Hi Wade
True ; to pass Zoran's acidic approach (AKA if it's not measurable it does
not exist) felt almost better than to see all of theories coming up with
practical results .
By the way you can bump to him in Anaheim ; send him some mail if you really
wish to be swamped by spreadsheets .
READ BELOW IF I MANAGE TO AD ANYTHING :
> ----------
> From: Zoran Raich[SMTP:[log in to unmask]]
> Sent: Wednesday, 3 February 1999 9:04
> To: [log in to unmask]
> Subject: Re: [TN] SMT dispensing systems
>
> Hi Wade,
> We are using a Camalot Gemini dispensing system. I being the most
> sceptical
> out of Paul Klasek and myself am now very much sold on dispensing systems
> for Sodder Paste as you guys call it. It has totally blown away my
> theories
> a had about screen printing. To give you some of the Benefits,
>
> The good Points
>
> 1: Solder Paste Wastage calculated at 4% (Call me inefficient but we had
> approx 30% with screen printing in past lives).
Can be less still with better syringe (individual light (4 pumps) monitoring
> 2: Safety, operators prefer handling the tubes and is very clean with no
> mess.
We went with 55cc EFD barrels , 2 mounted on back plane (on 644 pump 2in
back , 2 in front ) ,
200gr each ; 2 of them on back plane last a shift . Paradise .
> 3: Deposition Control. Ever had the problem where you wanted to control
> the
> amount paste using stencils but could not or were limited due to fine
> pitch
> constraints and aspect ratios (Stepped stencils can be a nightmare). It is
> interesting to note that coplaniarity issues with finer pitch devices and
> other IC's are negated dut to the dot height which eliminates dry joints
> due
> to coplaniarity issues.
This was one of major considerations (fill in hole) for example pth
anchoring of SMD connectors ;
(those which should be on Steve's toshiba's mother of all boards ) ;
the flexibility of tuning (hunting down the illusive balls) is tremendous ;
you can dispense even on solder mask to let it creep up the fillet .
> 4: Flexibility with regards to frequent design changes, no need to buy
> multiple stencils through alpha, beta and production stages.
> 5: No cleaning of stencils or blocked apertures.
> 6: Paste rheology does not play a large role in dispensing negating the
> need
> to constantly work the paste during screen printing to achieve a
> consistant
> deposit (Mainly concerned with fine pitch)
> 7: Change over time between one product and another is as quick as loading
> another program and changing the Conveyor width (Again no need for stencil
> maintenance).
>
> The Not so Good Points
>
> 1: The solder paste is more expensive for -400 to +500 Mesh (Negated
> somewhat buy the wastage and waste handling but unfortunately the
> accountants do not see this, I suspect it has something to do with the
> personality bypass operation they have after certification)
I think he says in the end it's the same on paste cost .
> 2: Programming can take a couple of hours to do depending on how many
> components there are but compared to stencils if you take into account the
> time taken to do aperture reductions and adjustements creating gerbers
> emailing them to stencil suppliers liasing with stencil suppliers creating
> purchase orders la di da etc etc programming time is not as bad as you
> first
> imagine.
Gosh ; he's really picky ; how often do you run new board in ?
And 2 hours beats 2 days on stencil > if yo're that fast , (+ COST ; and
pray you dead right on those apertures ) .
> 3: You do sometimes get a missing dot here and there but nothing of major
> concern a little process control fixes this.
We started with 325-500 mesh stock ; which we still unfortunately have to
use up ;
the 400-500 ind NC-SMQ90 is faultless (running a week on three shifts
without cartridge and needle clean)
We did validate it already ; did not implement in production yet .
> 4: It is slower than screen printing, you can not put them in front of
> fast
> chip shooters and utilise SMT machines effectively. We are on our boards
> getting approximately 70% of tact. So please note it depends on your
> production requirements, we are currently looking at buying the fastest
> dispenser on the market to take up this issue.
This one is again in the take-off stage :
We have here My19 with hydra and getting one more as it's from our opinion
ideal (almost 300 feeders) for a real bad mix as we have ;
and once we deploy dual needles ( on 0805's or 603's ; 1206's ; you name it
> keep them in parallel orientation on design) ;
the Gemini will comfortably keep up with My . Presently because this place
is for god (chairman) knows what reason going through the stratosphere;
we have little time to tingle the toys ; we can't get on them because
production sits on it 24 hours 7days almost .
That should change this year as Zoran goes to NW (kick tyres as he says) of
XYFlex for this application .
With him around I have more time for battles with R&D's dreams (they call it
manufacturable designs) .
> These are some of the more obvious things which may be food for thought
> and
> help in the decision making process.
