I have a unit that I damaged by accidently grabbing a 48V POE barrel plug from under my desk. It now cycles similar to yours.
Have you tried holding the little internal pushbuttons again? Otherwise, I'm out of ideas.
Dell VEP/VMWare Edge/Velo Cloud SD-WAN/VeraCloud VEP1400/VEP1400-X firewall units
- Thread starter j_h_o
- Start date
Notice: Page may contain affiliate links for which we may earn a small commission through services like Amazon Affiliates or Skimlinks.
Have you tried holding the little internal pushbuttons again? Otherwise, I'm out of ideas.
Yep, tried pushing the one close to the barrel, the one close to the ethernet ports, both at the same time with no success.
Let's see when my clamp arrives, maybe I can re-flash one of the bios chips
Let's see when my clamp arrives, maybe I can re-flash one of the bios chips
I guess these were the best 7€ I invested 
I did a dump of the top-layer eeprom, as well as the bottom one just in case (and if anyone needs), I wanted to cross-flash for testing, but at the end of the day since both versions are 16777216 bytes I checked what was the BIN file used with the Dell firmware package, and lo and behold, same size.
Plugged everything back and... it boots !
I did a dump of the top-layer eeprom, as well as the bottom one just in case (and if anyone needs), I wanted to cross-flash for testing, but at the end of the day since both versions are 16777216 bytes I checked what was the BIN file used with the Dell firmware package, and lo and behold, same size.
Code:
# flashrom -p ch341a_spi -w /home/arnaud/VEP1445/VEP1400-X-BIOS-3.48.0.9-23.bin --progress !820
flashrom 1.4.0 on Linux 6.12.74+deb13+1-amd64 (x86_64)
flashrom is free software, get the source code at https://flashrom.org
Found Winbond flash chip "W25Q128.V" (16384 kB, SPI) on ch341a_spi.
Reading old flash chip contents... [READ] 1% complete... [READ] 2% complete... [READ] 3% complete... [READ] 4% complete... [READ] 5% complete... [READ] 6% complete... [READ] 7% complete... [READ] 8% complete... [READ] 9% complete... [READ] 10% complete... [READ] 11% complete... [READ] 12% complete... [READ] 13% complete... [READ] 14% complete... [READ] 15% complete... [READ] 16% complete... [READ] 17% complete... [READ] 18% complete... [READ] 19% complete... [READ] 20% complete... [READ] 21% complete... [READ] 22% complete... [READ] 23% complete... [READ] 24% complete... [READ] 25% complete... [READ] 26% complete... [READ] 27% complete... [READ] 28% complete... [READ] 29% complete... [READ] 30% complete... [READ] 31% complete... [READ] 32% complete... [READ] 33% complete... [READ] 34% complete... [READ] 35% complete... [READ] 36% complete... [READ] 37% complete... [READ] 38% complete... [READ] 39% complete... [READ] 40% complete... [READ] 41% complete... [READ] 42% complete... [READ] 43% complete... [READ] 44% complete... [READ] 45% complete... [READ] 46% complete... [READ] 47% complete... [READ] 48% complete... [READ] 49% complete... [READ] 50% complete... [READ] 51% complete... [READ] 52% complete... [READ] 53% complete... [READ] 54% complete... [READ] 55% complete... [READ] 56% complete... [READ] 57% complete... [READ] 58% complete... [READ] 59% complete... [READ] 60% complete... [READ] 61% complete... [READ] 62% complete... [READ] 63% complete... [READ] 64% complete... [READ] 65% complete... [READ] 66% complete... [READ] 67% complete... [READ] 68% complete... [READ] 69% complete... [READ] 70% complete... [READ] 71% complete... [READ] 72% complete... [READ] 73% complete... [READ] 74% complete... [READ] 75% complete... [READ] 76% complete... [READ] 77% complete... [READ] 78% complete... [READ] 79% complete... [READ] 80% complete... [READ] 81% complete... [READ] 82% complete... [READ] 83% complete... [READ] 84% complete... [READ] 85% complete... [READ] 86% complete... [READ] 87% complete... [READ] 88% complete... [READ] 89% complete... [READ] 90% complete... [READ] 91% complete... [READ] 92% complete... [READ] 93% complete... [READ] 94% complete... [READ] 95% complete... [READ] 96% complete... [READ] 97% complete... [READ] 98% complete... [READ] 99% complete... [READ] 100% complete... done.
