javispedro1 wrote on 2026-08-08, 21:31:

I would order one of those with fries, please :) […]
Show full quote

Heathen wrote on 2026-08-08, 14:16:

might be tempted to spin up an internal picogus

I would order one of those with fries, please 😀

More seriously, one of my complains is still all the emulated hardware from the 8051 or whatever new thing is on the book386. I can understand why some people may prefer the real OPL3 , but for all the other hardware, I don't understand the reason, since I think that by now we are emulating it better on a RaspberryPi than these random 3rd party clones...
A single RPi could probably handle GUS|SB + trackpad + keyboard + USB mice + ethernet|wifi and likely still have cycles to spare? And probably draw more useful information in the "new" display while doing it...

One of the small dreams I had is to report the charge of the battery to the x86 side ... which can't be done due to the closed 8051 fw. As far as features go, I know this one is pointless, but then one could add an APM BIOS and get the fancy "battery" icon in win95.

Im not sure how much the 8051 (I assume you mean the STC8G2K64S4, I think the LCD controller has another one embedded! maybe the CL5420 does too???) is really doing in terms of emulated hardware anymore? I don't think its doing the keyboard stuff like the one in the pocket (and don't peripheral, external keyboards have things like 8051s or another micro in them anyway? OK the scanning on the pocket seemed to have issues...), I think (but not sure) its mainly just sniffing the keyboard data for the Fn whatever, monitoring loads of voltages and +5V current, sending things to the OSD in the LCD controller, and turning a few things ON/OFF based on keyboard input. It looks like the only real part of the 'PC' that its involved with now is the trackpad/com2 EDIT and even that is via the M5113?
Like you say with the battery monitor data, its really quite separate in many respects.

As such I think most it is pretty much sitting 'alongside' most of the 'PC'.
Im seriously tempted to do a respin of the pico, or buy one if someone else does it... Ive not used one but hats off to those that have developed it, the feature set is very tempting...
having said all that, I like that the pocket and book are pretty much real hardware, I think Id rather go in the direction of more 'real' hardware than emulation, so adding 8 bit mono soundblaster (yep the 'DSP' is another 8051!!!) is a tempting route for getting sampled sound, or failing that the es1868 I think Id still see as period correct hardware - I never had issues with them back in the day....

Was taking some power measurements while charging the batteries of the BOOK386 (see picture). Nothing that special. But what I see when the battery of the BOOK386 is at 100% it still uses around 0.8A/10W.

Maybe not an issue but I was using an USB powerbank to charge the battery and that means it will deplete the powerbank if you don't disconnect it when it's at 100%. So be careful with that. 😉

As promised hereby the covers for the BOOK386. There are 4 options to choose from. I went with the second option in the zip file. Looked good with the dark translucent PETG filament and is also easier to remove. I also included a 3mf file with all the 4 options included. This is configured for a Bambu Lab X1C btw. So it needs to be slightly modified for other 3D printers.

Photo shows it a bit whiter than it is in IRL (because of the flash) but it looks very good.

Jules_nerd wrote on 2026-08-09, 16:01:

I don't think its doing the keyboard stuff like the one in the pocket (and don't peripheral, external keyboards have things like 8051s or another micro in them anyway? OK the scanning on the pocket seemed to have issues...), I think (but not sure) its mainly just sniffing the keyboard data for the Fn whatever, monitoring loads of voltages and +5V current, sending things to the OSD in the LCD controller, and turning a few things ON/OFF based on keyboard input.

Why do you think so? Who is mapping keyboard wires into PC scancodes ? As in the Pocket386, you need someone to do this mapping, and it is again wrong in many details.
Most annoying one is that the new "Right" Alt key gives the same scancode as the Left Alt, making the new key kinda pointless, all international keyboard layouts are still unusable.
I have not checked the wiring to see if the issue is from the wiring, which would make this more difficult to fix, but as it stands now as long as we cannot reprogram these chips these annoyances are impossible to fix.

Jules_nerd wrote on 2026-08-09, 16:01:

It looks like the only real part of the 'PC' that its involved with now is the trackpad/com2 EDIT and even that is via the M5113?

