LattePanda ALPHA DC Power insights

A place to talk about all the Operating Systems, Software & Hardware
User avatar
ccs_hello
Posts: 286
Joined: Sat Oct 13, 2018 9:16 pm

Sat Dec 29, 2018 5:16 am

This is a continuation of my previous thread (DC Power Experiment...)
https://www.lattepanda.com/topic-f6t16887.html

Supplementary info:
Power (PWR) LED is blue-colored
D13 LED (appears to be coming from Arduino Leonardo's D13 output) is red-colored.

When DC supply power is low, none of the PWR or D13 LED would lit.
When I played with the DC power and gradually increased the voltage level, first thing to notice is Blue LED started to blink at a low frequency (about one short blink every 3 second or so.) the Red LED stay unlit or extremely vary faint/barely noticeable.

I further increased the DC supply voltage, the Red LED started initially very faint, pretty useless at this point. (No change on Blue LED blinking frequency.)
When DC voltage increased high enough, the Red LED started to fast blink for about 5 second or so.) After that, the Red LED stayed constantly lit. (Blue LED is still in its slow-blink pattern.)

One might notice the Red LED is still quite dim, increase the DC supply voltage to make sure the Red LED is in its normal brightness (about the same level as the Blue LED when it lit.)

ONLY at this point, the system is ready to be powered up.

Press the POWER button for about 8 seconds or so, until you see both the Blue and Red LEDs stayed ON all the time (No more blinking.)
At this time, if you see both LEDs stayed ON and no interruption, the DC power side of the necessary condition is (almost) met. If you see other issues (e.g., no Display, etc.) that's a different level of problems.
** Why I said "almost"? It's because power supply stability (lack of big capacitor, USB wire gauge/length, etc.), the operation condition sits in marginal satisfactory region. I will at least increase the DC voltage by 0.7V on the safe side.


=== continue on next post ===

User avatar
ccs_hello
Posts: 286
Joined: Sat Oct 13, 2018 9:16 pm

Sat Dec 29, 2018 5:31 am

Now I also had tested the USB type-C power IN configuration and learned quite a few things:
1) Comparing the USB type-C PD sink (external power supply feeding the power into LP Alpha's USB type-C port)
with the DC power feed (the four-pin connector I explained in the earlier "DC Experiment" thread)
I would add at least 0.5V to the minimum required supply voltage. I.e.,
in DC (4-pin) method, I would try min. 10.5V and up to 15.4V as its power source. < -- I recommend 12V, since I can also use that to supply power to external PCIe bus type of extension schemes. (See https://www.lattepanda.com/topic-f6t16946.html )

If using type-C port to feed DC power in, I would recommend starts from 11V and probably up to 15.5V.

2) If possible, add a separate capacitor to handle CPU Speed Boost and many instantaneous power demand (4700uF, 20V.)

3) When using USB type-C as the method to supply power to LP Alpha, there are a few negotiation sequences in between the USB type-C PD 2.0 power supply and the LP Alpha's USB type-C port in PD sink mode.
There are garden variety of voltage level negotiation schemes and mistakes can happen. Most often the negotiation will get into some classic power schemes (e.g., BC 1.2 or DCP 5V 1.5A mode). Note that these are 5V only, and LP Alpha will not have any LED lit and would appear to be dead.

=== more descriptions and experiments on (3) in next post ===

User avatar
ccs_hello
Posts: 286
Joined: Sat Oct 13, 2018 9:16 pm

Sat Dec 29, 2018 5:55 am

CAVEAT
Note that my LP Alpha came and worked fine from the very start. So I have to improvise to see if I can create the conditions
and/or guess what might go wrong, so the actual problem is unknown to me.

I had tested the LP Alpha supplied PD 2.0 45W power supply with integrated USB type-C cable.
My unit (can be manufacturer variation) supports quite a few power negotiation schemes.
This is not a good thing (may negotiated to a 5V scheme and LPA will be as dead as ice cold).

I tried an iClever 30W unit, it also has far too many schemes supported. Not great.
(model IC-UC01)

I tested Apple 61W type-C power supply (for Macbook Pro), that one works (only support two schemes: PD 2.0 and Apple 5V 2.4A mode) this is works out well (very little chance for negotiation mistakes.)
Example of such unit: https://www.ebay.com/itm/Genuine-Apple- ... 2450434938 < -- not this one
about $30 or higher (note many are clones)
WARNING: don't buy genuine one, see this post
https://www.lattepanda.com/topic-p27575.html#p27575

I tested an iClever 45W unit, that unit works fine (due to very few schemes are supported.) <-- note it is version 1 of the 45W model
This is version 1 (note the part number is IC-WD11W )
https://www.amazon.com/iClever-Delivery ... B07194RXTS

(Note I have no experience with ver 2, a $17 unit. Don't know what's changed inside or spec wise.)

