Post Reply 
 
Thread Rating:
  • 0 Votes - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Possible Shield Requirements Analysis
05-25-2014, 03:41 AM (This post was last modified: 05-25-2014 03:42 AM by steamboat.)
Post: #21
RE: Possible Shield Requirements Analysis
What about at least one LED on the board to check the live status.

All robots with legs have the problem that the legs can be blocked by it self or by any kind of external forces. This is what we should detect to avoid any kind of damage.

For this we can check the rotation of the crank or the motor. If it will not count after a certain amount of time something is wrong and we should stop moving.

The best thing would be to have a current sensor for each motor or at least for the complete unit. I have this one in my quadrocopter but it is quite expensive and we don't need to monitor such "big amps"... CIRUS Sensor

Rainer
http://www.steamboating.de
Visit this user's website Find all posts by this user
Quote this message in a reply
05-28-2014, 02:59 AM
Post: #22
RE: Possible Shield Requirements Analysis
(05-23-2014 06:59 AM)steamboat Wrote:  What features have you planed for the power system? At least I would like to see the voltage over BT on the Android.

Right now I have "on demand"* voltage monitoring on:
5v rail
3v3 rail
Battery

*I say on demand because the voltage divider is inherently going to leak power so I have these so we can switch on monitoring and make readings and switch them off.

I also have diagnostic led indicators for:
5v Regulator
3v3 Regulator
Battery
PWR (Raw power from battery for servo/motor)
AUX (Raw power from battery for aux sleep mode stuff)
Leonardo Power

These leds in most cases can be switch off by the brainstem

(05-23-2014 06:59 AM)steamboat Wrote:  My which would be (but not all which have to become true in live)
- an integrated BT module for all kind of Android communication.
- a temp sensor
- a real time clock chip with battery
- a gyroscope
- some kind of on/off or sleep button/functionality to keep shellmo alive for some days in sleep mode.
- a SD card slot for storing data and/or sending sound and music.
- an amplifier to play music
- an amplifier for a mic.

This is a good list.

I want to cover BT integration for power and data to eliminate those wires.

Temp sensor is easy and cheap.

RTC is a must for scheduling.

Accelerometer, audio, mic I think would be cheapest to just add cheap Chinese boards. It would increase the part count substantially I think, but a audio and mic are must haves to get the "behavior" so I am thinking on it..might just use add on boards.

SD slot, if I can find space and route it.

(05-23-2014 06:59 AM)steamboat Wrote:  With this Shellmo could tell you how hot it was at home when you are coming back or at night. It could say hello. Could say good morning when it is time to get up or play your favorite music. It could try to walk straight ahead with the gyro.

This is exactly what I have in mind.


(05-23-2014 06:59 AM)steamboat Wrote:  Maybe the Gyro is not necessary because Shellmo will never go long distances - but on other hand this are fun features. I still have one of this as spare part for my quadrocopter Arducopter 2.6 but never used it in a land vehicle and even did not program it by myself...

What about a camera?

I think camera would be nice, but makes more sense for the distant future and some sort of raspberry pi integration.
Find all posts by this user
Quote this message in a reply
05-28-2014, 03:33 AM
Post: #23
RE: Possible Shield Requirements Analysis
(05-25-2014 03:41 AM)steamboat Wrote:  What about at least one LED on the board to check the live status.

Maybe this is the heart led... Wink

I am thinking on the current sense.

Hall effect for keeping an eye on the rotation of the leg gears.
Find all posts by this user
Quote this message in a reply
05-28-2014, 06:36 PM
Post: #24
RE: Possible Shield Requirements Analysis
Good morning James,

I don't know how much such boards will cost in the US and how easy it is for you to source it in small quantity's. I could try to make it on my CNC mill but maybe the legs of the Atmel Chip are to small.

Maybe we could start with an easy board try this out and improve it step by step. Normally I am not able to make "THE GREAT THING" at the first time. And with modern rapid technologies this even is not necessary...

If prototypes are not to expansive, we can try to sell the prototypes at ebay with a printed RepWaler just for the production costs.

