mctom vs M1

Post Reply
User avatar
mctom
Posts: 1594
Joined: Wed Nov 11, 2020 4:44 am
languages_spoken: english, polish
ODROIDs: OGA, XU4, C2, M1
Location: Gdansk, Poland
Has thanked: 185 times
Been thanked: 199 times
Contact:

mctom vs M1

Post by mctom »

It is not late, I just wanted to test and try a few things, and provide a bit broader view than "first impressions"..

But then again, nothing wrong with first impressions really.

As a N2+ veteran it was striking that M1 is significantly slower, but I think this has finally become general knowledge - it was not meant to be faster at all.
Not everything is quite "there" on the software side of things, but again - it was to be expected and I don't mind, I can wait and explore hardware side of things in the meantime.

So what is going to follow is mainly a hardware analysis of new M1.

Power supply
As it has been already known at the release, the broad input voltage range (7.5V ~ 15.5V) makes this board very well suited to battery operation. Not only for portable devices, as bulky M1 form factor is not particularly suited for that, but for UPS action against transient power outages or graceful shutdown in such circumstances. Batteries (such as 3x 18650 in series) may be connected to M1's power input directly, which improves power conversion efficiency and simplicity.
Still, these batteries would require balancing and charging circuits, and some would argue it's nice to have them on board. I disagree, as different battery types require very different solutions, and there's hardly one that would work in all scenarios.
What M1 actually could have, however, is an extra connector (or at least pads) parallel to DC jack - for connecting extra batteries, voltage monitoring circuit, or more reasons that I'll mention later. Or as a main power input, if M1 is to be integrated into a bigger device. Barrel jacks, I imagine, are too easy to detach during transportation.
On the bright side, I've never heard any whining from power circuit yet - an improvement over N2+ that sometimes annoyed my girlfriend with younger ears.

SATA
Excellent addition that is surprisingly rare on the market - When I think about it, original Banana Pi had one, and Odroid HC series, can't think of anything else. An optional HDD mount kit for 2,5" drives is of course warmly welcome at competitive price.
All this hints that the original concept was to support most of the 2,5" drives only, with lack of 12V on respective power connector. In fact, I noticed that too independently, and I accept the reasoning behind this decision - that additional step-up and step-down circuitry for that obscure functionality would add to product cost.
Yet clearly there are people who want that feature anyway, with an example of @mad_ady already committed to execute a hardware mod on his own. That, of course, requires a user to guarantee 12V +-5% input power, but that's easy enough with recommended power adapter.
What I feel would be the most appropriate solution, is to add a 2.54mm jumper pins that connects 12V SATA rail to DC input, whatever that is, at the user's discretion.
One would argue that inexperienced users could do a lot of damage to their equipment, but let's face it - shorting two pins on GPIO header may destroy most SBCs and the world has not ended. Also, inexperienced users will most likely use recommended 12V PSU, which is completely fine in this scenario.
And finally, at least providing extra pads parallel to DC jack would help perform the mod cleanly, possibly without detaching M1 PCB from the heatsink.

M.2 drive
It's an absolutely fantastic feature that made me question the point of SATA connector. I never used NVMe drive before (I owned one for a short while, as i assumed it's compatible with SATA).
The drives are cheap, with 128GB unit I bought for less than $20. I imagine there are users who prefer 4TB storage options for (whatever), and that's when SATA comes in handy.
One thing to note when using M.2 is that power consumtion rises by about 2-3W, which is almost equal to the whole rest of the board idling. That is a lot of heat to deal with, but I believe that depends on the drive itself.
I think I recall someone voicing an opinion that drives with shorter physical formats (less than 80mm, that is) may be slightly cheaper.
Maybe, but for me an inability to mount other PCIe devices here is a major waste of potential. As little as an extra PCB hole, where a user-provided standoff would go, would already be acceptable for advanced tinkerers.
And to think about it, for system integrators that could have been a major opportunity, if they could plug in a WiFi interface, extra NPU, internal USB controller or whatever.
Perhaps I'm stretching my argumentation, but then again it's a low effort enhancement I'd really like to see in a future PCB release.
(EDIT: Okay, now I can see why this is impossible - SoC is in the way :/)
Also, the bundled M.2 screw is terrible - it has a tiny cruciform drive that made me fetch precision screwdrivers - and damaging screw head while attempting to loosen it. If there's a possibility to replace them with something easier to handle, I think many would appreciate it. Or maybe it was just me who got that small awkward screw tightened with a pneumatic wrench.

