SC126, Z180 motherboard

[Steve Cousins]

My Z180 motherboard is progressing well.

This design is a result of Richard Deane's design challenge discussed here.

I'm close to being satisfied with the design and a v1.0 production run.

It runs the Small Computer Monitor and RomWBW. 

I have some simple I2C code and hardware working, so the I2C bus looks good.

RomWBW is running with mass storage via SD card on one of the SPI / SD Card ports.

The Real Time Clock is also working with RomWBW.

The bus sockets are a superset of the RC2014 bus.

[jopil]

Hello Steve. Good place here to start considering some H&S extensions of the Z80 platform and beyond. Steve, If I send to you a Locomotive Basic 1.1 rom image will you have some time to check if it is working with SC111/SC119 & SC126 boards? John

[Steve Cousins]

Hi John

I'm happy to look at such things, but the first step would be to have a quick look to see if it appears compatible. Can you point me to any suitable documentation.

Steve

[jopil]

Here you go Steve:
https://www.cpcwiki.eu/index.php/Locomotive_BASIC

http://www.cpcwiki.eu/index.php/Technical_information_about_Locomotive_BASIC

John

[Karl Albert Brokstad]

I am sorry you didnt see my reply on this forum.

I did reply from email.

I am sure locomotive basic can be run if all prerequisite hardware is available.

But so far the rc2014 platform has no standard audio and video standards.

Maybe a standard z80 system with the right hardware, and modification to the code, it can run.

I hope someone find a solution.

Karl

[Steve Cousins]

John 

I've had a little look at Locomotive BASIC for the Amstrad. Nice looking BASIC.

I don't believe there is any chance it will do anything sensible on any of our RC2014 and related systems. It is far too specific to Amstrad's hardware and firmware to work elsewhere without extensive modifications. Even the most basic console text in and out do not share the same BIOS entry points as the Microsoft BASIC we normally use, so I assume it would also be necessary to port the Amstrad's BIOS to the RC to make it work.

The Microsoft BASIC shipped with RC2014 systems only requires a few simple BIOS character in and character output functions to work on the RC. Well, that is assuming you are running the version of BASIC Grant modified to remove the other hardware specific features, such as cassette tape save and load.

I quick search did not turn up the source code for Locomotive BASIC so this feasibility study has come to an early end.

Perhaps someone out there knows the Amstrad and Locomotive BASIC well enough to offer a better-informed opinion.

Steve

[Alan Cox]

On Friday, 5 July 2019 12:59:35 UTC+1, Steve Cousins wrote:

John 

I've had a little look at Locomotive BASIC for the Amstrad. Nice looking BASIC.

It's in copyright and the rightsholders do appear to care about it, their CP/M port and wordprocessor rights.

Perhaps someone out there knows the Amstrad and Locomotive BASIC well enough to offer a better-informed opinion.

There are far better options anyway - there is a BBC BASIC that is freely redistributable but not free software, includes a CP/M port and is smaller than Microsoft Basic despite all the extra features.

http://www.bbcbasic.co.uk/bbcbasic/z80basic.html

[Fredrik Axtelius]

Z80 source for BBC Basic

https://github.com/davidgiven/cpmish/tree/master/third_party/bbcbasic

--frax

--
You received this message because you are subscribed to the Google Groups "retro-comp" group.
To unsubscribe from this group and stop receiving emails from it, send an email to retro-comp+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/retro-comp/882b0333-1038-40d5-a8b2-b36f70f2555d%40googlegroups.com.

[Richard Deane]

David Given claims that bbc basic z80 software is now public domain having 9been released by Russell but I can find no evidence of that as David has not published the emails and Russell's site makes no mention of it being released into public domain.

[Mark T]

Links to the source code

http://www.bbcbasic.co.uk/bbcbasic/z80basic.html

http://cowlark.com/2019-06-14-bbcbasic-opensource/index.html

Mark

[Alan Cox]

I have some simple I2C code and hardware working, so the I2C bus looks good.

 

The i²c seems to have one small glitch. You drive the SCL line directly, but even without multiple masters (which you specifically don't support) a slave is entitled to hold the clock line low if it needs time to respond (clock stretching in the protocol), and it looks like those will get into a fight with the SC126 ?

Alan

(still waiting for the demux board testers.. hoping it's not gotten lost in the China post)

[Steve Cousins]

Hi Alan

It appears the clock stretching feature has been introduced in recent decades. It wasn't there when I was a lad :-) 

I don't think it is a serious problem unless the SC126 software is really pushing things - which seems to be your favourite activity.

I'd have made the I2C interface's clock line (SCL) open collector but it would have cost another chip (or a diode and possibly a poor low level). Design is just another word for compromise.

Steve

[Steve Cousins]

Alan

My recent experience of turn around from JLCPCB to UK is as follows:

DHL: 6, 8, 7, 6, 5, 5, 11, 4, 5 days from order to delivery.

GB Special: 13, 10, 12, 17, 17, 13, 11 days from order to delivery.

I usually end up ordering late in the evening so that appears to make the turn around in days appear slightly poorer than it might be.

The occasional poor turn around with DHL is usually due to holidays in China.

I usually get customs charges when using DHL, but not usually with GB Special delivery. 

Tracking is brilliant with DHL but pretty useless with GB Special delivery. You would expect some guidance as to progress and delivery but with GB Special delivery you tend to get nothing until the day before delivery.

So sit tight, the package will probably just turn up when you are least expecting it. I've placed many orders with JLCPCB and none have got lost.

