Jump to content
43oh

Recommended Posts

For the analog components that are there on things like the Educational BoosterPack, a simple resistor ladder divider should work fine to scale down the voltage.  It would be nice to at least have an option like that.  I can give you an op-amp circuit that will scale 3.3V down to 1.8V.  For me it'd be really handy.

 

The cape identification EEPROM would also be very handy for allocating the pins and configuring drivers.

Cool, I wouldn't mind seeing the op amp circuit you have in mind!  I understand the cape EEPROM et al, but I'm wondering what use it is here ... since the pins & drivers may vary widely based on what kind of boosterpack you install.  I suppose it won't hurt if I can fit it in though.  Could at least move all the pins' defaults to GPIO when the .dtbo gets loaded.

Link to post
Share on other sites
  • Replies 51
  • Created
  • Last Reply

Top Posters In This Topic

Top Posters In This Topic

Popular Posts

I haven't read this full thread, but I did get a board for a few minutes today to play around. I haven't yet understood how everything is wired.  It seems like the analog pins are not yet connected on

I got 'em!   I must say, for a chinese PCB fabber Elecrow is freaking fast.

Touched up & added some images:   ( http://spirilis.net/junk/beaglebone/bbblpv10/bbblpv10_top.png )   ( http://spirilis.net/junk/beaglebone/bbblpv10/bbblpv10_bottom.png )   I am going to

Posted Images

Cool, I wouldn't mind seeing the op amp circuit you have in mind!  I understand the cape EEPROM et al, but I'm wondering what use it is here ... since the pins & drivers may vary widely based on what kind of boosterpack you install.  I suppose it won't hurt if I can fit it in though.  Could at least move all the pins' defaults to GPIO when the .dtbo gets loaded.

I think you've got a good point on the EEPROM- some BP's need SPI, some I2C, some just GPIO.  You could certainly add a footprint for the device that a user could set if they're making a more permanent project though - so at least the option is there.  It'd be interesting to see if you could add a cheap MCU and some jumpers.  The MCU would have to emulate the EEPROM and respond based on the jumpers that are set.  That's a bit more complicated than a simple adapter though :)

Link to post
Share on other sites

I think you've got a good point on the EEPROM- some BP's need SPI, some I2C, some just GPIO.  You could certainly add a footprint for the device that a user could set if they're making a more permanent project though - so at least the option is there.  It'd be interesting to see if you could add a cheap MCU and some jumpers.  The MCU would have to emulate the EEPROM and respond based on the jumpers that are set.  That's a bit more complicated than a simple adapter though :)

Yeah I think if I can stuff one in (the routing was a bit of a rats nest for the v1.0 version so I'll see how easy it would be to fit, especially with some analog adapter circuitry stuffed in) the user could either take it or leave it.  Need to pop some jumpers or dip-switches to set the EEPROM address though.

Link to post
Share on other sites

For the analog components that are there on things like the Educational BoosterPack, a simple resistor ladder divider should work fine to scale down the voltage.  It would be nice to at least have an option like that.  I can give you an op-amp circuit that will scale 3.3V down to 1.8V.  For me it'd be really handy.

 

The cape identification EEPROM would also be very handy for allocating the pins and configuring drivers.

@@jkridner Can you send me the opamp circuit you have in mind for this?  I'd like to get started on it soon (if it's a TI chip, I'll sample it and make a test board first...)

Thanks!

Link to post
Share on other sites

Reposting what Jason posted on IRC-

http://electrotap.com/2004/11/01/op-amps-part-1-scaling-sensor-voltages/

Part 2 is important: http://electrotap.com/2005/01/11/op-amps-part-2-non-inverting-op-amps-and-voltage-followers/

 

My take so far:

 

A suitable 3.3V-to-1.8V adapter that is agnostic to source impedance of the signal would probably be a 3.3V (rail-to-rail I guess) Voltage Follower op-amp circuit whose output then passes through an accurate (e.g. 1% or 0.5% tolerance) but reasonably strong voltage divider, e.g. 10K+12K for a 3.3-to-1.8V shift.  Sitara ADC shouldn't have any trouble reading that.