USB
Nothing wrong with USBs functionally, they all work as far as I can tell - and feel more stable than in N2+. But a sole fact I experienced no problems doesn't prove anything, I guess?
The problem, a major one to me, is that all USB connectivity is provided on one side of the device only. That's one of the things I like about XU4 - an extra port on the opposite side from the rest. It is very handy, especially when someone plans to expose a few ports to the user, and use the rest internally.
Again, that is nothing an end user could possibly complain about, but since the board is meant for integrator businesses, I was hoping for at least one USB port exposed anywhere but on the "user" side of the board (which to me is where Ethernet and HDMI are).
Ideally, a 10-pin header would provide access to one or two USB2.0 ports for any internal affairs - and there's a well established standard for that, often found on PC motherboards.
I wouldn't mind if these headers were connected in parallel to existing USB2.0 ports - although I'm not sure if that's viable in terms of high speed routing.

Speaker amp
I'm really excited about this little feature and really wanted to try it out - alas, the connector type, which may or may not be some sort of established industrial standard, is something I cannot connect with anything I've found around my work bench.
I'm stressing again that perhaps that's the connector type it is supposed to be for any reason, but then again PC speakers used to be connected via 2.54mm headers since forever and that worked well.
So, er, again... How about additional 2,54mm header pads for tinkerers? :)

The Mysterious Button and silkscreen marking in general
There is a tact switch next to what I believe is SPI memory chip, and next to eMMC. Honestly, I thought that's a reset button, as many SBCs have one. There's "RCY" marking next to it, that could as well be the marking of neighboring speaker connector - not sure.
My complaint here is: "RCY" doesn't ring any bell, never heard of that before. After I read on the M1 product page that's a "SPI recovery switch", I guess that's what "RCY" was meant to stand for - recovery.
What I would suggest here is to add both "Speaker" and "Boot recovery" markings on the silkscreen layer. "Audio" jack is labeled with much smaller letters and that works for me, and I can't see why not use clearer markings in that tight space, using smaller font.
I'd argue that "SPI recovery" is not the best way to describe this functionality, both in the docs and on PCB itself. SPI is an interface for connecting many ICs together, what would that mean to "recover" it? If I didn't read the forums for past year I'd be very confused about that. Perhaps "Boot recovery" is not the best description either, but at least hints about what would one use this button for.
Similarly "BACKUP BATTERY" - it doesn't really tell the user what this battery is for. Long ago I used to believe "BACKUP" connector on XU4 was for a rechargeable battery to serve as UPS. Perhaps "RTC battery" would be a better description.
CSI, Display, SATA data and power, and SD card connectors don't have any markings at all - that wouldn't hurt to add them as well.
USB2.0 ports are marked as USB3.0 ports - something that ought to be corrected, but also hints that N2+ indeed served as starting point for PCB design, huh? :)
But in order to stay positive, I do highly praise the fact all components are marked on the PCB - in an unlikely event of fixing my M1 this will render the operation at least possible to attempt. I wasn't able to fix Odroid HC2 mainly because that board didn't have enough space for component designators.
Yet again I stand in admiration how clean the design is - both schematics and layout. Mr Ruppi Kim is someone I look up to in this regard.

LEDs
I designed PiStackMon and after building a few prototypes I realized how hard is to match resistors to LEDs of different models and brands. Eventually I've written on Github readme that no LEDs are explicitly recommended and you're free to deal with it on your own. ;)
That's why I'm not surprised LEDs on M1 are not matched perfectly on the first production batch. However, green NVMe status LED is much brighter than the rest - and way too bright for my taste. So bright, in fact, that I found it worth mentioning here.

And that's all for now - I'll allow myself to expand on this post whenever I spot something interesting later on.
To conclude - this board looks to me like a very solid piece of engineering and there are no major complaints from my side - just a few cosmetic errors and ideas for improvements.

