Tuesday, May 11, 2021

 Carette compressor first problem after 4 years.

Guess what the compressor had its first problem after 4 years.
It seems the non-return valve malfunctioned and blocked the inflow of the high pressure into the oil water seperator.
The end-result was a mini explosion when the steel tube removed itself from the ferrule after a too high pressure buildup 💣

the non-return valve before it failed

tube removed itself under the too high pressure



The culprit was the green rubber inside the non-return valve, it sheered off a small piece and that piece blocked the ingress of the high pressure air.

the failed non-return valve



The repair :
  • ordered a spare ferrule set and used it to reconnect the steel tube to the non-return valve.
  • removed the green rubber part so it no longer can block the inflow
All is fine again. No leaking even with the green rubber removed... 😎




Wednesday, March 3, 2021

RMA  CPU AMD 3700x after 16 months of use.

symptoms:

  • daily BSOD (blue screen of death), frequency of crashes increased in the last months
  • crashes during gaming and heavy loads
  • spontaneous reboots
  • CPU temps were normal under 100% load from stress benchmark cpu-z (60C) and 29C at idle.
  • NO WHEA errors.

Diagnostics:
  1. Clean install of windows10
  2. Part swapping: you need to remove the things that are NOT the problem:
    • changed the memory to QVL approved for the mainboard
    • disabled/enabled XMP profile
    • removed the 5700XT, swapped with RX580 and 6900XT
    • changed the PSU from 750W to spare 850W
    • downgraded the mainboard BIOS to the  version when I bought it.
    • swapped the 3700X with a 3600: problem gone....
    • swapped the 3700X with a 5800x: problem gone... (the 5800x is really hot and I returned it)
    • moved the 3700X to my other PC and the problem moved with it!
  3. installed OCCT v8 beta
    • the memory test indicated no problems even after 24 hours
    • used their CPU test with heavy load on the cpu cache and it flagged 2 cores as bad after 5 min.
      • select CPU test, Small data set, mode extreme, load type variable, auto on the rest
    • after a few hours the pc would BSOD or reboot during the stress test due to many errors.

workarounds:
  • fixing cpu multiplier (36x) at normal speed in the BIOS or in Ryzen Master : super stable
  • fixing cpu multiplier (39x) in the BIOS or in Ryzen Master : super stable
  • fixing cpu multiplier (40x) in the BIOS or in Ryzen Master : errors after a few hours
  • XMP on/off made no difference
Solution:

  • RMA the CPU at AMD: https://www.amd.com/en/support/kb/warranty-information/rma-form

Kudos to AMD for a very fast RMA process!
  • Mailed for a RMA on Thursday and mentioned all the diagnostics that I already did.
  • received a prepaid DHL return label on Friday
  • sent CPU back also on Friday by DHL Pickup service.
  • CPU received on Monday at noon in the AMD RMA center in the Netherlands, Europe
  • Confirmation Wednesday noon that the CPU passed inspection and RMA accepted
  • Email on Wednesday 5pm that replacement CPU was shipped to me!
  • Friday morning: brand new CPU 3700X

Tuesday, October 1, 2019

BIGTREETECH SKR-E3-DIP-V1.0 with ABL

###I NO LONGER USE THIS BOARD AS IT WAS SUPERUNSTABLE###

It seems the board has a lot of interference when using hardware PWM and it messes with the Bltouch servo signal.

The Hardware:

From BigTreeTech an upgrade board for my Creality CR10 3D printer that I bought on Aliexpress for less than 25$.

Out with the 8-bit board based on 1284p sanguino and in with 32-bit ARM.

Focus on getting the BLTouch 3.1 working...
Drivers BigTreeTech TMC2208 V3...

Visual inspection of the board and dodgy soldering of course! The Hotend connectors are not even connected properly... You can see the hole where the pin should have been...


The Drivers:

The main reason I bought this board was to be able to print more silent.
TMC2208 V3 preconfigured for UART: check J2 where it has a 0-ohm resistor soldered.





Assembled it looks like this:













Firmware:

Vanilla Marlin 2.0.X not working with the TMC2208 at the moment. BUG in TMCSTEPPER.

12/10/2019:
Marlin compiled fine but the bltouch v3.1 does not work reliably, so it is pointless to invest any more time is this board for the moment.

13/10/2019:
Update from BigTreeTech: use software PWM for fans and heaters to make Bltouch more reliable. First tests seem to confirm this. More testing needed of course... It's a 25$ board after all...

26/6/2020: threw the board out as it is very unstable





Wednesday, October 17, 2018

ESXI restore config error: Missing Manifest File

Reminder for self:

[[email protected]:/tmp] vim-cmd hostsvc/firmware/restore_config /tmp/configBundle.tgz force
(vim.fault.InvalidBundle) {
   faultCause = (vmodl.MethodFault) null, 
   faultMessage = <unset>, 
   text = "Missing Manifest File"
   msg = "Received SOAP response fault from [<cs p:06f94a48, TCP:localhost:80>]: restoreConfiguration
fault.InvalidBundle.summary"

}

The reason for this error is a different releaselevel in the Manifest.txt file

[[email protected]:/tmp] more Manifest.txt 
RELEASELEVEL=VMware ESXi 6.5.0 Update 2
UUID=00000000-0000-0000-0000-0secretnumber here
KERNELOPTS=no-auto-partition

USEROPTS=

change the releaselevel to the current release and try again!

