Author Topic: Goldfish - A super thin USB-C Pro Micro clone  (Read 21668 times)

0 Members and 1 Guest are viewing this topic.

Offline dr_derivative

  • Thread Starter
  • Posts: 292
  • Location: Cork, Ireland
  • Not a real doctor
Re: Goldfish - A super thin USB-C Pro Micro clone
« Reply #50 on: Wed, 03 April 2019, 17:41:49 »
Thanks for the reply! That would be great. I'm generally profficient in electronics but I haven't done any PCB prototyping yet so would appreciate any help you could provide.

I'd love to run the GB - if there's still interest of course. Trouble is I'm in EU and the vast majority of the mk scene seems to be in the US, so not sure how many orders I could get. Plus PCB manufacturing is probably more expensive here. Will see about that.

Sent from my OnePlus 6T with a rubbish gimmicky bloody annoying unreliable in-display fingerprint reader using Tapatalk

I'm afraid I can't help much with the GB since I'm EU too.

If you want to chat about the board feel free to PM me on Discord, my name is in my signature.
GH/Reddit/DT: dr_derivative | Discord: dr_derivative#3618 | IG: dr_derivative_mk
MK Ireland Discord: https://discord.gg/2KdxJzR

Offline ju6ju8Oo

  • Posts: 118
Re: Goldfish - A super thin USB-C Pro Micro clone
« Reply #51 on: Mon, 21 October 2019, 18:45:16 »
Is the rev C production ready?

Offline dr_derivative

  • Thread Starter
  • Posts: 292
  • Location: Cork, Ireland
  • Not a real doctor
Re: Goldfish - A super thin USB-C Pro Micro clone
« Reply #52 on: Tue, 22 October 2019, 14:25:52 »
Is the rev C production ready?

I don't think anyone has tested Rev. C, but it should be OK.

 The only major change from Rev. B is that the pins at the bottom are reordered to match the Elite-C. And I know other people have used Rev. B without problems.
GH/Reddit/DT: dr_derivative | Discord: dr_derivative#3618 | IG: dr_derivative_mk
MK Ireland Discord: https://discord.gg/2KdxJzR

Offline ju6ju8Oo

  • Posts: 118
Re: Goldfish - A super thin USB-C Pro Micro clone
« Reply #53 on: Thu, 24 October 2019, 07:10:46 »
I want to give it a try. Any tutorial for hand soldering the goldfish? I’m a complete newbie in SMT soldering

Offline dr_derivative

  • Thread Starter
  • Posts: 292
  • Location: Cork, Ireland
  • Not a real doctor
Re: Goldfish - A super thin USB-C Pro Micro clone
« Reply #54 on: Thu, 24 October 2019, 14:59:19 »
There's nothing specific for the Goldfish, but there are lots of SMT soldering guides and stuff on youtube if you search for it.
GH/Reddit/DT: dr_derivative | Discord: dr_derivative#3618 | IG: dr_derivative_mk
MK Ireland Discord: https://discord.gg/2KdxJzR

Offline ju6ju8Oo

  • Posts: 118
Re: Goldfish - A super thin USB-C Pro Micro clone
« Reply #55 on: Thu, 24 October 2019, 18:04:55 »
How does goldfish compare to elite C, other than the mid mount type C?

Offline Findecanor

  • Posts: 4362
  • Location: Stockholm, Sweden
  • Dans öl tarvö... "Pöngö"
Re: Goldfish - A super thin USB-C Pro Micro clone
« Reply #56 on: Thu, 24 October 2019, 20:18:10 »
* The Elite-C repurposes the RAW pin for PB0, which makes it incompatible with a couple of keyboard PCBs out there that expect it to be VBUS power. Goldfish does not.
* It has a physical reset switch, which is almost as thick as the USB-Type C port. Goldfish is thinner here.
* Has a power LED like the Pro Micro (Version 2 onwards). Goldfish has no LED.
* The current version: version 3, has a bug that makes auto-detection of Master/Slave on split keyboards not work. Previous versions worked, and the bug is supposed to be fixed in version 4. (Edit: Fixed in version 3.1)
Goldfish should support this auto-detection with revision B onwards.
« Last Edit: Sun, 08 December 2019, 08:12:46 by Findecanor »
I'm getting a bit fed up of spam-bots now ...

