Thread Tools
Jan 23, 2014, 01:24 PM
Registered User
Quote:
Originally Posted by eimulis
how about bigger cameras? Is Storm32 BGC able to lift biger DSLRs?
Any news? What is input power upper limit, 4S, 6S?
Sign up now
to remove ads between posts
Jan 23, 2014, 04:38 PM
Registered User
hexakopter's Avatar
"They allow for a maximum voltage of 18 V, and hence operation at up to 4 S." from his webside. (http://www.olliw.eu/2013/storm32bgc/)

Best reguarts, hexakopter
Jan 23, 2014, 05:53 PM
OlliW
Thread OP
as hexacopter mentioned, the voltage range is 2S to 4S

current-wise no experiments with the DFN packages were yet reported, but from thermal considerations one expects a maximum current of 1.9 A, hence a maximal practical current of 1.5 A per motor appears very realistic. This spec is en par with that of the upcoming 32-bit AlexMos controller(s).

1.5 A should be plenty to run gimbal motors up to the 5000 size
(the rctimer web page give resistances of >= 14 Ohm for the 5000 and 4000 sizes, so, with the PWM set to 100% the current would be I = U/R, hence I = 1.06A for R = 14 Ohm ... I've never seen somebody reporting a PWM setting of 100%, so in practice the currents will be even smaller than 1 A per motor => I would thus bet that with the DFN packages gimbals with this motor sizes could be run, but as said before, only a test would be a final proof, but it appears very reasonable to do this test )
Last edited by OlliW; Jan 23, 2014 at 07:01 PM.
Jan 27, 2014, 11:19 AM
Registered User

testing this weekend


Hi Olli - I should get a chance to spend some time testing it next weekend.

I notice though - you are all 3.3v on the board ? All my IMUs are 5v breakouts. Is there a 5v header I can tap into or will I need to short out the 5v rectifier on the IMU to test ?

stu
Jan 27, 2014, 11:40 AM
OlliW
Thread OP
Hey Stu,

nice to hear from you

good question on the MPU modules!
the internal voltage of the board is 3.3V; the board does NOT provide 5V
so, it feeds 3.3V to the MPU module(s) at the respective I2C port(s)

hence, in principle it would be best to short out the 5V regulator on the MPU module, e.g. by attaching the Vcc cable to a point after the voltage regulator such as cap etc. pp. (that's what I'm doing with one of my modules)
however, I also have a mpu module running flawlessly with the Vcc cable connected to the standard Vin pin of the module...

so, I would suggest to first try it the simple way and to connect the Vcc cable to the envisaged pin on the MPU module, only if this doesn't lead to stable operation I would consider shorting out the regulator

(if you check the electric scheme of the board, you will notice that I've put 2.2k resistors on the I2C line, which together with the resistors on the MPU module make kind of an "aggressive" pull up... this is still within the specs but I found this to improve the signal shapes and hence the stability of the I2C quite a lot, and maybe this also helps with the 5V vs 3.3V thing...)

Cheers, Olli
Jan 27, 2014, 01:28 PM
Registered User
Quote:
Originally Posted by OlliW
Hey Stu,

nice to hear from you

good question on the MPU modules!
the internal voltage of the board is 3.3V; the board does NOT provide 5V
so, it feeds 3.3V to the MPU module(s) at the respective I2C port(s)

hence, in principle it would be best to short out the 5V regulator on the MPU module, e.g. by attaching the Vcc cable to a point after the voltage regulator such as cap etc. pp. (that's what I'm doing with one of my modules)
however, I also have a mpu module running flawlessly with the Vcc cable connected to the standard Vin pin of the module...

so, I would suggest to first try it the simple way and to connect the Vcc cable to the envisaged pin on the MPU module, only if this doesn't lead to stable operation I would consider shorting out the regulator

(if you check the electric scheme of the board, you will notice that I've put 2.2k resistors on the I2C line, which together with the resistors on the MPU module make kind of an "aggressive" pull up... this is still within the specs but I found this to improve the signal shapes and hence the stability of the I2C quite a lot, and maybe this also helps with the 5V vs 3.3V thing...)

