• Please review our updated Terms and Rules here

Pocket 386

kubino said:
Seems like building one is the only option at the moment. I went this route and completed 2 units. Also have 3 more partially assembled, because minimum order was 5 PCBs. If I succeed I might have 3 spares.
Click to expand...
A few things made me decide to not make any more Femto PicoGUSes for sale. I sold my last few at VCF Midwest this year to people who asked me directly about them.
- Assembling them is a pain. The IDC pins on the keyed connector need to be bent and alignment is touchy as you're soldering the connector on, and it has to be held in a vise when doing so. It's fine if you're making a couple for yourself but in a batch of 50 or more it gets exhausting. I originally used a non-keyed connector that was far easier to solder but after someone fried their Femto after plugging it in upside down I decided to go with a keyed connector.
- The issue with needing to cut JP3 really made me feel bad about the project. I felt awful about shipping a product that had the potential of damaging people's equipment.
- The Pocket 386 (at least mine) is very unstable with regards to keeping its CMOS settings and RTC. It seems to get even worse when a Femto is plugged into it. I don't want people to have a bad experience get even worse when using the PicoGUS.
 
polpo said:
- Assembling them is a pain. The IDC pins on the keyed connector need to be bent and alignment is touchy as you're soldering the connector on, and it has to be held in a vise when doing so. It's fine if you're making a couple for yourself but in a batch of 50 or more it gets exhausting. I originally used a non-keyed connector that was far easier to solder but after someone fried their Femto after plugging it in upside down I decided to go with a keyed connector.
Click to expand...
That's the main reason I wondered if more would be available for purchase; I've never been very good with soldering tiny components, but I absolutely understand your reasoning.

FWIW, I've had zero issues with my Pocket386 model and the standard ISA PicoGUS connected to an ISA breakout board from the same AliExpress supplier as the Pocket386, so I'm very happy with that :)
 
polpo said:
A few things made me decide to not make any more Femto PicoGUSes for sale. I sold my last few at VCF Midwest this year to people who asked me directly about them.
- Assembling them is a pain. The IDC pins on the keyed connector need to be bent and alignment is touchy as you're soldering the connector on, and it has to be held in a vise when doing so. It's fine if you're making a couple for yourself but in a batch of 50 or more it gets exhausting. I originally used a non-keyed connector that was far easier to solder but after someone fried their Femto after plugging it in upside down I decided to go with a keyed connector.
- The issue with needing to cut JP3 really made me feel bad about the project. I felt awful about shipping a product that had the potential of damaging people's equipment.
- The Pocket 386 (at least mine) is very unstable with regards to keeping its CMOS settings and RTC. It seems to get even worse when a Femto is plugged into it. I don't want people to have a bad experience get even worse when using the PicoGUS.
Click to expand...
With respect to the overall design, what would need to be done, engineering-wise, to fix the J3 issue?
 
As a potential fix to the forgetful CMOS settings, I've been considering replacing D11 (and possibly also R2) with 0-Ohm links/solder blobs, and removing D12 without putting anything in it's place. This will make it so CMOS power input is directly connected to the battery. However, I'm not sure that this is where the problem actually comes from, so can anyone with more experience chime in?

On the schematic from 8086cpu, these components are at B3 on page 1 and they deal with deriving the CMOS backup voltage from the main 5V power rail when the laptop is on, and the battery while turned off.
 
I came to the same conclusion. My guess was between the diode forward voltage drop and R2, the voltage when VCC is 0 is under whatever threshold is "okay" for the M6117C.

Did you end up doing this mod, solder-bridging R2 at least? I was going to keep the diodes in there, so VCC and V3.7V didn't back-feed voltage to each other, since it looks like that's why they might have added those. That's at least what I was thinking anyway.

EDIT: Took a YOLO shot at bridging R2 and .... it's definitely not keeping CMOS settings. I have an "early model", i'm not sure if it's an issue with the reworked model, or if I should just remove D12/D11 and just solder in a coin cell.
 
Last edited:
brittman914 said:
I came to the same conclusion. My guess was between the diode forward voltage drop and R2, the voltage when VCC is 0 is under whatever threshold is "okay" for the M6117C.

Did you end up doing this mod, solder-bridging R2 at least? I was going to keep the diodes in there, so VCC and V3.7V didn't back-feed voltage to each other, since it looks like that's why they might have added those. That's at least what I was thinking anyway.
Click to expand...
I did remove the two diodes and replace D11 with a solder link, but I did not also replace the resistor. Unless replacing R2 is also necessary, this does not solve the problem as it still occurs for me. Best way to force a CMOS corruption/reset for testing is by turning the laptop off and on rapidly.

Maybe the real culprit is digital noise on the VCMOS line causing corruption? Checking the BIOS with AMI BCP shows the setting to auto-reset CMOS upon detecting curruption is enabled, probably this is what's happening and it's just silent about when this happens.

