>
> Hi, Roman
>
> > * Tube is held vertically in the machine, so chips are gravity
> > fed out the bottom one by one.
>
>     45 degress can give you better results. Some tubes have enough space for
> the chip to jam if held vertical.
>
> > * The chip can only fall one chip length, vertically, into
> > a very simple jig, with a solenoid "gate" at the bottom stopping
> > it falling any further.
>
>     Should work fine...
>
> > * The chip is secure enough in the jig that a second solenoid
> > clamps 5 sharp "pins" onto the appropriate legs, 2 on one
> > side of the chip and 3 on the other, giving Vdd,Vss,Vpp,SDA,SCL.
> > This is only 2 simple mechanical levers and one solenoid.
>
>     A motor and spring loaded "bed of nails" contacts will give more
> reliable connections without the impact of the solenoid.
>
> > * Programming is automatic, and after programming is complete
> > the pins are released, the bottom solenoid "gate" opens and the
> > chip simply falls into a tray, finished. The gate solenoid is
> > fast with an opto sensor, to eject only one chip. Another and
> > probably better option is a simple side ejector mechanism that
> > throws the chip out one side, and gravity feeds in the next chip.
>
>     You can use another tube and a second solenoid to make sure just one
> chip fall out.
>
> > I think this device could be built at quite low cost, I have
> > a few solenoids in my junk box. Each tube of fresh PICs is simply
> > inserted, press the start button, and voila 30 programmed chips
> > sitting in the tray.
> > Any suggestions?? :o)
>
>     More ??? :-) Please send us some pictures when you are done. I am
> already sick and tired of seing my production personal programming some
> 12c508's one by one...

>Recently we've been programming PICs by the 100, which
>is a PITA with a picstart. :o)
>
>I know production programmers are available, but since
>the main chips we use (16c505) program in a few seconds,
>even using a gang programmer will not save much time
>because of the time to insert 10 chips into 10 sockets,
>then remove them again. There is just as much labour
>needed and not much time saved.
>
>I have an idea for a auto-feeder programmer I would like
>to bounce off the list. :o)
>
>* The DIP chips are supplied in a standard plastic "tube".
>
>* Tube is held vertically in the machine, so chips are gravity
>fed out the bottom one by one.
>
>* The chip can only fall one chip length, vertically, into
>a very simple jig, with a solenoid "gate" at the bottom stopping
>it falling any further.
>
>* The chip is secure enough in the jig that a second solenoid
>clamps 5 sharp "pins" onto the appropriate legs, 2 on one
>side of the chip and 3 on the other, giving Vdd,Vss,Vpp,SDA,SCL.
>This is only 2 simple mechanical levers and one solenoid.
>
>* Programming is automatic, and after programming is complete
>the pins are released, the bottom solenoid "gate" opens and the
>chip simply falls into a tray, finished. The gate solenoid is
>fast with an opto sensor, to eject only one chip. Another and
>probably better option is a simple side ejector mechanism that
>throws the chip out one side, and gravity feeds in the next chip.
>
>* If a program error occurs, the units beeps etc, and operator
>removes the suspect chip.
>
>I think this device could be built at quite low cost, I have
>a few solenoids in my junk box. Each tube of fresh PICs is simply
>inserted, press the start button, and voila 30 programmed chips
>sitting in the tray.
>Any suggestions?? :o)
>-Roman
>
>--
>http://www.piclist.com hint: The list server can filter out subtopics
>(like ads or off topics) for you. See http://www.piclist.com/#topics
>
>
>
>
>
>
>

>
> Awhile ago someone else on the list drew up plans for a simple
> programmer like that, except it was based on a rotating wheel.
>
> The wheel is slotted so the chip falls into it with pins on either side.
>  It spins in the direction opposite the pins so the pins can run by a
> set of programming contacts, past a crusher and into another tube (or a
> tray, or whatever)
>
> It wouldn't be hard to adapt that with a gate (or even a moving ramp)
> that puts the chip at the bottom into the good or bad bin.
>
> Furthermore a wheel can hold and program several chips at once - load 6
> chips, move them into the 6 programmer positions, program, spit them all
> out while loading the next six.
>
> I suspect a rotating wheel is going to be easier (mechanically) to
> manage that a moving carriage.  A simple motor with opto sensing will
> tell you what position the wheel is in.
>
> You add another plate to the wheel to increase the thickness for wide
> dips, or (better yet) create different wheels for different chips (to
> account for chip length).
>
> This method would reauire only one moving part and one motor at its
> simplest.  Add a solenoid to gate the outgoing chips into proper bins.