 

So here's my plan...

 

I don't think I can reasonably fit the circuits into my existing LP board b/c the routing for the rest of the digital pins gets out of hand, I only have 2 layers to work with (for a reasonable cost from Elecrow et al, plus my PCB software only allows 2 signal layers unless I pay more), so I am going to do a hybrid approach...

BBB digital-only LP cape with some pins broken out in the middle for the AIN/AGND stuff, then a special "Analog Adapter" boosterpack that will supply the J4 ADC pins as well as introduce a long set of jumpers to allow you to manually connect J1 pins to J4 (so ADC inputs on the left side of traditional 20-pin MSP430 BoosterPacks can function).

 

The goal will be to fit all of that in a 5cm-by-5cm footprint so I can use the cheapest PCB fab option for this "Analog Adapter" board.  Mainly rides on how big the op-amp ICs will be.  Also gotta shop around for an opamp, probably sample it from TI and build out a toner-transfer board to test it.

 

But at the end of the day, it should make a pretty kick-ass accurate analog solution for the boosterpacks to work on the BB!

Link to post
Share on other sites

Hmm, doing some shopping for op-amps...

 

LMV844 seems decent: http://www.ti.com/product/lmv844

It's about $3.66/pc in qty 10 from Mouser.  4-channel, so each board would need 2 of these.  Available in SOIC-14 or TSSOP-14.  Rail-to-rail, appears to have good noise characteristics (I am a noob at this stuff but just reading the numbers, Sitara ADC SNR is 70dB and this chip's CMRR is 112dB, seems good) with plenty of bandwidth so I assume it can fully support signals that test the limits of the Sitara's 200Ksps ADC.

 

Still looking though.  Lots of specs and prices to sift through.

Also think I should invest in Op Amps for Everyone 4th ed. someday to school myself...

Link to post
Share on other sites

TI or someone has a free PDF on opamps. I am pretty sure it is TI, but you could ask oPossum, I am pretty sure he'd know if no one else. I'd give you the link myself, except I dotn remember the dahmed name of the pdf :/

 

By this, its a kind of fundamentals "book" as  I recall.

Link to post
Share on other sites

Op amps for everyone PDF ( over 500 pages ) http://www.ti.com/lit/an/slod006b/slod006b.pdf

HANDBOOK OF OPERATIONAL AMPLIFIER APPLICATIONS http://www.ti.com/lit/an/sboa092a/sboa092a.pdf
 
oooo just realized the last one is by bob pease ( rip ). As for the first link im pretty sure it is legit since TI seems to be hosting the download. Anyhow theres tons more out there too all I did was search google for "TI opamps PDF"
Link to post
Share on other sites

Ok now I think I have a straight set of search params...

 

http://www.ti.com/lsds/ti/amplifiers-linear/precision-amplifier-products.page

Supply Voltage Min set at 3V, # channels 4, Slew Rate at least 1V/uS, CMRR at least 80dB, sort by price.

 

LMV554 in TSSOP-14, $1.32/pc in qty 10.  Concern here is load capacitance, 100pF is a listed concern, but the ADC sampling capacitance on the Sitara is listed 5.5pF.  So I'm not really worried.

LMV654 in TSSOP-14, $1.43/pc in qty 10.  I like this one better, they mention very low THD at output loads as low as 2kohm and I was thinking of 22kohm for my output load.  Not much more money either, 12MHz bandwidth vs. 3MHz for the previous one.

LMV774 in TSSOP-14, $1.82/pc in qty 10.  Another decent one I guess (3.5MHz gain bandwidth product).

 

Skipping ahead to the first SOIC-14...

OPA4171 in SOIC-14, $1.82/pc in qty 10.  3MHz bandwidth, sounds like a good overall opamp.  No immediate concerns from reading through the datasheet (with the shoestring amount of info I know about these things).

 

So, LMV654 or OPA4171.  Comes down to higher precision vs. ease-of-soldering (but less board space too).  Guess I'll have to see.