EDIT1:

Micro USB port
I think the only reason for this port is Android development, or something along these lines. Users may flash OS images on all possible media with other methods, as microUSB in OTG mode is not the fastest of them, or offering any other advantage, compared to LAN for example.
It's not a secret that port has caused EMC problems too.. ouch :?
I already said M1 isn't the best choice for portable builds, but if someone attempted a compact design based on M1, due to its low power consumption, they would probably get rid of the heatsink in the first place. And then discover their design will end up 1mm thicker just for the sake of this one micro USB connector of questionable usefulness in the target application.
I know, M1 isn't supposed to be disassembled prior to integration, but! I want to do this anyway and will see that USB port removed :evil:
Anyway, "what could have been done" in this case, is a horizontal microUSB connector on a top layer. Not as solid mechanically, but should be good for development, I guess. I assume it has to be close to its sister USB3 port due to high speed routing considerations.

EDIT2:

Microphone input
I skimmed through RK809 datasheet, which is a PMIC chip, or should I say, a "company" chip to SoC, as it not only regulates all sorts of voltages, but also provides some extra functionalities, such as audio interface.
And yes, it does have microphone input, which has been left floating.
Okay, I remember the first Banana Pi, which must be the only SBC I've seen to date with an actual built-in microphone. And that's ridiculous.
Yet again, "what could have been done" is another pin header or a place for the user to solder one themselves. This time for a mic input, along with 4 recommended passive components.

Minimum DC power voltage
I studied schematics and found out that PMIC senses DC input voltage, which must exceed some threshold for PMIC to turn on. So my understanding is, after turn on the input voltage is completely ignored.
Nevertheless, the hardware configured boundary is right above 5.1V, which in practice means it's not configured at all. I suspect that feature may have caused more trouble than was worth it.
The actual limit for minimum input voltage, I believe, is the max duty cycle of 5V buck converter. It's not stated explicitly in its datasheet, but minimum off time combined with maximum frequency yields some 84% maximum duty cycle. Which translates to minimum input voltage being close to 6V, plus some for resistive losses.
The reason why I'm digging into this is to see if just two battery cells are enough to power up M1. Two cells combined have some 8.4V max, and are often discharged down to 5V or 6V. If discharged slowly enough, there's no significant difference in cutoff voltage in terms of total energy delivered. Moreover, if M1 is known not to work with 5V input at all, this serves as the natural cell protection against too deep discharge.
The question is whether I could stretch the minimum voltage of 7.5V down to some 6.4V. The answer is yes.. maybe... ;)
I'll set SmartPower3 to 6.4V and see if it works okay in these conditions, and modify this section if it fails at any time.
Last edited by mctom on Tue Apr 19, 2022 5:32 pm, edited 3 times in total.
These users thanked the author mctom for the post (total 2):
brad (Wed Apr 20, 2022 11:51 am) • egil (Fri Apr 22, 2022 6:04 pm)
Punk ain't no religious cult, punk means thinking for yourself!

Maintainer of PiStackMon

User avatar
tobetter
Posts: 9820
Joined: Mon Feb 25, 2013 10:55 am
languages_spoken: Korean, English
ODROIDs: Many
Location: Paju, South Korea
Has thanked: 626 times
Been thanked: 1654 times
Contact:

Re: mctom vs M1

Post by tobetter »

+1 since no thanks button.

User avatar
mctom
Posts: 1594
Joined: Wed Nov 11, 2020 4:44 am
languages_spoken: english, polish
ODROIDs: OGA, XU4, C2, M1
Location: Gdansk, Poland
Has thanked: 185 times
Been thanked: 199 times
Contact:

Re: mctom vs M1

Post by mctom »

You got me - actually I hesitated to write this post, hoping someone would fix thanks buttons by now :lol:
These users thanked the author mctom for the post:
neal (Wed Apr 20, 2022 10:12 am)
Punk ain't no religious cult, punk means thinking for yourself!

Maintainer of PiStackMon

hominoid
Posts: 663
Joined: Tue Feb 28, 2017 3:55 am
languages_spoken: english
ODROIDs: C2, C4, XU4, MC1, N1, N2, N2+, HC4
Location: Lake Superior Basin, USA
Has thanked: 81 times
Been thanked: 283 times
Contact:

Re: mctom vs M1

Post by hominoid »

mctom wrote:
Mon Apr 18, 2022 10:31 pm
Or maybe it was just me who got that small awkward screw tightened with a pneumatic wrench.
I started with the right size screw driver and still almost rounded it the first try. I couldn't believe how tight it was.

megauti
Posts: 44
Joined: Wed Apr 13, 2022 2:38 am
languages_spoken: english
ODROIDs: Odroid M1
Has thanked: 1 time
Been thanked: 3 times
Contact:

Re: mctom vs M1

Post by megauti »

mctom wrote:
Mon Apr 18, 2022 10:31 pm

Speaker amp
So, er, again... How about additional 2,54mm header pads for tinkerers? :)
This is great request! I would ask for same thing.
I also ask for 1 more, USB pin-out for tinkerers as well.
The major reason why I love Orange Pi Zero 2 so much is that it has both USB & Audio header pads.