But.. how? It doesn't even look like the M5113 is programmable at all, and the datasheet doesn't mention anything about trackpad. I do not know what is connected to that serial input, but I bet it is one of the 8051s with closed program ( not sure which one 😜 ) , and almost guarantee it is not real hardware . I'd wish it was , because then I could maybe find a spec for the protocol or even the correct drivers...

Well, I started the post with 'Im not sure' and whilst I have spent a bit of time & looked at it a fair bit, 'I 'm not sure' is very much how I hope people read it!
Im a hardware engineer by trade and an analogue one at that!! I wouldn't even know where to start on things like writing drivers! for sure I worked alongside guys who did programming, so I get some of what they do & I designed digital hardware as well, though never for PC hardware & so Id be very,very surprised if you weren't a great,great deal more competent than I on this kind of thing! 😀

Ive only the schematics to go from, which could be wrong & are definitely missing bits, like the IDE/CF PCB & the keyboard,
I was hoping (!) that he'd improved things & gone for a standard keyboard controller as the connections on the pocket schematics for the 8051 to the keyboard (all the KEY_ROW* & KEY_line*) are no longer there in the book schematics,
The only connections I can see on the 8051 that probably go to the keyboard are KBTxD and KBRxD (i think KBDATA i s going to the PS2 port?? not sure, I cant find that connector on the schematics). So i assumed the connector J1 with these signals goes to another PCB with another micro/keyboard controller on it??
I suppose without details of what's on this keyboard PCB the protocol between it and the 8051 could be anything , maybe totally non standard..
maybe he's put another 8051 on the keyboard PCB with similar code as the pocket! 😀

Ive no more to go on than that, so if you've tested the keyboard and its the same as the pocket, faults and all... then Im sure you are correct.

Yep,I agree, it looks like the track-pad is connected to the 8051, and the 8051 has connections to the COM2 port on the M5113, so my presumption is as you say, that the 8051 is doing the trackpad to RS232 mouse translation work, so yes that bit of code in the 8051 is part of the 'PC', for want of a better way of putting it..
RE-reading what I put I can see your interpretation - what I meant was that, yes I do think the trackpad goes via the 8051, but then on to the COM2 on the M5117.
Though I thought the rest of that 8051 looks kinda separated from most of the operation of the 'PC' - though as above if i got that wrong on the keyboard then I wont be surprised...
All the trackpad display, OSD, PSU monitoring etc etc looks separated,so no 'emulation' there...

Yep, its a shame the 8051 code isn't public, I wouldn't have a clue what to do with it, but it seems there are those on these forums that certainly do!

more on the m5117 to follow in another post...

Jules_nerd wrote on 2026-08-10, 16:15:

Well, I started the post with 'Im not sure' and whilst I have spent a bit of time & looked at it a fair bit, 'I 'm not sure' is very much how I hope people read it!

Sorry, don't read my above questions in accusatory tone, read them in "I'm frustrated and I really want to know" tone 😀
I should open my device and have a look, but I'm hesitant yet on whether to keep it or sell it ...

I think that the right Alt & Ctrl keys having the same keycodes as their left counterpart is going to affect gaming too, so maybe enough people will complain so that the designer fixes this in a later revision or something 😜
Interestingly the menu key (which was previously used by the mouse emulator in the pocket386) now produces a keycode (e0 5d, which I think it's actually the correct one), so I could remap that to a right alt.

I have no idea whether one can really even reprogram these chips (i.e. they are not factory protected?), or resoldering new ones would be required.

I dont know how BIOS setup on the book is involved with the RS232, Im just asking the Q, are the BIOS settings for these ports relevant? are they just a hangover from a copied BIOS?