Cheers, Olli
depends alot on the IMU, but *most* breakout boards have on board 3v3 regulators as again *most* imu chips are 3v3 or 1v8 only. Check the breakout design all the ones I have run just fine on 3v3... they are generic MPU6050 breakout boards.
Jan 28, 2014, 05:26 AM
Registered User
Will try it. I thought 3.3 wouldn't be enough for the rectifier to work, but will try that first.

Also handy hint generally : if you get a few of the 99p/99c retractable usb leads, the cable they use is ideal for IMUs. You have a twisted pair in the middle for the signal, and gnd and vcc cables on outside. I find I get no errors using cable even with no pull ups generally (on other boards I'm using etc).

It's also really thin and flexible of course.
Jan 28, 2014, 05:28 AM
Registered User
This kind of thing

http://pages.ebay.com/link/?nav=item...d=231003124029
Jan 28, 2014, 05:52 AM
OlliW
Thread OP
Quote:
Will try it. I thought 3.3 wouldn't be enough for the rectifier to work, but will try that first.
well... it wouldn't work as intended... on these modules you usually find regulators with 3.3V output voltage, but they need at least about 0.3V higher voltages at their inputs (for LDOs)... so, then you feed it with 3.3V at the input it (1) will output a voltage, which is (2) about 0.3V lower and (3) not regulated. So, it doesn't work as intended, but it seems to work for us (in most cases) since luckily enough the ca. 3.0V output is sufficient for the I2C to work well enough (the MPU can go down to 2.4V, so that's not the problem)

so, the module is working although the regulator is not working

however, what has not yet been really tested is the consequences of having an unregulated voltage at the MPU chip... this might lead to larger noise, and there is a power supply dependency of e.g. the gyro bias, etc. pp...


BTW: in my test and micro gimbal setups I'm using open wiring without any shielding of lengths up the several 10 cm and the i2c error rate is exactly zero... and o323bgc runs the I2C even at 800kHz
(I think what is to seldomly considered in the i2c error problematic is two things, improper pull ups and improper code)
Last edited by OlliW; Jan 28, 2014 at 05:58 AM.
Feb 04, 2014, 03:56 PM
OlliW
Thread OP
have you seen this, http://www.basecamelectronics.com/downloads/, the description of the new AlexMos 2.40b6 firmware:
Quote:
...
* Follow mode was completely rewritten:
- speed can be adjusted in wide range by the "SPEED" setting
...
- Follow for the axes PITCH+ROLL and YAW can be activated independently
- Follow for ROLL axis may be permanently enabled or disabled regardless of camera PITCHing
...
* Up to 5x more precision with the new PID (increase I-term until oscillation starts and than reduce a little back to get the maximum from it)
...
hihi, o323bgc had ALL of this from the very begining

yeah... I've beaten AlexMos... just for a short period of time, but at least...

(I'm quite sure that his new is algorithm is identical to mine, the coincidences can't be a fluke)

EDIT: I have to add this to the feature list lol
Feb 04, 2014, 05:56 PM
Registered User
Waiting ever so patiently
Feb 06, 2014, 06:26 PM
OlliW
Thread OP
new board STorM32 BGC v1.2 released
hardware supports now also Spektrum and Futaba S-bus
Feb 06, 2014, 06:43 PM
Flying Wood For Fun
irun4fundotca's Avatar
ohhh dear

what a tease
Feb 06, 2014, 06:46 PM
Registered User
Quote:
Originally Posted by OlliW
new board STorM32 BGC v1.2 released
hardware supports now also Spektrum and Futaba S-bus
Just happy if it will work with my Taranis
Feb 09, 2014, 12:03 AM
Registered User
Hi all

I got this board 3 days ago, after waiting for almost a month lol
Then after soldered all the pins, and placing it on my "homemade" gimbal.. here is the result..

