lv Serial Cable

[knc1]

I know the feeling, still haven't decided what I want to be when I grow up.

[knc1]

The photos displayed in-line here are thumbnails of the full image.
I am including links with each thumbnail so you can grab the full image if desired.
Note: If you "pay by the byte" - the full images are about 1Mbyte @ about 1,500 ppi.

I had a few questions in my mind about that "build a slip-in tab" for the lamp power socket described in the original thread on this subject.

Now that I got my Chatty Kathy to hold still for a photo session, here is what I found by taking a closer look.

The Kindle-3 (3g, keyboard) connector for the cover light power:

j14 top, full image

Removing the connector from the case/motherboard assemble reveals the two rows of four connector pads. These are the pads shown in-use in the lead photo of the old thread on this subject:

j14 pads, full image

I was able to get enough light on the board that there should not be any question as to which pads are connected to the "ground" plane.

Note that the two sets of four pads are different sizes (widths) and are on different centers.
Also note the two registration holes for the connector in the motherboard, these are significant.

The connector uses "grasshopper leg" spring contacts with the pads on the motherboard.
Holding the connector in alignment with the pads confirms that the contacts are spaced on the center lines of the "outside" row of pads.

j14 align, full image

Also significant in that image is the two little registration posts on the connector, which fit into the motherboard registration holes when mounted.

Getting an oblique side view of that connector shows that the "knees" of the "grasshopper leg" contact springs will only contact the front row of pads, just "inside" of a line between the registration posts/holes.


j14 contacts, full image


After a six month pause, next post in sequence here:
https://www.mobileread.com/forums/sho...3&postcount=23

[geekmaster]

The pins in the bottom connector are (from the bottom):

Power Enable (when shorted to GND, turns on power to top connector for LED book light)
GND
RxD
TxD

AFAIK the kindles route the signals so shorting adjacent pins does not hurt. Shorting RxD and TxD is just a loopback connection. Shorting RxD to GND is a "break" signal. Shorting GND to the bottom enable pin enables power to the top slot, which supplies power to the LED book light in the official amazon lighted case.

Some kindle models have a 1K resistor in series with RxD and TxD, which can cause problems if your USB serial adapter supplies too much current through a status LED that pulls up that signal line.

This is documented in other posts.

[knc1]

Originally Posted by geekmaster

This is documented in other posts.

That might well be so,
but it wasn't what is shown as the lead photo in the old thread on this subject.

I will continue to confirm these things for myself,
I quit believing what I read on the 'net when DARPA let it loose to the public.

[geekmaster]

Huh? I thought that EVERYTHING on the interwebs was all TRUE!

[knc1]

Originally Posted by geekmaster

Huh? I thought that EVERYTHING on the interwebs was all TRUE! ;-)

I thought that as soon as I made that post that I might have offended you or seemed very ungrateful for the information.

No offense intended and thanks for the info.

It is my own policy on these open forum sites to consider that I am writing for the least experienced user that might be reading the posts.

Which means that from time to time I offend the experts. Such is life.

But keeping in mind that the reader with a PhD in Computer Science might not be familar with electronics - I am taking the "old school path" here.

* Look up prior work
* Look up documentation
* Make a visual inspection
* ....

Which I am sure you recognize for what it is.
Other readers might not.

Again, thanks for the information.
WA8GUW

[geekmaster]

Originally Posted by knc1

Again, thanks for the information.
WA8GUW

No Troubles, Mate!

Regarding the pinout, that was from memory, which ain't what it used to be!

EDIT: Okay, you shamed me into doubting myself and actually looking it up instead of trusting my memory. I was right. Look here:

https://www.mobileread.com/forums/sho...10&postcount=2

What they identify as 1.8v is actually a pullup resistor floating an input at 1.8v, which enables power to the top connector when grounded. The amazon cover metal clips short the bottom two pins, which enables power to the top connector.

How dare I doubt myself?

[knc1]

Originally Posted by geekmaster

The pins in the bottom connector are (from the bottom):

Power Enable (when shorted to GND, turns on power to top connector for LED book light)
GND
RxD
TxD

I gimp'd that photo a bit so we could "see through" the masking:



Click for: j14 masked, full image


From this transformed photo, I expect that tomorrow I will find that the outside and inside rows of pads are indeed connect together one-to-one.