Offline ErgoMacros

  • Posts: 292
  • Location: SF Bay Area
Re: Goldfish - A super thin USB-C Pro Micro clone
« Reply #57 on: Thu, 24 October 2019, 21:03:45 »
Re:Soldering...
There are also a lot of SMT practice boards at Amazon and eBay. Couple buck a piece.
Today's quote: '...“but then the customer successfully broke that.”

Offline ctrlshiftba

  • Posts: 57
    • ctrlshiftba
Re: Goldfish - A super thin USB-C Pro Micro clone
« Reply #58 on: Fri, 25 October 2019, 06:45:39 »
* The current version: version 3, has a bug that makes auto-detection of Master/Slave on split keyboards not work. Previous versions worked, and the bug is supposed to be fixed in version 4.
Goldfish should support this auto-detection with revision B onwards.

They mentioned on off the clack they did find a software fix for this. Not sure if you need a branch of QMK or it’s in master.

Offline ju6ju8Oo

  • Posts: 118
Re: Goldfish - A super thin USB-C Pro Micro clone
« Reply #59 on: Fri, 25 October 2019, 06:46:32 »
I made this BOM at Mouser. Could somebody be so kind to verify?

Show Image


the number of 0u1 should be 6?

Offline Findecanor

  • Posts: 4362
  • Location: Stockholm, Sweden
  • Dans öl tarvö... "Pöngö"
Re: Goldfish - A super thin USB-C Pro Micro clone
« Reply #60 on: Fri, 25 October 2019, 10:13:25 »
They mentioned on off the clack they did find a software fix for this.
It's a workaround, not a fix. The workaround disables suspend-mode, and a USB device that does not suspend is not USB compliant.
I'm getting a bit fed up of spam-bots now ...

Offline ju6ju8Oo

  • Posts: 118
Re: Goldfish - A super thin USB-C Pro Micro clone
« Reply #61 on: Tue, 29 October 2019, 21:30:41 »
The schematic is referencing goldfish-rescue. But it's not included in the repo. 
« Last Edit: Tue, 29 October 2019, 21:49:16 by ju6ju8Oo »

Offline ju6ju8Oo

  • Posts: 118
Re: Goldfish - A super thin USB-C Pro Micro clone
« Reply #62 on: Wed, 30 October 2019, 02:06:44 »
Do I need a hot air gun for the atmega32U4 QFN? Or is it easier to switch it to QFP?
« Last Edit: Wed, 30 October 2019, 02:26:58 by ju6ju8Oo »

Offline yui

  • Posts: 95
  • Location: 127.0.0.1 (in azerty)
Re: Goldfish - A super thin USB-C Pro Micro clone
« Reply #63 on: Wed, 30 October 2019, 04:34:54 »
in theory you could reflow it in a toaster oven, i never tried but there is quite a few videos of peoples doing it successfully (i would try on cheaper components 1st), and the very cheap e-bay hot air guns have some issues with temperatures and mine had it's heater filament melt at a weak spot and some peoples also manage to do it with soldering irons. so depending on budget and if you already have the controller you could go that route, if you do not the QFP should be easier but you should check if it is compatible, with it being wider than QFN.
vi vi vi - the roman number of the beast (Plan9 fortune)

Offline dr_derivative

  • Thread Starter
  • Posts: 292
  • Location: Cork, Ireland
  • Not a real doctor
Re: Goldfish - A super thin USB-C Pro Micro clone
« Reply #64 on: Wed, 30 October 2019, 17:45:22 »
The schematic is referencing goldfish-rescue. But it's not included in the repo. 

Sorry about that, I just pushed a fix ;D

Do I need a hot air gun for the atmega32U4 QFN? Or is it easier to switch it to QFP?

QFP won't work because it's quite a bit bigger than the QFN.

