Completed DIY "laptop" / portable PC (7.39 litres, 17'' screen, 5700x, RTX2070)

timginter

Cable-Tie Ninja
Original poster
Apr 21, 2019
199
80
I don't really know how to explain this... I'm not sure if I'm even right or not...
Well anyways here is an attempt:

A current limited voltage supply will not usually supply its max current and max voltage at the same time.
If the load is low (high resistance), it will supply the max voltage and let it draw whatever current it wants as long as its below the set max current.
If the load is high (low resistance) and tries to draw more than the set max current, the voltage regulator will lower its voltage so that the updated current draw will be less than max current.

As an example let's take Vmax=12.6V and Imax=6A.
A 5Ohm load would get 12.6V and draw 2.52A
A 1Ohm load would only get 6V and draw 6A.
If the load got 12.6V instead, it would be draw 12.6A.

However with a step-up converter taking a 12V input, the load will always get at least 12V because a step-up converter can't supply an output voltage lower than its input voltage.


The problem is that in this case we have an unregulated 12V charger for battery voltages <12V, a CC cycle only from 12V-12.6V and a CV cycle finally at 12.6V.
I'm not sure if depending on ocp as a regulator for battery charging at the lower voltages is a good idea.
(On the other hand it might allow for pulse charging :p)


One minor problem with having a low 12V rail is that fans won't spin at max rpm. (And maybe some harddrives will not like it?)
Thanks for detailed explanation, really good points since the description of the item is not very clear. I found CCCV boards marked as buck/boost, boost-only and buck-only, the one I linked was one of the only CCCV boost with a sensible delivery time. Have I misunderstood the linked item? Wouldn't it provide a constant 12.6V 6A at output?

Tried to set it up with a multimeter before I replied, but it was all over the place... set it to 12.60V without load OK. Then switched the multimeter to current and it was all over the place, from 0 to 4A, jumping around. I used the battery as a load and current was more or less stable at 4A, but couldn't adjust it with the second potentiometer. This one looks the same, it's listed as a charger. Maybe I got a faulty board? Or maybe it's not CCCV - tough to tell. I'll return it and have a look around for a board marked as CCCV charger which I can actually test. May have to be patient with delivery

Speaking of stepping... Lemme just steps.. away and let people learn their lesson... Again... In the name of fun..
There's no need for posts like that
 
Last edited:

ruleh

Trash Compacter
Jan 19, 2021
34
24
If your battery voltage is lower than the converters input voltage, the current limiting will not work.
It might still be usable, if the charge current for a fully empty battery is still less than 6A.
Not sure why it jumps around like that though. Short circuit protection? Maybe it's input is cutting out for some reason?

A buck-boost converter should have working current limiting but those can have other quirks.

EDIT:
I expect the other module to behave the same way.
 

ruleh

Trash Compacter
Jan 19, 2021
34
24
Variable CV supply means a CV supply with output voltage you can adjust. Constant just means it'll try to maintain that set voltage.

A CCCV module would always need a buck. If it says step up then it's probably buck boost.

Ahh that makes sense. However now I am somewhat more confused than before. I think I misunderstood something, just not sure what.
 

timginter

Cable-Tie Ninja
Original poster
Apr 21, 2019
199
80
Ahh that makes sense. However now I am somewhat more confused than before. I think I misunderstood something, just not sure what.
I'm confused now, too. To charge the battery I need constant voltage at 12.6V and constant or limited current below - as long as the current doesn't go over 8A it's OK, BMS will take care of the rest. I found this video demonstrating exactly the same Voltage Converter 400W DC-DC Step-up Boost Converter Constant Current Power Supply Module LED Driver - it does seem to be a constant current constant voltage converter.

EDIT:
From the video, short circuit with multimeter to set current doesn't work with that CCCV - that would explain why it was all over the place.
If I understand the description correctly, the more charged the battery, the lower the current draw - that could explain why I couldn't set it over 4A. All in all, looks like it will do the job - constant voltage with a limited current, I just need the BMS to discharge the battery to 10.5V and then I'll be able to use it as a load to set the current (don't have anything else)

EDIT2:
I followed the "turn RV2 potentiometer around 30 times counter-clockwise" and it worked. Using the battery as a load, CCCV's output started at 0.9A and I could adjust it up to 4.3A - seems that's how much my battery at 11.3V will pull. So it looks it's Capped Current Constant Voltage ;) Thanks, @ruleh, sparked s really interesting conversation
 
Last edited:

timginter

Cable-Tie Ninja
Original poster
Apr 21, 2019
199
80
Big milestone! :D GxR-ONE arrived, all tested and boots. I finally put the bottom aluminium sheet and connected all the cables.