[READ] 0% complete... [READ] 12% complete... [READ] 25% complete... [READ] 37% complete... [READ] 50% complete... [READ] 62% complete... [READ] 75% complete... [READ] 87% complete... [READ] 100% complete... [READ] 0% complete... [WRITE] 50% complete... [WRITE] 0% complete... [WRITE] 2% complete... [WRITE] 0% complete... Erase/write done from 0 to ffffff
Verifying flash... [READ] 1% complete... [READ] 2% complete... [READ] 3% complete... [READ] 4% complete... [READ] 5% complete... [READ] 6% complete... [READ] 7% complete... [READ] 8% complete... [READ] 9% complete... [READ] 10% complete... [READ] 11% complete... [READ] 12% complete... [READ] 13% complete... [READ] 14% complete... [READ] 15% complete... [READ] 16% complete... [READ] 17% complete... [READ] 18% complete... [READ] 19% complete... [READ] 20% complete... [READ] 21% complete... [READ] 22% complete... [READ] 23% complete... [READ] 24% complete... [READ] 25% complete... [READ] 26% complete... [READ] 27% complete... [READ] 28% complete... [READ] 29% complete... [READ] 30% complete... [READ] 31% complete... [READ] 32% complete... [READ] 33% complete... [READ] 34% complete... [READ] 35% complete... [READ] 36% complete... [READ] 37% complete... [READ] 38% complete... [READ] 39% complete... [READ] 40% complete... [READ] 41% complete... [READ] 42% complete... [READ] 43% complete... [READ] 44% complete... [READ] 45% complete... [READ] 46% complete... [READ] 47% complete... [READ] 48% complete... [READ] 49% complete... [READ] 50% complete... [READ] 51% complete... [READ] 52% complete... [READ] 53% complete... [READ] 54% complete... [READ] 55% complete... [READ] 56% complete... [READ] 57% complete... [READ] 58% complete... [READ] 59% complete... [READ] 60% complete... [READ] 61% complete... [READ] 62% complete... [READ] 63% complete... [READ] 64% complete... [READ] 65% complete... [READ] 66% complete... [READ] 67% complete... [READ] 68% complete... [READ] 69% complete... [READ] 70% complete... [READ] 71% complete... [READ] 72% complete... [READ] 73% complete... [READ] 74% complete... [READ] 75% complete... [READ] 76% complete... [READ] 77% complete... [READ] 78% complete... [READ] 79% complete... [READ] 80% complete... [READ] 81% complete... [READ] 82% complete... [READ] 83% complete... [READ] 84% complete... [READ] 85% complete... [READ] 86% complete... [READ] 87% complete... [READ] 88% complete... [READ] 89% complete... [READ] 90% complete... [READ] 91% complete... [READ] 92% complete... [READ] 93% complete... [READ] 94% complete... [READ] 95% complete... [READ] 96% complete... [READ] 97% complete... [READ] 98% complete... [READ] 99% complete... [READ] 100% complete... VERIFIED.I assume you were able to reprogram without lifting any pins? There can often be issues with the chip select pin or the programmer attempting to power other ICs on the PCB.I guess these were the best 7€ I invested
I did a dump of the top-layer eeprom, as well as the bottom one just in case (and if anyone needs), I wanted to cross-flash for testing, but at the end of the day since both versions are 16777216 bytes I checked what was the BIN file used with the Dell firmware package, and lo and behold, same size.
Plugged everything back and... it boots !
Are you now 100% updated? Do you still have the NVMe problem or going to stop while you're ahead? I would think you're feeling invincible at this point.
You're correct, I did not desolder the chips, I simply used the IC clamp that came with the CH314A.
Since I cleared cmos and nvram I had to go again through the nvram configuration, but it looks like it is booting now with the nvme mounted instead of the wifi card (for good measure I initially booted with the wifi card just in case).
Still have no idea how I managed to so much corrupt the bios / nvram previously, but at least the box is 100% usable again.