TP-32 Ground
TP-77 For testing that control signal

The Rx pin is clearly connected to that chip and what is most likely a pull-up resistor above.
The Tx pin may also be connected to that chip, and is connected to what is most likely a pull-up resistor above.

EDIT: Had a re-think of the above. If the SoC and this mystery chip are sharing the serial lines, those two resistors are in series with the serial line from the SoC to provide a bit of isolation between the two devices.

Those two little SMD thingies in the top left of the photo would be the pull-up resistors.
(Although SMD caps, diodes, inductors, etc. all look pretty much alike to the eye.)

That control lead on the pad and TP-77 might also be connected to that chip. It wouldn't need to be routed under the chip to reach a via, it already has a via on its path.

Which poses an interesting question of what the chip is and what it is doing on the console operator's serial port. ? ? ?

Is nothing sacred any longer?
The operator's console is wire-tapped?


As my eyes have aged to almost useless condition, digital photography has more than compensated for that loss. 1,500 pixels per inch make looking at this stuff easy.

[knc1]

After a six month delay, I now need to finish my K3 serial port adapter connector.

Looking closely at the picture of the J14 connector cover in the prior post in this photo sequence:
https://www.mobileread.com/forums/sho...6&postcount=17
It can be seen that it is a two part molding.

Lifting on the outside edge, the "spacer" part will swing in its pivot:


Click for: j14 half open, full image

When the "spacer" part is in an up-right position, it will lift out of the pivot point:


Click for: j14 open, full image

From the "spacer" part, the detailed dimensions may be taken and added to a sketch of the part. This will give the required dimensions of the mating tab to be made.

[geekmaster]

The K4 and K5 have 1K series resistors (not pull-ups) on the Rx and Tx lines. Perhaps the K3 does too.

[knc1]

Originally Posted by geekmaster

The K4 and K5 have 1K series resistors (not pull-ups) on the Rx and Tx lines. Perhaps the K3 does too.

Ah...

Originally Posted by March 3rd, 2012

EDIT: Had a re-think of the above. If the SoC and this mystery chip are sharing the serial lines, those two resistors are in series with the serial line from the SoC to provide a bit of isolation between the two devices.

That gimp'd photo shows the top layer buried traces.

[geekmaster]

The 1K resistor is why I needed a 22K pull up on my level shifter circuit, so a logic zero would have a low enough voltage after the voltage division. Those 1K resistors protect the serial port from overvoltage to some extent.

[knc1]

So after opening up the two part connector:
https://www.mobileread.com/forums/sho...1&postcount=24

Make a sketch of the connector cavity, the actual tab will need to be a bit smaller than the cavity for movement clearances.
(I.E: so you can slide it in and out.)

The dimensions shown are in inches.

Sorry, I had to compress the dxf file to get it past the MR attachment restrictions. You'll have to gunzip it.

Next, add the contact points to the sketch of the cavity, decide on what shape and size to make the mating insert tab.

[knc1]

With the connector cavity outlined, then the contacts are outlined.

We find that our tab needs to interface with 0.046 wide contacts on 0.060 centers.

These are "grasshopper leg" pressure contacts.
So next find the depth of the contact points.

[knc1]

And we find that the total depth of each grass hopper knee is 0.140 inch. With the contact point centered in the raised portion.

The contact point (center of knee) for the front pair is 0.170 inch in from the connector edge and for the rear pair is 0.310 inch in from the connector edge.

The knees are 0.030 inch high (in the 0.040 inch high slot).

Since the "inside, front" (shown right of center in the sketch) is GND, we will also want to stagger the engagement points on the tab so that GND gets connected before the CMOS Tx and Rx contacts.

The "outside, front" (shown at the right side of the sketch) is the light control sense lead, we will not be bringing that out on our connector tab.

For reference:
https://www.mobileread.com/forums/sho...3&postcount=23
And keep in mind that the sketch is of the "cover" part of the connector, shown "upside down" (board side up).
(Which will be its orientation when using the Kindle - since the Kindle is upside down for the photographs.)