I did try altering the parallel port BIOS settings on the pocket & analyzing the hardware/schematics (which are different to my pocket386 hardware, there are some optimizations on the logic) and the settings didn't seem to do anything,
I'll happily try similar tests on the Book386's serial ports when it arrives.... its late! 🙁

My understanding (I guess Im asking have i got this right??) from looking at the schematics & the M5113 datasheet:
The M5113 supports hardware config (page 13-20) and also has a register that can be programmed to set up serial and parallel port address etc
For hardware config (page 20)there are internal pull up 20k resistors on the relevant pins that are active/latched on reset, so if they weren't connected, you'd get some default config, or you can use some hardware that forces a particular state on reset. I presume that this config persists until the config registers are written to by BIOS at startup or potentially some other program?

The pins are dual purpose, many seem to be internally assigned to UART1, UART2 and DMA after being read during reset - so the config at start up does depend on the behavior of what's connected to these pins at during reset .
I can see no specific hardware in the book 386 that forces these pins to a particular state at startup.
Some are disconnected (just internal pull ups), some go to the MAX232, one to a DMA line, some to the 8051 (UART2/com2) .

So unless the BIOS writes to the config registers, youd be stuck with these settings.... do the BIOS settings actually work?

If anyone wants to verify...
again, Ill have a go, when mine turns up...
That the manual etc states that the book supports different parallel port modes (PS2, EPP, ECP) makes me think the designer has at least implemented that part of programming the control register???
EDIT I guess what Im thinking here is, if the M5117 config isnt explicitly programmed by the BIOS, the hardware option for setting the config at reset looks rather ill defined on the book386 , could this be responsible for the weird trackpad/random mouse driver behavior?

javispedro1 wrote on 2026-08-10, 16:33:

Sorry, don't read my above questions in accusatory tone, read them in "I'm frustrated and I really want to know" tone :) I shou […]
Show full quote

Jules_nerd wrote on 2026-08-10, 16:15:

Well, I started the post with 'Im not sure' and whilst I have spent a bit of time & looked at it a fair bit, 'I 'm not sure' is very much how I hope people read it!

Sorry, don't read my above questions in accusatory tone, read them in "I'm frustrated and I really want to know" tone 😀
I should open my device and have a look, but I'm hesitant yet on whether to keep it or sell it ...

I think that the right Alt & Ctrl keys having the same keycodes as their left counterpart is going to affect gaming too, so maybe enough people will complain so that the designer fixes this in a later revision or something 😜
Interestingly the menu key (which was previously used by the mouse emulator in the pocket386) now produces a keycode (e0 5d, which I think it's actually the correct one), so I could remap that to a right alt.

I have no idea whether one can really even reprogram these chips (i.e. they are not factory protected?), or resoldering new ones would be required.

NP, I guess we cant do anything on this... Ive a vague recollection tat someone was asking about metering the keyboard connections to work out whats going on, but if we cant program a new 8051 on the pocket, or whatever is used n the book, is there any point? if its likely to go anywhere I dont mind having a crack at that next time i open it up...

Even though mine hasn't turned up yet & now others are posting about issues Im beginning to wonder how much of an upgrade it is! 😁
well actually there are some reliability probs on the pocket, that without my mods....

Oh I i got it! the keyboard connector, for external PS2, is right next to where i was looking... its EDIT P1, though I was wrong, its no longer connected to the 8051, the schematics kinda still have it there, but its been re-purposed to #COVOXEN and on the PCB thats all its connected to.

Yoghoo wrote on 2026-08-09, 00:00:

Will take a look at the internals. But that probably will have to wait until next weekend.

If it helps, I happened to notice looking at the schematics, the PS2 mouse connector "P2" is directly connected to the M6117, there really isnt much hardware on the 2 lines (IRQ12 & MOUSECLK ) to go wrong...
IRQ12 has a 4k7 pull up R44, a 47pF cap to ground (EMC?) C7 & its connected to the co-processor PCB connector and the M6117.
MSCLK similarly just has edit this isnt 47k , its 4k7 47k pull up R43 &47pF C8 to ground the on to the M6117 - the sort of mention of it on the schematic next to the 8051 is out of date, its not connected there on the PCB design we have...

You can prob tell quite a lot from metering it out without even opening it up... obv you have to open it up to check continuity to the M6117....