Alternatively, perhaps C106 on the actual motherboard is the wrong value, as it looks awfully small for a 22uF cap...
 
wishindo said:
I did remove the two diodes and replace D11 with a solder link, but I did not also replace the resistor. Unless replacing R2 is also necessary, this does not solve the problem as it still occurs for me. Best way to force a CMOS corruption/reset for testing is by turning the laptop off and on rapidly.

Maybe the real culprit is digital noise on the VCMOS line causing corruption? Checking the BIOS with AMI BCP shows the setting to auto-reset CMOS upon detecting curruption is enabled, probably this is what's happening and it's just silent about when this happens.

Alternatively, perhaps C106 on the actual motherboard is the wrong value, as it looks awfully small for a 22uF cap...
Click to expand...
Could be, I'm betting that's exactly what it is though. I think my next plan is yank the diodes and just stick a CR2032 rechargable in there and call it a day....then it doesn't matter what happens on VCC. Positive on VCMOS side of either D11/D12, and ground somewhere close to C106
 
So presently, if the device stays plugged in, the CMOS settings stay saved, which is delightful. I've ordered a small CR2032 cell and i'm just going to remove the diodes and add the coincell + to one of the diode inputs and call it a day, but this certainly works well enough for now

also, regarding that capacitor, I'm also betting a different/larger one would help alleviate voltage "spikes?" that might not be helping here.
 
brittman914 said:
Could be, I'm betting that's exactly what it is though. I think my next plan is yank the diodes and just stick a CR2032 rechargable in there and call it a day....then it doesn't matter what happens on VCC. Positive on VCMOS side of either D11/D12, and ground somewhere close to C106
Click to expand...
Before you do that, try this BIOS with the auto-reset option disabled. I think that's fixed it for me!

Change log from AM1: More options unhidden, 'Initialize CMOS in Every Boot' disabled in BIOS Options.
 

Attachments

wishindo said:
Before you do that, try this BIOS with the auto-reset option disabled. I think that's fixed it for me!

Change log from AM1: More options unhidden, 'Initialize CMOS in Every Boot' disabled in BIOS Options.
Click to expand...
I will be trying this version out latter today. Thank you.
 
wishindo said:
Before you do that, try this BIOS with the auto-reset option disabled. I think that's fixed it for me!

Change log from AM1: More options unhidden, 'Initialize CMOS in Every Boot' disabled in BIOS Options.
Click to expand...
I had to flash back to the BIOS that shipped. Using your edited BIOS my CF card would no longer be detected once powered off or on a restart. I would have to go back into the settings and save for the CF card to be detected again.
 
Hey guys...has anyone had a problem with a sticking key on their Pocket386? I have an occasionally sticking left arrow key right out of the box that I can wiggle it and it will pop back (at least it does for now ;)) but it's a concern and a nuisance at the very least. Other than returning it (and possibly seeing this issue again in another), are there any simple fixes? Has anyone popped the keys off these keyboards to see what mechanism this Pocket386 keyboard has? Any experience, direction or expertise is welcomed and thank you!
 
spaceboyjeffy said:
Hey guys...has anyone had a problem with a sticking key on their Pocket386? I have an occasionally sticking left arrow key right out of the box that I can wiggle it and it will pop back (at least it does for now ;)) but it's a concern and a nuisance at the very least. Other than returning it (and possibly seeing this issue again in another), are there any simple fixes? Has anyone popped the keys off these keyboards to see what mechanism this Pocket386 keyboard has? Any experience, direction or expertise is welcomed and thank you!
Click to expand...
Seems the KB isn't as good as the Book8088 so I use the PS/2 breakout with an external keyboard and HP optical mouse. I think the stock KB is not made for high-mileage, and especially not gaming
 
What is a good program to image the CF? I have one of these coming in the mail soon and I'd like to have a backup image before I tinker with it. I've tried to image bootable CF hdds that I've configured before but when I test the backup the backup is never bootable...
 
dukeofurl said:
What is a good program to image the CF? I have one of these coming in the mail soon and I'd like to have a backup image before I tinker with it. I've tried to image bootable CF hdds that I've configured before but when I test the backup the backup is never bootable...
Click to expand...
Win32 Disk Imager works well for saving/loading image files.
 
dukeofurl said:
What is a good program to image the CF? I have one of these coming in the mail soon and I'd like to have a backup image before I tinker with it. I've tried to image bootable CF hdds that I've configured before but when I test the backup the backup is never bootable...
Click to expand...

I suggest USB Image Tool:


It can make an image of anything attached to USB, regardless of what it is, SD, CF, etc... it's completely media and format agnostic. I've been able to backup and restore images on CF that had DDO partitions on them that Windows didn't even recognize but this program had absolutely no trouble reading it and creating a perfect image. It's my favorite agnostic imaging tools.
 
A US MSX machine would be interesting to have but I don't like the laptop form factor (a console to connect to a TV would be better).
 
Plasma said:
I see it has a mechanical keyboard and 4:3 screen. Looks like they are making some improvements.
Click to expand...
Looks like more of the things that are external add ons on the pocket 386 are integrated into the case on the msx. Though it seems from the pictures that the case might be bulkier compared to the 386.
 
You must log in or register to reply here.
Back
Top