• Please review our updated Terms and Rules here

SBC6120 "Mini" Front Panel

3D printing is where it kinda all goes off the rails if you don't have your own printer or access to one cheaply. I priced out JLCPCB's 3D printing service (which I'm heard is one of the cheapest) and ya, it's ~$50 for all the parts and then another ~$25 in shipping given the size. Versus the less than a dollar in PLA filament and 12+ hours of clock time (no Bambu Labs X1 here yet) it took me to print my own.
FWIW I did a rough BOM at JLCPCB (not sure whether I included the correct # of spacers) and my total came out to $23.57 using the 8111X Resin (https://jlc3dp.com/help/article/199-8111X---Photosensitive-Resin). That'll result in white parts but a good surface finish. Resin parts are remarkably "finished" compared to typical filament printing. Unfortunately I couldn't get the shipping cost below $31.35. So $55 all-in unless one can amortize shipping across more projects. For another ~$10 one could go with a PLA case in blue, red, or black rather than resin. I'm guessing that you went with PLA for all parts in your estimate. Resin prints reportedly can be dyed in a manner similar to cloth; I have no experience with that but I suppose one could get a DEC-like case color that way :->. Also markers can be used, reportedly ...

Note that JLCPCB has some Christmas coupons with various expiration dates next week that may be applicable to ones order.
 
I have just ordered the PCBs from JLCPCB. The "dress panel" is white with black silk screen. This is for 5 each: $27.43 + $31.07 shipping.

I also ordered the 3D printed parts from them. I used black Nylon for durability and temperature tolerance. It is expensive: $92.43 + $39.47 shipping.

The AliExpress M102 switches seller only allows you to order 10 per customer.

I like the interactive BOM.
 
Sigh. So what do you plan to do? Alternative/additional seller?
I have now ordered a bunch of parts from AliExpress. On the M102 order page of this particular seller you can only buy one lot of 10. But when you go to the checkout page you can increment the number of lots from 1 to 2. Lets see what happens. Of course there are other seller of this M102 switch, but you can't be sure that they are exactly the same. At least everything on AliExpress was cheap.
 
Ordered some more knobs to try out, looking for something that has the aluminum top like the original panels. The one I'm using now was simply what I had lying around from past projects.

These guys are 15mm so same as what's on their now.


and the MF-A01 which is listed as 20mm


Also added the last two Keystone 8568 (on clearance) to my Newark order which looks the closest of the three and is 17.8mm.

 
Keystone 8568 came in today. Nice, but really to big for this. At a minimum would require the front panels labels to be extended and spaced out more for the indicator labels and lines to line up better. Also kinda spendy at ~$12 regular price, Newark had them on clearance for $2.28, but I don't see that happening again.

PXL_20231222_224509922.PORTRAIT.jpg
PXL_20231222_224527557.PORTRAIT.jpg
PXL_20231222_224638246.PORTRAIT.ORIGINAL.jpg
 
Slightly OT, but is there a known source of the 6120 still available?

UTSource still lists some for $~40 USD. I bought a bunch from them years ago and only one out of the batch didn't fully work (booted, but wouldn't run Adventure), but no idea what the stock looks like now.


Bunch of seller on eBay for around the same price also.
 
I have now ordered a bunch of parts from AliExpress. On the M102 order page of this particular seller you can only buy one lot of 10. But when you go to the checkout page you can increment the number of lots from 1 to 2. Lets see what happens. Of course there are other seller of this M102 switch, but you can't be sure that they are exactly the same. At least everything on AliExpress was cheap.
Any news? I still need to order mine.
 
Any news? I still need to order mine.
I got all the parts including the "on-on" toggle switches which only sold in single lots of 10. I incremented the lot number to two on the checkout page and got all 20.

I have also received the 3D printed parts from JLCPCB. I chose the expensive Nylon but it is not suitable for the base plate. It is not rigid enough so I will print my own in PLA. The top might be OK, but it too is less rigid than I like.
All other Nylon parts are fine.

I won't have the spare time now to build the front-panel and a new SBC6120-RBC as I am pre-occupied with grandchildren and family visits. I will have time from mid February.
 
Thanks for the update Tom! Enjoy your "pre-occupations" :->. I found that if I used the buy-it-now link, rather than adding to my cart, that there were no limitations. And there is a price-sale going on until the 4th. $2.35 USD for ten. So I purchased 5 sets and received free shipping as well. $12.48 total for 50 MTS-102 switches; such a deal (although we'll see about duds among them). Bought some extra MTS-112 switches in a separate order.
 
I got my parts on order today. I'm looking forward to doing this build. I think the AliExpress seller "fixed the glitch" that allowed you to change the quantity in the cart. I had to do two separate transactions in order to get enough switches on order.
 
I completed my build of the SBC6120 "Mini" Front Panel. I'm quite happy with how it turned out, certainly a much more visually interesting system with the Blinkenlights and Switches...
SBC6120_Blinken Small.jpeg
I had a couple of notes from my build, first because I wasn't familiar with the machine, it took me a bit of a think to figure out which switches go where. The switch registers are all toggles of course, and most of the rest of the switches are momentary, but the HALT switch is also a toggle, the 13th one.

Second, when I printed the Shaft Extender with the default settings on my printer it was too fragile. I had to print a second one with a higher infill (40%). Seems to be strong enough so far.

