Searching \ for '[PIC] to use ICSP or not for mass production' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: www.piclist.com/techref/microchip/devprogs.htm?key=icsp
Search entire site for: 'to use ICSP or not for mass production'.

Exact match. Not showing close matches.
PICList Thread
'[PIC] to use ICSP or not for mass production'
2005\12\14@211857 by Chen Xiao Fan

face
flavicon
face
I change the subject.

> -----Original Message-----
> From: Olin
> Sent: Wednesday, December 14, 2005 9:14 PM
> To: Microcontroller discussion list - Public.
> Subject: Re: [EE] bulk erase of 18F parts when running at 3.3V?
> can'tbedone?!
>
> Chen Xiao Fan wrote:
> > ICSP is anyway not a good solution for mass production.
> > We switched to off-line programming since last year.
>
> I disagree with this blanket statement.  ICSP has a number of
> advantages for production, although like any feature, it's not free.

Okay I agree with you that my statement is too general.
I think you are right to say ICSP has its place and it can
be used if it is carefully designed and it is allowed in the
design.

{Quote hidden}

Still I would like to share our experience.

We also believed ICSP is the way to go four years back
when flash PICs starts to getting cheaper. After three
years of using ICSP and lots of problem with Promate II
ICSP modules, we came to a conclusion that it is not so
cost effective as off-line programming using production
quality programmers (not in the sense of Microchip's
definition, Promate II and III are not of production
quality IMHO). There are some other potential
problems with ICSP as well (testing pins takes valuable
board space, pogo pins contact problems, Promate II/III
problems, etc).

For higher quantity (25kpcs per lot for Microchip QTP), it
is cheaper to go to Microchip or other programming house.
Microchip may also provide small quantity service (say
5k or 10k) now that competitors like Silicon Labs
offer this service at very low cost.

For low quantity mass production maybe ICSP can be a
good choice.

Anyway, this is only our experience and different
companies may have different cost structures and
different preferences.

Regards,
Xiaofan

2005\12\15@055617 by Gerhard Fiedler

picon face
Chen Xiao Fan wrote:

>> All in all, ICSP as part of the production test process can
>> be very cost effective if the circuit is designed for that
>> up front.

> For higher quantity (25kpcs per lot for Microchip QTP)

For some projects, batches of 100 is "production", whereas in other
projects that's just a "prototype run" :)

I guess it depends a lot on the numbers and the whole production process.

Gerhard

2005\12\15@074151 by olin piclist

face picon face
Chen Xiao Fan wrote:
> We also believed ICSP is the way to go four years back
> when flash PICs starts to getting cheaper. After three
> years of using ICSP and lots of problem with Promate II
> ICSP modules, we came to a conclusion that it is not so
> cost effective as off-line programming using production
> quality programmers (not in the sense of Microchip's
> definition, Promate II and III are not of production
> quality IMHO).

I'm curious what the problems were.  I know the Promates are rather
expensive but always thought they were robust, although I've never used one.
I've heard people complain that they are awfully klunky, require various
adapters, and until recently weren't scriptable, but I hadn't heard of them
being unreliable before.  Of course everyone should be using a ProProg
anyway http://www.embedinc.com/proprog ;-)

> There are some other potential
> problems with ICSP as well (testing pins takes valuable
> board space, pogo pins contact problems,

I guess it depends on what the other functional testing requirements are
already.  If you have to make a fixture with pogo pin pads for other reasons
anway, adding ICSP isn't a big deal incrementally.

> For low quantity mass production maybe ICSP can be a
> good choice.

I've got customers using it for high volumes (100K - 3M) too.  In this case
a small RF transmitter needs to be calibrated, and an IR LED and a few other
things checked.  Also a serial number needs to be written to the 10F202
controller.  That essentially requires a programmer anyway.  Writing the
program is a trivial addition given all that.

We're also using ICSP on many other products.  One of the benefits is that
the lag time from new code to new units produced with that code is much
lower.  It also allows for easily upgrading the code in existing units, at
least at the factory or in the lab.