User avatar
mctom
Posts: 1594
Joined: Wed Nov 11, 2020 4:44 am
languages_spoken: english, polish
ODROIDs: OGA, XU4, C2, M1
Location: Gdansk, Poland
Has thanked: 185 times
Been thanked: 199 times
Contact:

Re: mctom vs M1

Post by mctom »

hominoid wrote:
Tue Apr 19, 2022 8:33 am
I started with the right size screw driver and still almost rounded it the first try. I couldn't believe how tight it was.
I'm sorry to hear that but I'm glad it wasn't just me who struggled.
megauti wrote:
Tue Apr 19, 2022 9:42 am
This is great request! I would ask for same thing.
I also ask for 1 more, USB pin-out for tinkerers as well.
I also try to keep my requests somewhat realistic and at minimum cost, but we'll see where it goes. :)


I'm pleased someone actually got through my wall of text ;) EDIT2 is out!
Punk ain't no religious cult, punk means thinking for yourself!

Maintainer of PiStackMon

megauti
Posts: 44
Joined: Wed Apr 13, 2022 2:38 am
languages_spoken: english
ODROIDs: Odroid M1
Has thanked: 1 time
Been thanked: 3 times
Contact:

Re: mctom vs M1

Post by megauti »

The question is whether I could stretch the minimum voltage of 7.5V down to some 6.4V. The answer is yes.. maybe... ;)
I'll set SmartPower3 to 6.4V and see if it works okay in these conditions, and modify this section if it fails at any time.
I agreed with this opinion. For tinkerers, 12V is not prefered.
This is a photo of one of bot that I have made.
Due to USB & Audio pinout, I could create a bot with minimal size.
Image

User avatar
mctom
Posts: 1594
Joined: Wed Nov 11, 2020 4:44 am
languages_spoken: english, polish
ODROIDs: OGA, XU4, C2, M1
Location: Gdansk, Poland
Has thanked: 185 times
Been thanked: 199 times
Contact:

Re: mctom vs M1

Post by mctom »

Impressive build! :)

Yeah, this board is definitely not meant to be a small one, or maker friendly.
That's why most of my comments focus on making it even better for those who demand maximum connectivity and flexibility otherwise. Some of those features would benefit makers too.

I'm fiddling with my own idea for this board, but what can I say.. I did a few 3D designs and it appears to be bigger than I anticipated.
For now I assume that nothing portable will come out of that, which is sad because I wanted to leverage its low power consumption.
On the other hand, low power consumption is a plus for something that works 24/7.
Knowing myself I'll have another genius idea after a few more showers...
Punk ain't no religious cult, punk means thinking for yourself!

Maintainer of PiStackMon

L67GS
Posts: 792
Joined: Wed Apr 22, 2020 3:02 pm
languages_spoken: English, Jibberish, Pig Latin
ODROIDs: M1 8GB -w- MIPI-CSI Camera Kit, XU4, C1+,(3) C0's, and a whole big pile of accessories, VU7A Plus,, ect....
Location: Great Lakes Region, U.S.A
Has thanked: 236 times
Been thanked: 102 times
Contact:

Re: mctom vs M1

Post by L67GS »

I think the M1 has merit as a maker board. Certainly the size and voltage are limitations but there's lots of people making large projects like lawnmowers and stuff.
With larger bots in mind the M1 has low current draw, a full spread of GPIO with analog inputs and PWM, MIPI camera interface, NPU, WIDE voltage range, and onboard sound.

User avatar
odroid
Site Admin
Posts: 39119
Joined: Fri Feb 22, 2013 11:14 pm
languages_spoken: English, Korean
ODROIDs: ODROID
Has thanked: 2514 times
Been thanked: 1382 times
Contact:

Re: mctom vs M1

Post by odroid »

I've tested the minimum working voltage roughly.
It didn't boot or barely boots with a SmartPower3 at 6.0V output.
Once I changed the voltage to 6.1V, it booted well for several times of cold boot test.
Therefore, I think 6.4Volt might be fine in many use cases if there no significant voltage drop via DC power cable/connector nor batteries' internal resistance.

