All,
I've seen comments suggesting the need to get into the modules at the last minute and probe the bottom of the PCBs. To me that suggests extremely serious problems with any process that would require such drastic last minute troubleshooting. In my humble opinion the old ways of doing things are long gone and didn't work with AO-40, resulting in two spectacular failures - both preventable.
Present and future projects need to be carried out in a much more structured way. I've attempted to do that with the Eagle receiver construction and testing but even my best efforts are woefully inadequate for Eagle and unthinkable for a geostationary payload with a design life of 15 years.
One very small example of what I'm talking about is that you should never probe a surface mount component or a via with a test probe. You can cause latent defects which will lead to premature failures. Test points need to be designed into the PCB from the beginning. All it takes is one person doing a seemingly harmess act to cause critical damage to a payload that might represent a multi-million dollar effort and many man years of work.
I hope I can get some of my concerns across at the Symposium. I'm looking forward to seeing you all...
73, Juan - WA6HTP
On 10/18/07, Rick Hambly (W2GPS) w2gps@cnssys.com wrote:
Dick,
The current design has too much unnecessary hardware and access to the inside of the modules is unnecessarily restrictive. Why not have a full depth machined chassis with a screw-on top and front plate? That would eliminate the PEM standoffs and lots of hardware plus it would allow top only access to PCBs. Tolerances for board mounting could be more closely managed as could overall chassis stiffness. It would also allow for maximum flexibility in the use of front panel space. Boards could be inserted from the top or front as desired. We could choose to use front loading slots for the PCB with fewer screw-in machined standoffs to maximize usable board area.
Rick
W2GPS
AMSAT LM2232
*From:* eagle-bounces@amsat.org [mailto:eagle-bounces@amsat.org] *On Behalf Of *Dick Jansson-rr *Sent:* Wednesday, October 17, 2007 7:29 PM *To:* 'Chuck Green' *Cc:* 'AMSAT Eagle' *Subject:* [eagle] Re: Updated Module
Chuck:
Indeed the machined baseplate width is just 141mm and the cavity is 126mm wide. I would NOT recommend any PCB to be that full width, that it be no wider than 140mm, so theoretically you could have a PCB that is 140x180mm. Perhaps that can be another iteration. (I did not deal with the PCB at all in this round.) Not tonight honey, I've slugged at this beast for six hours today, enough already!
Dick Jansson, KD1K
kd1k@amsat.org
kd1k@arrl.net
-----Original Message----- From: Chuck Green [mailto:greencl@mindspring.com] Sent: Wednesday, 17 October, 2007 21.46 To: Dick Jansson-rr Cc: Bob Davis; AMSAT Eagle Subject: Re: [eagle] Updated Module
Hi Dick,
This is a good step forward.
I had in mind making the PCB wider so that it extends over the entire
length of the base plate on each side. It seems to me that you could
then make the PCB mounting points less intrusive into the cavity below
the PCB. It looks like I still have that option even if you don't
change the mounting points (true?). And I think that would also better
support the PCB and allow for more thermal contact between the PCB and
the base.
I see that you have more mounting points per linear distance for the
cover than for the PCB. Is this based on vibration modeling?
Also note that the depth of the cavity below the posts is 6.35mm, save
for in the very center where there is an attachment for the connector
plate that is located 0.75mm below the PCB for a space in the center
that is 10mm wide by 6mm deep from the connector plate. I was loathe
to locate such a piece here but felt the need for a third attachment
for the connector plate.
Please reconsider this. It will cause considerable restriction on where
connectors can be located that are also soldered to the PCB. Connector
space/flexibility is a *major* issue.
Chuck
Via the Eagle mailing list courtesy of AMSAT-NA Eagle@amsat.org http://amsat.org/mailman/listinfo/eagle