Steve

On Wednesday, 7 August 2019 19:59:36 UTC+1, Alan Cox wrote:

[Richard Lewis]

DHL is about the same to the US. Ironically domestic OSH Park (I think out of Texas?) takes longer (2-3 days) even via Fedex. I guess they don't do their own pcb fab and send off batch orders to whichever fab  shop they use

Richard

[Alan Cox]

> It appears the clock stretching feature has been introduced in recent decades. It wasn't there when I was a lad :-)

Chuckle.

> I don't think it is a serious problem unless the SC126 software is really pushing things - which seems to be your favourite activity.

I'm trying to do the full 400Kbit fast mode while writing some initial i2c drivers at this point. I actually had to add some small delays 8). My touchscreen/display does however use clock stretching in i2c mode.

> I'd have made the I2C interface's clock line (SCL) open collector but it would have cost another chip (or a diode and possibly a poor low level). Design is just another word for compromise.

I guessed as much. My touchscreen display also has a UART mode so I might have to use the other UART for that.

Alan

[Mark T]

Problem with dhl is even if its below threshold to need to charge import duty they still declare it, results in a very small duty charged but still a huge processing fee charged by dhl.

[John Kennedy]

Got my SC126 kit - Seattle - and just finished building it. Wow, it's really good! And the documentation deserves an award.

[Richard Lewis]

Just now finished mine scavenging parts from other boards: I would like to honor SC109 & SC119 who gave up their lives for this project. I especially enjoyed sacrificing the "blue" ones (they will not be missed).

It's nice that I have something flat to throw in my bag. 

-Richard

[Steve Cousins]

Richard, that's murder :-)

and you shouldn't discriminate based on colour!

[Alan Cox]

My boards finally arrived. I do need to take one small change as while it works electrically I misjudged the space and covered part of the i2c adapter. I think I will also take the time to add a mux so that I can use devices that don't isolate MISO properly and rotate the connectors so you can keep devices flat over the CPU and ROMs if you are putting it in a flat case as I intend.

Alan

[Richard Lewis]

I'm loving this board (platonically of course). It needs a cool looking custom case. Been looking for an excuse to get into 3D printing. 

[Greg Holdren]

My SC126 kit came in the mail to today. Unfortunately I wont be able to start it until tomorrow. I'm amazed at how nice it fit in that little flat box. :)

Greg

[John Kennedy]

I housed my kit in a DIY terminal. I used a Raspberry Pi Zero to handle the USB keyboard and HDMI display, but I did need to solder up a 3.3v to 5v shifter on a scrap of board.

Works really nicely!

[Steve Cousins]

Reminds me of the first computer I ever used. Or more accurately the first terminal I ever used. I was low life, not fit to actually see the computer. Rumour among us plebs was the computer was big and blue - but no one knew for sure :-(

Just realised the text is white. It needs to be green to really look right.

Steve

[John Kennedy]

Some big 360 system perhaps?

When I want green I emit the right escape characters :-) Or swap in a bigger Pi and run Cool Retro Term.

[Alan Cox]

On Thursday, 15 August 2019 23:20:01 UTC+1, Steve Cousins wrote:

Reminds me of the first computer I ever used. Or more accurately the first terminal I ever used. I was low life, not fit to actually see the computer. Rumour among us plebs was the computer was big and blue - but no one knew for sure :-(

Just realised the text is white. It needs to be green to really look right.

Or amber, and with scanlines.

Looks really good although we'd have thought glowing blue keys was sci-fi movies only 8)

Alan

[Steve Cousins]

In "Buck Rogers in the 25th Century" the classic Commodore PET was making a come back after 400 years. Proof that green text will ultimately prevail. Admittedly it doesn't look green in this picture. Cassette tape storage also seems to have had a revival. I guess data compression will be so advanced in 400 years time that cassette tapes will be fast enough.

[John Kennedy]

I can connect it to the actual DEC VT520 I was lucky to find when I need amber scanlines :-). That brings me back to an internship with a VAX/VMS environment.

It's surprisingly tricky to find a cheap keyboard that doesn't glow! Thankfully I can turn it off.

[Alan Cox]

On Thursday, 15 August 2019 23:48:11 UTC+1, Steve Cousins wrote:

In "Buck Rogers in the 25th Century" the classic Commodore PET was making a come back after 400 years. Proof that green text will ultimately prevail. Admittedly it doesn't look green in this picture.

The early chicklet PETs were black and white.

It is making a comeback already. The big thing in UI features right now is 'dark mode'.

Alan

[Greg Holdren]

Hows does one get to the I2C and SPI ports?

I see there are API calls to IO ports provided by SCM to test and clear set bits to and from a IO port. SCM API calls 17h to 20h. I know SCM doesn't utilize these ports in CP/M but WBW does. Haven't look at WBW FW yet.

It is either that or bit bang in User Land I take it. I want to play with some MCP2515 CAN interface chips.

Thanks,

Greg

[Steve Cousins]

Hi Greg,

Looks like I need to add some documentation about the use of I2C and SPI.

Currently, both require bit-banging from SCM or stand-alone CP/M, as neither have suitable APIs. I think the same is true of RomWBW's HBIOS.

I have a sample program for SCM to do some basic I2C I/O using a PC8574 (simple 8-bit I/O port chip) - see attached.

Using the second SPI port is a problem when the included micro SD card adapter is connected as this adapter does not tri-state its serial output when not selected. See Alan's post above: https://groups.google.com/d/msg/retro-comp/TfIagxG5b7M/z084okySAQAJ

Steve

[Greg Holdren]

Thanks for the sample. I do have some of those port expanders on hand so I can try it out. I'll work something up for the SPI port code with some chips I have too.

Yeah, I remember the post. It looks like Alan added a chip to tristate the MISOs. Have not seen the detail of what he did. Can't read the part number from the picture either. Maybe a HCT125.  It seems like the CS line can be used to tri-state MISO between the SD card adapter and the SPI port connector.