With some patience and a lot of flux you should be able to solder the QFN with a regular iron. You can look up videos of drag soldering if you're unsure of the technique.

The only exception is the grounding pad on the bottom, but you can get away with leaving it disconnected. There's plenty of other ground pins and an atmega is never going to generate enough heat to really need thermal relief. If I were running a GB or something I'd make sure they were connected properly of course, but I've always done it this way for prototyping and never had any problems.
GH/Reddit/DT: dr_derivative | Discord: dr_derivative#3618 | IG: dr_derivative_mk
MK Ireland Discord: https://discord.gg/2KdxJzR

Offline ju6ju8Oo

  • Posts: 118
Re: Goldfish - A super thin USB-C Pro Micro clone
« Reply #65 on: Thu, 31 October 2019, 02:46:21 »
first time making the BOM list..no confidence
please give me some feedback on the BOM list. thank you!!

Samsung   CL05B104KO5NNNC
Samsung   CL05A475MP5NRNC
Samsung   CL05C220JB5NNNC
Samsung   CL05A105KQ5NNNC
UNIOHM   0402WGF5101TCE
UNIOHM   0402WGF1002TCE
UNIOHM   0402WGF220JTCE
Schottky   DFLS1100Q-7
Microchip   ATMEGA32U-MU
Epson Timing      FA-238 16.0000MB-C3
Hirose   CX70M-24P1
« Last Edit: Thu, 31 October 2019, 03:22:41 by ju6ju8Oo »

Offline dr_derivative

  • Thread Starter
  • Posts: 292
  • Location: Cork, Ireland
  • Not a real doctor
Re: Goldfish - A super thin USB-C Pro Micro clone
« Reply #66 on: Thu, 31 October 2019, 16:54:48 »
first time making the BOM list..no confidence
please give me some feedback on the BOM list. thank you!!

Samsung   CL05B104KO5NNNC
Samsung   CL05A475MP5NRNC
Samsung   CL05C220JB5NNNC
Samsung   CL05A105KQ5NNNC
UNIOHM   0402WGF5101TCE
UNIOHM   0402WGF1002TCE
UNIOHM   0402WGF220JTCE
Schottky   DFLS1100Q-7
Microchip   ATMEGA32U-MU
Epson Timing      FA-238 16.0000MB-C3
Hirose   CX70M-24P1

Looks good to me
GH/Reddit/DT: dr_derivative | Discord: dr_derivative#3618 | IG: dr_derivative_mk
MK Ireland Discord: https://discord.gg/2KdxJzR

Offline Findecanor

  • Posts: 4362
  • Location: Stockholm, Sweden
  • Dans öl tarvö... "Pöngö"
Re: Goldfish - A super thin USB-C Pro Micro clone
« Reply #67 on: Fri, 01 November 2019, 08:16:32 »
... atmega is never going to generate enough heat to really need thermal relief.
BTW. I noticed on the Teensy 2.0 that there is a ground pad on the bottom with 16 vias spaced in a grid connecting it to the top under the µC. I suppose that was meant for cooling, but maybe it is superfluous.
I'm getting a bit fed up of spam-bots now ...

Offline vvp

  • Posts: 795
Re: Goldfish - A super thin USB-C Pro Micro clone
« Reply #68 on: Fri, 01 November 2019, 09:16:57 »
BTW. I noticed on the Teensy 2.0 that there is a ground pad on the bottom with 16 vias spaced in a grid connecting it to the top under the µC. I suppose that was meant for cooling, but maybe it is superfluous.
Maybe it was meant to allow soldering with iron (instead of hot air or oven). I have seen a video of heating the centre pad from the other side of PCB. The idea was that heat will travel  through vias to the centre pad. I did not try it myself though. I tried other approach: I placed a "huge" through hole pad under the QFN centre pad so that I can reach to it directly from the other side (using a tip thin enough to fit into the through hole pad).