Im sure you'd get that from the schematics yourself , I was just looking to see if the 8051 plays any roll with thePS2 mouse and happened to notice...

Jules_nerd wrote on 2026-08-10, 17:22:

If it helps, I happened to notice looking at the schematics, the PS2 mouse connector "P2" is directly connected to the M6117, t […]
Show full quote

Yoghoo wrote on 2026-08-09, 00:00:

Will take a look at the internals. But that probably will have to wait until next weekend.

If it helps, I happened to notice looking at the schematics, the PS2 mouse connector "P2" is directly connected to the M6117, there really isnt much hardware on the 2 lines (IRQ12 & MOUSECLK ) to go wrong...
IRQ12 has a 4k7 pull up R44, a 47pF cap to ground (EMC?) C7 & its connected to the co-processor PCB connector and the M6117.
MSCLK similarly just has 47k pull up R43 &47pF C8 to ground the on to the M6117 - the sort of mention of it on the schematic next to the 8051 is out of date, its not connected there on the PCB design we have...

You can prob tell quite a lot from metering it out without even opening it up... obv you have to open it up to check continuity to the M6117....

Im sure you'd get that from the schematics yourself , I was just looking to see if the 8051 plays any roll with thePS2 mouse and happened to notice...

Disassembled the laptop (not very enjoyable) and measured the 5V and ground of the PS/2 mouse port. They are okay. Cleaned up some flux around the ALI chip. Made some photos from the front and back.

Will do some more measurements tomorrow. I am not exactly a hardware repair guy so if you can give some pointers how and where to measure that would be great.

Was there anything in particular the rest of us should keep in mind when taking these apart?
Don't loose sight of the fact you've bought a new product that should work , so you've every right to get back the your vendor/supplier and ask them to sort it for you....

Well youre really looking for continuity of the connector & tracks, making sure there are no short circuits the right value parts are fitted.
It sounds like you've done the first step, a visual inspection of the tracks, parts & solder joints.
Next is to use a DMM to check the continuity of tracks and the connector. If you have a miniDIN PS2 connector, great, if not you could just push thin (don't force a thick wire in that might spread out the connector ) wires into the connector for now, then if all measurements below turn out OK then double check the PS2 connector by dismantling a mouse and use your meter to bell out which pins are connected to which colour wires in the mouse. Hopefully you could unplug the wire in the mouse to ensure the circuits in the mouse are disconnected and wont give you false readings.
you are looking for continuity from pin1 of the PS2 to the end of R44 & c7 nearest the connector, then on to pin 155 of the M6117(see datasheet)
also check the resistance between pin 1 +5V (pin4 on the PS2) most DMMs should give you about 4k7 here - the resistor R44 (yes the M6117 is in circuit as well, but most meters don't supply enough voltage for that to matter).
Do the same checks for pin5 of the PS2 connector - continuity to R43 & C8 , then on to pin34 of the M6117 & check for 4k7 to +5V(pin 4)
It s probably worth checking with a meter set to resistance that there are no shorts to adjacent pins the M6117, or to GND, looking at the PCB design there is a GND flood that goes right under the capacitors, so short here is possible...

In theory if a really large value capacitor(s) has been fitted in error it could stop it working, but its not easy to measure & it wouldn't be high on my list of possible issues...

TLDNR, i think this design would benefit from swapping the PSU capacitors to ones with low ESR, designed for switching regulators...

@Yoghoo, Im going to order some parts for other projects anyway, & i want to add some low ESR replacements for my book, before it arrives (and the pocket for that matter). The standard ones fitted are 7mm high in the datasheet, Im guessing there isnt space for taller ones? how does it look to you?

Id been meaning to check out the ESR of the caps in the pocket, but since you sent the pics of the book....
The following is largely theory, my Book386 is still not in the UK, the customs notice I got was from the Netherlands... - I was wrong there...

