Where the heck did you learn all of this? Lol!
@Zerowin - Thanks! That may be fine with 1.18, provided it matches what you get from ASProgrammer. Yes, I will be able to tell if all looks OK, and if all three of those match (which we should expect) You’re welcome!! I learn all this just from helping people with BIOS and doing BIOS for myself over many years
Results - EN25Q32Achip1.bin - Matches your originally sent file, which you said did not verify (correct)
Difference - Original dump.bin + EN25Q32Achip1.bin = same/same @ 0x291885h = DB >> @ 0x291885h in compiled BIOS EN25Q32Bchip1.bin + CH341ver1.18chip1.bin = @ same location ^^ = FB << These, when compiled with other chip, are broken BIOS, so these are failed dumps and it would seem your original + EN25Q32Achip1.bin are actually valid
Confusing! EN25Q32Achip1.bin you said verified correctly, yes? If yes, then how did your original dump.bin fail to verify, files are same/same? Maybe it was just a failed read during it’s verification process, or did you maybe make a few dumps originally and some failed but some passed and you forgot which passed?
This really important to sort out, so you know which software/version works properly, so when you write it will be OK. Please make one more dump using CH341A v1.31(1.4) (CH341AFree)" (for this one, just choose correct type “25” and size “4MB/32”) If it fails to verify, it’s OK, don’t send to me, just let me know.
However, thinking it over now, your BIOS is broken/bricked anyway, so maybe my findings are valid and EN25Q32Bchip1.bin + CH341ver1.18chip1.bin = broken/brick BIOS is proper dump of this chip (which we wanted anyway), and I’m only seeing the broken BIOS I think this is correct now, maybe? (NOT, see thoughts below!) So, when I fix BIOS for you, the write should be done back to this chip with 1.18 or ASProgrammmer and EN25Q32B ID
But, also really hard for me to accept above to be OK/True, since Original dump.bin + EN25Q32Achip1.bin both look more like a working BIOS when compiled (ie these are not corrupted at the main BIOS volume, and it can be expanded properly) This does not mean BIOS is OK, it only means the others look far more broken than these, so I would assume these and the software used here would be proper one to use. So hard to be sure, I guess you may just have to try writing back several times to this 4MB chip until you find which writes OK, verified, AND then boots. The 2MB chip can be written back using 1.29 since it seems to be fine there.
Wait a minute, before I fix anything… What caused this failure anyway?? Did you flash in wrong model BIOS somehow? I see BIOS Version A08 in the files you sent me, but only A01-A03 have been released for the model you specified. I was wondering why structure didn’t look right as I went to rebuild using stock volumes!
The original dump file may have been ok, I just assumed not. When I first started with this, I only had one or two versions of CH341A Programmer, and a few driver versions for it. None of the drivers initially installed without causing some kind of error, and both tool versions said there were no working drivers installed when opened even though it showed the programmer as connected. One of the first two or three reads verifies as OK, but I had bad reads also, and saved copies of each one in different locations and with different names as I went along. Since I initially had problems, I started looking more closely at the chip until I had a good readable picture of it to identify which chip it was. After that I looked for CH341A Programmer versions that actually supported that chip, which is when I found the tools download with several versions in it from this site. The drivers here also installed without error, although some versions still said it was not installed when opened. All of those newer versions made bad dumps but I kept copies of them as well. By the time I picked one to upload here, I just assumed it was bad. The original instructions I found for the CH341A Programmer said it would not detect chip ID at all, which is clearly wrong.
Once I have the second chip written to and verified, I can flip the MB back over and reassemble the laptop excluding hard drive and keyboard. I should be able to flash chip1, remove the SOIC connection from the programmer socket, apply ac power and see if it boots. If not, disconnect, reconnect the SOIC and try again, so that part is rather forgiving.
Using CH341Aver1.31 failed to verify when read. I am surprised 1.18 produced a verified read, but that’s why I suspected it was a false positive in the first place. But of course that makes me wonder why EN25Q32A and B both read in ASProgrammer and had different dump files but didn’t error for one atleast. Both values can’t be contained in the chip at the same location, so one must not have matched. This is why all of this is a bit over my head these days
Honesly I don’t know what happened to the Laptop. I was working on it for a friend who just bought another used, as this one has been down for years, and it was his only computer in the house. He doesn’t know anything about computers at all, so it’s possible he tried to load a wrong bios version, or put the wrong model number in on dell’s site. I though most executable updates from dell and HP all checked to make sure they were compatible with the machine they are being loaded on before writing so exactly that couldn’t happen. As far as physically writing the chip, he wouldn’t even know how to get it apart, so I’d rule that out. If a kid got hold of it though, who knows?
Again, thanks for your help, anything to try to fix it is better than leaving it a clumsy large paperweight
---------------
Sorry for the “Guest” posts, forgot to login the last two replies .
Yes, I assumed original was not OK too, based on your comments about it not verifying It’s OK about the guest, I know it’s you, merged
There is only one driver to install for all the CH341A software versions, it’s in the package I linked you to previously (I think, if not here - http://s000.tinyupload.com/index.php?fil…695330485827902) Some older versions wont work on Win10, but most should, and all would use that same driver.
Both chips need written to when I fix BIOS for you, the BIOS is split in the middle across both chips.
Yes, something went wrong with the BIOS versions, and flashing or programming etc And usually you can’t “Flash” wrong BIOS either, at least not using official/standard flash methods (only with programmer or FPT and for that you’d have to know what you were doing and possibly unlock things first etc)
So, are you 100% sure it’s 3531 Model? If yes, I guess we can just toss the old BIOS crap out and put in the correct BIOS across both chips. I can’t tell what model BIOS this is you’ve dump off it, only that it’s BIOS A08 which was never made for 3531
Well I’m an idiot. I originally talked to him in text messages, and he searched the model number under the battery (P28F, which could be several different models) and told me that Inspiron 15 3531 is what the model was, and I never double checked it. Sorry for misleading you, it’s actually a 3521. Here is the link to the service tag searched on Dell’s site, Service Tag specific. Now it makes sense why there was an A08 version in there to begin with. Sorry for the confusion Lost_N_Bios. My humble apologies sir.
I originally talked to him in text messages, and he searched the model number under the battery (P28F, which could be several different models) and told me that Inspiron 15 3531 is what the model was, and I never double checked it. Sorry for misleading you, it’s actually a 3521. Here is the link to the service tag searched on Dell’s site, 
My humble apologies sir.