Offline dr_derivative

  • Thread Starter
  • Posts: 292
  • Location: Cork, Ireland
  • Not a real doctor
Re: Goldfish - A super thin USB-C Pro Micro clone
« Reply #69 on: Fri, 01 November 2019, 15:35:23 »
... atmega is never going to generate enough heat to really need thermal relief.
BTW. I noticed on the Teensy 2.0 that there is a ground pad on the bottom with 16 vias spaced in a grid connecting it to the top under the µC. I suppose that was meant for cooling, but maybe it is superfluous.

Yeah, it's definitely superfluous for an atmega. Other chips using the same package could need it though. The designer of the teensy might well have used a standard footprint which included those vias.

BTW. I noticed on the Teensy 2.0 that there is a ground pad on the bottom with 16 vias spaced in a grid connecting it to the top under the µC. I suppose that was meant for cooling, but maybe it is superfluous.
Maybe it was meant to allow soldering with iron (instead of hot air or oven). I have seen a video of heating the centre pad from the other side of PCB. The idea was that heat will travel  through vias to the centre pad. I did not try it myself though. I tried other approach: I placed a "huge" through hole pad under the QFN centre pad so that I can reach to it directly from the other side (using a tip thin enough to fit into the through hole pad).

I've seen both those approaches before too, I should try them some time. How well did the big through hole pad work for you?
GH/Reddit/DT: dr_derivative | Discord: dr_derivative#3618 | IG: dr_derivative_mk
MK Ireland Discord: https://discord.gg/2KdxJzR

Offline hanya

  • Posts: 103
  • Location: Japan
Re: Goldfish - A super thin USB-C Pro Micro clone
« Reply #70 on: Fri, 01 November 2019, 20:35:56 »
Hirose CX70M-24P1 USB Type-C connector could be replaced by Hirose CX90M-16P.
CX90M-16P does not have TX and RX pins that USB 2.0 does not require. And also no through hole pins, wider pin pitch.
Less pins helps you to decrease manufacturing cost too.
PFU HHKB JP, Sanwa MA-TB38 trackball

Offline dr_derivative

  • Thread Starter
  • Posts: 292
  • Location: Cork, Ireland
  • Not a real doctor
Re: Goldfish - A super thin USB-C Pro Micro clone
« Reply #71 on: Sat, 02 November 2019, 05:48:31 »
Hirose CX70M-24P1 USB Type-C connector could be replaced by Hirose CX90M-16P.
CX90M-16P does not have TX and RX pins that USB 2.0 does not require. And also no through hole pins, wider pin pitch.
Less pins helps you to decrease manufacturing cost too.

Yeah, I definitely could. When I designed this nearly 2 years ago the only type-C port without USB 3.0 pins was the HRO Type-C-31-M-12 which isn't mid mount. I know there are way more options now. If I get some time I might have a look at what's available to replace the current connector.

The only issue I have with the CX90M-16P is that it doesn't sit low enough in the PCB. With a ~1.6mm PCB the CX70 is almost exactly centred while the CX90 sticks out way more on top.
GH/Reddit/DT: dr_derivative | Discord: dr_derivative#3618 | IG: dr_derivative_mk
MK Ireland Discord: https://discord.gg/2KdxJzR

Offline ju6ju8Oo

  • Posts: 118
Re: Goldfish - A super thin USB-C Pro Micro clone
« Reply #72 on: Wed, 13 November 2019, 09:56:01 »
Which soldering tip do you recommend for the usb c connector?
I have a T12-BC2 but it might be too big.

Offline dr_derivative

  • Thread Starter
  • Posts: 292
  • Location: Cork, Ireland
  • Not a real doctor
Re: Goldfish - A super thin USB-C Pro Micro clone
« Reply #73 on: Thu, 14 November 2019, 05:14:49 »
Which soldering tip do you recommend for the usb c connector?
I have a T12-BC2 but it might be too big.