Greg

[Alan Cox]

Rev 1 was just a 138 to split the two CS lines - both of which start low - into 4. Enough to use it with polite devices.

Rev 2 adds a mux on miso so that the input follows CS. I will build and test that once sufficiently recovered.

Alan

[Alan Cox]

Untested as I said....

[Greg Holdren]

Alan,

Looks like it will work. Nice way to expand the SPI ports using a binary weighted CS value to get 4 vs a 2 bit CS value using "One Hot" outputs.

My idea to fix the SD card adapter was just to add a non-inverting tristate buffer to the MISO line. But.... there is already a LVC125 on the adapter! I followed the MISO line back to pin 11 so pin 12 is the input from the SD card. Pin 13 is the tri-state control line. This line is hardwired to ground which always enables the MISO driver. As a fix I lifted pin 13 disconnecting to ground and connected the actual pin on the device to CS  on the SPI interface connector. This will enable the MISO output when CS is active. This should fix the issue with the shipping adapter. Drive G: (SD card) still works under WBW bootrom.

I don't have another SPI device ready to interface to the second port now so if someone can verify that the second port works with the SD adapter that would great.

Greg

[Alan Cox]

https://hackaday.io/project/167445-spi-fix-for-sc126

With a couple of blue wires this version seems to do the trick so I've updated the pcb and schematics and uploaded it to hackaday.

Alan

[Tom Szolyga]

Hi Greg,

Would you post a picture of your modified board?

Thanks,

Tom

[Greg Holdren]

Tom,

Here is it. I cut the ground pin on the pin but it was still connected to ground so it must be connected on the inside area of the land pattern too. So I had to lift the pin up.

Greg

[Steve Cousins]

Hi Greg,

That's a nice little mod for those wanting to use the second SPI port.

I may produce an alternative to this rather compromised micro SD card adapter I currently supply. My 'plan' (well vague concept really) is a board with one or two full-size SD card sockets. These are big enough to be easy to assemble. The board could be connected directly across both SPI connectors or be on Dupont cables for panel mounting. I like what Alan is doing with the extra decoding of the chip select signals so I could do something similar to include an additional SPI port. No shortage of projects to do!

Steve

[Greg Holdren]

Depending on the circuit maybe a few tri-state buffers that the gates are jumperable to the CS line. There are many options to go with the design and in adding more ports.

Greg

[Alan Cox]

On Friday, 6 September 2019 10:07:24 UTC+1, Steve Cousins wrote:

Hi Greg,

That's a nice little mod for those wanting to use the second SPI port.

I may produce an alternative to this rather compromised micro SD card adapter I currently supply. My 'plan' (well vague concept really) is a board with one or two full-size SD card sockets. These are big enough to be easy to assemble. The board could be connected directly across both SPI connectors or be on Dupont cables for panel mounting. I like what Alan is doing with the extra decoding of the chip select signals so I could do something similar to include an additional SPI port. No shortage of projects to do! 

If you do then the thing I realise I now need to add for rev 3 is a 3v3 pin to each port and a 5v to 3v3 supply. With an HCT chip as the mux it's otherwise already friendly with anything that is 3v3 but 5v I/O tolerant as far as I can tell.

Alan

[Greg Holdren]

Alan,

The SD card and the LVC125 is 3.3V and there is a 5V to 3.3V reg on the SPI CD card that Steve ships with the SC126. In general HCT (5V) <==> LVC (3.3V) is ok. LVC has 5V tolerant Inputs. The LVC Voh (output high)  should meet the HCT Vih (input high).

Greg

[Alan Cox]

On Saturday, 7 September 2019 22:38:31 UTC+1, Greg Holdren wrote:

Alan,

The SD card and the LVC125 is 3.3V and there is a 5V to 3.3V reg on the SPI CD card that Steve ships with the SC126.

Yes but there isn't one on the SPI ethernet card and some of the other SPI devices I'm attaching to my SC126. For those having a single central 3v3 source will be useful. I did four ports because I needed them. I came close to borrowing an LED as well for 8 8)

Alan

[Greg Holdren]

I pronounce you SC126 and CPLD...

Now I can do some serious development.

Greg

[Nick Brok]

Hi Steve,

The sc126 has come in last Thursday and it works fine... But I find some little issues.

Running your SC-monitor the system hangs after pressing to much keys on the keyboard while outputting something.

I had this problem in Scrumpel also. Solution: check receiver overrun bit 6 and reset bit 3 in ascicntla register. (This is a known bug on some Zs180 chips)

ROMWBW works fine, I was curious how this works. (That's the reason I bought your kit) Your kit is a nice design, good documented. My advice on your PCB design: use copper-pour on side 1 connected to VCC and on the other side connected to ground. Then the pathway from the decoupling c's are as short as can be, which is recommended.

In 3 hours I've build the kit without mistakes. I use a RS232 converter and I did replace all 2K2 resistors with resistors of zero Ohms (simple wire ;-) ) The pull-up resistors of 100K I didn't use them.

I've still a question:

How can I install a new version of ROMWBW on it? The delivered flash utility in CP/M Z-dos doesn't recognize the used FLASH chips.

Greetings,

Nick de pe1goo

[Richard Deane]

Don't forget to enable flash write by setting jumper correctly. I use the flash program ok.

[Nick Brok]

This did the job.... I can access the FLASH... :-(  Not logical reaction of the program... could better be FLASH write protected.....

Thanks anyway... it works.

Op zaterdag 14 september 2019 09:44:28 UTC+2 schreef Richard Deane:

[Steve Cousins]

Hi Nick

I'm glad your SC126 is up and running and thanks for the tips.