A bit nervous before I booted... but holy crap! :D It works off of a battery :D

Voltage meter is not very visible in low light, it read 11.3V when I shut down the rig. I'll wait for the BMS, properly charge the battery with balancing and then test to see how low it will go, max to 10.5V discharge.

Really happy with how all works, feeling like a kid at Christmas :D Now to mod a server PSU for THE JUICE!
 
Last edited:

timginter

Cable-Tie Ninja
Original poster
Apr 21, 2019
199
80
Have to say, it doesn't look half bad!
Thanks :D A bit rough, but I'm quite pleased. Some cable management left to do.

Was the battery fully charged? How long did it get you to 11.3?
HobbyKing discharged it for delivery, came at 11.4V so not too bad on power draw either. Manual said to charge before first use, though, but I don't have the BMS yet - I'd rather not charge or discharge it too much. Just wanted to test if it works at all, really surprised :D
 

Choidebu

"Banned"
Aug 16, 2017
1,199
1,205
Not too bad a VDrop then. Watchout for those chinese voltmeters. My battery alarm consistently showing .5V more than my multimeter.
 
  • Like
Reactions: timginter

Valantar

Shrink Ray Wielder
Jan 20, 2018
2,201
2,225
Watching this thread, just because it's a really cool project. Sadly I don't think I have anything useful to add in terms of guidance - most of this is well and truly beyond my areas of """expertise""" when it comes to computers and electronics. But I'm really looking forward to seeing how it pans out!

A few questions:
- are you planning to integrate the keyboard into the aluminium plate somehow?
- why not use laptop hinges? Those can be had on ebay for cheap, and should be a lot more compact than other solutions, while otherwise being well suited to the task.
- the current design requires bottom ventilation, which is problematic when placed on any flat surface (unless you have very tall legs it's going to be very restrictive. Have you considered removing the fans from your heatsinks and instead adding side fans with ducting to direct airflow through the heatsinks? This would mean smaller fans of course, but even something like a handful of Noctua NF-A4x20s should be able to cool a PC like this without getting too noisy. This could also allow you to use a much thicker CPU heatsink - something like the L9x65 might fit.
- have you considered buying a cheap HDMI switch and running your two display outputs through it? You can get decent ones for very little money on AliExpress or Ebay, and this would let you have both connected at all times, with just a simple button to switch between which output is connected to the display.
 

CC Ricers

Shrink Ray Wielder
Bronze Supporter
Nov 1, 2015
2,234
2,557
If anything else you can also search for "friction hinges" with general purpose hinges that retain their position when you leave it alone.

Hopefully you get all the battery power management squared away but it looks like you're on the right track there. Now I fully understand why the GxR One comes with that XT90 connector, since a lot of RC LiPo batteries come with it. I'm just trying to power a low-wattage SoC board and small monitor... your project is on a whole different level from mine.
 
  • Like
Reactions: timginter

timginter

Cable-Tie Ninja
Original poster
Apr 21, 2019
199
80
Watching this thread, just because it's a really cool project. Sadly I don't think I have anything useful to add in terms of guidance - most of this is well and truly beyond my areas of """expertise""" when it comes to computers and electronics. But I'm really looking forward to seeing how it pans out!
All comments are welcome :) I learned a lot from this project, hopefully at some point I'll be able to give back, too.

A few questions:
- are you planning to integrate the keyboard into the aluminium plate somehow?
I wanted the aluminium cut and the Microsoft Wireless keyboard set inside, but that was when I planned the hardware behind the screen. It took most of the "bottom box", though - there aren't short and small enough keyboards to sensibly integrate. With hardware mounded underneath it wouldn't fit anyway now.
Initially I wanted to convert a laptop keyboard+touchpad combo to USB, but that's more difficult than I thought. I'm hoping to go back to that idea when everything else is done. For now it's just a wireless combo, though.

