LowPowerLab Forum

Hardware support => Projects => Topic started by: WhiteHare on January 08, 2017, 01:44:09 PM

Title: AB18x5 Breakout Board
Post by: WhiteHare on January 08, 2017, 01:44:09 PM
On a different thread (https://lowpowerlab.com/forum/low-power-techniques/stuck-in-listen-mode-listen-mode-bug-with-workaround/msg16974/#msg16974) I offered to design a simple breakout board for the AB18x5 RTC, so that anyone who wants to try it out will have a way to do so.  For those who don't already know, that AB18x5 RTC uses very little power, making it a good choice for battery based projects.

Attached is the first pass on the breakout board, which should fit easily into any 2.54mm pitch breadboard.  A few open issues:

1.  I wasn't sure if connecting the oscillator directly to pins 15 and 16 might somehow (?) impair its performance.  Also, it's not really necessary.  So, I left them disconnected, but I added some jumper pads so you can connect them after-the-fact if you want to.

2.  I didn't include a GND copper pour.  Not sure if in this case it would help or cause problems, so I left it out.  However, if someone thinks it should have one, I could add it.

3.  The AB18x5 comes in two different versions: one uses I2C, and the other uses SPI.  Rather than do two different boards (with two different silk screens to label the pins), I just numbered the pins.  That way you can use the same board for either chip.

I'll wait a bit for comments.  Then I'll finalize the design and ship off the files to a fab.  After I receive the board and verify that it works, I'll then post the gerber files.

Title: Re: AB18x5 Breakout Board
Post by: TomWS on January 08, 2017, 05:15:34 PM
A square pad on pin 9 is a bit confusing, especially on the bottom side where there are no pin labels.

EDIT: Also there is no ground signal to the center pad and, on the AB1815 it is the ONLY ground pin.  Consequently this makes 17 pins on the device and you only have 16 on the fanout.  Also if you look at the datasheet for recommend layout, they suggest four small square solder paste areas on the larger pad (this is to keep your chip from floating away when you reflow).  You don't show the cream layer so can't tell from what you have in the design - the same is true for the LGA pads, the cream layer (and stencil) should be smaller than the pad itself.

Just my $0.02...

Tom
Title: Re: AB18x5 Breakout Board
Post by: WhiteHare on January 09, 2017, 12:58:21 PM
Thanks!  That's great feedback.

How about Version 2 below?  I used a via to connect the pad under the AB18x5 to a new copper pour on the bottom, which in turn connects to "Pin 17" (in actuality Pin 15 now) on the header.
Title: Re: AB18x5 Breakout Board
Post by: WhiteHare on January 09, 2017, 01:02:36 PM
More attachments below, including the gerber files.

As for the cream layer, I think I'll try the painting method that you and Joe have developed instead, since this is quantity 1 production.
Title: Re: AB18x5 Breakout Board
Post by: ChemE on January 09, 2017, 08:45:00 PM
Might as well label the pins on the bottom of the board too since there is room.
Title: Re: AB18x5 Breakout Board
Post by: perky on January 09, 2017, 09:42:10 PM
I think the AF pin might require a 47pF cap to Vss if you want this to be a purely digital breakout board.
Mark.
Title: Re: AB18x5 Breakout Board
Post by: WhiteHare on January 10, 2017, 11:38:43 AM
Thanks, guys!  Here's version #3:

Title: Re: AB18x5 Breakout Board
Post by: WhiteHare on January 10, 2017, 11:41:46 AM
Attached are the version 3 gerber files.

Anyone have any more comments/suggestions?
Title: Re: AB18x5 Breakout Board
Post by: WhiteHare on January 11, 2017, 10:44:43 AM
As there have been no further comments, I've sent the files to the fab.

As for the components BOM, the oscillator is this one:  https://www.digikey.com/product-detail/en/abracon-llc/ABS07-120-32.768KHZ-T/535-11937-1-ND/3724046

and the capacitor is an 0603 size.  In general, I default to X7R for all capacitors, unless there's good reason not to.
Title: Re: AB18x5 Breakout Board
Post by: WhiteHare on January 22, 2017, 05:53:38 PM
According to package tracking, I should be receiving my AB1815 breakout boards tomorrow.

@ChemE: Have you found a worthwhile library for the AB1815, or were you planning to build your own from scratch?