******************************************************************
Embed Inc, Littleton Massachusetts, (978) 742-9014.  #1 PIC
consultant in 2004 program year.  http://www.embedinc.com/products

2005\12\15@191115 by Xiaofan Chen

face picon face
On 12/15/05, Olin Lathrop <spam_OUTolin_piclistTakeThisOuTspamembedinc.com> wrote:
>
> I'm curious what the problems were.  I know the Promates are rather
> expensive but always thought they were robust, although I've never used one.
> I've heard people complain that they are awfully klunky, require various
> adapters, and until recently weren't scriptable, but I hadn't heard of them
> being unreliable before.  Of course everyone should be using a ProProg
> anyway http://www.embedinc.com/proprog ;-)
>

We are still using Promate II for some old design (16C621A/622A and
16F819). It needs an ICSP module. The major problem with Promate II
is that the socket module often got contact problem. We just got the
base unit replaced early this year due to some failure but it is 5 years
old (I bought it in 2000). The test department bought the Promate III
but it is not compatible with DOS which the old test gig is using. So
I got this Promate III as my development programmer. It seems to
be good (better than ICD2 anyway) but I have encounter USB driver
stability problem and switched to RS232. However people in Microchip
Forum has reported reliability problem with Promate III ICSP and I
recommend your ProProg to him. Please refer to the following thread.
http://forum.microchip.com/tm.asp?m=125019

In our case, I was told that the production like off-line programming better
because it raises the throughput even though we need to buy programming
adapter which itself costs more than ProProg. However it makes sense
to utilize the capability of the gang programmer in-house.

>
> I've got customers using it for high volumes (100K - 3M) too.  In this case
> a small RF transmitter needs to be calibrated, and an IR LED and a few other
> things checked.  Also a serial number needs to be written to the 10F202
> controller.  That essentially requires a programmer anyway.  Writing the
> program is a trivial addition given all that.

I wish that our product got that higher quantity so that our
purchasing department
and we will have a good time negotiating the price. ;-) I think in that case,
Microchip will be very happy to provide SQTP (serial quick turn programming)
service and you get customized printing on the chip as well. I only have one
design which reaches 100k quantity per year. And I guess it is the only MCU
based design in the whole company which reach 100k quantity.

> We're also using ICSP on many other products.  One of the benefits is that
> the lag time from new code to new units produced with that code is much
> lower.  It also allows for easily upgrading the code in existing units, at
> least at the factory or in the lab.
>

Yes I agree with you that ICSP does have its merit and I like it at
the development
stage as well. To upgrade the code, we use bootloader in some other MCU
design.

Regards,
Xiaofan

2005\12\15@193347 by Andre Abelian

flavicon
face
Xiaofan,

Why do you worry about connection problem just call microchip they will
replace it for you. Do not forget whey are one of the best in the world.
I use promat 3 usb and all my devices are programmed in ICSP I never
had any problem with  ICSP or usb. Did you talk to them?

Andre

Xiaofan Chen wrote:

{Quote hidden}

2005\12\15@194304 by Harold Hallikainen

face picon face

> We are still using Promate II for some old design (16C621A/622A and
> 16F819). It needs an ICSP module. The major problem with Promate II
> is that the socket module often got contact problem. We just got the
> base unit replaced early this year due to some failure but it is 5 years
> old (I bought it in 2000). The test department bought the Promate III
> but it is not compatible with DOS which the old test gig is using. So
> I got this Promate III as my development programmer.


We got a Promate II several years ago while our Promate was out for
repair. I still have the Promate, but it doesn't do a whole lot any more.
Production uses the Promate II programming DIP chips out of board. Newer
stuff is SMT with in circuit programming with the ICD-2. On the Promate
and Promate II, we were able to do DOS batch files to drive the
programmer. The production people just put the chip in the socket, hit a
key to select which product they're doing, and wait for it to be done. No
digging through MPLAB!

Harold



--
FCC Rules Updated Daily at http://www.hallikainen.com

2005\12\15@200014 by olin piclist

face picon face
Harold Hallikainen wrote:
> Newer stuff is SMT with in circuit programming with the ICD-2.
> On the Promate and Promate II, we were able to do DOS batch files to
> drive the programmer. The production people just put the chip in the
> socket, hit a key to select which product they're doing, and wait for
> it to be done. No digging through MPLAB!

