Use bare wire, without of an insulation. You may use larger flexible wire if you split the thick wire strand into single wires.
I believe the most hard work will be de-soldering of existing jumper pins. You may heat the pin from bottom side until the top plastic start to melt and you will be able to pull the pin by a tweezers. Don't pull to early and too strong to not pull out the PCB traces. After you clear the holes from old jumpers, place new ones.
Step one. Prepare a thin bare wire from a larger flexible wire strand:
06-01-2019, 05:47 AM (This post was last modified: 06-01-2019, 05:52 AM by vecnar.)
(06-01-2019, 04:40 AM)Nikolay_Po Wrote: Use bare wire, without of an insulation. You may use larger flexible wire if you split the thick wire strand into single wires.
I believe the most hard work will be de-soldering of existing jumper pins. You may heat the pin from bottom side until the top plastic start to melt and you will be able to pull the pin by a tweezers. Don't pull to early and too strong to not pull out the PCB traces. After you clear the holes from old jumpers, place new ones.
Step one. Prepare a thin bare wire from a larger flexible wire strand:
Step two. Bend the wire to insert into the holes:
Step three. Insert the wire:
Step four. Solder it using a neutral soldering flux:
Step five. Cut the tails.
Thank you for such a detailed instructions Nikolay!
I do not have single stranded cable of good size. I have stranded cable wires but they are too thin if using single strand. Can I twist them and create a single one or am I better off using old posts and bend them into u shape?
Please see picture of cables and board with one jumper more or less cleaned.
By the way, on this video with SSD testing was achieved the speed of 262MB/s which is significantly lower than 500MB/s of Kingston A400 SSD. I can make a conclusion that the speed of interfaces in this SATA test was 3Gbps.
I have done some bad soldering attempt. It was hard to get the solder out of the holes and i damaged a few o rings, detached from pcb. I also couldn't find wire cutters and i sanded a bit with a file at the rear. My soldering is probably worse than jumpers, hard to tell without any tools.
Please see pictures attached and dmesg log. Dmesg log is after successful reboot with limiting one port to sata 3, probably bad soldering as other one is not doing that.
I turned on while hdd cage was outside the case and the same issue, after reboot hard drives came back. The only difference i saw led flashing only once during startup on sata controller board where before i saw at least 3-4 quick flashes when hdds were not listed. So probably i made it even worse now but anyhow i need to get another one to rule it out.
Once hard drives are recognized i am only getting 63 MB/s when copying from one drive to another with "rsync --progress" command, same as before. I am also getting slow speed when copying from cifs share 18 MB/s, but when copying from cifs to sd card i get 30MB/s. I used hdparm -t --direct /dev/sda1 and got 179 MB/s and 180MB/s for /dev/sdb1.
Do you think below is a good choice to go for, i will only have 2 drives and would like write speeds between drives to be around 100 MB/s. As below is pcie version 1 link throughput is 250 MB/s so should be probably getting something close to 100 MB/s. https://www.ebay.ie/itm/Controller-Pcie-...Sw5cNYSZAr
I placed hdd case inside the case with ground wire connected to cable coming out of ground pin on pcb. But couldn't reproduce it yet. I will find out tomorrow as previously leaving it for longer not connected to electricity triggered it for sure.
I haven't connected ground wire from pcb directly to case, i had this wire attached before to pcie card but now i placed it between pcb and mounting bolt below pcie card.
06-02-2019, 05:02 AM (This post was last modified: 06-02-2019, 05:06 AM by Nikolay_Po.)
This is good enough soldering:
All the jumpers are conducting the signal to corresponding SATA connectors pins? This is right time to check by a DC Ohmmeter. You may loose O-ring if it is on inactive side of PCB. If other side with a PCB trace is OK, the connection will be OK.
You have decreased the stray capacitance by replacing larger jumper by smaller. If they are conducting right and not shorted between each other, this is all you may done here.
Looking into the dmesg output...
(06-02-2019, 05:02 AM)Nikolay_Po Wrote: This is good enough soldering:
All the jumpers are conducting the signal to corresponding SATA connectors pins? This is right time to check by a DC Ohmmeter. You may loose O-ring if it is on inactive side of PCB. If other side with a PCB trace is OK, the connection will be OK.
You have decreased the stray capacitance by replacing larger jumper by smaller. If they are conducting right and not shorted between each other, this is all you may done here.
Looking into the dmesg output...
Thank you Nikolay,
I had some paste laying around with soldering iron so I assume it was flux as it was non conductive. Cleaned it after with methylated spirit and a tooth brush, something I had at home.