I was wondering why SCM sometimes hangs. I'll look into supporting the overrun and reset bits, as you suggest.

Thanks for the PCB design advice. My approach to copper pour on SC126 was this: I placed a decoupling capacitor as close to each IC's Vcc pin as possible. I manually routed a very short thick track from the capacitor to the Vcc pin of its IC. I also manual routed a thick track from the capacitor to other Vcc nodes. I don't see how a copper pour for Vcc will help much with the connection from each capacitor to the Vcc pin of its IC (the critical path for noise) but I agree it could reduce the impedance of the Vcc path elsewhere. As for the ground return, I manually route a track as direct as possible from the capacitor to the GND pin of its IC. This track will often be relatively long which is not ideal but can't usually be avoided. I then use a copper pour to improve this return path. My feeling is that using both sides for a ground fill allows a really good, nearly complete ground plane. If I use one side for a Vcc pour then both will suffer from being incomplete with large areas not filled. As the Vcc path from the IC to its decoupling capacitor is nearly idea already, I see the longer return path as the one that needs a good copper pour. Is this trade-off bad? 

I assume I am correctly understanding the point about the serial resistors: You are using an RS232 adapter board which is powered from SC126. Thus there is no need for protection resistors. I assume the RS232 adapter card being connected means you are not concerned about pull up resistors being fitted to set the state of unterminated inputs. Out of interest: Do the RS232 adapter boards you use work okay with the 2k2 current limit resistors fitted?

Steve

[Mark T]

Hi Steve,

I prefer your method of power connections. I’ve tried the ground fill on one side and vcc fill on the other but as you mentioned this gives an incomplete ground on two layer boards. I normally ground fill both sides and then stitch the two sides together with vias, also try to route other traces to leave a solid ground fill from the ic package ground to its decoupling cap on one side of the board. 40 or 32 mil trace to connect vcc between packages. Connection between cap vcc and ic vcc is not something I worry about as only a few mm. If there is any resistance or inductance in the vcc traces then maybe it just helps to isolate noise between devices.

Mark

[Nick Brok]

Hi Steve,

My rs232 module is 5V and don't need any resistors to limit current (Look at my schematic diagram of scrumpel7 v1.3b at github) Those resistors are only needed for a 3,3V device to limit the current. With the 2K2 resistors the communication is not stable. (This is what I had with the SIO boards from Spencer (rc2014). I use all in and outputs. But the pull-up resistors can stay in place.

Copper pour:

What Mark says is only useful on high frequencies < 50MHz so your approach isn't that bad after all. The reason I suggested the two copper pours was that I saw long lines to the VCC and GND of some ic's. For HF VCC and GND are the same because the impedance of the power-supply is a very low resistance. The only one disadvantage of my method is the capacity (about a few pF) you introduce, but as already told this is only a problem at high frequencies. More important is keep the clock lines as short as possible. When using multi-layer (more then 2)  then inductance and capacity can cause more problems.

Here the code of my serial input-routine on Scrumpel (Z8S180) I use no interrupts.

serrst  in0     a,(ascicntla0)
        and     a,11110111b
        out0    (ascicntla0),a
        ret
rxa     in0     a,(ascistat0)    ; Check the overrun bit
        bit     6,a
        jr      z,rxa1
        call    serrst
rxa1    bit     7,a              ; Check the receiver empty bit
        jr      z,rxa
        in0     a,(ascirx0)
        ret

Probably you'll create a new release of SCM soon. How can I update this in the system itself? With romwbw you can use the flash utility in CP/M. Otherwise I have to use my prom-programmer.

Op zaterdag 14 september 2019 21:38:48 UTC+2 schreef Steve Cousins:

[Richard Deane]

To program Rom, could you boot Romwbw monitor, boot from disk, switch Rom select jumper and then program SCM Rom from running Romwbw?

[Marten Feldtmann]

Ist the source code for the S6 confguration available ? I can not find it.

Marten

Am Donnerstag, 20. Juni 2019 11:06:55 UTC+2 schrieb Steve Cousins:

My Z180 motherboard is progressing well.

This design is a result of Richard Deane's design challenge discussed here.

I'm close to being satisfied with the design and a v1.0 production run.

It runs the Small Computer Monitor and RomWBW. 

I have some simple I2C code and hardware working, so the I2C bus looks good.

RomWBW is running with mass storage via SD card on one of the SPI / SD Card ports.

The Real Time Clock is also working with RomWBW.

The bus sockets are a superset of the RC2014 bus.

[Steve Cousins]

I'm part way through an update to SCM (which seems to be progressing very slowly) which is why the source code on my website has not been up to date. 

However, I've had several inquiries now and it will be a while before the full release is complete so I've just added a link to the latest work in progress version.

Source code and tools

Porting guide

General documentation

It is probably not very clear on my website, but the source code for the full release of SCM (v1.0) is part of the SCWorkshop v0.2 download. I've just added SCWorkshop v0.3 which includes the source code for the interim release of SCM (v1.1).

Steve

[Marten Feldtmann]

So, my SC126 has been put together and it worked out of the box :)

[Steve Cousins]

Marten, that's what I like to hear :)

Steve

[Richard Lewis]

I've been setting up only 4-layer boards now with SMD and TH mixed parts with a standard stack-up. Also been careful not to break-up the ground plane on inner layer 1 by spacing apart vias and ensuring for TH parts that i put thin traces between pins connecting them to gnd so I don't create slots in the ground plane. 

Hi-speed single/clk/bypass-caps/GND

--------------------------------------------------

GND Only

--------------------------------------------------

VCC Only

--------------------------------------------------

Signal (passives and support IC's only)/GND

On OSH Park charges $10/sq inch which get expensive ($160 quoted for 3x 4x4" 4-layer board). Think JLPCB was $28 for 5. Quality from Oshpark it waaaay better. 