I did notice from your pictures the capacitors fitted to the main buck regulator... these appear to be Cheng VT and VH series (not 100% sure on the VH, there is less markings).
If so, these are general purpose caps not primarily intended for high frequency regulators...
The ESR of the 220uF output caps at 120Hz calculates to be about 1.2Ohm (Tan delta .20), they have a max ripple current 120mA(at 120Hz), which doesn't sound great, but the ESR is likely to be quite a bit lower at 180kHz that the regulator is operating, Id guess at maybe 600 to 120mOhm??.. with 3 of them in parallel they should handle things not too badly... ripple current with 12V input is about 345mA & its less than half that when running off battery....
so I guesstimate 15 to 70mV ripple?? any track resistances etc will increase this, but the ceramic 1uF bypass caps will help knock this down a little (and esp any additional inductive spikes) , but 10uF would likely help more!!
so the steady state isnt tooo bad
However.... when the input to the regulator swaps between 12V to battery (the lower the charge on the battery the worse this is), there will be a period where the duty cycle needs to change & the regulator wont contribute as much current through the inductor as it needs, under these circumstances the output capacitors have to source more current than steady state, here the ESR will be more of a factor, leading to a dip in the +5V supply voltage, I think this is likely to contribute to the issue mentioned in the manual where it resets on unplugging the +12V with low battery...

unfortunately the input caps have to cope with a much higher ripple current in a buck converter, depends on the circumstances but say around half the output current....
The 100uF caps on the input side have a ripple current rating of 86mA!!! ESR approx 2.2Ohm @120Hz...(tan delta 0.16) again 3 of them in parallel...
Now, without the stylized 'C' in the dark printed area of the caps Im not 100% sure Ive got the right manufacturer/series/data.... so I could be wrong.... But
say its running at 2A (much depends on circumstances might be near half that) on the +5V rail, that gives about 300mA per cap on the input side
The 1uF bypass caps will help out a bit again, but a decent physical size 10uF might be better...
Also the reservoir capacitor in the supply will source some of this current, but with the mosfet switch resistance and cable resistance Id expect a fair bit of it to come from the 100uF caps..
Hard to say how much but if over 25% is from the input caps they will have a reduced life.

I wouldn't worry too much if you don't have the capability to change them, SMD electrolytics can be a bit of a pain to work with.... and I reiterate, this is just theory, I haven't measured a real one yet...
and ultimately you can just use it and wait until the input caps go.... and change them after that, if needed...

more to follow on this issue in the pocket....

Jules_nerd wrote on 2026-08-11, 15:08:

@Yoghoo, Im going to order some parts for other projects anyway, & i want to add some low ESR replacements for my book, before it arrives (and the pocket for that matter). The standard ones fitted are 7mm high in the datasheet, Im guessing there isnt space for taller ones? how does it look to you?

I don't have time today to take a look so it's still disassembled. But I think higher ones will definitely not fit. It's really crammed between the motherboard and the cards above.

When the Book486Dx will be released I will buy it!

Jules_nerd wrote on 2026-08-10, 16:52:

I did try altering the parallel port BIOS settings on the pocket & analyzing the hardware/schematics (which are different to my pocket386 hardware, there are some optimizations on the logic) and the settings didn't seem to do anything,
I'll happily try similar tests on the Book386's serial ports when it arrives.... its late! 🙁

Interesting, I also remember it did nothing in the Pocket386, but it definitely seems to be doing something in the Book386!
I tried changing COM2 's IO base to 0x2E8 and not only it did change the BIOS data area (i.e. what the software uses to figure out where the port is), but the touchpad was then also responding on 0x2E8! (and no longer on 2F8!).
I then tried changing COM1 to 0x3E8... and the BIOS decides to swap COM1&2, for some reason. BDA now claims COM1 is now in 0x2F8 and COM2 is on 0x3E8 . (not sure why, maybe some compatibility thing).
But the physical serial port does indeed respond @0x3E8. And nothing responds @0x3F8 anymore.

I tried the same on the Pocket to be sure I remember it correctly. It definitely does nothing. It does not change the BDA nor change the actual settings.

So this seems to work better on the Book than on the Pocket, and I guess the BIOS can indeed configure the m5113 ?
Since I think this is the same BIOS binary, my wild guess is that this BIOS maybe have been designed for a companion m5113 to begin with?