- why not use laptop hinges? Those can be had on ebay for cheap, and should be a lot more compact than other solutions, while otherwise being well suited to the task.
If anything else you can also search for "friction hinges" with general purpose hinges that retain their position when you leave it alone.
Hinges I'm using now are the cheapest I could find with a sensible delivery time :D Laptop hinges are usually specific for a particular model, same with friction hinges (constant torque hinges) - they have a set force at which they stay open. Depending on the chassis/layout/hardware/battery your N.m/kgf.cm will change so adjustable torque hinges were the only ones which really worked with the level of "moddability" I wanted. I'm hoping to find some prettier hinges, though - the ones on the prototype are ugly as hell.

- the current design requires bottom ventilation, which is problematic when placed on any flat surface (unless you have very tall legs it's going to be very restrictive. Have you considered removing the fans from your heatsinks and instead adding side fans with ducting to direct airflow through the heatsinks? This would mean smaller fans of course, but even something like a handful of Noctua NF-A4x20s should be able to cool a PC like this without getting too noisy. This could also allow you to use a much thicker CPU heatsink - something like the L9x65 might fit.
Legs are around 10mm high, but from what I tested so far you don't need clearance at the bottom (my RTX2070 is actually lower than the Noctua on the CPU). For a taller GPU, you can always extend the metal rods and create more space between the hardware and the bottom sheet (TinkerCAD drawing has RX 9600XT dimensions). I'm not a fan of blower-type fans in general, in my case either GPU would blow hot air on the CPU or the other way around. When I thought about the design, I was considering heatsink fans + smaller fans on the sides to create better air flow, but with the case being open there's enough air circulation for the fans to pull fresh air, didn't have to force air like in a "conventional" PC case (worked well to keep power draw low, too).

- have you considered buying a cheap HDMI switch and running your two display outputs through it? You can get decent ones for very little money on AliExpress or Ebay, and this would let you have both connected at all times, with just a simple button to switch between which output is connected to the display.
Do you mean an external monitor + the laptop screen? Definitely an option. I have 2 external monitors, but I use DisplayPorts with them - laptop screen can stay connected on HDMI and I can use the other two at the same time.

Hopefully you get all the battery power management squared away but it looks like you're on the right track there. Now I fully understand why the GxR One comes with that XT90 connector, since a lot of RC LiPo batteries come with it. I'm just trying to power a low-wattage SoC board and small monitor... your project is on a whole different level from mine.
Thanks, hope something from my trial-and-errors will come in handy in your project :) GxR-ONE may come with XT90 because that's one of the few sensible connectors which can handle that type of current :D After soldering and assembling KPPX/KPJX I just fell in love with XT90 - with anti-spark and the 4-pin versions there's an XT90 for everything :D
 

Valantar

Shrink Ray Wielder
Jan 20, 2018
2,201
2,225
All comments are welcome :) I learned a lot from this project, hopefully at some point I'll be able to give back, too.


I wanted the aluminium cut and the Microsoft Wireless keyboard set inside, but that was when I planned the hardware behind the screen. It took most of the "bottom box", though - there aren't short and small enough keyboards to sensibly integrate. With hardware mounded underneath it wouldn't fit anyway now.
Initially I wanted to convert a laptop keyboard+touchpad combo to USB, but that's more difficult than I thought. I'm hoping to go back to that idea when everything else is done. For now it's just a wireless combo, though.



Hinges I'm using now are the cheapest I could find with a sensible delivery time :D Laptop hinges are usually specific for a particular model, same with friction hinges (constant torque hinges) - they have a set force at which they stay open. Depending on the chassis/layout/hardware/battery your N.m/kgf.cm will change so adjustable torque hinges were the only ones which really worked with the level of "moddability" I wanted. I'm hoping to find some prettier hinges, though - the ones on the prototype are ugly as hell.


Legs are around 10mm high, but from what I tested so far you don't need clearance at the bottom (my RTX2070 is actually lower than the Noctua on the CPU). For a taller GPU, you can always extend the metal rods and create more space between the hardware and the bottom sheet (TinkerCAD drawing has RX 9600XT dimensions). I'm not a fan of blower-type fans in general, in my case either GPU would blow hot air on the CPU or the other way around. When I thought about the design, I was considering heatsink fans + smaller fans on the sides to create better air flow, but with the case being open there's enough air circulation for the fans to pull fresh air, didn't have to force air like in a "conventional" PC case (worked well to keep power draw low, too).