Due to many various power supply related factors, we will keep the official power source specification at 7.5V~15.5V to have some margins.
These users thanked the author odroid for the post:
mctom (Fri Apr 22, 2022 8:36 pm)

brad
Posts: 1595
Joined: Tue Mar 29, 2016 1:22 pm
languages_spoken: english
ODROIDs: C2 C4 HC4 N1 N2 N2+ H2 H2+ M1 (64 bit ftw)
Location: Australia
Has thanked: 178 times
Been thanked: 251 times
Contact:

Re: mctom vs M1

Post by brad »

I think the M1 will attract a different type of maker in time to come, it has a whole new level of IO that does not rely on USB expansion (even though the M1 also provides some nice designed USB hosts)

M1 has the PCIE exposed and this is serial and sends or receives data fast, usb on the other hand is mostly a polling protocol, it buffers and send data in chunks and this increases latency (it is mostly insignificant but hinders certain low latency use cases). I'm looking forward to seeing the M1 running a realtime or low latency kernel, as this would really open up some possibilities for robot control and audio processing.

P.S. I have taken a likening to your keyboard without usb question viewtopic.php?f=215&t=44236 as shutting down USB completely might significantly help with low jitter for realtime projects.

User avatar
mctom
Posts: 1594
Joined: Wed Nov 11, 2020 4:44 am
languages_spoken: english, polish
ODROIDs: OGA, XU4, C2, M1
Location: Gdansk, Poland
Has thanked: 185 times
Been thanked: 199 times
Contact:

Re: mctom vs M1

Post by mctom »

L67GS wrote:
Wed Apr 20, 2022 7:23 am
I think the M1 has merit as a maker board. Certainly the size and voltage are limitations but there's lots of people making large projects like lawnmowers and stuff.
With larger bots in mind the M1 has low current draw, a full spread of GPIO with analog inputs and PWM, MIPI camera interface, NPU, WIDE voltage range, and onboard sound.
Fully agreed, I'm just a maker of a different sort. I like my builds small, or at least tightly packed. But still, builds have to start from somewhere, and M1 with all its options is great for prototyping.
By the way you mentioned a few things I wanted to explore more thoroughly and cover in my future rants.
odroid wrote:
Wed Apr 20, 2022 9:41 am
I've tested the minimum working voltage roughly.
It didn't boot or barely boots with a SmartPower3 at 6.0V output.
Once I changed the voltage to 6.1V, it booted well for several times of cold boot test.
Therefore, I think 6.4Volt might be fine in many use cases if there no significant voltage drop via DC power cable/connector nor batteries' internal resistance.

Due to many various power supply related factors, we will keep the official power source specification at 7.5V~15.5V to have some margins.
Thank you, you did this before I had a chance really. :)
I believe the actual "boot success" voltage threshold will vary between the devices and depend on the power load from peripherals.
I agree declaring 7.5V as a minimum value is a right thing to do. I'd never blame anyone if anything goes wrong below it.
Nevertheless, my estimation of "around 6V" turned out to be so accurate it made my day. :)
brad wrote:
Wed Apr 20, 2022 12:49 pm
M1 has the PCIE exposed and this is serial and sends or receives data fast, usb on the other hand is mostly a polling protocol, it buffers and send data in chunks and this increases latency (it is mostly insignificant but hinders certain low latency use cases).
Hmm.. Don't want to ruin your day, but USB has "serial" even in its name. ;) I could agree it's mostly a polling protocol, but devices still can request attention from the hub. Constant data streams are also possible (bidirectional only since USB3 though).
I never looked into USB latencies, maybe except of audio equipment when I recorded music, and for that it was totally fine. :)
I do believe PCIe is much faster in many regards, as it never had to deal with weak links between hardware.. which are the users, of course, that can plug, unplug, tie knots on a cable, use low quality cables and so on.
brad wrote:
Wed Apr 20, 2022 12:49 pm
I'm looking forward to seeing the M1 running a realtime or low latency kernel, as this would really open up some possibilities for robot control and audio processing.
That is a really good call. Are there any other SBCs with low latency kernel? Even Ubuntu Studio used that until a few years ago.
brad wrote:
Wed Apr 20, 2022 12:49 pm
P.S. I have taken a likening to your keyboard without usb question viewtopic.php?f=215&t=44236 as shutting down USB completely might significantly help with low jitter for realtime projects.
Wow, you're not the first one to say that. It's the next thing I have on my roadmap. Already got my FPGA eval board running with M1, two laptop keyboards to play with, and read HID via I2C specs. Yesterday I got some connectors that fit the keyboards, and weekend is close. ;)
Punk ain't no religious cult, punk means thinking for yourself!