> Here comes the ASCII art, let's hope it works.
>
> Chips fall in here
>                    |
>                    |
>               /---------\
>              /     |     \
>             /      |      \
>             |      |      |
>             |====     ====|   <-- crusher/programmer goes here
>             |      |      |
>             \      |      /
>    ^         \     |     /
>    |          \---------/
> Direction          ^
> of rotation        |
>                Chips fall
>                out here
>
> two exit slots could be used so that the good/bad testing could
> easily be done, with the addition of a single solenoid.  I have shown
> a 4-slot machine, that does 1 PIC at a time.  It could easily be
> compressed to a 1-slot machine, or expanded to a 4x6 slot machine,
> where PICs fall in 6 times, then there are 6 crushers, and it then
> tristates all crushers and drops the PICs out sequentially into the
> proper bin, while moving the next set of PICs into position, and
> setting up loading the set after that into its slots.
>
> This technique, in its basic form, takes 1 stepper, and no solenoids.
>  In its advanced form, it takes 1 stepper and 1 solenoid.  In either
> form, it can be set to do gang programming, purely using the addition
> of slots and crushers.

> > The crusher is a non-moving or sprung component.  The PIC moves
> > past it, which squishes the pins against the wheel.  The wheel
> > stops at the crusher position while the PIC is programmed.
>
> This is where the wheel idea has problems. The crush
> pressure is quite considerable, +4kg from the side, but
> to push the chip into a v-shaped groove to crush is still
> over 3kg measured by my crude test. Assuming your wheel
> is 2 inch radius this is 7lbs @ 2 inches, give it 100%
> overhead for stepper reliability and you get 28 lb @ 1
> inch or 448 oz/in stepper motor needed. You will need
> gearing on the stepper, probably about 10:1 even with
> a larger stepper. Using rubber rollers on the side of
> the main wheel reduces this by maybe half, but still
> requires a large stepper with gearing, ie toothed belt
> + 2 pulleys.

> Roman Black wrote:
>
> > I have an idea for a auto-feeder programmer I would like
> > to bounce off the list. :o)
>
> How about something cheap like this.
>
>
> --
> Best regards
>
> Tony
>
> mICros
> http://www.bubblesoftonline.com
> .....salesKILLspam.....bubblesoftonline.com
>

>Brendan Moran wrote:
>
>
>
>>Once again, I think you've missed something.  I was suggesting again
>>that you use the .1" edge connector as the crusher.  Place an opto
>>sensor near the wheel, and stop it when the slot gets to the crusher
>>point, where programming will occur.
>>
>>
>
>My apologies, I did miss something! :o) I had imagined
>your wheel with the PIC mounted axially, ie around the
>outside which made me think about problems lining up the
>contacts and the crush pressure.
>
>Now you have clarified that the PICs are radially mounted,
>like a "spoke" it sounds like a great idea. Feeding should
>be fully reliable, as only one PIC can fall into the slot.
>Then contact positioning will be reliable as all the contacts
>are moved into range simultaneously. Thanks for the picture.
>:o)
>
>I'm almost converted to the "wheel" design now, shame I went
>out and bought solenoids today. :o) The only thing that
>still puts me off the wheel idea is the amount of torque
>required. I have a car windscreen wiper motor which puts
>out mega torque at about 1 rev/second, this might be good
>to turn the wheel directly. Deserves a think anyway! :o)
>-Roman
>
>--
>http://www.piclist.com#nomail Going offline? Don't AutoReply us!
>email spamBeGonelistservspamBeGonemitvma.mit.edu with SET PICList DIGEST in the body
>
>
>
>
>
>
>

> Now you have clarified that the PICs are radially mounted,
> like a "spoke" it sounds like a great idea. Feeding should
> be fully reliable, as only one PIC can fall into the slot.
> Then contact positioning will be reliable as all the contacts
> are moved into range simultaneously. Thanks for the picture.
> :o)
>
> I'm almost converted to the "wheel" design now, shame I went
> out and bought solenoids today. :o) The only thing that
> still puts me off the wheel idea is the amount of torque
> required. I have a car windscreen wiper motor which puts
> out mega torque at about 1 rev/second, this might be good
> to turn the wheel directly. Deserves a think anyway! :o)