The main Maple page states that the Maple board can "supplies up to 800mA @ 3.3v" Can this supply be used to power other external circuitry, like the Arduino supplies 5V and 3.3V. If this is possible, how does one do it?
Maple supplying to other circuitry
(15 posts) (6 voices)-
gbulmer
ModeratorThe main Maple page states that the Maple board can "supplies up to 800mA @ 3.3v"
I believe that is an error.
This thread discusses the part used to provide 3.3V http://forums.leaflabs.com/topic.php?id=228#post-1743
and it says...MCP1703 3.3V regulators. That's a very generic 3.3V, 250mA regulator...
Can this supply be used to power other external circuitry, like the Arduino supplies 5V and 3.3V. If this is possible, how does one do it?
Yes you can draw some current, but much less than 800mA. I guess maybe 100mA at 3.3V, depending on what the microcontroller is doing.
That is available from one of the two 6-pin headers, it has the label "VCC".
If you have a rev5 board, you may also be able to draw current from the neighbouring :VDDA" header pin.
Be careful what you connect to VDDA as that supplies the analogue to digital converters, and is intended to have low electrical noise.HTH
(full discolosure: I am not a member of LeafLabs staff.)
Posted 4 years ago # -
gbulmer
Moderatorposlathian - the STM32F103x8/xB manual says the STM32F will itself 'eat' upto 50mA at 72MHz (Table 13), and a GPIO can sink or source 25mA upto a maximum of 150mA (Table 7), so it depends on what folks do with Maple.
Aren't you concerned that putting something drawing 500mA (say a DC motor) onto the combined logic+analogue supplies might destroy the low-noise performance of the ADCs?
Posted 4 years ago # -
gbulmer
Moderatorit is true that the noise performance of the ADC's will decrease when tying the supplies together. However, this is not permanent damage ...
I understand and agree.
... so its really just a matter of managing design tradeoffs in a specific case.
I understand. This is the nub of the issue. The 'just' is hard without accurate information.
Also 450mA (allowing 50mA for the STM32F) is 350mA smaller than the
Supplies up to 800mA @ 3.3v
*STILL* claimed on page http://leaflabs.com/docs/specs.html
which is wrong and unclear how to get 450mA (50mA for STM32F) on rev 5 (no mention of connecting both regulator outputs or its implications),
wrong for rev3, which only supplied 200mA (50mA for STM32F), and
wrong for USB which should be limited to 500mA maximum.So, yes, I agree 'its really just a matter of managing design tradeoffs in a specific case' but with inaccurate and incomplete data, that is practically impossible.
IMHO, it would be helpful to correct this error (which has been identified in these forums before).
State that the rev 5 can supply a maximum of 500mA from the two regulator outputs, how to achieve that, and that the STM32F consumes some of that. Further, state that rev3 only made the output of one 250mA regulator available. I'd suggest identifying in the spec that the ST Micro's datasheet defines a maximum value of 50mA for the STM32F103RB.
I believe in the UK, it would be legal for a purchaser to return and demand a full refund with the force of law for their Maple on the basis of this error. In the USA who knows how much further it could go?
Posted 4 years ago # -
mbolivar
LeafLabsgbulmer,
*STILL* claimed on page http://leaflabs.com/docs/specs.html
which is wrong and unclear how to get 450mA (50mA for STM32F) on rev 5 (no mention of connecting both regulator outputs or its implications),Thank you for the pointer. I had previously fixed the corresponding 0.0.10 docs in this regard (http://static.leaflabs.com/pub/leaflabs/maple-docs/0.0.10beta/hardware/maple.html#technical-specifications ), and have now updated the 0.0.9 version as well. I did not include information on connecting the regulator outputs, since, as you say, there are further implications, and I didn't want to replace erroneous documentation with more of the same.
I will fully describe the relevant information by the time the next official release is made.
And of course, if the error has caused anyone to buy a Maple who otherwise would not have, we would be happy to accept a return for a full refund.