At the very least I'd prefer to start with some simple known good sketches so as to test that the hardware is working.
Title: Re: AB18x5 Breakout Board
Post by: ChemE on January 22, 2017, 09:40:19 PM
I've been playing with my RFM69CW code but I'll have a look tomorrow to see what I can find.  I was absolutely planning on writing my own 'cause that's just the way I roll, but I do like to start from something functional.  Glad to hear they are en route, enjoy your jellybeans!  Lemme know how much you need for one (one is slated to come to me right?).

EDIT: Possible starting points linked below.  Some are vastly more complete than others.
https://github.com/Snowda/AB1805 - minimal but easily taken in
https://github.com/dantudose/AB1805 - in between to minimal
https://github.com/nigelb/Arduino-AB1815 - seems quite complete

I don't see in any of these anything allowing for us to store data in RAM or use the chip as a WDT.  These all just seem to focus on using it as a RTC.
Title: Re: AB18x5 Breakout Board
Post by: WhiteHare on January 23, 2017, 03:24:34 AM
LOL.  Well, they cost so little that I don't mind mailing you one for free.  In fact, I thought you had already ordered yours.  But before this gets out of hand, were you planning to solder on the parts yourself, or were you hoping I'd do that too?  I wouldn't want you to miss out on the fun part.   ;)
Title: Re: AB18x5 Breakout Board
Post by: ChemE on January 23, 2017, 08:00:36 AM
LOL, I thought part of what I was paying for was your advanced soldering expertise.  I don't plan to learn how to hand-solder 0.5mm pitch stuff.  I'm going to let a souped up toaster oven handle that for me so I can concentrate on code.
Title: Re: AB18x5 Breakout Board
Post by: WhiteHare on January 23, 2017, 08:48:32 AM
In that case, I'll simply wish you good luck.  The files are already posted if you ever want to order one for your toaster oven.
Title: Re: AB18x5 Breakout Board
Post by: WhiteHare on January 23, 2017, 10:18:35 PM
Attached photo shows how it turned out.  Some of the soldered connections look a bit iffy, but it's really hard to tell for sure just by looking at it.  If it doesn't respond as expected, then I'll attempt some in situ remediation.
Title: Re: AB18x5 Breakout Board
Post by: ChemE on January 23, 2017, 10:52:52 PM
Very curious to see if it functions.
Title: Re: AB18x5 Breakout Board
Post by: WhiteHare on January 24, 2017, 05:14:17 AM
The I2C version (AB1805) might be slightly easier to solder because it does not rely on the bottom pad for an electrical connection, as the SPI version does.  That's because on AB1805, the thermal pad appears to be redundant with pin 9 (both being labelled a VSS pin).  On the SPI version (which is the one shown in the photo), if that thermal pad connection is not electrically sound, then it's a complete re-do, as I see no way to remediate it short of complete removal and starting over.  Fortunately, the same board would work for either chip, except that the silkscreen on the bottom would be wrong.  That's partly why I left the pin numbers on the top silkscreen.

Anyhow, I'd say this chip is much harder to solder than the BQ25504 chip of the same size.  That's because the pin connections on the BQ25504 go up the sides of the chip a bit more, and so it's easier to see if each soldered connection is a good one.  On the AB1815, it's just very hard to see.  If I had a photo of what a properly soldered AB1815 looks like, it might be easier to judge.

Title: Re: AB18x5 Breakout Board
Post by: WhiteHare on January 24, 2017, 08:32:58 AM
Very curious to see if it functions.

Me too!  I'm supposing that if I can read and write to the Seconds register (0x01), then that's a good enough test, at least for now.
Title: Re: AB18x5 Breakout Board
Post by: WhiteHare on January 24, 2017, 12:47:00 PM
Just now looking at the rfm69 chip to see how it was soldered onto the HopeRF module, and it's in the same league: just looking at it, I don't see a way to tell whether it was adequately soldered or not.  I know that it was correctly soldered, though, because it functions correctly.  Maybe the *only* way to tell is by using the chips and seeing if they work as expected.

I think this chip may be just too difficult for me.  I'm going to shelf it and maybe revisit it at a later date.
Title: Re: AB18x5 Breakout Board
Post by: WhiteHare on January 24, 2017, 03:04:49 PM
In case some future reader of this project has interest in a breakout board for the AB1805 (I2C), instead of the AB1815 (SPI), attached are files for the same board, but with bottom silkscreen labels appropriate for the AB1805.
Title: Re: AB18x5 Breakout Board
Post by: ChemE on January 24, 2017, 05:52:58 PM
I'm not even sure how you would go about soldering the ground pad with a hand-held iron.  Bummer that you've hit a snag.  If I do get a working reflow oven knocked together I can send your board through if you'd like.  That project isn't on the immediate horizon but I plan on doing it this year. 
Title: Re: AB18x5 Breakout Board
Post by: WhiteHare on January 24, 2017, 06:35:33 PM
Actually, I am using a reflow/toaster oven for this project.  I only recently made the switch. 