Testing storm32 bgc (1 min 22 sec)


Im using rctimer 5010/150t for the pitch, 5208/180t for the roll and yaw..
The battery, 3S.

As you can see, i changed 3 different modes.
1. Auto yaw, pitch and roll lock
2. All axis locked
3. Auto yaw and pitch
Its all done with 2 tactile switches which was extremely easy to place

Well i havent record something "real", maybe in the next few days i will post the videos..

Thanks for watching
And thanks Olli, this is great..

Regards
Erick
Last edited by erickwesz; Feb 11, 2014 at 01:19 AM.
Feb 09, 2014, 08:35 AM
Registered User
What size motors can this controller take?

Can it take something as big as the Blackmagic Production Camera 4K?

That is 1.7 kg for the camera alone.
Feb 09, 2014, 09:04 AM
Registered User
When a commercial version will become available?
Feb 09, 2014, 09:56 AM
Registered User
Ordered my boards now to order the components, hoping DigiKey has it all.
Feb 09, 2014, 10:02 AM
OlliW
Thread OP
Quote:
What size motors can this controller take?
well, I can't say much more than before, there are not enough testers and results out there and experience is hence limited... I would think however that Erick's great results are very encouraging

IMHO the question is actually not well posed. It should be rather "what max currents and voltage can this controller take?". The point is, the motor can be as large as you want and produce as much torque as you want, as long as the resistance is such that the current flow at the given voltage is not too large. I've spend some time to research a bit the different motors lately, and - at least if the resistance values given on the web pages are correct - you can't say that a larger motor also means larger currents (or lower resistance, respectively). E.g. there are 2000 size motors with resistance of 10 Ohm, while there are also 5000 size motors with resistance of 17 Ohm or more...

(Erick's motors have 15 Ohm and 17 Ohm)

The experience so far indicates that motors with 15 Ohm or more are no problem for the board.
For the voltage the limits are known, max 18V, i.e. 4S.

So, if you're happy with 4S and your whatever-sized motors have resistance of 15 Ohm or more, you shouldn't have any problem balancing your 1.7 kg camera.

Anyway, let me repeat: This board has not be designed with high-current applications in mind. However, the board is open source and the electronic scheme hence public, so anyone can design her/his own board, and if she/he wants with current capabilities of 70A like some AlexMos boards (LOL)

Quote:
When a commercial version will become available?
I honestly can't tell. As stated before I myself don't have a financial interest in here, so commercializing it is up to whoever wants to do that. A couple of folks had contacted me about this, but I of course don't know how serious they are. One person seems to be really determined, but it's totally his decission if and when and how and what the price will be. We are in contact and I share information, but I'm not going to interfer with his decissions (I mean, he would take all the risks, not me LOL)

Quote:
order the components, hoping DigiKey has it all.
if you post the digikey parts numbers before you order, I'd look it over and confirm/correct it
Feb 09, 2014, 10:19 AM
Flying Wood For Fun
irun4fundotca's Avatar
my new imu/mpu came yesterday when I get home will start doing testing again
Feb 09, 2014, 11:52 AM
Registered User
I ended up going to Newark for components, haven't finalized the ordered yet so here it is in zipped format if you want to check it.

I was wondering if there would be enough people interested in doing a group buy if we could find a place that would assemble the boards? I just don't know how many would need to be interested so it would be worthwhile. Anyone with experience in getting populated boards ordered would be very helpful here.
Feb 09, 2014, 12:10 PM
OlliW
Thread OP
ähm... what would be the best way to look at a csv file?
(ASCII editor and Excel show something, but the content isn't really easy to digest )

BTW: I should adapt the electronic scheme and BOM: resistor R11 should be 1.5k instead of 2.2k (thanks ala42!)
(and I might change R1 to 10k, its value is really not very relevant, 10k makes it just easier for those who buy the parts themselves as this reduces the number of differnt R values one has to buy)