I tested for shorts yesterday and there were none between neighbouring pins. Sata connector pinout is the same on both connectors. Pins 2,3,5 and 6 are linking to soldered posts. pins 1,4 and 7 have no connections to soldered posts and looking online for pinout they are ground
There are two kinds of SATA interfaces start: sequence and simultaneous.
Fist, sequence start case is in earlier log:
Code:
[ 2.822172] ata1: SATA max UDMA/133 abar m512@0xfa010000 port 0xfa010100 irq 239
[ 2.824016] ata2: SATA max UDMA/133 abar m512@0xfa010000 port 0xfa010180 irq 239
[ 3.289330] ata1: SATA link up 6.0 Gbps (SStatus 133 SControl 300)
[ 3.295986] ata1.00: ATA-10: WDC WD40EFRX-68N32N0, 82.00A82, max UDMA/133
[ 3.307042] ata1.00: 7814037168 sectors, multi 16: LBA48 NCQ (depth 31/32), AA
[ 3.319425] ata1.00: configured for UDMA/133
[ 3.331341] scsi 0:0:0:0: Direct-Access ATA WDC WD40EFRX-68N 0A82 PQ: 0 ANSI: 5
[ 3.338871] sd 0:0:0:0: [sda] 7814037168 512-byte logical blocks: (4.00 TB/3.64 TiB)
[ 3.346751] sd 0:0:0:0: [sda] 4096-byte physical blocks
[ 3.354588] sd 0:0:0:0: [sda] Write Protect is off
[ 3.362123] sd 0:0:0:0: [sda] Mode Sense: 00 3a 00 00
[ 3.362305] sd 0:0:0:0: [sda] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA
[ 3.418361] sda: sda1
[ 3.425786] sd 0:0:0:0: [sda] Attached SCSI disk
[ 5.102569] systemd[1]: Mounted FUSE Control File System.
[ 5.108546] systemd[1]: Mounted Configuration File System.
[ 5.116147] systemd[1]: Started Journal Service.
[ 5.179134] systemd-journald[317]: Received request to flush runtime journal from PID 1
[ 5.695722] ata1.00: configured for UDMA/133
[ 5.699739] ata1: EH complete
[ 5.704375] sd 0:0:0:0: [sda] Write cache: disabled, read cache: enabled, doesn't support DPO or FUA
[ 6.004362] FAT-fs (mmcblk0p6): utf8 is not a recommended IO charset for FAT filesystems, filesystem will be case sensitive!
[ 6.029101] FAT-fs (mmcblk0p6): Volume was not properly unmounted. Some data may be corrupt. Please run fsck.
[ 6.775230] EXT4-fs (sda1): mounted filesystem with ordered data mode. Opts: user_xattr,usrjquota=aquota.user,grpjquota=aquota.group,jqfmt=vfsv0,acl
[ 8.824559] ata2: link is slow to respond, please be patient (ready=0)
[ 13.377369] ata2: softreset failed (device not ready)
[ 18.900596] ata2: SATA link up 6.0 Gbps (SStatus 133 SControl 300)
[ 23.914593] ata2.00: qc timeout (cmd 0xec)
[ 23.928554] ata2.00: failed to IDENTIFY (I/O error, err_mask=0x4)
[ 33.991567] ata2: softreset failed (device not ready)
[ 34.465590] ata2: SATA link up 6.0 Gbps (SStatus 133 SControl 300)
[ 34.480898] ata2.00: failed to IDENTIFY (I/O error, err_mask=0x100)
[ 34.495434] ata2: limiting SATA link speed to 3.0 Gbps
[ 39.939592] ata2: SATA link up 3.0 Gbps (SStatus 123 SControl 320)
[ 39.955188] ata2.00: ATA-10: WDC WD40EFRX-68N32N0, 82.00A82, max UDMA/133
[ 39.969728] ata2.00: 7814037168 sectors, multi 16: LBA48 NCQ (depth 31/32), AA
[ 39.981263] ata2.00: configured for UDMA/133
[ 39.992757] scsi 1:0:0:0: Direct-Access ATA WDC WD40EFRX-68N 0A82 PQ: 0 ANSI: 5
[ 40.002185] sd 1:0:0:0: [sdb] 7814037168 512-byte logical blocks: (4.00 TB/3.64 TiB)
[ 40.010183] sd 1:0:0:0: [sdb] 4096-byte physical blocks
[ 40.019003] sd 1:0:0:0: [sdb] Write Protect is off
[ 40.023064] sd 1:0:0:0: [sdb] Mode Sense: 00 3a 00 00
[ 40.023151] sd 1:0:0:0: [sdb] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA
[ 40.093354] sdb: sdb1
[ 40.112645] sd 1:0:0:0: [sdb] Attached SCSI disk
[ 40.235716] ata2.00: configured for UDMA/133
[ 40.246159] ata2: EH complete
[ 40.262243] sd 1:0:0:0: [sdb] Write cache: disabled, read cache: enabled, doesn't support DPO or FUA
[ 41.178347] EXT4-fs (sdb1): mounted filesystem with ordered data mode. Opts: user_xattr,usrjquota=aquota.user,grpjquota=aquota.group,jqfmt=vfsv0,acl
After both controller interfaces initialization at 2.82s, the first interface was successfully negotiated at full speed 6Gbps at 3.28s then the disk was attached at 3.42s. At 6.02s file system mounting of first interface HDD was completed without of errors.