To check the current releaselevel execute the following command:

[[email protected]:/tmp] vim-cmd hostsvc/firmware/backup_config
Bundle can be downloaded at : http://*/downloads/{some-drive-here}/configBundle-ESXI.lan.tgz


Download this file and check the layout of manifest.txt inside the tgz

Tuesday, July 3, 2018

APS CAM870 MK2 paintball co2 pin nozzle firing problem

The APS CAM870 mk2 marine edition paintball.

Was looking for a shotgun paintball for the collection with some realistic loading.
Here it is:


It uses an internal bolt system on Co2:



There are lots of reviews online so I won't be doing one!


Because mine does not work out of the box....


So what's the problem? Well the Co2 does not power the cartridge. When firing the only thing that happens is a puff of Co2 smoke from the back of the bolt!

Guess what: the firing pin broke on the first strike. This does not bode well for the future.

Here is a photo of a complete pin:
This is the part I found loose inside the cam870: the back of the pin.

.

I know I did not dry fire the gun, so or the pin left the factory already broken/weakened or it really did brake on the first strike. In both cases, this means that there is a serious flaw in the design of the pin.

I'm waiting for the shop to reply for my warranty claim...
Update 15/07/2018: received the replacement pin under warranty and it all works fine now.

Some chronos: 

4gram ball @170-190fps = 5.3 - 6.6 Joule = 3.9 - 4.9 fpe

be sure to add silicone spray into the co2 port when charging, the first shots were very low on fps due to leaking...

Friday, May 19, 2017

Repairing an ipad air 2 with spots on the left side of the display

Well after swapping the old battery in my ipad air 2 I got a nasty surprise....
Spotlights on the left side of the screen!
First thought was: I messed up the screen with my smd airstation heat while trying to remove the glued display :-(
So I ordered a new one, installed it and ..... still the same problem.


After some googling and some more you-tubing. I spotted the same problems with the Iphone 6 and 6S. Apparently it is the backlight filter fuse that fails.

One small problem, really small, even so small that I can't see it without a microscope?

The 1 euro cent coin has a diameter of 16.25mm


the top left and middle components are the failed filters, notice the fine split in the middle one

Next steps in the coming days will be to order the components and try to desolder the filters without killing the ipad air. I don't have any nano tweezers so it will be challenging...

Heat issues

Filter was replaced, both of them to be sure. But spots still are there. A new issue has occurred, the ipad heats up :-)





The cause is the failing zener diode. A diode test reveals 0.2v on the left (good) one and 1.5v (bad) on the other. So I replaced it and all works now.

The cause of all this was me forgetting to FIRST remove the battery connector before disconnecting  the display flex cable. It shorted the fuse and the zener diode.



Tuesday, May 2, 2017

Carette 300 bar 4500 psi compressor review test - Part2

This is a part2 of the review, goto part1

Adding a PMV to the standard Carette

You can find priority valves or PMV on the internet for less than 100$.
Some Bauer compressors use an OEM PMV which can be found in the US as model 211 priority valve. Just google it. I wasn't able to find it here in Europe.

Some sites where they sell it :
http://www.augustindustries.com/back-pressure-valve-val-back.asp
http://aeroconsystems.com/cart/back-pressure-regulators-priority-valves/model-211-priority-valve-regulator/
http://scubacompressor.com/shop/back-pressure-regulator-p-69.html

A good article on using a PMV in a compressor can be found at http://www.scubaengineer.com/high_pressure_compressor_filtrat.htm


So where do you put in on the Carette compressor? Ideally it should be put after the last filter but it is easier to put just behind the Oil water separator. Since almost all of the water is removed by the Oil water separator with the PMV set at 2000psi, that's just fine.




On the European model (the one I have) the fitting from the Oil-water separator to the filling hose is 1/4 BSPT.
The PMV has a 1/4 NPT fitting.
So you need:

  • a 1/4 female barrel BSP (to connect to the BSPT)
  • a 1/4 male BSP to male NPT (to connect to the PMV input)
  • a 1/4 male NPT plug or 1/4 NPT gauge (to connect to the PMV output1)
  • a 1/4 male NPT to male BSP (to connect from the PMV output2 to the bleed)
  • a 1/4 female BSP BLEED (to release pressure between the PMV and the tank you are filling)
  • a 1/4 male BSP to BSPT (to connect to the standard filling hose which is female BSPT)
  • 4 bonded seal washers
All these fittings can be found here in Europe on www.bestfittings.co.uk but should not be difficult to find elsewhere.
FYI the connections on the standard carette filter are metric m10x1.

The end result:
Priority valve fitted to the exit of the oil water separator