Do you mean an external monitor + the laptop screen? Definitely an option. I have 2 external monitors, but I use DisplayPorts with them - laptop screen can stay connected on HDMI and I can use the other two at the same time.
I meant for switching between the iGPU and dGPU on the integrated panel :) integrating a switch will allow you to keep both hooked up to the display controller, and just press a button to switch between them when you need to.

As for cooling, I didn't mean blower fans (those would still require bottom intakes), but rather smaller side-mounted axial fans, and I meant removing the stock fans from the heatsinks entirely. I don't doubt that the current design works decently, but it also leaves the case very open, and makes dirt and dust ingress easy. Some small ducts to direct airflow into the heatsinks would make a system like that quite effective (it would be absolutely terrible without them!) - but of course it also requires a significant change in thinking for the layout and design here. Just a suggestion :)

And I know laptop hinges are specced for a specific model of laptop, but if you could find some that fit your needs (likely from a rather large laptop) they would still be far more compact than your current ones while still likely to work well. Of course the battery etc being mounted to the display might mean that there's too much weight up there for that to work.
 

timginter

Cable-Tie Ninja
Original poster
Apr 21, 2019
199
80
I meant for switching between the iGPU and dGPU on the integrated panel :) integrating a switch will allow you to keep both hooked up to the display controller, and just press a button to switch between them when you need to.
Thanks. If I understand correctly, there doesn't seem to be any point in having both plugged in but using the iGPU? :D Initially, I was planning an APU build with an eGPU like a dock, similar to my laptop+eGPU setup, but the more I went through design and use cases, the less sense it made to me:
  • with a mini-ITX mobo, there's enough space in the chassis for a high-end dGPU,
  • with an eGPU (dGPU like a dock) I'd have to also carry around an extra power supply and the GPU/dock,
  • with the dGPU in the chassis, there's really no point in having an APU - better to go with a better/cheaper CPU at similar TDP and use the GPU all the time - even high-end cards have quite low power draw on idle and low loads,
  • with an APU, there's really no point in having the dGPU in the chassis unless you unplug its power before boot (access is really easy on the side) to only use the iGPU and reduce power draw. You're still carrying dead weight and you'd have to shut the rig down, plug the dGPU in and boot to use it,
  • with that motherboard there's still the option to go with an APU-only build (reduced power draw, weight, or with a larger motherboard), but I'd rather have everything in one package focused on performance rather than battery life.

As for cooling, I didn't mean blower fans (those would still require bottom intakes), but rather smaller side-mounted axial fans, and I meant removing the stock fans from the heatsinks entirely. I don't doubt that the current design works decently, but it also leaves the case very open, and makes dirt and dust ingress easy. Some small ducts to direct airflow into the heatsinks would make a system like that quite effective (it would be absolutely terrible without them!) - but of course it also requires a significant change in thinking for the layout and design here. Just a suggestion :)
I understood what you meant :) Without the heatsink fans, though, intake would be from top/bottom, but the side-mounted axial fans would work exactly like blower fans. Adding ducts would make a tight design even tighter and restrict airflow - sounds like a solution waiting for a problem ;)
As for dust, unless you'd use filters with air intake, it will still be better sitting on a desk/lap than a PC case on the floor. I've had my current eGPU standing on the desk for over 3 years and other eGPUs earlier - dust was never an issue. After some time I even stopped worrying about spilling anything on it and had it angled so the hot air was blowing over my coaster and my tea :D Seriously :D
In the end it's what you want to achieve, then how to achieve it - open case gives excellent air flow and cooling and makes the design simple at the same dust/dirt risk as a laptop. No vent-holes on top mean sensible protection from tea/coffee. With an open chassis, holes in the bottom are not as functional/crucial as on a laptop, but they look cool (pun not intended) :D They also give an option to make the box as low as possible, even to the point of the aluminium plate touching hardware.
I appreciate the suggestion :) It's very sound from a theoretical point of view, and similar to why I drilled the vent holes in the bottom, but in the end an open chassis solved the problem better than I expected.