Thanks again to @hideehoo for making this available!
 
I am getting ready to build Hideehoo's mini front panel. I got the parts a while ago, but decided to build a new SBC6120-RBC rather than using my existing one. When I built the existing one about 3 years ago, it worked straight away. This time I am having strange problems:
  • At first POST stopped showing a value 2 indicating some problem with the UART. I swapped it to a new but untested UART, but the problem stayed the same. I tried the know good UART from my original SBC6120-RBC, but the problem stayed the same. I tried another HD-6120, but the problem remained the same. I swapped in the GALs from the good board, the problem remained the same. I hooked up the scope to check the UART related logic and finally realised that the /SKIP signal never made it to the HD-6120 causing the TSF IOT to never skip. A multimeter confirmed that between /SKIP on pin 8 of U19 and pin 4 of the HD-6120 was open circuit. On closer examination of the PCB with a magnifying glass there was a tiny break in the trace seemingly caused by a small scratch. As the board were tested by JLCPCB (flying probe test), this must have happened some time after testing (or their testing is less comprehensive than I thought). I patched up the trace, and on power up the POST now showed a value of 1 which is what is expected. But not all is well.
  • Unfortunately I only see gibberish being sent to the terminal in a continuous stream. I checked the bit rate generated by the 4.9152 MHz oscillator and the 74HC4040 ripple counter and that the correct x16 frequencies are routed to the TRC and RRC BAUD_CLK inputs of the 6402 UART. I checked the actual bit timing of the sent characters on the TRO pin 25 of the UART and they match 9600 bps as jumpered. I tried a different UART, but the problem remains the same. It almost looks as if the software in the EPROMs is writing the gibberish to the terminal. I burned the HEX files from the EEPROM directory in "sbc6120-rbc-edition-archive-10may2018.zip" to a pair of ST M27C256B-10F1 EPROMs (I also tried the slower AMD AM27C256-150DC but the problems is still the same). IIRC this is the same as I used before. I am running of a very stable bench supply set to 5V with 1.5A current limit.The SBC6120 is only drawing 180 mA from the supply. I used the 5 MHz oscillator for the HD-6120.
  • Occasionally I still get a POST 2 but the problem occurs after two TFS IOTs somewhere in the keyboard input test. I need to analyse this further.
What is going on here? Has anyone see something similar?
 
Last edited:
Oh well, probing the UART receive related logic quickly showed that the console cable I used created some loop-back scenario which the SBC6120 did not deal gracefully with resulting in a never ending stream of gibberish.

During my testing of the new board I used the cable from Hideehoo's BOM obtained from AliExpress. Please save yourself some time and grief and don't use that cable unmodified with an SBC6120.

I made a simple 9 wire IDC cable with a 10 pin IDC socket on one side and a male DB9 on the other with pin 1 of each connector on the red wire of the cable.

With this newly made cable I got the standard SBC6120 output and command prompt. With an IDE to Compact FLASH adapter I then successfully booted OS/8.

So all along the only real problem was the scratched PCB trace which killed all /SKIP functionality.

The other problems were self inflicted by using the DB9 to IDC cable from AliExpress Console Cable unmodified. In Hideehoo's notes he actual writes "May need to rewire". Better just make a cable from scratch.
 
Oh well, probing the UART receive related logic quickly showed that the console cable I used created some loop-back scenario which the SBC6120 did not deal gracefully with resulting in a never ending stream of gibberish.

During my testing of the new board I used the cable from Hideehoo's BOM obtained from AliExpress. Please save yourself some time and grief and don't use that cable unmodified with an SBC6120.

I made a simple 9 wire IDC cable with a 10 pin IDC socket on one side and a male DB9 on the other with pin 1 of each connector on the red wire of the cable.

With this newly made cable I got the standard SBC6120 output and command prompt. With an IDE to Compact FLASH adapter I then successfully booted OS/8.

So all along the only real problem was the scratched PCB trace which killed all /SKIP functionality.

The other problems were self inflicted by using the DB9 to IDC cable from AliExpress Console Cable unmodified. In Hideehoo's notes he actual writes "May need to rewire". Better just make a cable from scratch.

Ya, you'll definitely need to rewire, I can update the text. The BOM link is the cheapest way I've found found to source that DB9 connector with the shrouded hood and screws. I also used it to make a custom cable.

1711393716075.png
 
After my adventure with the damaged /SKIP track on the SBC6120 PCB I have finally found time to built Hideehoo's nice mini front panel. It is relatively straight forward and took about 2 hours to assemble. Everything worked on first power up.

Congratulations to Chris Tersteeg (Hideehoo) for a job very well done! Thank you very much for sharing your design on Github!

Here are some photos (click to see full size):

PXL_20240329_073957585.jpg
SBC6120 mounted on base plate with power and serial cable and CompactFlash adapter.


PXL_20240329_075302484.jpg
First power on - running OS/8.


PXL_20240329_082603941.jpg
Fully assembled and running.

Here is a short video of it running SPACWR.BAS:


It was fun to build the system, and I get a lot of joy running it. There is no heat and no noise unlike my PDP-8/e or my LAB-8/e. The whole system is very light-weight and can easily be carried around. It is powered by a cheap plug-pack.
 
Back
Top