[Bill Shen]

It is always good to follow best practices, but having worked during the Dark Age of pc board design and fabrication and built numerous prototype boards with poor power & ground distribution, I would also say that parts from 1970's and 1980's work just fine in 2-layer board with simple power & ground distribution scheme.  In fact, (this is a bit controversial) the designs of those era will generally work without ANY bypass capacitors--give that a try on any RC2014 boards.  I'm not suggesting getting rid of bypass caps, but only suggest that placement of bypass caps is not important.  I'll push bypass out of way to make room for other components.  For 4 layer board with dedicated power & ground planes, I'm completely comfortable to go without any bypass caps.  This comment is specifically for 5V digital TTL/NMOS/CMOS from 1970's and 1980's; ECL still need bypass cap and analog circuits absolutely need meticulous component placement.
  Bill

[Tom Storey]

On Sunday, September 15, 2019 at 5:07:06 PM UTC+1, Steve Cousins wrote:

I'm part way through an update to SCM (which seems to be progressing very slowly) which is why the source code on my website has not been up to date. 

However, I've had several inquiries now and it will be a while before the full release is complete so I've just added a link to the latest work in progress version.

Source code and tools

Porting guide

General documentation

It is probably not very clear on my website, but the source code for the full release of SCM (v1.0) is part of the SCWorkshop v0.2 download. I've just added SCWorkshop v0.3 which includes the source code for the interim release of SCM (v1.1).

Steve

Hi Steve,

Why not use a versioning tool, e.g. git with github/gitlab, rather than distribution via a ZIP file?

Although it takes a little getting used to, there are several advantages to this approach:

* Full history of all changes (if you ever break something you could go back to a working state without having to remember and revert everything you changed by hand)

* Different branches can be used to store stable vs "work in progress" code

* These websites often include bug reporting tools which can feed directly into the development process (e.g. the hang bug), 

* Once someone has downloaded your repo, it is trivial to keep it up to date

* People can make changes and submit them to you for review and integration

Its a very different mindset, but once you get in to it you wonder how you ever did without it. :-)

[Steve Cousins]

Hi Tom,

Thanks for the advice. Yes, I can see at least some of the benefits you listed and have often considered making the necessary change of mindset. I think the problem is I just have other things to do. I've gone and got hooked on PCB design and now making kits, so the software projects have not really progressed. When I do want to work on the software it just seems easier to stick to the methods I already use. I guess I'm just getting old :) Perhaps I'll give it a try at some point. 

At the moment SCM source and documents are in flux and a bit confusing. If you are interested in v1.0 (the long term stable version for most hardware) then the docs and source are all there. If you need (or want v1.1) then things are not so clear. I'll have to try to improve that situation soon.

Steve

[Tom Storey]

On Wednesday, September 18, 2019 at 12:49:54 PM UTC+1, Steve Cousins wrote:

Hi Tom,

Thanks for the advice. Yes, I can see at least some of the benefits you listed and have often considered making the necessary change of mindset. I think the problem is I just have other things to do. I've gone and got hooked on PCB design and now making kits, so the software projects have not really progressed. When I do want to work on the software it just seems easier to stick to the methods I already use. I guess I'm just getting old :) Perhaps I'll give it a try at some point. 

At the moment SCM source and documents are in flux and a bit confusing. If you are interested in v1.0 (the long term stable version for most hardware) then the docs and source are all there. If you need (or want v1.1) then things are not so clear. I'll have to try to improve that situation soon.

Steve

Old habits die hard, as they say. I hear what you say about time - theres not enough of it. Im rennovating a house, so all of my projects which I so desperately want to work on are on the back burner. :-( Things are gradually returning to normal.

But, you can also use it for your PCB designs! You can check in binary files as well as plain text files. Happy to give you a 5 minute intro to git if you ever want to know the basics.

[Tom Szolyga]

Hi Steve,

I assembled my SC126 from a bareboard.  It powered up and worked correctly on the first try.  I like the two FLASH ROMs with alternate systems.  It is very easy to power down, change a jumper and boot up in the other system.  My next step is to try SD Card operations.

Great board!  Thanks!

Tom

[Karl Albert Brokstad]

Starting at first try? What did you expect? This is not RC#$%& its SC126..... 

[Steve Cousins]

Tom, 

Actually you can just hold the reset button, change the jumper, and then release the reset button. Or you can just change the jumper while it is running then hit reset. 

Steve

[Francis Pierot]

Another plan is to build a case, and have a front switch linked to the jumper. One side labelled SCM, the other WBW. And a reset button f course, although I tend to put this button on back side to avoid mistakes.

[Francis Pierot]

My RC2014 Pro with 1000 soldering points worked at first try. It's not easy, nobody would say that. But if you spend time in docs (admittedly dispatched here and there) you figure out what switch or jumper goes where, how you configure RAM and ROM and SIO and PiZeroTerminal and so on, and when you power the thing up in the end, it also can work at first. It's what I lived.

If your point was that's it's not an easy kit, I agree. But it's designed to work at first try :-)

[Nigel Kendrick]

On my Searle-ish build with a CH340G-based USB-serial adapter, I had to drop down to 510R to get comms working.

Notes at https://github.com/linker3000/Z80-Board#usb-serial-adaptor-notes

-- Nigel

[Richard Lewis]

For the RC2014 Pro... besides soldering the wrong header on the RAM card which I had to the snip off and getting a few resistors wrong, mechanically it was more or less successful (had to look at several different photos on Tindie to get it right). Afterwards it still took 3 days to get CP/M running on it and ROM Basic never worked. 