And I know laptop hinges are specced for a specific model of laptop, but if you could find some that fit your needs (likely from a rather large laptop) they would still be far more compact than your current ones while still likely to work well. Of course the battery etc being mounted to the display might mean that there's too much weight up there for that to work.
Yeah, the battery behind the screen and the size of the plate (longer lever) are two main things. Laptop parts are often a pain to work with - the only info you often get is the model it fits, but no proper specs. Hinges are really expensive, too - there are some options on AliExpress at a sensible price, but delivery time was too long for the prototype.
Some of the laptop hinges are adjustable, though - that may be an option in the future. The beauty of the design is that I can replace what I want at any point, even down to hinges and aluminium plates ;)
 

timginter

Cable-Tie Ninja
Original poster
Apr 21, 2019
199
80
Looks great, can't believe I missed it. Thanks for the link!

I was planning the same trapezoidal design at the beginning, but it blows up dimensions a lot to fit all of the components and limits you with hardware. Looks much better than my Mad-Max design, though :D Really like the keyboard and the WQHD screen

EDIT:
Just read through that thread - really cool ideas and design, shame the project focused on design and renders but stopped before a working prototype. Gave me an idea to use a Type Cover Wireless Bluetooth Keyboard with Trackpad - would fit under the screen when hinges are closed.
 
Last edited:

Valantar

Shrink Ray Wielder
Jan 20, 2018
2,201
2,225
Looks great, can't believe I missed it. Thanks for the link!

I was planning the same trapezoidal design at the beginning, but it blows up dimensions a lot to fit all of the components and limits you with hardware. Looks much better than my Mad-Max design, though :D Really like the keyboard and the WQHD screen
That's more along the lines of what I was thinking of in terms of cooling: closed-off top and bottom, mesh sides/front (whichever is more convenient), axial fans blowing through the case, with ducting ensuring air flows through the relevant heatsinks rather than the path of least resistance (much like desktop workstations lay out their cooling). In fact I would advocate for the bottom plate to be pushed up flush against the heatsinks in such a design, as that would further fence in airflow and ensure it passes through the heatsinks. I have no doubt your solution works, though I would think some sort of active exhaust would help ensure air circulation through the case - open cases can often create a lot of internal turbulence and little actual exchange of air inside of the case, so some sort of side mounted fan would help keep proper circulation going. Though of course it might just be open enough that the semi-passive dissipation out of all the sides is enough to keep it cool :)

As for my comments on switching between the GPUs, I thought you were still going with an APU and powering the dGPU only when needed? I've probably missed something - skimming 6 pages will do that :p If you're using the dGPU for everything there is obviously no point in any type of internal switching.

As for laptop hinges, I think this is mostly down to different approaches to DIY - I would just have bought a few and tested them, given how cheap and available they are. But then I would have zero idea of how to calculate the necessary specs either. This, for example, looks both sturdily built and somewhat adjustable, though carrying a thick aluminium plate, large display, battery and related circuitry might still be asking too much. And dirt cheap, obviously. And there are hundreds and hundreds of options. It might even be a slimmer/more elegant solution to use two pairs of laptop hinges (should double their strength/resistance, no?) than something not meant for that kind of use. But then, as I said, I like to just try stuff and see if it works :p
 
  • Like
Reactions: timginter

timginter

Cable-Tie Ninja
Original poster
Apr 21, 2019
199
80
That's more along the lines of what I was thinking of in terms of cooling: closed-off top and bottom, mesh sides/front (whichever is more convenient), axial fans blowing through the case, with ducting ensuring air flows through the relevant heatsinks rather than the path of least resistance (much like desktop workstations lay out their cooling).
That's another part to why it's a shame that project didn't go past drawings and renders - lovely graphics, but I'd love to see a working prototype. I started with EXACTLY the same design, but I run into so many issues and limitations that I re-thought it instead of finding solutions to problems I didn't need to have :D Now my design is smaller, fits up to RX 6900XT with any AMD CPU up to 95W (in theory) and is adjustable/moddable in all important dimensions. I also have the option to mount smaller side-fans if I need to, but can only do it when I notice thermal throttling instead increasing power draw "by design".

Also, that project doesn't have a battery which is one of the biggest challenges when you have limited space and want to make a DIY laptop legal on an aeroplane.

