Page MenuHomeAleph Objects Inc

USB Drive read intermittent
Open, NormalPublic


Once you turn on the machine it doesn't read the USB drive. You have to unplug the USB stick and plug it back in to get it to read. It seems intermittent like something else might be causing it, but we can't pin it down. And we're not sure if it happens on every machine since it seems intermittent.

Event Timeline

robert created this task.Apr 10 2019, 10:36 AM
robert triaged this task as Normal priority.
robert created this object with edit policy "LulzBot Hardware Products (Project)".

The machine on my desk is currently doing it. This is with 108 FW too.

@robert you should be able to read the usb after powering on the machine with it plugged in, there is a delay for it to recognize the USB however it shouldnt take very long. maybe 10 seconds

mine is only taking about 4 seconds to recognize

@tutley this one on my desk is doing it consistently. It wont read even after a couple of minutes. Not til I pull it out and plug it back in does it read it.

@mjpelletier any ideas on why this is happening?

Make sure you have the newer, REV C, version of the USB Reader board. The REV B boards, which I believe were used in some early Quiver machines, had a crystal oscillator. I found that the crystal oscillators did not turn on reliably (some taking more than 10 seconds or not turning on at all) and redesigned the board with a MEMS oscillator module. I have not seen a boot up issue with the new design but, if you have a REV C board, please send it to me so I can analyze the problem.

@robert any updates on this one?

robert added a comment.EditedApr 15 2019, 8:12 AM

@tutley Kent took that machine. I think he said it had the right rev board, but it may have been the USB itself. But Im not sure it if its doing it with the other USB stick since it was doing it with 2 different ones before. @kent is it still doing it with the good USB stick?

kent added a comment.Apr 15 2019, 9:25 AM

I have it on my bench. I did verify that the revC USB board is installed. It seems like it is specific to this USB drive, since other USB drives are recognized on boot up ok.

kent added a comment.Apr 15 2019, 9:30 AM

if you have a REV C board, please send it to me so I can analyze the problem.

@mjpelletier I will remove the board and get it to you. I'm going to take a quick video of the issue first.

@kent you should send the specific USB drives its failing with as well

Still seeing individual USB sticks fail on boot up. Once you switch it out with another it works. So we're just red tagging the USB sticks.

I was having this problem with a machine I was working with on Friday. I ended up changing out the USB stick for a different one and it started reading correctly shortly after boot up.

We're having an issue now on a couple machines where right in the middle of the print we'll have a USB read error kill the print. Not sure if its the USB sticks or something else. Hard to replicate so its hard to tell whats causing it.

tutley added a subscriber: paulette.May 1 2019, 9:36 AM

@robert i saw one of these that @paulette had red tagged,
The issue was that one of the LCD ribbon cables was not fully plugged in. I would check this if you run into more of these

So we're still seeing this issue. I had one today that would not read the USB stick on boot up, but it read a different USB on boot up. The one that will not read on boot up will read if you remove it and re-insert it while the machine is on. I put a red tag on the USB stick that did not read on boot up.

MikeR added a comment.May 8 2019, 2:43 PM

@paulette are you seeing this on many machines?

@MikeR No, it's pretty sporadic. But still something we thought worth looking into.

robert added a comment.May 9 2019, 2:01 PM

Dan just had another machine doing this too.

I took a look at some that were failing out for this reason and believe it to be a USB quality issue more than anything. I snagged a batch of USB drives from calibration that @paulette mentioned they were failing on machines in T7205. While the gcode is the same between files, I have had a few fail for random USB drive errors that are difficult to reproduce but do appear localized to those drives.

Perhaps an alternative source may bring more consistent results?

@sherylb @kent @DaniAO we've similar issues before with low quality SD cards. Can we please get feedback from the USB drive vendor on this?

Steven assigned this task to kent.May 14 2019, 7:12 AM
kent added a comment.May 20 2019, 11:41 AM

The part number for these USB sticks is EL-MS0543 and here is a link to the supplier's product page:

I have 9 drives that have been tagged by production since start of quiver production. I'm running badblocks on one of the drives that production was having issues with now to see if it flags any bad sectors.

kent added a comment.May 21 2019, 8:41 AM

I put in a purchase request for 100 alternate flash drives for testing.

kent reassigned this task from kent to jperry.May 21 2019, 10:50 AM
kent added a subscriber: jperry.