For the SC126 it took about 2 hrs to go from bare board to Zork. The major problem with Steve's boards is that it's impossible to learn hardware debugging with them because I never had one fail. 

-Richard

[Greg Holdren]

I bought some of these SD adapters on Amazon. These look the same as far as layout and components to the ones Steve sends out with the SC126. The seller provided schematics for the adapters.

Alternatively, the lifted DO (MISO) signal pin 13 for tri-state enable DO, can be connected to pin 8 of the same device.

Greg

On Sunday, September 1, 2019 at 9:01:00 PM UTC-7, Greg Holdren wrote:

Alan,

Looks like it will work. Nice way to expand the SPI ports using a binary weighted CS value to get 4 vs a 2 bit CS value using "One Hot" outputs.

My idea to fix the SD card adapter was just to add a non-inverting tristate buffer to the MISO line. But.... there is already a LVC125 on the adapter! I followed the MISO line back to pin 11 so pin 12 is the input from the SD card. Pin 13 is the tri-state control line. This line is hardwired to ground which always enables the MISO driver. As a fix I lifted pin 13 disconnecting to ground and connected the actual pin on the device to CS  on the SPI interface connector. This will enable the MISO output when CS is active. This should fix the issue with the shipping adapter. Drive G: (SD card) still works under WBW bootrom.

I don't have another SPI device ready to interface to the second port now so if someone can verify that the second port works with the SD adapter that would great.

Greg

On Sunday, September 1, 2019 at 5:27:25 AM UTC-7, Alan Cox wrote:

On Sunday, 1 September 2019 07:59:49 UTC+1, Alan Cox wrote:

Rev 1 was just a 138 to split the two CS lines - both of which start low - into 4. Enough to use it with polite devices.

Rev 2 adds a mux on miso so that the input follows CS. I will build and test that once sufficiently recovered.

Alan

Untested as I said....

[Marten Feldtmann]

Steve,

what I do not understand ... why did you not connect /int1 and /int2 to the bus as you did with I2C ?

[Steve Cousins]

Hi Marten,

The honest answer is that I just forgot to bring them out to the bus connector.

Steve

[Francis Pierot]

This Sunday has been soldering day for my SCZ180, SC126 form. I had a few connectors changes to cope with because I've designed a case for this beautiful baby, one with leds and inputs and switches on the front and back panels instead of on board. A computer must blink!

I cannot express how great I find the documentation. It matches the quality level of Steve's tutorials on his site.

I have a few minor suggestions, one for the board itself (certainly not worth a new version :-)  and three for documentation.

On the hardware side, I choosed to "export" the LEDs and resistors for power to a lighted power button (with integrated led and resistor), and status/port 0x0D leds and resistors to a little board which will be fixed to the front panel. Replaced the power switch by a 3 pins header, for the LED I simply soldered a pin at the (-) hole of the LED1 and another at the bottom hole of R11, to connect my (-) and (+) pins on the button. In the same move, I dispatched the status LEDs and resistors. This is where my suggestion comes. Currently, the resistors pins are spaced by a distance which doesn't match positions of a pin header, so I had to cut lone pins and solder them one by one, and it's been delicate - whilst not critical - to have them correctly aligned. It would've been easier if they'd been separated by a more practical distance or maybe use a resistor battery.

On the documentation, it would be nice to have the number of soldering points. (Say, with all optionnal pins). Kind of indication of the time it will takes. I thought I would be fast seeing the low number of components but I had underestimated the number of ICs sockets and bus headers and pin headers for peripherals.

Now, the Assembly and the Circuit Explained docs are really something!

But I found my self switching back and forth between Assembly and the Circuit Explained docs. Why not merge them, so the explanations of Circuit Explained comes handy at the time we solder the related part? It's really great to understand what you're soldering and why it is there when you're a beginner!

Last, one of the nicest things are the Quick Test steps in assembly guide. It really gives confidence along the process. There's only one information that I missed. At the last Quick Test step after the resistor netword soldering, the very last 2 tests are on pins 11 and 12 of the Z180 socket. You'd assume the marked angle is pin 1, then 2 on its right etc. Not sooo fast, Junior. Zilog don't want its Z180 babe to be for any rookie wannabee. Pin numbering starts at the middle of a side and goes counter clock wise. Rock'n roll baby. Actually I found the information easily on Zilog doc but it's still surprising  (http://www.zilog.com/appnotes_download.php?FromPage=DirectLink&dn=PS0140&ft=Product%20Specification%20(Data%20Sheet)%20%20&f=YUhSMGNEb3ZMM2QzZHk1NmFXeHZaeTVqYjIwdlpHOWpjeTk2TVRnd0wzQnpNREUwTUM1d1pHWT0=). So I think, Assembly Guide would ease things at this step with a little image to tell where the pins actually are. (Or tell they're 2nd and 3rd from the right on top side.)

Apart from this, I don't think I've ever had a kit with such a fantastic assembling and technical documentation. It made by day, litterally :-)

Unfortunately I cannot tell if my board works yet - although I certainly have total confidence in it - because I have not populated it with the ICs. I'd rather avoid any accident while handling the board and working on the casing so I prefer to keep sensible components safe until launch day.

I'll let you know when it's done, but it's something to draw the case you want on paper, and something else to punch the right holes at the right place with the right tool in a metal plate. Work is in progress, and I'm awaiting some boards I ordered a little late.

[Steve Cousins]

Hi Francis,

Thanks for the feedback. I'll give those ideas some thought,

The resistor spacing I think was chosen just to look nice. As indeed was the use of separate resistors rather than a resistor network. You can't beat nice visible resistor colour codes for a real retro look.

Keep the feedback flowing. I think there is quite a bit of room for improvement so perhaps there will be a version 2. Of course, you can't make everyone happy as some ideas are mutually exclusive.