You'll have to purchase a separate USB C to C cable (male type-C on both ends.)
https://www.amazon.com/dp/B077S2KJCP (about $6)

Will it work? I don't know. Just some ideas for you.

User avatar
ccs_hello
Posts: 286
Joined: Sat Oct 13, 2018 9:16 pm

Sat Dec 29, 2018 6:03 am

BTW, there is a third type of DC power input thru that mysterious 10-pin (laptop battery) connector.
It appears to be the only one that can take lower voltage (say 7.4V native 2S Li-Po) or even lower (typically 2S at 6-point-something (say 6.3V) voltage will shut itself down.)
I have no idea on how it works (min, max, charging the battery directly thru that port, charge-n-power simultaneously, under-voltage protection, charging circuit self-contained, how OS supports battery fuel gauge, ...) all kind of questions. DC power path is and has never been an easy feast.

User avatar
ccs_hello
Posts: 286
Joined: Sat Oct 13, 2018 9:16 pm

Sat Dec 29, 2018 6:24 am

Add another balancing view

Instead spending time/energy on fighting the USB type-C (similar to USB 1.0 era's situation, the type-C standard is like its infancy, many gaps to be fixed) related issues, why not just use the 4-pin DC Power in as the method to supply power.

Type-C has other IMHO much more interesting features so why not (free it up and) consider these:
- supply power to attached devices (charge a phone, connect to a USB3 type-C hard drive)
- use type-C to HDMI cable as another monitor
- use LP's streaming cable (or other similar schemes such as DisplayLink)
- etc.

User avatar
ccs_hello
Posts: 286
Joined: Sat Oct 13, 2018 9:16 pm

Sat Dec 29, 2018 6:28 am

(Add another link sharing my debugging method:
https://www.lattepanda.com/topic-f6t17266.html )

User avatar
ccs_hello
Posts: 286
Joined: Sat Oct 13, 2018 9:16 pm

Tue Jan 08, 2019 12:13 am

Now I can confirm that LattePanda Alpha's minimum DC supply voltage is as follows:

1. on 4 pin DC-in connector, min is 10V and stable at 10.5V
Below that, the PCB shuts itself down.

2. on type-C connector as PD sink(as reciving power from external power adapter), min is 10.5V and stable at 11V
Below that, the PCB shuts itself down.

3. on 10 pin Li-Po battery connector, min is 5.2V (not recommended, I would call) stable at 6.4V (the cut-off voltage of 2s Li-Po battery pack.) < - - see https://www.lattepanda.com/topic-f23p26723.html

Note that "momentary push the Power button to trigger the Red LED blinking when supply voltage is lower than 10V"
only works in case (3), but not in cases (1) and (2).

User avatar
ccs_hello
Posts: 286
Joined: Sat Oct 13, 2018 9:16 pm

Tue Jan 08, 2019 12:41 am

I'd like to emphasis again:

This 10-pin Li-Po in connector is meant for
well-protected (BMS + balancer at minimum) battery pack, not for raw Li-Po/Li-Ion/LiFePO4 cells.
(e.g., latter setup will drain below the minimum Li-Po cell voltage which will significantly/irreversibly degrade the life expectancy of cell.)

Also charging without proper CI (then later CV) charging circuit is outright dangerous.

User avatar
LattePanda
Site Admin
Posts: 2397
Joined: Wed Jan 20, 2016 9:09 am

Wed Jan 09, 2019 5:34 am

ccs_hello wrote:
Tue Jan 08, 2019 12:41 am
I'd like to emphasis again:

This 10-pin Li-Po in connector is meant for
well-protected (BMS + balancer at minimum) battery pack, not for raw Li-Po/Li-Ion/LiFePO4 cells.
(e.g., latter setup will drain below the minimum Li-Po cell voltage which will significantly/irreversibly degrade the life expectancy of cell.)

Also charging without proper CI (then later CV) charging circuit is outright dangerous.
I think this post is really important. So reposted these info in docs - http://docs.lattepanda.com/content/alph ... -batteries

And link the users, who are interested in this post back to here! ;)
Enjoy Tinkering with LattePanda ! :lol:

User avatar
hylo
Posts: 2
Joined: Wed Feb 06, 2019 11:28 pm

Fri Feb 08, 2019 7:15 pm

ccs_hello wrote:
Tue Jan 08, 2019 12:41 am
This 10-pin Li-Po in connector is meant for
well-protected (BMS + balancer at minimum) battery pack, not for raw Li-Po/Li-Ion/LiFePO4 cells.
(e.g., latter setup will drain below the minimum Li-Po cell voltage which will significantly/irreversibly degrade the life expectancy of cell.)
Thanks for sharing this information. It is very helpful.
I found below LiPo 3.7V battery. I see some circuit on top of the battery but I'm not sure it's well protected.

https://www.aliexpress.com/item/1-2-4-P ... 3c00F55s2C

Do you think it would be okay to connect the two of them in a series to use them as a 2S LiPo?

Post Reply