I don't have any experience with buying assembled boards.
Feb 09, 2014, 12:24 PM
Registered User
I did a very quick look at Seeed Studio and very roughly to assemble 100 boards would cost about $19.00 plus components. So if there were 100 people wanting to do this it's actually not too bad, but I think it could be a bit of a
cowboys herding cats (1 min 7 sec)
situation.
Feb 09, 2014, 12:51 PM
OlliW
Thread OP
as regards your csv, I did it now the old fashioned way

some improvements are possible (I have not looked at the resistors and all the small capacitors)
in that cases I give the hbe/farnell order numbers, not sure how they relate to Newark (though both are element 14)

items 1-8: OK
item 9: incorrect, should be: Art. 1349646, NXP, 1PS76SB40, DIODE SCHOTTKY,SOD323
items 10-13: OK
items 14-21: not checked
item 22: not sure, I'll know it as 1206 size, see Art. 1770137, BOURNS, CAY16-103J4LF, WIDERSTANDSARRAY, 1206, 10K
item 23: OK (maybe you'll find one with a lower ESR, mine are 2.1 Ohm)
item 24: OK, but maybe get some with 50V
item 25: probably OK, mine are NPO
item 26: puh ESR 1.2Ohm, that's bad: Art. 1672609, AVX, TRJC336K020R0300, KONDENSATOR, 33UF, 20V, BAUF. C
items 27-28: not checked
item 29: hm... doesn't look right... it should have 4 wings, and two little nops on the bottom (open the data sheet, its more like the longer version shown there)
items 30-34: OK

I'm missing the 10uf/6.3V tantal

if you buy the parts but don't get them assembled, I wouldn't buy the MPU6050, I find it hard to solder, and, 14$ is ugly expensive (aliexpress is something like 2-3$, I think)... you'll be better off by buying a second MPU module and pluging it to the I2C#2 port (this on-board MPU makes sense "only" for comerzializing it, i.e., with cheap sources)
Last edited by OlliW; Feb 09, 2014 at 01:03 PM.
Feb 09, 2014, 03:38 PM
OlliW
Thread OP
v1.2 rev1 posted, sole change: value of R11 changed to 1.5k
Feb 10, 2014, 08:04 PM
Registered User
Quote:
Originally Posted by OlliW
v1.2 rev1 posted, sole change: value of R11 changed to 1.5k
the 1.1 boards can be made very very soon. Does this change need to be made to the 1.1 boards?
Last edited by witespy; Feb 10, 2014 at 08:17 PM.
Feb 10, 2014, 09:21 PM
Registered User
Hi guys

Olli asked me to do some test on my gimbal, i've sent the result to him. Well for anyone who wonder how hot the drivers are
on 200 PWM, the drivers get hot even for just a couple of mins.
on 150 PWM, just warm...
on 100 PWM, not so much different with 150PWM

After the test, i set my PWM with my 'best' values and turn-on the gimbal for almost an hour..
the PWM for pitch: 115, roll: 151 and yaw: 170... just now i touched the TCs, and it doesn't get hot at all... just warm...

FYI i'm using rctimer 5010/150t for the pitch and rctimer 5208/180t 17.2ohm for roll and yaw.. with 3s battery

Now good luck for you who want to use larger motor, the TC's seems to do well enough..
Last edited by erickwesz; Feb 11, 2014 at 01:33 AM.
Feb 10, 2014, 09:58 PM
Registered User
Quote:
Originally Posted by OlliW
hey obelix6620000
thanks for this info
you may however assume that I know these standard things very well
btw, this type of the code is found already in multiwii (brugi reused it)
this technique is nice, but is - IMHO - not the best, I think it can be improved....
(it's the thing I named somewhat mysteriously the "Kalman-filter mimicing" thing in the first post)
cheers Olli
If using Kalman-filter, but not adding sensors, if would not help but only complicate the code.
Feb 11, 2014, 01:35 AM
OlliW
Thread OP
@witespy: COOL

the R11 thing applies only to v1.2, not v1.1,
for v1.1 I think the values are all OK (one might change - maybe - the values of R8,R9 to 1k instead of 2.2k - never made my mind up what would be better)

@erick: THANKS for the tests, and the reporting!