Since I cleared cmos and nvram I had to go again through the nvram configuration, but it looks like it is booting now with the nvme mounted instead of the wifi card (for good measure I initially booted with the wifi card just in case).
Still have no idea how I managed to so much corrupt the bios / nvram previously, but at least the box is 100% usable again.
I can also confirm that Virtium VL41A4G63B-C3SA DDR4-3200 works. Of course, the system still runs at 2400MT/s.
There is no public source for the 610 BIOS. Dell only supports the 620/640/680. Everything else is under the control of VMware and would require a subscription running their proprietary OS. What are you hoping to accomplish?
Depending on your internet bandwidth, you'll get much better performance out of a 640. I'm running 2.5Gbps Comcast through the 10G ports on mine, with Suricata running on the external interface, and it's running about 60% CPU. Not bad at all.
I assume that people are getting these 610 models either free or cheap off eBay. There are also 510, 520, and 540 models out there. I wish we could unscramble this thread. However, at this stage, I think we just need to start over with a clear divide.
Does anyone have an LTE model with the top-side m.2 socket populated? If so, can you post a photo of the empty socket? I'm trying to determine which additional components may get populated. There does already appear to be 3.3V power. I found this regulatory tear-down, but the images aren't great.
I originally noticed the USB traces leading to the mini PCIe socket. I had to research what part would bridge the gap. You could just jumper each trace with a small wire, but assuming you have hot air, it's probably comparable effort. I can confirm that if you populate L31, the internal USB does work.
Only later did I notice a similar choke at L30. The correct footprint for both locations is 0805. I had already used a 1206 for L31. I used WCM-3216 90 Ohm. I also ordered WCM-2012 67 Ohm, but have not tested. It's strange that L30 was populated when there's no connector, but L31 was omitted. Dell intentionally disabled USB for this slot. If you're clever, you could just move L30 to L31. I'm now looking at using both locations.
The gaps in the grill on the back plate are not accidental. There's space for 3x SMB antenna studs. You just have to drill your own holes. The back plate can be temporarily removed to make modification easier. There's also ample room inside the case. Through creative use of AliExpress adapters, I have both an NVMe drive and a Mixtile Z-Wave/Zigbee radio. The radio only requires the mini PCIe USB signals. Unfortunately, the Mixtile module is currently impossible to purchase in quantity 1. Same goes for the similar module from Z-Wave.me. Most people use USB dongles. I went for the module because I got it cheap and I prefer the external antennas. It's also very tidy inside the case.
I originally noticed the USB traces leading to the mini PCIe socket. I had to research what part would bridge the gap. You could just jumper each trace with a small wire, but assuming you have hot air, it's probably comparable effort. I can confirm that if you populate L31, the internal USB does work.
Only later did I notice a similar choke at L30. The correct footprint for both locations is 0805. I had already used a 1206 for L31. I used WCM-3216 90 Ohm. I also ordered WCM-2012 67 Ohm, but have not tested. It's strange that L30 was populated when there's no connector, but L31 was omitted. Dell intentionally disabled USB for this slot. If you're clever, you could just move L30 to L31. I'm now looking at using both locations.
The gaps in the grill on the back plate are not accidental. There's space for 3x SMB antenna studs. You just have to drill your own holes. The back plate can be temporarily removed to make modification easier. There's also ample room inside the case. Through creative use of AliExpress adapters, I have both an NVMe drive and a Mixtile Z-Wave/Zigbee radio. The radio only requires the mini PCIe USB signals. Unfortunately, the Mixtile module is currently impossible to purchase in quantity 1. Same goes for the similar module from Z-Wave.me. Most people use USB dongles. I went for the module because I got it cheap and I prefer the external antennas. It's also very tidy inside the case.
Code:
root@gc:~# lspci -s 04:00.0
04:00.0 Non-Volatile memory controller: Micron Technology Inc 2550 NVMe SSD (DRAM-less) (rev 01)
root@gc:~# lsusb
Bus 001 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
Bus 001 Device 002: ID 1a86:55d2 QinHeng Electronics USB Dual_Serial
Bus 002 Device 001: ID 1d6b:0003 Linux Foundation 3.0 root hub
Last edited:
Yeah and I am trying the most to recover the edge 610 vep1400 so I am looking for the bios from people who already have the model so they can share the files.