At the weekend - which starts in 5 hours here and will last for 4 days ;-) - I also will prepare a discussion base for the Shellmo module sizes including space for the PCB. What about batteries? Do you think we should stick with the 4x AA rechargeables?

Rainer
http://www.steamboating.de
Visit this user's website Find all posts by this user
Quote this message in a reply
05-29-2014, 01:49 AM
Post: #25
RE: Possible Shield Requirements Analysis
(05-28-2014 06:36 PM)steamboat Wrote:  I don't know how much such boards will cost in the US and how easy it is for you to source it in small quantity's. I could try to make it on my CNC mill but maybe the legs of the Atmel Chip are to small.

Right now the PCB is going to cost $19us for 10 boards. I say add 15-20 for components and maybe an hour of time a piece for building it out. So really your looking at maybe $25 max for a single shield as I have it designed right now. I don't care to recoup any labor and expect not to sell these as I am not sure I am that comfortable in my design skills to go commercial Wink

The good thing is you wont have to populate everything, just what you need as its mostly just a shield anyways. Something like the RTC can be soldered by hand.

(05-28-2014 06:36 PM)steamboat Wrote:  Maybe we could start with an easy board try this out and improve it step by step. Normally I am not able to make "THE GREAT THING" at the first time. And with modern rapid technologies this even is not necessary...

We could start with an easy board, but then there are plenty of them out there. I agree, this board is complex, but I am taking all the risk here anyways so don't worry. Almost all of this is really an integration from a bunch of spark-fun schematics anyways, so while there is a lot of room for error, it will be in the integration, not at the individual part level.

(05-28-2014 06:36 PM)steamboat Wrote:  If prototypes are not to expansive, we can try to sell the prototypes at ebay with a printed RepWaler just for the production costs.

At the weekend - which starts in 5 hours here and will last for 4 days ;-) - I also will prepare a discussion base for the Shellmo module sizes including space for the PCB. What about batteries? Do you think we should stick with the 4x AA rechargeables?

4x AA seems alright to me. I don't think these thing are going to be doing a lot of moving through out the day. Something I do want to think more on before I send this design off.

Re:ebay, I mean whatever you think, if I could get a hot air rework station out of all this that would be lovely, but I am not thinking really about that.

Have a nice weekend. Sounds good. The PCB is large by the way. The only way to incorporate data+power+gnd takes a lot of space. I think the board is 8x10cm which is HUGE, but looks like it can fit on the things we have been building and will look better than a rats nest of wires.

I am going to try to send out for PCB in 2 weeks. So in 5 weeks from now I should have 10 from China and can make a couple and test them out and build the software for brainstem.
Find all posts by this user
Quote this message in a reply
05-29-2014, 02:03 AM
Post: #26
RE: Possible Shield Requirements Analysis
$19 for all 10 = 1,90 per board?

So really your looking at maybe $25 max for a single shield

That sounds great to me - I would take 2 of it - or will take 3 from the pre series if necessary to make the 10.

ebay, I mean whatever you think, if I could get a hot air rework station out of all this that would be lovely

Selling on Ebay for me would be only a way not to have to much electronic junk in my room and to finance the next boards ;-)

Rainer
http://www.steamboating.de
Visit this user's website Find all posts by this user
Quote this message in a reply
05-29-2014, 02:13 AM
Post: #27
RE: Possible Shield Requirements Analysis
(05-29-2014 02:03 AM)steamboat Wrote:  $19 for all 10 = 1,90 per board?

10 pcs 10cm x 10cm 2Layer Green PCBs with 100% e-test.

I have made 3 orders from itead studio in the past, total of 5 different 5cm x 5cm boards and they always turn out great.
Find all posts by this user
Quote this message in a reply
05-30-2014, 10:11 PM (This post was last modified: 05-30-2014 10:12 PM by steamboat.)
Post: #28
Servo current
Hello James,

I have the new servo mount for Macropede in print. While printing I thought about how to measure the servo current. What about to use the logic chip in the servo as current probe?!