I'm not sure though if this indicates that the TC's in SO08 package (the ones you have) would also handle motors with even lower resistance like 10 ohms... the DFN's might do... I have TC's in DFN package coming... let's see...

@aband: well, I guess I disagree. the Kalman filter adds TWO great things to a complementary filter, namely it adjusts the gains dynamically (and not only instantaneously) and second it "measures" the reliability of the estimates. This as you say makes it more complicated to calculate, but would also improve results as compared to a complimentary filter, even whithout further sensors. BTW: I have not said that I would implement a Kalman filter
Feb 14, 2014, 03:58 AM
Registered User
@erick

I've seen your video testing the STorM32. Nice gimbal!

What camera is mounted on the gimbal?


thx
Gerhard
Feb 14, 2014, 08:37 AM
Registered User
Quote:
Originally Posted by Gerhard34
@erick

I've seen your video testing the STorM32. Nice gimbal!

What camera is mounted on the gimbal?


thx
Gerhard
Hi Gerhard

Thanks, that is a homemade gimbal... a bit heavy since i use aluminium...

Anyway the camera is Sony Nex 5n...

Regards
Erick
Feb 14, 2014, 08:49 AM
OlliW
Thread OP
your Nex, it provides IR remote control, right?

I'm just looking a bit ahead in future
Feb 14, 2014, 09:29 AM
Registered User
Quote:
Originally Posted by OlliW
your Nex, it provides IR remote control, right?

I'm just looking a bit ahead in future
Yes...
I notice that in my board, there is also IR pin ready to be use... can't wait to use it...

Anyway, i had really much fun with the bluetooth, now i can tune my gimbal while walking with it as i told you... later i will record some video of it...
Feb 14, 2014, 03:59 PM
Flying Wood For Fun
irun4fundotca's Avatar
nice im waiting for my blutooth module to arrive

Feb 14, 2014, 04:40 PM
OlliW
Thread OP
uff... soldering the MPU and the DFN packages with an ersa30 and a pan only took me quite some effort now and several tries... but eventually a v1.1 was assembled and working LOL

I can't be sure if my soldering really established an as good as possible thermal link between the DFNs and the print board, but nevertheless, I guess, it will be interesting to run some current/power tests on this board now... I need some bigger motors though... let's see...

Feb 15, 2014, 09:50 AM
OlliW
Thread OP

Power test of board with TC in DFN packages


ok, here a result

I did a test for at least 30 min using a rewound CF2822 with ca. 15 Ohms connected to the Pitch channel, a fully charged 4S lipo (15.98V currently) and with maximal PWM of Vmax=255:
the motor gets probably too hot, you can't touch it for longer than ca. 2-3 secs with your fingers
the pitch-TC's feel warm at best then touching them with a finger from above
snuffling with the nose at the board tells however that there is some heat around, indeed the PCB bottom side gets hand warm

of course, this is just one motor... with three motors on each channel more heat will be produced, but neverthless heat management seems to be quite OK

these results could be compared to Erick's finding that with motors of simliar resistance the TC's in SO08 package got too hot at "only" 3S and Vmax= 200.

so, I herewith conclude:

with the DFN packages the board can well handle motors with resistances of 15 Ohm or more even at 4S and full power of Vmax=255.

considering that Vmax=255 is rarely used, it seems probable that the board can also handle motors with 10 Ohm resistance (but proof is still lacking)
Feb 16, 2014, 01:20 PM
bacon and eggs
remvideo's Avatar
if anyone used an IR trigger with an NEX and can PM me please... need help
Feb 17, 2014, 04:15 AM
Registered User
Quote:
Originally Posted by remvideo
if anyone used an IR trigger with an NEX and can PM me please... need help
http://sebastian.setz.name/arduino/m...ra-ir-control/

Arduino based and gives shutter and video control for the NEX series (as well as other IR controlled cameras). Works well!