@dagio Can you capture a log of your boot process? Does the primary BIOS fail repeatedly and then cycle to the secondary BIOS? Can you make it back to DiagOS? Member @Brood copied the secondary BIOS to the primary using an external USB programmer. You might be able to use the software updater if the file is formatted correctly.
A very late reply, but I didn't see this answered. You want 3-pin Chinese 1.25mm. I don't know if the Dell sockets are technically JST or Molex, but my AliExpress cables fit just fine. If the wire colors don't match up, it's pretty easy to lift the plastic finger, remove the contact, and then slide back in. The 1.25mm size is the smallest that still offers both male and female on the end of a wire.
Last edited:
Unfortunately I don't have the log right now but yeah the boot initially start from primary vep1400-x twice and then after some time it boots vep1400 from the backup bios and yeah I am able to make it to the DiagOs.
Did you make any progress on this? I want to control the fans on my Edge 640 running OPNsense. I am able to load the i2c and i2cbus kernel modules, but no iicN devices are created.
You forgot the CC2650F128RSM (Bluetooth/Zigbee) in the top front left corner. It looks like its got its own debug header.
Code:
<platform>VEP1400</platform>
<bluetooth>
<uartport>ttyS1</uartport>
<reset>
<tool>i2ctool</tool>
<i2ctool-cmd>
<i2c-bus>/dev/i2c-1</i2c-bus>
<i2c-addr>0x31</i2c-addr>
<i2c-reg>0x1</i2c-reg>
<i2c-bytes>1</i2c-bytes>
<i2c-cpld-bit-val>0x04</i2c-cpld-bit-val>
<i2c-cpld-bit-onoff>0</i2c-cpld-bit-onoff>
</i2ctool-cmd>
</reset>
</bluetooth></config>
I continue to love my 640 running OPNsense, and last night I finally figured out how to utilize the QAT for OpenVPN! Adding a note to this thread since I feel like a fool for taking so long to figure this out due to conflicting information - it really is quite simeple.
You will find lots of info stating that only the paid version of pfsense supports the QAT - this is not true!
First things first: I run ESXi and pass through all the network adapters AND the QAT to the virtual machine running OPNsense.
Under System > Settings > Miscellaneous you will find a drop down menu to set hardware acceleration to QAT.
I did this, and verified that the qat driver was loaded by FreeBSD, but I could see no effects at all. FreeBSD did not create the /dev/crypto special device node for user application access to QAT (more on this later) and vmstat did not show any QAT interrupts being processed.
It turns out that within your OpenVPN instance you must set the instance type to DCO. This is an option to perform the crypto in the kernel instead of userspace - avoidiing two transitions between kernel and user space in the processing of each VPN packet - a nice optimization. But as a bonus, the FreeBSD kernel crypto functions use QAT if available! No need for the /dev/crypto device nodes.
Having set the OpenVPN instance type to DCO instead of TUN, I have verified QAT interrupts being processed when I VPN and reduced CPU load.
Hope this helps someone!
You will find lots of info stating that only the paid version of pfsense supports the QAT - this is not true!
First things first: I run ESXi and pass through all the network adapters AND the QAT to the virtual machine running OPNsense.
Under System > Settings > Miscellaneous you will find a drop down menu to set hardware acceleration to QAT.
I did this, and verified that the qat driver was loaded by FreeBSD, but I could see no effects at all. FreeBSD did not create the /dev/crypto special device node for user application access to QAT (more on this later) and vmstat did not show any QAT interrupts being processed.
It turns out that within your OpenVPN instance you must set the instance type to DCO. This is an option to perform the crypto in the kernel instead of userspace - avoidiing two transitions between kernel and user space in the processing of each VPN packet - a nice optimization. But as a bonus, the FreeBSD kernel crypto functions use QAT if available! No need for the /dev/crypto device nodes.
Having set the OpenVPN instance type to DCO instead of TUN, I have verified QAT interrupts being processed when I VPN and reduced CPU load.
Hope this helps someone!