Posted 4 years ago # -
gbulmer
Moderatormbolivar -
I had previously fixed the corresponding 0.0.10 docs in this regard (http://static.leaflabs.com/pub/leaflabs/maple-docs/0.0.10beta/hardware/maple.html#technical-specifications ), and have now updated the 0.0.9 version as well.
Excellent.
I did not include information on connecting the regulator outputs, since, as you say, there are further implications, and I didn't want to replace erroneous documentation with more of the same.
IMHO, it is better to be conservative, simple, accurate and published than either impossibly optimistic, or detailed and complete but unpublished :-)
Thank you.
Posted 4 years ago # -
mm666
MemberI still think the current specs are misleading... The poor SOT23 linear voltage regulators can only dissipate around 500 mW (and even less if not soldered to a decent size copper plane). So if the input voltage is 12V, the 3.3V regulators have to 'burn' 8.7V. 8.7V times 250 mA is 2.175W, much more then the 500 mW. The regulators will burn very fast if someone will draw 250mA while supplying 12V to the input...
At 12V input, the max. current will be around 60 mA... after that, the regulators will start to burn...
Since there are two regulators, the max. current at 12V input will be around 120mA.
Please correct me if I am wrong!
MarkPosted 4 years ago # -
jessb
LeafLabsHey Mark, my reading of the datasheet indicates that the regs should be able to output 250mA at all accepted input voltage, but it's possible that either my reading of the datasheet is wrong or the datasheet is simply incorrect in this regard. Neither would surprise me, considering the problems with this chip we've been having. I've filed a ticket with Microchip asking if my reading of the datasheet is correct, and I'll try in the next few days to do a test and see what happens. Thanks for pointing this out.
Posted 4 years ago # -
gbulmer
Moderatormm666 - I believe you are correct.
I had a quick look at a microchip 3.3V SOT-23 linear regulator.
one of these: http://www.microchip.com/ParamChartSearch/params.aspx?branchID=9004&mid=11&lang=en&pageId=79There is only one rated at higher than 6 volts (the MCP1802), which runs upto 10V
It's datasheet states
Power dissipation = 0.25W (250mW for a SOT-23-5, which should be better than a SOT-23)It goes on to state:
Output Current (Continuous) ..................... PD/(VIN-VOUT)mA
So, if the input voltage were 10V, I think that means:
Output current = 0.25/(10-3.3) 0.037ASo, unless someone at LeafLabs can correct that calculation then I believe, the maximum specified current from each regulator at 10V is 37mA.
Posted 4 years ago # -
jessb
LeafLabsArgh, sorry for dropping the ball on this, it fell off my radar. You're absolutely correct. We're actually using the MCP1703 component (here: http://ww1.microchip.com/downloads/en/DeviceDoc/22049a.pdf ) in the SOT-23A package. For this chip operating at room temperature (25C) it looks like the maximum power dissipation is actually at about .37W, so the max current draw is about 42mA, but yeah.
I'm working on putting together a blog post with some information about various things related to the power circuitry; in the meantime we'll make sure the docs are all up-to-date. Thanks, guys.
Posted 4 years ago # -
gbulmer
Moderatorjessb - Thanks you for the update on the actual part Maples uses being the MCP1703.
I apologise. I had failed to find the MCP1703 with the microchip product selector, and I had assumed that the SOT-23 voltage regulators would be similar, which was incorrect.
The MCP1802 has a much lower maximum operating temperature, but a better power dissipation. So my calculation was close because of confounding errors :-(
Having said all that, current falls reasonably quickly with rising temperature, so at 40C (which I experienced when I lived in Houston), the current falls to about 38mA.(I remember 40C vividly because at that temperature water from the air started to condense on me because my body temperature was lower. I'm a Brit. We haven't evolved for that sort of thing. Yeuk!
I also remember experiencing 42C, on my first working day, when we walked to lunch. I said "this is ... hot", and one of my Houston-native colleagues said "this is pretty hot, but some days it gets really hot". The next day he came to my office and cheerfully told me "I checked, and yesterday was the hottest day on record". You've got to love folks like that :-)Posted 4 years ago #
Reply
You must log in to post.