 

I'll sample both of those and maybe sample the single & dual-opamp versions too for my own projects & experiments later.

 

edit: stuffed in OPA4171, OPA171, LMV654, LMV651, PGA112 into the cart.  Curious about the PGA112 (binary-scale programmable gain amplifier) for rolling my own thermocouple amplifier or toying with the idea anyhow.  Wouldn't mind doing that by hand instead of paying Maxim ~$8 for the MAX31855....

Link to post
Share on other sites

I guess the LMV654 is the better choice simply owing to the fact that its bandwidth is higher, and after reading Bruce Carter's book most of the way through I understand you need a *lot* of headroom in bandwidth to get accurate waveforms through.  At ~200KHz (though it's really 100KHz we care about since the Sitara ADC is 200Ksps) the open loop gain is around ~40dB, so that appears to be plenty of headroom.

 

Accuracy at the rails might be slightly off (up to ~60-75mV at the rails with Rload=10K), but I think we can live with it... and I might consider upping the output impedance to using 100K+120K resistors instead.  Or 50K+60K if suitable sizes are available.

Link to post
Share on other sites
  • 2 weeks later...

Ok guys, finally making progress here.

 

I got the digital cape ready to roll.  BBB LP v3.0 I'm calling it... It should clear up the conflicting I/O ports, and this time the analog stuff has been ditched in favor of a non-standard-location header which will be used for an "Analog Adapter BoosterPack" which I'll get started on soon.  This gives me more board space to deal with the analog stuff.

 

This cape includes some extra features--a mounting hole which should provide easier access to the S2 button, a pair of 0805 SMD pads for I2C pullup resistors (+ jumpers to enable/disable) and it properly supports an EEPROM (CAT24C256, per the SRM's recommendation).  The EEPROM is all SMD components on the bottom side, using an SOIC-8 version of the chip + 2x3 SMD headers for configuring A0-A2 + an SMD resistor array (x4) for the A0-A2 pullups.

 

OSHpark mockups:

post-4-14264691118479_thumb.png

post-4-14264691119321_thumb.png

 

Schematic: DipTrace Schematic - BeagleBone_LaunchPad_draft3.pdf

 

Elecrow/Seeed/ITead gerbers: Elecrow_BBBLP_v30_10x10.zip

OSHpark gerbers: OSH_BBBLP_v30.zip

 

Gerbers and DipTrace design files available via Github: https://github.com/spirilis/bbblp

 

Going to submit an order today from Elecrow.

 

edit: oops, forgot pullup resistors for I2C2 for the EEPROM identification.  Fixing & telling Elecrow to hold the order...

edit: Updated the files in this post, waiting for elecrow to get back to me on sending the new gerbers.

Link to post
Share on other sites
  • 2 weeks later...

New boards in!

 

post-4-14264691121755_thumb.jpg

 

post-4-14264691122685_thumb.jpg

 

I won't personally be able to test these for a while, but I'll send some over to @@bluehash for the store.  Only glitches on these boards I can see is the S2 button hole isn't aligned quite that well, and the silkscreen for the I2C/SPI jumper shows the Sitara port name instead of the P9_x pin naming scheme I used everywhere else on the board.  I also don't have the HW to test the EEPROM but I did "design by datasheet" for that one... haha (and the EEPROM isn't strictly necessary, probably not all that useful for this sort of application anyhow, just tossed in for completeness sake).  A revision 3.1 will be used for the next batch fixing those minor issues.

 

New boards have I2C pullup resistor pads & jumpers to enable/disable them.  That's for the LaunchPad's I2C bus, not the EEPROM bus (though it has R4 & R5 on the back to provide pullups for the EEPROM cape I2C bus too).

 

Analog is broken out to J4 but it's not intended to be used by your boosterpacks here, I did that in preparation for an opamp-based "analog adapter" boosterpack I'll be working on soon to properly level-shift the 3.3V analog inputs down to the Sitara's 1.8V levels.

 

@@cubeberg you also showed some interest in the board?

Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.


×
×
  • Create New...