Note that interestingly the Pocket386 BIOS setup allows you to change the other serial port settings (IIR mode or not, Duplex True or False), while the Book386 just shows N/A there and doesn't allow you to change them at all.
Also both BIOSes have the issue that if you do "echo hello > com1" it will just hard lockup (or crash if using emm386).

Thanks for checking that out! sorry it seems it was a wild goose chase!
But at least we do know that all work properly.. 😀
Yeah the com2 connections from the M5113 to the 8051 are just Rx & Tx I think, so I think the 8051 doesn't care what address its at...
edit: sorry! 🤣 I get what you mean now! you were saying not only does the BIOS think its changed the address, it definitely has because it works!!:-D

I think the m5117 does support ir modes etc, though I presume the book doesn't have hardware to do it..
Yeah I think your right, how could it work if the BIOS wasn't designed for a M5117??? it looks like it was a fairly common chip...
unless perhaps it somehow was adopted as a standard way of setting multi I/O chips up?edit: nope, I checked this out, it has some similarities with other multi IO chips in the range but that's all
Anyway at least we know it works!

So.. unfortunately I'm out of ideas about the trackpad/mouse driver problems people are experiencing, I cant think of any other options than the 8051 code?

Yoghoo wrote on 2026-08-11, 18:05:

I don't have time today to take a look so it's still disassembled. But I think higher ones will definitely not fit. It's really crammed between the motherboard and the cards above.

thanks for the info, thats helpful. 😀

EDIT, Deleted a whole load of rubbish about possible hardware conflict, I don't think a hardware conflict on COM1 is likely, or not on the pocket at least... Its always passed a variety of different hardware check utilities & Ive used it a lot to transfer data using Zterm to a classic mac, never had any issues (well once I fixed a problem with that tiny B2B connector on the serial card) So you can probably ignore the rest of this post!!

javispedro1 wrote on 2026-08-11, 21:28:

Note that interestingly the Pocket386 BIOS setup allows you to change the other serial port settings (IIR mode or not, Duplex True or False), while the Book386 just shows N/A there and doesn't allow you to change them at all.
Also both BIOSes have the issue that if you do "echo hello > com1" it will just hard lockup (or crash if using emm386).

Incidentally, I assume that on the BOOK you can see BIOS config options for LPT2? In the soundcard manual's description of the COVOX it mentions that a real LPT2 hardware must exist if your COVOX software looks for a parallel port. If the BIOS allows you to set a different address for LPT2 it may 'break' covox software that does these checks; Whilst BIOS may be fine setting the 'real' LPT2 address in the M5113, the COVOX has its own, separate address decoding is fixed in hardware to the default address.

BTW I tried "echo hello > com1" on both my pockets, one has a serial card, the other does not, FWIW it only hangs on the one with the serial card installed, it seems DOS just ignores if you echo to one that doesn't exist?

Jules_nerd wrote on 2026-08-11, 13:28:

Was there anything in particular the rest of us should keep in mind when taking these apart?
Don't loose sight of the fact you've bought a new product that should work , so you've every right to get back the your vendor/supplier and ask them to sort it for you....

Missed this part. There are 2 cables which you have to detach to separate the top part. Cables are not very long. Also there are a lot of screws. But the most annoying thing was to get the daughter boards from the motherboard. I needed much more force to get them off than I liked. Also needed to pry them a bit loose with a screwdriver. Battery is under one of them so I hope it lasts a long time. To remove the motherboard you need to get 2 daughter boards off as there are a couple of screws under them. You need to slide one side up a bit to get it out of the case because of the power button.

I talked with the seller and at first he wanted to give me a PS/2 mouse for free. But after some talking (as I know it's not a mouse problem) I said just send me the ISA extension kit for free and call it a day. He agreed. So I am pleased with the outcome. Even if I can't get it to work it's not that important to me as I have a dozen serial mouses.

Does sound a bit of a pain to get apart...

Well, there really isn't much to go wrong on the PS2 mouse port in this design, so long as that input on the M6117 is working...
Did you get a chance to check it out?