Second interface behaved not as good as first. Only at 8.82s it start negotiation. At 13.37s a soft reset happened. At 18.90s 6Gbps link at SATA2 was established. But at 23.92s timeout error was triggered. It was negotiated to 6Gbps again at 34.46 but failed again at 34.48 with "polling device detection hint" error (device absent equivalent, IMHO). After two failures the speed was decreased to 3Gbps at 34.49s. Then file system mounting went OK and was finished for SATA 2 at 41.17s.
Because the drives was attached not simultaneously, with a delay of 40.112645-3.425786=36.68s, there were no chances for RAID to bring up.
IMHO, a severe hardware problem. What if you start the drives completely separate?
Try to record a logs for only one ATA1 interface connected (only one HDD connected) while changing the cable and the drive. Will be there a difference between the cables and between the drives?
The trial when both drivers were up and the RAID was detected is different. Can't analyze right now. It is better to collect additional information by making the test with cable/drive changing.
Also repeat the test with cable and drive change for ATA2 interface. Record a logs.
Probably these tests and logs will give more information.
(06-02-2019, 07:50 AM)Nikolay_Po Wrote: There are two kinds of SATA interfaces start: sequence and simultaneous.
Fist, sequence start case is in earlier log:
Code:
[ 2.822172] ata1: SATA max UDMA/133 abar m512@0xfa010000 port 0xfa010100 irq 239
[ 2.824016] ata2: SATA max UDMA/133 abar m512@0xfa010000 port 0xfa010180 irq 239
[ 3.289330] ata1: SATA link up 6.0 Gbps (SStatus 133 SControl 300)
[ 3.295986] ata1.00: ATA-10: WDC WD40EFRX-68N32N0, 82.00A82, max UDMA/133
[ 3.307042] ata1.00: 7814037168 sectors, multi 16: LBA48 NCQ (depth 31/32), AA
[ 3.319425] ata1.00: configured for UDMA/133
[ 3.331341] scsi 0:0:0:0: Direct-Access ATA WDC WD40EFRX-68N 0A82 PQ: 0 ANSI: 5
[ 3.338871] sd 0:0:0:0: [sda] 7814037168 512-byte logical blocks: (4.00 TB/3.64 TiB)
[ 3.346751] sd 0:0:0:0: [sda] 4096-byte physical blocks
[ 3.354588] sd 0:0:0:0: [sda] Write Protect is off
[ 3.362123] sd 0:0:0:0: [sda] Mode Sense: 00 3a 00 00
[ 3.362305] sd 0:0:0:0: [sda] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA
[ 3.418361] sda: sda1
[ 3.425786] sd 0:0:0:0: [sda] Attached SCSI disk
[ 5.102569] systemd[1]: Mounted FUSE Control File System.
[ 5.108546] systemd[1]: Mounted Configuration File System.
[ 5.116147] systemd[1]: Started Journal Service.
[ 5.179134] systemd-journald[317]: Received request to flush runtime journal from PID 1
[ 5.695722] ata1.00: configured for UDMA/133
[ 5.699739] ata1: EH complete
[ 5.704375] sd 0:0:0:0: [sda] Write cache: disabled, read cache: enabled, doesn't support DPO or FUA
[ 6.004362] FAT-fs (mmcblk0p6): utf8 is not a recommended IO charset for FAT filesystems, filesystem will be case sensitive!
[ 6.029101] FAT-fs (mmcblk0p6): Volume was not properly unmounted. Some data may be corrupt. Please run fsck.