My ProProg (http://www.embedinc.com/proprog) can be operated like that too,
but it also has a button right on the unit.  The supplied software has
the -LOOP command line option which causes the same HEX file to be
programmed every time the button is pressed.


******************************************************************
Embed Inc, Littleton Massachusetts, (978) 742-9014.  #1 PIC
consultant in 2004 program year.  http://www.embedinc.com/products

2005\12\15@205959 by Xiaofan Chen

face picon face
On 12/16/05, Olin Lathrop <olin_piclistspamKILLspamembedinc.com> wrote:
> Harold Hallikainen wrote:
> > Newer stuff is SMT with in circuit programming with the ICD-2.
> > On the Promate and Promate II, we were able to do DOS batch files to
> > drive the programmer. The production people just put the chip in the
> > socket, hit a key to select which product they're doing, and wait for
> > it to be done. No digging through MPLAB!
>
> My ProProg (http://www.embedinc.com/proprog) can be operated like that too,
> but it also has a button right on the unit.  The supplied software has
> the -LOOP command line option which causes the same HEX file to be
> programmed every time the button is pressed.
>

Actually I mean real DOS, the 16bit DOS. Promate II supports DOS but not
Promate III. I do not think ProProg will work with real 16bit DOS.

Regards,
Xiaofan

2005\12\15@210524 by Xiaofan Chen

face picon face
On 12/16/05, Andre Abelian <.....aabelianKILLspamspam.....dnfcontrols.com> wrote:
> Xiaofan,
>
> Why do you worry about connection problem just call microchip they will
> replace it for you. Do not forget whey are one of the best in the world.
> I use promat 3 usb and all my devices are programmed in ICSP I never
> had any problem with  ICSP or usb. Did you talk to them?
>
> Andre
>

Yes Microchip replaced the Promate II base unit for us so it is getting better.
Their support is excellent.

As for Promate III, it does not support real DOS and I do not think they will
support DOS. I guess it should be better than Promate II in terms of
ICSP but our old test gig is using real DOS, the 16bit DOS. Promate II
does support 16bit DOS.

As for Promate III, the first day my computer reboots for no reason after
installing the usb driver (Win XP SP1, MPLAB 7.10 that time). I may want to
try again with new version of MPLAB but I am happy with the RS232.


Regards,
Xiaofan

2005\12\16@065634 by olin piclist

face picon face
Xiaofan Chen wrote:
> Actually I mean real DOS, the 16bit DOS. Promate II supports DOS but not
> Promate III. I do not think ProProg will work with real 16bit DOS.

You're absolutely right, and I can't imagine ever supporting DOS or other
prehistoric operating systems.  All my PIC programmer programs are 32 bit
executables intended to run from a command line.


******************************************************************
Embed Inc, Littleton Massachusetts, (978) 742-9014.  #1 PIC
consultant in 2004 program year.  http://www.embedinc.com/products

2005\12\16@073820 by Xiaofan Chen

face picon face
On 12/16/05, Olin Lathrop <EraseMEolin_piclistspam_OUTspamTakeThisOuTembedinc.com> wrote:
> Xiaofan Chen wrote:
> > Actually I mean real DOS, the 16bit DOS. Promate II supports DOS but not
> > Promate III. I do not think ProProg will work with real 16bit DOS.
>
> You're absolutely right, and I can't imagine ever supporting DOS or other
> prehistoric operating systems.  All my PIC programmer programs are 32 bit
> executables intended to run from a command line.
>

I can understand that. Our testing department starts to migrate to Windows
some time ago but there are still some DOS based test gigs.

Regards,
Xiaofan

2005\12\16@074759 by Wouter van Ooijen

face picon face
> Actually I mean real DOS, the 16bit DOS.

WISP.EXE will run on an 8088 with MSDOS, but its range of target PICs is
limited.

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


More... (looser matching)
- Last day of these posts
- In 2005 , 2006 only
- Today
- New search...