That tip should be fine. You can use the same drag soldering technique as for the microcontroller and it should be fine. For the through hole pins on the back I'd use a similar technique too, just be careful not to accidentally bridge the pads since they are very close.
GH/Reddit/DT: dr_derivative | Discord: dr_derivative#3618 | IG: dr_derivative_mk
MK Ireland Discord: https://discord.gg/2KdxJzR

Re: Goldfish - A super thin USB-C Pro Micro clone
« Reply #74 on: Sat, 23 November 2019, 12:08:06 »
Hello!
I came here from google, I like how small this board is and that itll not be that expensive at the quantity I want to make (like 10), but before I go ahead and buy all the components (kinda sad that its not for sale), one thing is bugging me: How did you solder the through hole parts of the connector to 1.6 mm FR4? The pins looked really short to me, so I downloaded the STEP files and measured only 0.33 mm between the SMD pin bottom and the bottom of the pins... The Hirose website says max PCB thickness is 0.8 mm, so how does that work?  :confused:

I would highly appreciate any photos of finished ones... Pretty sure the pins dont reach the other side! Sorry for the trouble...

Offline dr_derivative

  • Thread Starter
  • Posts: 292
  • Location: Cork, Ireland
  • Not a real doctor
Re: Goldfish - A super thin USB-C Pro Micro clone
« Reply #75 on: Sun, 24 November 2019, 09:05:52 »
Hello!
I came here from google, I like how small this board is and that itll not be that expensive at the quantity I want to make (like 10), but before I go ahead and buy all the components (kinda sad that its not for sale), one thing is bugging me: How did you solder the through hole parts of the connector to 1.6 mm FR4? The pins looked really short to me, so I downloaded the STEP files and measured only 0.33 mm between the SMD pin bottom and the bottom of the pins... The Hirose website says max PCB thickness is 0.8 mm, so how does that work?  ???

I would highly appreciate any photos of finished ones... Pretty sure the pins dont reach the other side! Sorry for the trouble...

You're right, the pins don't reach the other side with a 1.6mm PCB. If you solder the pins the surface tension will pull the solder into the hole (look up capillary effect) to make contact with the pin. This also happens with normal through-hole soldering; the solder will flow all the way through the hole to the other side.

Otherwise, if you'd prefer, there's no reason why you can't use an 0.8mm PCB. There really is no reason why you'd have to stick with the standard 1.6mm for this board :D
GH/Reddit/DT: dr_derivative | Discord: dr_derivative#3618 | IG: dr_derivative_mk
MK Ireland Discord: https://discord.gg/2KdxJzR

Re: Goldfish - A super thin USB-C Pro Micro clone
« Reply #76 on: Mon, 25 November 2019, 07:01:29 »
Yeah, I do know that solder will in fact wick and will reach them. But is that good practice? I already ordered everything, just curious at this point.

Offline dr_derivative

  • Thread Starter
  • Posts: 292
  • Location: Cork, Ireland
  • Not a real doctor
Re: Goldfish - A super thin USB-C Pro Micro clone
« Reply #77 on: Mon, 25 November 2019, 15:41:58 »
Yeah, I do know that solder will in fact wick and will reach them. But is that good practice? I already ordered everything, just curious at this point.

Probably not if you're assembling them by machine. You might end up with some bad joints and they'll be more difficult to inspect and stuff.

But in this case for manual assembly; if it works, it works Ż\_(ツ)_/Ż
GH/Reddit/DT: dr_derivative | Discord: dr_derivative#3618 | IG: dr_derivative_mk
MK Ireland Discord: https://discord.gg/2KdxJzR

Offline equalunique

  • Posts: 367
Re: Goldfish - A super thin USB-C Pro Micro clone
« Reply #78 on: Sat, 30 November 2019, 22:35:51 »
On twitter I saw this "ShiroMicro UZU Edition" microcontroller that looks similar to Goldfish: https://twitter.com/elfmimi/status/1200590989267623936?s=19

Sent from my Ono-Sendai Cyberspace 7 using Tapatalk

Offline Findecanor

  • Posts: 4362
  • Location: Stockholm, Sweden
  • Dans öl tarvö... "Pöngö"
Re: Goldfish - A super thin USB-C Pro Micro clone
« Reply #79 on: Sat, 30 November 2019, 23:40:04 »
I checked the ShiroMicro's schema yesterday. It's pretty basic in comparison. It has a diode for master/slave detection and a mid-mounted port. No additional pins, no voltage regulator, no LEDs and no over-current protection.
One interesting detail though: to enable the RAW pin, you'd need to bridge a solder-jumper which shorts it to VBUS. Another jumper shorts PD6 to ground - for hard master/slave assignment maybe?