Regards
Len
Last edited by len1286; Feb 18, 2014 at 04:27 AM.
Feb 17, 2014, 11:33 AM
Flying Wood For Fun
irun4fundotca's Avatar
here's that link fixed
http://sebastian.setz.name/arduino/m...ra-ir-control/
Feb 18, 2014, 04:28 AM
Registered User
Quote:
Originally Posted by irun4fundotca
Thanks! Why didn't I see that Edited post.

There's also the gentLED lineup, used extensively by the KAP community. http://www.gentles.ltd.uk/gentled/

Len
Feb 18, 2014, 05:42 AM
OlliW
Thread OP
hey folks,

all this is excellent info... thx for it
however, this thread is not on IR remotes in general but the STorM32bgc... please consider moving with non-STorM related discussions to the respective thread or to open a new one

Thx
Feb 20, 2014, 10:50 AM
OlliW
Thread OP
hey everyone

The v1.2 board has been tested now... and everything (including the on-board IMU ) works exactly as intended, I couldn't spot any mistake, neither in the layout, pin assignments, scheme, nor print-ons.

I've also created a wiki for the documentation: www.olliw.eu/storm32bgc-wiki. I choose a wiki because it allows anyone to contribute and edit (you just have to register)... so, I hope that STorM users make plentyful use of that

cheers, Olli
Last edited by OlliW; Feb 20, 2014 at 12:03 PM.
Feb 20, 2014, 11:24 AM
Flying Wood For Fun
irun4fundotca's Avatar
done

looks good Olli
Feb 24, 2014, 03:25 AM
Registered User
Hi guys

This is another demo of the "on-the-fly" switches with the newer firmware version (0.20). I think Olli will release it soon.
STorM32 BGC 4 modes (1 min 27 sec)


And this is the "live" test
STorM32 BGC live test (1 min 52 sec)


Thanks for watching
Regards

Erick
Feb 24, 2014, 08:53 AM
Registered User
I am redoing my parts order and one thing I am not sure of is the LED specification, forward voltage. I'm trying DigiKey this time because too many items from Newark were overseas orders and would take 2 weeks min to get here.
Feb 24, 2014, 09:09 AM
OlliW
Thread OP
the forward voltage is irrelevant... it is essentially (nearly) determined by the color, so that's not under our control... the current is of more interest, try to find some with 5mA or less, but if you don't such, it's no issue to go with 10mA

digikey or mouser should both do well on the shipping thing, I guess
Feb 24, 2014, 09:55 AM
OlliW
Thread OP
as regards sourcing parts a major issue is to get a list of item numbers...

maybe you all could help each other... , e.g. those with (some) knowledge on item numbers could fill up the table I've created in the wiki: http://www.olliw.eu/storm32bgc-wiki/...ectronic_parts

if you happen to have the data available in some electronic format suitable for direct download to a supplier, you also can post it there or add a link to the page.


Last edited by OlliW; Feb 24, 2014 at 10:07 AM.
Feb 24, 2014, 10:21 AM
Registered User
Is there a big advantage to the DFN8 package over the SO08 package for the TC4452? Mouser has the DFN8 and DigiKey doesn't.
Feb 24, 2014, 10:29 AM
OlliW
Thread OP
well, the DFN8 packages do have - IMHO - a big advantage in how much current they can handle (since they are much better cooled by the PCB board)
otherwise they don't have any advantages (and the disadvantage of being more difficult to solder)
I don't know what your application is, so I can't give advice... if it's just a matter of sourcing I would think play it safe and go with the DFN
Feb 24, 2014, 10:39 AM
Registered User
Damn just entered all the digikey numbers and someone else was editing the same page and now I don't know how to resolve it


Quick Reply
Message:

Thread Tools