Maintainer of PiStackMon

brad
Posts: 1595
Joined: Tue Mar 29, 2016 1:22 pm
languages_spoken: english
ODROIDs: C2 C4 HC4 N1 N2 N2+ H2 H2+ M1 (64 bit ftw)
Location: Australia
Has thanked: 178 times
Been thanked: 251 times
Contact:

Re: mctom vs M1

Post by brad »

mctom wrote:
Wed Apr 20, 2022 9:41 am

Hmm.. Don't want to ruin your day, but USB has "serial" even in its name. ;) I could agree it's mostly a polling protocol, but devices still can request attention from the hub. Constant data streams are also possible (bidirectional only since USB3 though).
I never looked into USB latencies, maybe except of audio equipment when I recorded music, and for that it was totally fine. :)
I do believe PCIe is much faster in many regards, as it never had to deal with weak links between hardware.. which are the users, of course, that can plug, unplug, tie knots on a cable, use low quality cables and so on.
Yes serial maybe, free flowing stream maybe not ;)
brad wrote:
Wed Apr 20, 2022 12:49 pm
I'm looking forward to seeing the M1 running a realtime or low latency kernel, as this would really open up some possibilities for robot control and audio processing.
That is a really good call. Are there any other SBCs with low latency kernel? Even Ubuntu Studio used that until a few years ago.
[/quote]

Ubuntu studio was really good for x86 and even x64 later on. They have realtime kernels for some rpi modules and some other arm boards, I've had a RT kernel running on C2 (limited by usb2.0 connectivity problems) and C4 (similar only USB conectivity unless try to use GPIO). N2 it can run but difficult to work with 2 different types of cores for realtime applications. Now the M1 has PCIE and that opens up a new world of possibilities.

ruppi
Posts: 151
Joined: Fri Feb 22, 2013 11:13 pm
languages_spoken: korean
Location: South Korea
Has thanked: 0
Been thanked: 17 times
Contact:

Re: mctom vs M1

Post by ruppi »

Hi mctom,

Thank you for your good comments as the person who designed most of the ODROID series of hardware and PCBs since the establishment of HARDKERNEL.

Power Supply
- There is no space to add extra 2Pin 2.54mm Pitch pad around DCJACK, so we will add it below to the M.2 connector.
Personally, I don't like adding extra pads in addition to the primary connector because the pattern itself is not neat.
Please refer to the attached picture below.

SATA
- I totally agree with you, but I don't think I can handle the RMA risk.

M.2 drive
- Because there is no space to add nuts because the high-speed signals are passing by, it is difficult to support 44mm x 22mm shorten NVME.
Please let me know if you have a nut that meets the NVME standard but can be recommended at an affordable price

USB
- Adding a 20-pin header is unrealistic in terms of signal quality, even though there is no physical space.

Speaker amp
- I will add 2.54mm 2pin dummy header pin.
- We used a 1.25mm connector to sell the accessories to our company and make them compatible.

The Mysterious Button and silkscreen marking in general
- Since RESET is abbreviated as RST, I used the RECOVERY abbreviation as RCY when I was a software developer a long time ago, so I implicitly used it.
Anyway, i will mark it as RECOVERY without abbreviation.
Additionally, we will add additional silk screens such as SATA, SATA POWER, SPEAKER, etc
Please refer to the attached picture below.

LEDs
- Let's lower the brightness of the NVME Status LED.

Micro USB Port
- No suitable space could be found with no signal quality or EMI issues. so I don't have an idea.

Microphone Input
- Please understand that more modifications are needed to extract the microphone input of the RK809, which can affect the deployment or other signals, making it difficult to implement in reality.
In some cases, EMC certification is required again, resulting in additional costs. For your information, in the case of M1, the pure material cost is almost the same or slightly lower in sales.