The HK servos use the ME9926:
http://www.force-mos.com/db/pictures/mod...828957.pdf

Other servos also need some "switch" which will have an internal resistance... At the free running HK servo I can measure 24 mV between D and S. Breaking it with the finger a little gives 26 to 27 mV - can't test high torque because the new gears are glued to it and have to dry out first...

All I need is an amplifier which can bring this some mV to a detectable voltage - don't know what to choose here in modern times ;.)

It would make me happy even if we can measure only one turning direction with this.

Rainer
http://www.steamboating.de
Visit this user's website Find all posts by this user
Quote this message in a reply
05-31-2014, 12:44 AM (This post was last modified: 05-31-2014 12:45 AM by jmccartney.)
Post: #29
RE: Possible Shield Requirements Analysis
(05-30-2014 10:11 PM)steamboat Wrote:  All I need is an amplifier which can bring this some mV to a detectable voltage - don't know what to choose here in modern times ;.)

I think something like this would do what we want
MAX4372T

I got the idea from this thread:
avrfreaks
Where a user describes it use as:
Use a .1 ohm 1% resistor (or .05 1% if I can find it for a reasonable price in a 0805) between servo supply (+6V) and servo. I'll then have a MAX4372F connected across that resistor. Datasheet here: http://pdfserv.maxim-ic.com/en/ds/MAX4372F-MAX4372T.pdf specifically I'd use a MAX4372FEUK-T (SOT-23) (or MAX4372HEUK-T if .05 ohm resistor). I'd then feed the AVR's AREF with a precision 5V signal. The MAX4372F multiplies the voltage across the resistor by 50, so that would give the ADC a range of 0-1A, which should be way more than enough. I'm going to run some tests to see if I can get current over half an amp - and if I can't I'll probabaly go with the MAX4372H (multiplies voltage by 100). Oh, I should mention that with a .1 ohm resistor I'll see a 1V/20 drop across the resistor, which I feel is small enough not to be worried about.

I would have to order the parts and actually try it out before I felt comfortable putting it on the PCB..(current sense is out of my element a bit).. Also we would not want to have more than one or two individual servos with this because of: routing issues, cost (1.50us) per chip, use of an analog line, space constraints. Maybe two servos would have current sense, and then have current sense on the entire PWR bus so we know the total draw of all servos/motors.. IDK. There might be a lower cost newer solution, this thread was 2005 i think, so i will have to look through more parts..
Find all posts by this user
Quote this message in a reply
05-31-2014, 02:22 AM (This post was last modified: 05-31-2014 02:46 AM by steamboat.)
Post: #30
RE: Possible Shield Requirements Analysis
(05-31-2014 12:44 AM)jmccartney Wrote:  I think something like this would do what we want MAX4372T
Yes! That's it. Two servos would be enough and we could "sell" it as an option too.
In the data sheet of this current-sense amplifier they even use a PCB trace as Rsense!
"My" data sheet is from 2011

Here is something similar with a Allegro acs756 current sensor

[http://diydrones.com/profiles/blogs/allegro-acs756-current-sensor-for-arducopter]Arducopter forum[/url]

HEre you can buy one of this:
http://www.emmeshop.it/product.php?id_product=402

Some data about the Allegro - but this is using Hall effect from a near by copper path - sounds like lots of try and error in the board layout?
http://www.allegromicro.com/en/Products/...CS756.aspx


This one is using a TI INA-169
http://diydrones.com/profiles/blogs/atto...tor-sensor

see here -> http://www.ti.com/product/ina169

PS:
I can source the INA 139 and 169 in a "stupid" all people pharmacy price like shop (Conrad) for 2,24 Euro
The Data Sheet is from 2000 and revised 2005

Can't find the MAX4372T in this shop.

But found a LTC6101BCS5 for 2,59
and a LTC6101AIMS8 for 3,22

In the other German "all people shop" Reichelt I also can't find any of this...

Rainer
http://www.steamboating.de
Visit this user's website Find all posts by this user
Quote this message in a reply
Post Reply 


Forum Jump:


User(s) browsing this thread: 1 Guest(s)