Please post pictures of the case when you get to an appropriate point.

Steve

[Francis Pierot]

Hi Steve,

I received the expansion board SC113 that I want to put with my SC126 in the case.

I've been optimistic and didn't what an engineer should do, that is measure things to get sure it fts.

So, to make short, it doesn't fit, that's too wide for my case, and I must find a B-plan. I have a number of possible plans.

1) Give up the metal case I have, and build one from wood or plastic or pvc or whatever using the dimensions I need.

2) Get schematics from SC113 (easy, as you give access to them) and derive a 3 slots version (keeping the out expansion slot), make it build on an asian site then wait a few weeks and then check.

3) Not likely at my level, put the expansion board above the SC126 board, linked by the leftmost RC80 slot (sacrificing one on each board and almost giving up the out-expanding slot.). Gives up a few SC126 jumpers below or designing connectors away from board.

4) Not likely at my level, re design SC126 with 3 slots more.

5) Put only the SC126 in the case, and bring the expansion slot available from left case side so I can connect the SC113 in box of whatever.

6) Don't have yet but there must be one. At least.

Right now, I think I'll give a try at plan 2, so I can learn a few things. But plan 1 also talks to me as I love working with wood. The thing bothering me is heat dissipation, which is typically not good with wood or plastic cases.

Of course if you have a 3 slots expansion board in mind, I'll be glad to beta test it. Or, make the SC126 motherboard taller and redistribute components so it's not as wide (165mm) as it is currently.

I'll sort things out anyway, this post is more a list of things you can keep in mind when you'll design V2.

[Alan Cox]

How narrow do you need it to be and how many slots - would something like this do the trick ?

https://www.tindie.com/products/DTronixs/extender-card-for-the-rc2014-bus/

Alan

[Steve Cousins]

Hi Francis,

I assume where you have said "SC130" you mean "SC126".

I don't have any plans for a 3-slot version of SC113.

Number 6: Cut the SC113 to the desired length with a hacksaw. File the edge of the board to make a clean edge to the cut tracks. Check there are no shorts between tracks you have cut and filed. One problem with this method is you won't have a space-efficient place for a horizontal expansion socket

One disadvantage of the modular backplane approach is that the plug and socket which joint the sections uses up space that could provide a further card socket. Another is that all those extra connections is a potential reliability issue. The best way to get three extra sockets would be to extend the motherboard. Each extra socket requires the board to be extended 0.6 inches (15.24 mm). Three extras would add 45.72 mm which would take the board (currently 160.02 mm) to over 205.74 mm.

It would be relatively easy to take the SC126 design and add 3 extra slots. As long as you do not change the main component area it shouldn't be any more difficult than producing a three slot version of SC113. The PCB cost is however significantly more when working on a board greater than 102mm x 102mm.

SC126 only generates about 0.5 watts of heat. Adding typical expansion boards might push this to 1 watt or perhaps a little more. A few ventilation holes in just the back of the case should easily prevent this being a problem. If you have plans for more power hungry devices in the case then you may have to provide better ventilation.

Steve

On Monday, 14 October 2019 22:49:54 UTC+1, Francis Pierot wrote:

Hi Steve,

I received the expansion board SC113 that I want to put with my SC126 in the case.

I've been optimistic and didn't what an engineer should do, that is measure things to get sure it fts.

So, to make short, it doesn't fit, that's too wide for my case, and I must find a B-plan. I have a number of possible plans.

1) Give up the metal case I have, and build one from wood or plastic or pvc or whatever using the dimensions I need.

2) Get schematics from SC113 (easy, as you give access to them) and derive a 3 slots version (keeping the out expansion slot), make it build on an asian site then wait a few weeks and then check.

3) Not likely at my level, put the expansion board above the SC130 board, linked by the leftmost RC80 slot (sacrificing one on each board and almost giving up the out-expanding slot.). Gives up a few SC130 jumpers below or designing connectors away from board.

4) Not likely at my level, re design SC130 with 3 slots more.

5) Put only the SC130 in the case, and bring the expansion slot available from left case side so I can connect the SC113 in box of whatever.

6) Don't have yet but there must be one. At least.

Right now, I think I'll give a try at plan 2, so I can learn a few things. But plan 1 also talks to me as I love working with wood. The thing bothering me is heat dissipation, which is typically not good with wood or plastic cases.

Of course if you have a 3 slots expansion board in mind, I'll be glad to beta test it. Or, make the SC130 motherboard taller and redistribute components so it's not as wide (165mm) as it is currently.

[Francis Pierot]

Yes you guessed right, I was talking about the SC126, late hours made me mix with 130. (I've updated my post to fix that.)

I had thought about your solution 6 in fact but I'm reluctant to cut this baby. I'll probably order an SC130 (the actual 130 this time) and plug the 113 in it into another box to make a pure CP/M box. Piece of cake, keep it for later.

In the meantime, today I dived into EasyEDA. Downloaded your SC113 files, and derived a Tiny version with 3 slots (and in/out expansion sockets, so the out socket willl be available on the case left side).

It's my first EDA work ever. It was MUCH easier than I thought (although quite intimidating at first). Trivial schematics helps of course, but at first I had no idea of how to do what. Then I did a few mistakes but found them rather easily using the layer display/hide functionality. The biggest work was to move mounting holes, I had to redraw all related tracks. I used copy/paste, to discover later that I had to rename the pasted tracks (net-name) to ease checking. In the end I think I'm as close as possible to your work.

The only question I have left is, when you are sastisfied and you don't find any more problem visually in EasyEDA, is it time to order a sample, or do you have a few checking processes that can be done before?

Thanks for the advices!