@jperry can you please contact Kingston and find out if they can screen our logo onto these drives and if there is an MOQ and if there is a price break at quanity?

I contacted Kingston with this request and copied Kent on the email.

kent claimed this task.Jun 12 2019, 9:56 AM

I wrote all these drives with the NexCopy. 100/100 of the DTSE9 drives passed the copy+compare process. I'm running a vernier_dim-test print with each one now to check if there are any drives that aren't recognized or have errors during the print. So far, 8/8 of the drives I have done a test print with have passed with no issues.

kent reassigned this task from kent to DaniAO.Jun 18 2019, 8:11 AM

I was able to complete a vernier test print with all 99 flash drives that I tested (@DaniAO has the last one). There was an issue with two of the drives where it didn't read on the fist plug in, and threw a USB read error when the drive was removed. After the machine was power cycled in these two cases, I was unable to reproduce the issue, and was able to complete the test print without any further issues.

Based on this testing, I recommend removing the EL-MS0543 drive and replacing it with the OF-CP0418 drive.

@DaniAO I am assigning this to you for now, let me know if there is anything else I can do to help resolve this issue.

The one I tested showed no issues.

@sherylb I am not sure what else I can do for this.

Is this something we can to move forward with to change?

sherylb added a subscriber: tomc.Jun 18 2019, 11:56 AM

@jperry @tomc Josh, what is the lead time and price on the USB drives that Kent and Dani have tested?

Tom, how many are in stock of the old ones? If there are a number of them I'd like to test them use the ones that are good and get credit on all bad ones. This should be done via NCMR/Red Tag tickets that Kent/Josh can help process once we have numbers.

Once we get that number, we will need to have the BoM changed with the new part number and vendor if it is decided that this will be the change on this issue going forward. Kent, can you help run this process?

We have 662 on hand of the EL-MS0543 and were purchased at $3.83. The new version from Kingston is 4.99 and we only quoted up to 150 units. I emailed them for a new quoter and lead time.

kent added a comment.Jun 18 2019, 2:59 PM

I will create a change order for this.

@jperry can we get a part number for the new USB please? This is needed for the change order.


I created EL-MS0562: Kingston, 16GB, USB 2.0 Flash Drive

kent added a comment.Jun 18 2019, 3:05 PM

@DaniAO OF-CP0418 is the PN for the new drive.

I will delete EL-MS0562

Josh, we need lead time and pricing on the new proposed change. We also need to know if that part can be imprinted or laser etched with the Lulzbot logo for branding purposes.

I emailed Kingston to get volume pricing on 500 and 1000 pcs as well as the lead time. Kingston only offers limited custom runs of these up to 150 units. Anything outside of of those quantities they do not do any custom logos. Dani G and I decided to source a label from My Asset Tag that we can place on the side of the USB. I will create a part number and get some samples on order.

It doesn't have to be from My asset tag.. sticker giant may work too.

Will a sticker work for branding? If we add a sticker then we are adding labor costs to apply the sticker to the USB. We would order more than 150 units for production so this shouldn't be an issue with customizing a logo. For samples we do not need them customized I don't think. Dani or Planning would make that call.

I think the hang up is Kingston will do 150 MAX with the logo and we need more then that.. so the sticker was another option. But after talking to marketing.. the branding isn't required. It is preferred to get a USB drive that works.

Kingston will only do custom branding up to a total of 150 units. Anything over that quantity they do not do custom branding and we would need to order from a distributor for any quantities greater than 150 branded/ unbranded. The USB starts at $4.99 before any branding and I'm sure the cost would be higher than if we did it in house. I got a quote from Sticker Giant and each label would be $0.176.

Thank you Josh, I misinterpreted the 150 as a min to customize.

zachah added a subscriber: zachah.Jul 12 2019, 2:53 PM

Just had a Pro act up when I flashed to .135 with a USB drive inserted. The printer was stuck in a reset loop after flashing. I powered off the printer, took out the USB drive, then it powered on fine. I thought maybe it had corrupted the USB drive somehow but after resetting the printer and re-inserting the drive it seems to work fine. This is with the older green flash drives as well.