[ 6.775230] EXT4-fs (sda1): mounted filesystem with ordered data mode. Opts: user_xattr,usrjquota=aquota.user,grpjquota=aquota.group,jqfmt=vfsv0,acl
[ 8.824559] ata2: link is slow to respond, please be patient (ready=0)
[ 13.377369] ata2: softreset failed (device not ready)
[ 18.900596] ata2: SATA link up 6.0 Gbps (SStatus 133 SControl 300)
[ 23.914593] ata2.00: qc timeout (cmd 0xec)
[ 23.928554] ata2.00: failed to IDENTIFY (I/O error, err_mask=0x4)
[ 33.991567] ata2: softreset failed (device not ready)
[ 34.465590] ata2: SATA link up 6.0 Gbps (SStatus 133 SControl 300)
[ 34.480898] ata2.00: failed to IDENTIFY (I/O error, err_mask=0x100)
[ 34.495434] ata2: limiting SATA link speed to 3.0 Gbps
[ 39.939592] ata2: SATA link up 3.0 Gbps (SStatus 123 SControl 320)
[ 39.955188] ata2.00: ATA-10: WDC WD40EFRX-68N32N0, 82.00A82, max UDMA/133
[ 39.969728] ata2.00: 7814037168 sectors, multi 16: LBA48 NCQ (depth 31/32), AA
[ 39.981263] ata2.00: configured for UDMA/133
[ 39.992757] scsi 1:0:0:0: Direct-Access ATA WDC WD40EFRX-68N 0A82 PQ: 0 ANSI: 5
[ 40.002185] sd 1:0:0:0: [sdb] 7814037168 512-byte logical blocks: (4.00 TB/3.64 TiB)
[ 40.010183] sd 1:0:0:0: [sdb] 4096-byte physical blocks
[ 40.019003] sd 1:0:0:0: [sdb] Write Protect is off
[ 40.023064] sd 1:0:0:0: [sdb] Mode Sense: 00 3a 00 00
[ 40.023151] sd 1:0:0:0: [sdb] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA
[ 40.093354] sdb: sdb1
[ 40.112645] sd 1:0:0:0: [sdb] Attached SCSI disk
[ 40.235716] ata2.00: configured for UDMA/133
[ 40.246159] ata2: EH complete
[ 40.262243] sd 1:0:0:0: [sdb] Write cache: disabled, read cache: enabled, doesn't support DPO or FUA
[ 41.178347] EXT4-fs (sdb1): mounted filesystem with ordered data mode. Opts: user_xattr,usrjquota=aquota.user,grpjquota=aquota.group,jqfmt=vfsv0,acl
After both controller interfaces initialization at 2.82s, the first interface was successfully negotiated at full speed 6Gbps at 3.28s then the disk was attached at 3.42s. At 6.02s file system mounting of first interface HDD was completed without of errors.
Second interface behaved not as good as first. Only at 8.82s it start negotiation. At 13.37s a soft reset happened. At 18.90s 6Gbps link at SATA2 was established. But at 23.92s timeout error was triggered. It was negotiated to 6Gbps again at 34.46 but failed again at 34.48 with "polling device detection hint" error (device absent equivalent, IMHO). After two failures the speed was decreased to 3Gbps at 34.49s. Then file system mounting went OK and was finished for SATA 2 at 41.17s.
Because the drives was attached not simultaneously, with a delay of 40.112645-3.425786=36.68s, there were no chances for RAID to bring up.
IMHO, a severe hardware problem. What if you start the drives completely separate?
Try to record a logs for only one ATA1 interface connected (only one HDD connected) while changing the cable and the drive. Will be there a difference between the cables and between the drives?
The trial when both drivers were up and the RAID was detected is different. Can't analyze right now. It is better to collect additional information by making the test with cable/drive changing.
Also repeat the test with cable and drive change for ATA2 interface. Record a logs.
Probably these tests and logs will give more information.
Hello Nikolay,
I do not use raid, they are both ext4 formated and act as single disks.
I think my soldering added this additional error so i wouldn't focus on it too much. The main error was that on first boot neither of the drives were recognized and i do not think I was able to reproduce it with single drive.
I tested both ports on sata controller and after my soldering ata2, the right one which had 2 o ring detached from both sides of pcb is failing or causing errors when copying for both drives but ata1 only for 1 drive when copying.
I just need to order one sata controller now and go from there. Which one do you think is best to go for, both look to be low profile but black one has all ports from side which is probably best for cables in this case. I can detach mounting bracket from current one. https://www.ebay.ie/itm/Controller-Pcie-...SwaK1cNgTY
Step_one_single_wire.JPG (Size: 52.79 KB / Downloads: 390)