Minimum DC power voltage
- The reason why the M1 works at 6.4V is conservative because the power input range is considered as a margin. If you want to operate at a lower voltage, you can modify the resistance of the voltage divider at 11 pins (EN) of U5.
As a precaution, the input voltage to the EN pin should be 3.6 volts or less.

Image
BR,
Ruppi
These users thanked the author ruppi for the post:
mctom (Mon Apr 25, 2022 4:58 pm)

tkaiser
Posts: 752
Joined: Mon Nov 09, 2015 12:30 am
languages_spoken: english
ODROIDs: C1+, C2, XU4, HC1
Has thanked: 0
Been thanked: 20 times
Contact:

Re: mctom vs M1

Post by tkaiser »

mctom wrote:
Mon Apr 18, 2022 10:31 pm
I think I recall someone voicing an opinion that drives with shorter physical formats (less than 80mm, that is) may be slightly cheaper.
The M.2 slot is key M and there exist almost no quality NVMe drives shorter than 80mm. SATA is something different but since Ruppi decided to use the Combo PHYs in a different way there's no SATA available in the M.2 slot.

In fact with M keying only supporting the 2280 format isn't a big deal. And key M is the only reasonable choice for NVMe anyway.
These users thanked the author tkaiser for the post:
mctom (Mon Apr 25, 2022 4:58 pm)

User avatar
mctom
Posts: 1594
Joined: Wed Nov 11, 2020 4:44 am
languages_spoken: english, polish
ODROIDs: OGA, XU4, C2, M1
Location: Gdansk, Poland
Has thanked: 185 times
Been thanked: 199 times
Contact:

Re: mctom vs M1

Post by mctom »

Hello Ruppi,

Pleased to "meet" you and I'm glad you've applied so many of my suggestions already. :)

Some of my comment may be not applicable because the layout image on Wiki page is very hard to read with blue copper pour, so I didn't pay close attention to it.

Power Supply
Thank you for the additional PSU pin pads. It is in a strange place but clearly there's no other suitable place for it. I don't think it is a problem.
I know the layout suffers from adding the 2.54mm pin connector as it is relatively big, but that helps the end user with accessibility to the board features. Think of them as extra test points. ;)

SATA
Understood. Well, at least the new PWR IN pad will make the mod somewhat easier. I expect many future users asking for that.

M.2 drive
As I already corrected myself, I see no way to add another M.2 standoff to accommodate shorter PCIe devices without moving SoC and its traces away, and I'd never ask for that.
I'm not an expert on M.2 either, and I trust @tkaiser's opinion that it's not so important.

I suspect you were asking for suggested bolts for M.2 mounting. It depends on the suppliers you have available. As far as I know it's supposed to be any 3mm long M2 screw. Something like this looks much easier to handle, as this head looks like it accepts bigger screwdriver.

USB
I think you misunderstood my request - I never asked for 20-pin USB header.
I was thinking about 10-pin USB header that is often used for two USB 2.0 ports on motherboards. It's 9-pin actually, for keying.
Image
That extra feature would make sense only if SoC still had unexposed USB ports. Adding such connector in parallel to existing USB 2.0 data lines will probably break signal integrity, as you say. Oh well, maybe that's an idea for a future ODROID board.

Speaker amp
Thank you very much for this connector, and the reason behind 1.25mm is understood now. I never owned ODROID speakers so I wasn't aware of that.

Silkscreen marking
Thank you for accepting many of my suggestions, these look much better in my opinion.
I'd appreciate if you would like to consider a few more changes:
- GPIO header's pin numbers: 1, 2, 39, 40 - I think these are a bit too small and hard to notice, especially surrounded by via stitching. I already managed to connect my whole project upside down, even if that was entirely my fault.
- "Infrared" is OK, but since there's enough space over there, maybe "Infrared receiver" would be even better? Some may believe it's a bidirectional port.

Micro USB Port
I suspected it's going to be impossible to move it anywhere. Please tell me at least it's not glued so I could remove it with hot air if I needed to.

Microphone Input
I did not expect this to be so complicated, but then again it's not that important feature to redo certification for it. I assume the difficulties arise from many changes necessary in layout to route these signals out from RK809.

Minimum DC power voltage
As I already calculated (perhaps wrongly), the voltage divider on EN pin of PMIC is already set for operating at 5.1V typ. So changing that is not expected to yield any difference in operation.