BTW. The red "Uzu edition" is for the Uzu42 keyboard. The designer has also shown a ARM-based Pro Micro clone.
« Last Edit: Sun, 01 December 2019, 07:33:23 by Findecanor »
I'm getting a bit fed up of spam-bots now ...

Offline ju6ju8Oo

  • Posts: 118
Re: Goldfish - A super thin USB-C Pro Micro clone
« Reply #80 on: Mon, 02 December 2019, 05:41:39 »
Hirose CX70M-24P1 USB Type-C connector could be replaced by Hirose CX90M-16P.
CX90M-16P does not have TX and RX pins that USB 2.0 does not require. And also no through hole pins, wider pin pitch.
Less pins helps you to decrease manufacturing cost too.
second this! single side SMT would decrease the cost!

Offline ju6ju8Oo

  • Posts: 118
Re: Goldfish - A super thin USB-C Pro Micro clone
« Reply #81 on: Mon, 02 December 2019, 19:16:26 »
I checked the ShiroMicro's schema yesterday. It's pretty basic in comparison. It has a diode for master/slave detection and a mid-mounted port. No additional pins, no voltage regulator, no LEDs and no over-current protection.
One interesting detail though: to enable the RAW pin, you'd need to bridge a solder-jumper which shorts it to VBUS. Another jumper shorts PD6 to ground - for hard master/slave assignment maybe?

BTW. The red "Uzu edition" is for the Uzu42 keyboard. The designer has also shown a ARM-based Pro Micro clone.
The ShiroMicro's type C mount looks interesting too.

Offline ju6ju8Oo

  • Posts: 118
Re: Goldfish - A super thin USB-C Pro Micro clone
« Reply #82 on: Tue, 03 December 2019, 02:47:40 »
Is there a way to test if the goldfish is fully working? Maybe I should write a small arduino program to test each pin? or does it already exist?

Offline dr_derivative

  • Thread Starter
  • Posts: 292
  • Location: Cork, Ireland
  • Not a real doctor
Re: Goldfish - A super thin USB-C Pro Micro clone
« Reply #83 on: Sat, 07 December 2019, 09:23:40 »
Is there a way to test if the goldfish is fully working? Maybe I should write a small arduino program to test each pin? or does it already exist?


I haven't written any program like that, but it shouldn't be very difficult to do something like that in Arduino.
GH/Reddit/DT: dr_derivative | Discord: dr_derivative#3618 | IG: dr_derivative_mk
MK Ireland Discord: https://discord.gg/2KdxJzR

Offline ju6ju8Oo

  • Posts: 118
Re: Goldfish - A super thin USB-C Pro Micro clone
« Reply #84 on: Sun, 08 December 2019, 03:12:54 »
Another newbie question, what are the benefits of using a PowerDI123 verses 1N5819 sod-123?

Offline dr_derivative

  • Thread Starter
  • Posts: 292
  • Location: Cork, Ireland
  • Not a real doctor
Re: Goldfish - A super thin USB-C Pro Micro clone
« Reply #85 on: Sun, 08 December 2019, 11:49:14 »
Another newbie question, what are the benefits of using a PowerDI123 verses 1N5819 sod-123?

There aren't really any benefits in this case, I just happened to have a bunch of powerdi123 Schottkies lying around when I originally designed the board ;D

In general, PowerDI-123 is designed for better heat dissipation than a regular SOD-123 package. It has a much bigger pad on the cathode pin to transfer heat to the PCB more effectively.
GH/Reddit/DT: dr_derivative | Discord: dr_derivative#3618 | IG: dr_derivative_mk
MK Ireland Discord: https://discord.gg/2KdxJzR