TECHNET Archives

October 2018

TechNet@IPC.ORG

Options: Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
Yuan-chia Joyce Koo <[log in to unmask]>
Reply To:
TechNet E-Mail Forum <[log in to unmask]>, Yuan-chia Joyce Koo <[log in to unmask]>
Date:
Thu, 18 Oct 2018 10:46:33 -0400
Content-Type:
text/plain
Parts/Attachments:
text/plain (67 lines)
IMHO, part of issue supermicro is the original design that allow  
remote upgrade (V9 I guess if I read their spec correctly... not sure  
if all the after version also have such capability), independent of  
OS.  some chat group indicated the units can also self diagnostic  
upon fault - independent of OS.  I believe (personally) it possibly  
the combination of relatively open design (not tight enough) allow  
such a scheme take place.  Still waiting for more info comes to the  
light in the future.  The companies statements are very interesting,  
except Amazon sort of direct address the issue, all the others,  
including apple, a bit of wishy washy (consumer data not compromised  
etc.etc... Siri is not on super micro server is the only strong  
points in the statement IMHO).  The units do have strong point for  
easy upgrade/cut user installation, service cost (possibly can hire  
inferior support staff in the user company... specially, at current  
tight labour market... high quality personnel are hard to come by if  
affordable)... but look like expose some weak points on security...  
Supply chain is not the key issue, the design is (if they farm out  
detail design, like the gerber review - can't see how to insert a  
chip in the design if gerber was reviewed and signed off).  I might  
barking on the wrong tree.. let's wait until all the dirt come out  
(if ever).  IMHO.
jk
On Oct 18, 2018, at 9:54 AM, Ahmad, Syed wrote:

> Would it help disable embedded spy chips if we label all empty  
> areas on the boards with lasers?
>
> -----Original Message-----
> From: TechNet <[log in to unmask]> On Behalf Of Wayne Showers
> Sent: Tuesday, 16 October, 2018 10:44 AM
> To: [log in to unmask]
> Subject: Re: [TN] Incompatible Labels
>
> I can write a book or at least a pretty good paper on the pros and  
> cons of each of laser, label, and inkjet and they all have their  
> issues, but in order of best to worst.
>
> Laser onto a white silkscreen marking square, this is the best for  
> permanency. Legibility (HRI) is very good.  The barcodes,  
> especially 2-D (MRI) can be a little hard to get a good reading on  
> but a good scanner helps.
> Best is to have a copper pad, with solder mask and silkscreen over it.
> Issues: can (will) expose copper so, cannot cross traces.
>
> For a label, Brady Hi-Temp with thermal transfer and acrylic  
> adhesive backing.  Up to 4 SMT reflows have no impact and actually  
> help label adhesion.  They do NOT like wave solder.  The trick to  
> maintaining the printing is to actually send the labels through at  
> least one thermal profile (apply labels BEFORE SMT).
> * I did a mini-validation for a customer with 4 thermal excursions  
> and 10 wash cycles and maintained legibility throughout.  I also  
> did not lose any labels in the cleaner.
> PCB, .062, 4-layer, 1 oz inners, 1/2 oz outers with .001 inch  
> plating, green semi-gloss solder mask.
> Label: Brady B-719 Material
>
> Direct Inkjet.  Good process, but sensitive to (1) gloss and over- 
> cured solder mask, (2) PCB cleanliness, (3) proper bake on. (1)  
> Gloss and Over-Cured solder mask do not provide enough anchor  
> points for the printing and can (will) come off in a cleaner.  (2)  
> PCB cleanliness is critical as any dirt, debris or oil will prevent  
> the ink from adhering.  (3) Bake on time and temperature is  
> critical to getting a good set on the ink.  The 2-part inks are  
> less sensitive to (1), but (2) and (3) still apply and a new (4) is  
> in play as depending on application method, these inks tend to  
> smear at the edges.

ATOM RSS1 RSS2