The lone AB1815 SPI library on github doesn't come with an example script.  However, both of the I2C AB1805 github libraries for the AB1805 do come with example scripts.

So... I soldered an AB1805 onto one of the boards, ran the dantudose I2c example script, and voila, it produced the expected output:

Code: [Select]
12/01/16 16:56:01
12/01/16 16:56:02
12/01/16 16:56:03
12/01/16 16:56:04
12/01/16 16:56:05
12/01/16 16:56:06
12/01/16 16:56:07
12/01/16 16:56:08
12/01/16 16:56:09
12/01/16 16:56:10
12/01/16 16:56:11
12/01/16 16:56:12
12/01/16 16:56:13
12/01/16 16:56:14
12/01/16 16:56:15
12/01/16 16:56:16
12/01/16 16:56:17
12/01/16 16:56:18
12/01/16 16:56:19
12/01/16 16:56:20
12/01/16 16:56:21

Although one could do more extensive tests, I think that's pretty solid initial validation that the PCB Breakout board works. 

Project complete.   :)
Title: Re: AB18x5 Breakout Board
Post by: perky on January 24, 2017, 06:50:02 PM
Get yourself an adjustable heat gun, like this (UK only unfortunately):
https://www.machinemart.co.uk/p/evolution-hdg200-2000w-heat-gun/

If you get one make sure it has continually variable temperature, and continually variable speed down to a very low level. There are many with variable temperature (50degC - 650degC are common), but not so many with variable speed, they tend to come in either two or three speed settings, you'll need one that's very low speed for soldering delicate components. Oh and a reducing nozzle but they usually come with the gun.

This is probably the most useful tool I have in my toolkit, I can solder just about anything with this within reason ;)

Mark
Title: Re: AB18x5 Breakout Board
Post by: TomWS on January 24, 2017, 09:33:18 PM
This https://www.amazon.com/gp/product/B00ITMPQS2 works well if you want to unsolder a chip.  Resoldering is another story - the forced air will blow most components off their pads. Use this to remove a device's defective solder operation, fix the pads, etc, paint new paste, toaster oven reflow to repair.

Uh, you probably only get two retries before you've broken something... depending on how good your toaster heat profile is and whether your device is MSL3 and it's 'pink'...

Tom
Title: Re: AB18x5 Breakout Board
Post by: perky on January 25, 2017, 06:31:05 AM
There are ways around the air flow blowing components off pads, for example using a piece of gauze with different grid sizes can control it reasonably well. The hot air rework stations can go quite low in air flow but I've found they're considerably more expensive than heat guns.
Title: Re: AB18x5 Breakout Board
Post by: WhiteHare on January 25, 2017, 07:10:55 AM
I later discovered that some of my initial "floating away" problems were probably just me being too eager to remove the pcb from the oven.  The sliding metal tray that the pcb sits on tips down at an angle if I pull it out too far, and that caused some dislocation because everything was still molten.  Now I open the oven door and wait two minutes before moving anything.  Mea Culpa, and obvious in retrospect, so just a warning for other reflow oven beginners like me who might try this.
Title: Re: AB18x5 Breakout Board
Post by: TomWS on January 25, 2017, 08:55:16 AM
Now I open the oven door and wait two minutes before moving anything.
It takes my oven about 5 minutes to cool down to a touchable temperature after opening the door.  I use that time to start a cup of coffee.

Tom
Title: Re: AB18x5 Breakout Board
Post by: WhiteHare on January 25, 2017, 01:27:36 PM
The step that, at least for me, stands out as by far the most difficult in the entire process is the initial chip placement.  By that I mean the chip has to be set reasonably close to the ideal even if surface tension helps fine tune it during the reflow.  I'm  not yet sure just how close "reasonably close" needs to be, but... pretty close.   

I have fairly steady hands (probably better than average), and nonetheless my first placement would often not be close enough (based on my guessing of what was close enough).  So, I would then nudge it with a pair of tweezers until it finally ended up where I thought was "reasonably close".  However, it's very easy to either overshoot or undershoot or accidentally rotate it, all of which leads to more nudging.  Now, granted, eventually it gets to where it needs to be, but it can be frustrating at times.  Any of this sound familiar?