Ever since I wrote that post I run M1 at 6.4V and never had any issues. I test it because it is required in my build, but by any means I don't suggest changing the official voltage rating. Safety margins are always welcome.
Punk ain't no religious cult, punk means thinking for yourself!

Maintainer of PiStackMon

ruppi
Posts: 151
Joined: Fri Feb 22, 2013 11:13 pm
languages_spoken: korean
Location: South Korea
Has thanked: 0
Been thanked: 17 times
Contact:

Re: mctom vs M1

Post by ruppi »

Hi Mctcom,

Thank you for your understanding that I can't apply everything you want.
The picture below reflects what you requested additionally.
Image


BR,
Ruppi

User avatar
mctom
Posts: 1594
Joined: Wed Nov 11, 2020 4:44 am
languages_spoken: english, polish
ODROIDs: OGA, XU4, C2, M1
Location: Gdansk, Poland
Has thanked: 185 times
Been thanked: 199 times
Contact:

Re: mctom vs M1

Post by mctom »

In my opinion the board is much better with all these subtle changes. Thanks a lot!

Too bad I can't update the hardware with a single command. ;)

Hopefully the era of users asking what's the battery for is over. :)
These users thanked the author mctom for the post:
istanbulls (Tue Apr 26, 2022 4:50 pm)
Punk ain't no religious cult, punk means thinking for yourself!

Maintainer of PiStackMon

User avatar
istanbulls
Posts: 718
Joined: Tue May 14, 2019 10:18 pm
languages_spoken: Turkish
ODROIDs: ODROID; N2, C4, XU4, N2+, CH4
Location: Istanbul-Turkey
Has thanked: 430 times
Been thanked: 125 times
Contact:

Re: mctom vs M1

Post by istanbulls »

mctom wrote:
Tue Apr 26, 2022 4:18 pm
In my opinion the board is much better with all these subtle changes. Thanks a lot!

Too bad I can't update the hardware with a single command. ;)

Hopefully the era of users asking what's the battery for is over. :)
huh, what's the battery for?!! :roll: :lol:
These users thanked the author istanbulls for the post:
odroid (Tue Apr 26, 2022 8:00 pm)

User avatar
mctom
Posts: 1594
Joined: Wed Nov 11, 2020 4:44 am
languages_spoken: english, polish
ODROIDs: OGA, XU4, C2, M1
Location: Gdansk, Poland
Has thanked: 185 times
Been thanked: 199 times
Contact:

Re: mctom vs M1

Post by mctom »

There's so many options!
https://www.allacronyms.com/RTC
Punk ain't no religious cult, punk means thinking for yourself!

Maintainer of PiStackMon

brad
Posts: 1595
Joined: Tue Mar 29, 2016 1:22 pm
languages_spoken: english
ODROIDs: C2 C4 HC4 N1 N2 N2+ H2 H2+ M1 (64 bit ftw)
Location: Australia
Has thanked: 178 times
Been thanked: 251 times
Contact:

Re: mctom vs M1

Post by brad »

ruppi wrote:
Mon Apr 25, 2022 3:51 pm
Thank you for your good comments as the person who designed most of the ODROID series of hardware and PCBs since the establishment of HARDKERNEL.
@ruppi I have been using odroid's for many years since the C2, always look forward to seeing the designed by ruppi on the pcb :D

Your work is commendable!

@mctom the battery is what you make it :)

User avatar
rooted
Posts: 9447
Joined: Fri Dec 19, 2014 9:12 am
languages_spoken: english
Location: Gulf of Mexico, US
Has thanked: 758 times
Been thanked: 479 times
Contact:

Re: mctom vs M1

Post by rooted »

brad wrote:
ruppi wrote:
Mon Apr 25, 2022 3:51 pm
Thank you for your good comments as the person who designed most of the ODROID series of hardware and PCBs since the establishment of HARDKERNEL.
@ruppi I have been using odroid's for many years since the C2, always look forward to seeing the designed by ruppi on the pcb :D

Your work is commendable!

@mctom the battery is what you make it :)
Agreed, since I started with the C1 (non +), I've seen very minimal issues to none with the first shipped revision of every board since then which is quite good work.

The broken CEC on the original C1 was quite a problem for me though :)

Post Reply

Return to “General Topics”

Who is online

Users browsing this forum: No registered users and 0 guests