>
> Regards
>
> Zoran Raich
> ResMed Ltd
> www.resmed.com.au
>
>
To sum it up Wade ; yo're on perfectly right line ; local subs think the
same way ; on small runs the stencil pains are unbearable ;
and with US market ; as Steve said ; I'm surprised you don't have dispensing
lines by dozen's .
Resent this mail to Indium guys ; (William Jackson . MD ;
[log in to unmask] ; http://www.indium.com ; ) ;
if you don't get reply from Bill this week ;
ask for resmed's 55cc 400-500 mesh ' ( I can not fault that paste, and did I
try !) ; get from camelot (speedline mayhem now) the 644 head (last quote
I've heard was US $ 16K) ;
with 4 extra cartridges ( $ 6+K more ) to rotate in maintenance cycle and
yo're in business .
As you have likely already 642 head you can start with that ( no room for
dual needles ) ;
on mostly single 21 gauge > we do , ( buy few of them > operators will shear
few of them on learning curve ) .
Once you gain experience and establish dot sizing and spreads and component
library the programming of new board should be less than 2 hours .
To get there will take you about a month of clean time .
If you have $ 5K spare I can send you a fellow with more than decade
experience on Gem's ; or come myself ; or send another wise guy who started
it all up with me .
That is for ticket; food and place in the garage for sleepin' bag .
See you Wade
Paul Klasek
http://www.resmed.com
PS
good to see subs waking up ; I'll try to dig up for you old camelot's charts
of savings, line by line dispensing X printing .
It's an eye opener I validated and verified last year as factual ( no wool ;
no bull )
PPS
Bill , any chance to "lend" Wade few of those 55cc barrels (400-500) for all
of this good will ? Please .
They go usually to Advaced Material Technologies - Aust .
Thanks A lot
PPPS
If you've read the article from Chuck Bauer (smta) in Advanced Packaging (I
think > Steve's got a copy) ;
I'd have to clarify it as Chuck painted it with bit to loose brush strokes ;
some less correct than others ;
let me know if you need any details (lets digitize) .
NB
almost forgot ; auger is the only way .
As far as I see it ; there's only one on top = you got it ; II with stated
45K speed it is .
When I calculated the balance you need on average mix 4x dispensing speed
than p&p .
That is apples with apples = statedxstated ; or actualxactual .
the dual needle will get you to 60K with 644 ; design plays a role .
> > ----------
> > From: Wade Oberle[SMTP:[log in to unmask]]
> > Sent: Wednesday, February 03, 1999 3:13 AM
> > To: [log in to unmask]
> > Subject: [TN] SMT dispensing systems
> >
> > Dear Technetters,
> > I'm interested in who are the 'major players' in the SMT
> > dispensing business. In my opinion, Camalot and Asymtek are on the top
> > tier. I'd be interested to here other opinions and experience.
> >
> > Regards,
> >
> > Wade Oberle
> > 414-947-3477
> > [log in to unmask]
> >
> > ################################################################
> > TechNet E-Mail Forum provided as a free service by IPC using LISTSERV
> 1.8c
> > ################################################################
> > To subscribe/unsubscribe, send a message to [log in to unmask] with
> > following text in the body:
> > To subscribe: SUBSCRIBE TechNet <your full name>
> > To unsubscribe: SIGNOFF TechNet
> > ################################################################
> > Please visit IPC's web site (http://www.ipc.org) "On-Line Services"
> > section for additional information.
> > For technical support contact Hugo Scaramuzza at [log in to unmask] or
> > 847-509-9700 ext.312
> > ################################################################
> >
>
> ################################################################
> TechNet E-Mail Forum provided as a free service by IPC using LISTSERV 1.8c
> ################################################################
> To subscribe/unsubscribe, send a message to [log in to unmask] with
> following text in the body:
> To subscribe: SUBSCRIBE TechNet <your full name>
> To unsubscribe: SIGNOFF TechNet
> ################################################################
> Please visit IPC's web site (http://www.ipc.org) "On-Line Services"
> section for additional information.
> For technical support contact Hugo Scaramuzza at [log in to unmask] or
> 847-509-9700 ext.312
> ################################################################
>
################################################################
TechNet E-Mail Forum provided as a free service by IPC using LISTSERV 1.8c
################################################################
To subscribe/unsubscribe, send a message to [log in to unmask] with following text in the body:
To subscribe: SUBSCRIBE TechNet <your full name>
To unsubscribe: SIGNOFF TechNet
################################################################
Please visit IPC's web site (http://www.ipc.org) "On-Line Services" section for additional information.
For technical support contact Hugo Scaramuzza at [log in to unmask] or 847-509-9700 ext.312
################################################################
|