So, if there are any tips on that part of the process , I'd be interested to hear.  In comparison, the rest of it is relatively straight forward.  I'm starting to think that making some kind of simple but small x-y table with a fixed vacuum chip holder above it that simply plunges in the z-axis might be the way to go.  That way one could literally dial in the chip placement and remove imperfect hand motions from the equation.  Yet, I haven't read of others doing that, so there must be easier ways.
Title: Re: AB18x5 Breakout Board
Post by: WhiteHare on January 25, 2017, 05:04:35 PM
This is probably my last post on this thread, but I thought it might be helpful for others to see what a functioning soldered connection looks like on an AB1805 (see attached).  There are some excess solder balls near the corner of the chip that I  would have cleaned off if it hadn't worked, but since so far all appears to be working correctly, I'm leaving them there for now.  i.e. "if it isn't broken, don't fix it."  In any case, the solder that's where it counts, namely on the pads, is more visibly prominant and correct looking (at least to my eye) than in the photo that I posted earlier of my first attempt with the AB1815, where the connections looked more iffy.  I think going forward this is roughly about what I would be shooting for.

That said, I'd be curious as to how this compares to what others aim for in their soldered connections on this particular AB18x5 chip.
Title: Re: AB18x5 Breakout Board
Post by: perky on January 25, 2017, 05:27:44 PM
Good work WhiteHare!

BTW I think you've been way over-generous with the solder paste. That's probably the reason for the solder balls, but like you say if it ain't broke..!

Mark.
Title: Re: AB18x5 Breakout Board
Post by: TomWS on January 25, 2017, 05:57:00 PM
I agree with Mark about too much solder, but the other thing I see is there is solder outside of the device pads meaning your PCB pads are probably not stopped down enough.  On a part this small it is absolutely vital that all of the pads present an identical surface to each of the device's pads so that the surface tension pulls each pad equally, thereby aligning it.  You can do this if you keep the exposed pad not too much bigger than the chip pad and certainly don't have any 'wiring' exposed as this will tend to pull the chip off of its pads.

Tom
Title: Re: AB18x5 Breakout Board
Post by: perky on January 25, 2017, 08:03:19 PM
Indeed. But we don't want to discourage you from doing this stuff, it's admirable that you've done what you've done already so keep up the good work! ;)

Mark.
Title: Re: AB18x5 Breakout Board
Post by: Neko on April 04, 2017, 09:30:48 AM
Can I get one of these?
Title: Re: AB18x5 Breakout Board
Post by: grafisoft on February 14, 2019, 02:09:22 PM
I am interesting in use this RTC low power to design a IoT node, and i was finding a way to power management with RTC.

In case that i want use the RTC in VCC Power Switched application, What would be the transistor/mosfet that i could use? I need a transistor that can support 2A but with the minimun current consumption. For this application, the pull-up resistor have be on the order of megaohms.

So, anybody could recommend me a component part? or any other solution that it can manage a power switch.

Thanks
Regards,
Title: Re: AB18x5 Breakout Board
Post by: TomWS on February 14, 2019, 06:33:03 PM
In case that i want use the RTC in VCC Power Switched application, What would be the transistor/mosfet that i could use? I need a transistor that can support 2A but with the minimun current consumption. For this application, the pull-up resistor have be on the order of megaohms.

So, anybody could recommend me a component part? or any other solution that it can manage a power switch.
The Si2365 or practically any low Vgsth P-FET would work well in this application, using the transistor configured as in the AB18x5 datasheet section 3.2, VCC Power Switched.  I wouldn't go too much higher than 4-5M, though as the leakage current of the PSW pin might cause problems.

If you don't need tight timing, then the TI TPL5110 breakout that LPL offers might work for you.

Tom
Title: Re: AB18x5 Breakout Board
Post by: grafisoft on February 16, 2019, 07:04:03 AM
The Si2365 or practically any low Vgsth P-FET would work well in this application, using the transistor configured as in the AB18x5 datasheet section 3.2, VCC Power Switched.  I wouldn't go too much higher than 4-5M, though as the leakage current of the PSW pin might cause problems.

If you don't need tight timing, then the TI TPL5110 breakout that LPL offers might work for you.

Tom

Thanks Tom,

I will check your recomendations. The TPL5110 have a resistor time settings and it is the reason for i have discard this component.

Thanks.

Regards,
Title: Re: AB18x5 Breakout Board
Post by: TomWS on February 16, 2019, 09:03:09 AM
One thing to note is that the Drain/Gate resistor only supplies current when the Radio etc is turned on, which, presumably, is a short amount of time.  The rest of the time there is no current flowing through it so above 100K might be diminishing returns...