In the end, I'm pragmatic :) It has to work, give me the biggest freedom in hardware and layout, and be easy to build/modify. Looks are secondary. I actually love how rough it is - that Mad-Max DIY style while packing a big performance punch. I ditched mesh sides because it would have meant less freedom with cabling and a fixed height (less freedom with hardware).

In fact I would advocate for the bottom plate to be pushed up flush against the heatsinks in such a design, as that would further fence in airflow and ensure it passes through the heatsinks
I'm not sure that would work unless you like roasted nuts :D Your chassis would act like a heatsink extension, may get really hot :D

As for laptop hinges, I think this is mostly down to different approaches to DIY - I would just have bought a few and tested them, given how cheap and available they are. But then I would have zero idea of how to calculate the necessary specs either. This, for example, looks both sturdily built and somewhat adjustable, though carrying a thick aluminium plate, large display, battery and related circuitry might still be asking too much. And dirt cheap, obviously. And there are hundreds and hundreds of options. It might even be a slimmer/more elegant solution to use two pairs of laptop hinges (should double their strength/resistance, no?) than something not meant for that kind of use. But then, as I said, I like to just try stuff and see if it works :p
Same approach here and thinking along similar lines :) I have way too many "spares" like that laying around, actually - no point in returning when postage is more than the item.
As for hinges, the cheap one had a month lead time - too long when I was building the chassis. When I finish the battery and know exactly what spec I need, I can drill new holes in the prototype and see how different hinges work in practice (usually crash-course with reality). Now I have the prototype to learn on, I can look for something nicer and not worry about lead time.

Interesting discussion, @Valantar :)
 
  • Like
Reactions: Valantar

timginter

Cable-Tie Ninja
Original poster
Apr 21, 2019
199
80
I don't really know how to explain this... I'm not sure if I'm even right or not...
Well anyways here is an attempt:

A current limited voltage supply will not usually supply its max current and max voltage at the same time.
If the load is low (high resistance), it will supply the max voltage and let it draw whatever current it wants as long as its below the set max current.
If the load is high (low resistance) and tries to draw more than the set max current, the voltage regulator will lower its voltage so that the updated current draw will be less than max current.

As an example let's take Vmax=12.6V and Imax=6A.
A 5Ohm load would get 12.6V and draw 2.52A
A 1Ohm load would only get 6V and draw 6A.
If the load got 12.6V instead, it would be draw 12.6A.

However with a step-up converter taking a 12V input, the load will always get at least 12V because a step-up converter can't supply an output voltage lower than its input voltage.


The problem is that in this case we have an unregulated 12V charger for battery voltages <12V, a CC cycle only from 12V-12.6V and a CV cycle finally at 12.6V.
I'm not sure if depending on ocp as a regulator for battery charging at the lower voltages is a good idea.
(On the other hand it might allow for pulse charging :p)
Thanks, I think I finally understand what you meant. I found this topic on CCCV chargers and it looks like Constant Voltage, Limited (max) Current is how they are meant to work as the battery draws less current the more it's charged. There's also a great post on current-voltage relation and how to achieve "true" constant current by modifying voltage until battery is full
 
Last edited:

ruleh

Trash Compacter
Jan 19, 2021
34
24
It turns out that given the right (or really wrong) circumstances, a buck converter can also work as a boost converter in reverse, providing a high voltage on what should have been the input.
So far max measured current was ~17A (at 22V) and max voltage 40V (0A), which is a bit higher than the 18-35V rating.
It's probably not so good for the converter and has no real use anyways since this specific converter is not adjustable.
Oh well... too bad that bidirectional dc-dc converters aren't a thing on ebay and such.
Putting a high load on the 12V rail such that it drops to 12.3V causes 40V to appear at the battery V+ rail.
Code:
AC-DC ----------------------->|------------------------DC-ATX
(12.6V)        |                             |
               |                             | (out: 12.3V)
             BOOST                         BUCK
             CONVERTER                     CONVERTER
               | (out: 25.3V 1A)             | (in: 18-35V rated)
               |                             |
               -------------------------------
               |   40V when ac-dc at ~12.3V
               |
              BMS
               |
           6S BATTERY
           (18-25.3V)