[Steve Cousins]

Hi Francis,

Sounds like you've cracked it. I predict you will be addicted to PCB design within a few months.

The one thing to do before ordering is to ensure it passes the Design Rule Check (DRC). If your schematic is correct and the board design passes the DRC you can be quite confident all the tracks go to the right places and no routing is missing.

Enjoy your first parcel of PCBs. It is quite rewarding opening that first parcel and seeing a stack of professional looking PCBs.

Steve

[Steve Cousins]

One more thought:

I expect you have already found this, but I use the photo view of the board to spot aspects of the layout and screen print that I don't like the look of. It is worth spending a few minutes having a good look at both sides of the board. It is amazing what you spot doing this especially when you zoom in to look at the details.

Steve

On Wednesday, 16 October 2019 00:37:32 UTC+1, Francis Pierot wrote:

[Francis Pierot]

Thank you for suggestions about DRC checking and the photo view. I had not seen the Design Manager button! This is exactly what I had in mind.

I found a few net errors but most importantly, a hardly visible direct link between a +5V hole and GND! Regenerating copper areas fixed that.

Still, the DRC highlights a number of clearance issues in the little solder pads area near P1. They're also in your SC113 original file and I guess it's not an actual problem, as these pads are so close on purpose. I suppose making another design rule with a shorter clearance would make DRC happy.  I ignored the warnings.

And here am I, just ordered "my" first own-designed (so to say :-) PCBs.

I can testify that ordering a first pcb is both exciting and frightening!

By the way at the same time I ordered a few SC129. Was not well awaken this morning and soldered the IC sockets the wrong way. Yes, I mean on the back side, right were it's written "solder this side" :-)

I had to create a project in EasyEDA, open your files and save them in my project. Is this the right way to do?

[Steve Cousins]

Hi Francis,

Your question about opening files from one of my projects is an interesting one. 

I get a message when someone "Clones" one of my "projects" or "libraries". However, I have created a separate account and can't see how to clone a project. I feel I'm missing something 'obvious'. 

I don't seem to get a message when someone opens one of my files (and presumably then saves it in their own workspace). This again suggests there is a separate mechanism to clone things, which I haven't found.

Perhaps someone can point out what I have missed.

Loving the story about the soldering error. We all make mistakes.

I'm just curious here: Did you see my little series about designing circuit boards?

https://smallcomputercentral.wordpress.com/2019/07/26/design-part-1/

Steve

[Bill Shen]

As many years as I've been soldering I can tell you the most exciting mistake is having moisture under a surface mount component.  As you (perhaps I should say 'I') applied heat to the first pad, the moisture can flash steam and pop the SMT right off the board with a cloud of smoke and associated sound effect--instant heart attack.  Do you know a dropped solder gun can quickly melt a nice hole in the linoleum floor?  Do you also know a linoleum floor with burnt solder hole in it is a pilgrimage site for newly trained assemblers?
  Bill

[Francis Pierot]

Your steam explosion story made me laugh :-) But I'm not very good at SMT (and don't really intend to be at the moment, but maybe some day!) so I'm quite protected from this exploding special FX.

Linoleum also like clumsy smokers, but I stopped smoking long time ago. (Except flux.)

Francis

[Francis Pierot]

I've seen your series exist and just had a fast look, I miss the time to read them at the moment but I certainly will soon. This SC113 modification opened my appetite!

While I wait for the pcbs and a few connectors/components, I'm still working on the case. I drilled holes for the 8 status LEDs on front plate but I wasn't able to align all 8 holes correctly, and also have put the little board with leds too close from the on/off button. I'll sacrifice the plate and make other holes in it using better techniques. That's what prototypes are made for! I have a lot of LEDs, switches, buttons, dip swicthes and female headers to appear on the front plate. In my dream, the front place will look like a Gemini capsule control panel. Sort of.

I'm still dubitative about the metal plate. PVC would be easier to work at least for front and rear plates but I can't find a seller (I started to look for it on the web but no success yet.)

With EasyEDA I also discovered the time it takes to look for components on Mouser or Furnell when you don't know all the terms. But I learn.

I start your EasyEDA tutorial tomorrow, for today I had enough electronics (and I have to test a programmable power supply I soldered :-)

Francis

Le samedi 19 octobre 2019 15:18:32 UTC+2, Steve Cousins a écrit :

[...]

[Richard Lewis]

I had a similar story. Got a bunch of SMD components on the cheap from an Alibaba source, pasted up my board, meticulously placed the parts with tweezers then put the board in my reflow (Black&Decker toaster) oven. As soon as it hit around 120C (ramping to 235C) I heard popcorn like noises. Afterwards I opened up the oven and there were parts everywhere, some soldered onto the oven tray.

I then learned why Mouser/Digikey always packs their parts with silica gel and include humidity indicators... When SMD stuff has been sitting around for awhile supposed to bake them at 100C for 24hrs. Don't have that kind of patience. 

-Richard

[Alan Cox]

By the way at the same time I ordered a few SC129. Was not well awaken this morning and soldered the IC sockets the wrong way. Yes, I mean on the back side, right were it's written "solder this side" :-)

Should work providing you bend all the IC pins back the other way I think  ;-)

I never quite figured out how to clone something in EasyEDA except by accident and then I couldn't find out how to delete it. I've taken to Kicad 5 instead, in part because if relations between the West and China go terminally bad I'd like to still be able to use my design files.

Alan

[Alan Cox]

On Saturday, 19 October 2019 15:05:19 UTC+1, Bill Shen wrote:

As many years as I've been soldering I can tell you the most exciting mistake is having moisture under a surface mount component.  As you (perhaps I should

Wiring a big darlington driver in a robot backwards is IMHO a contender, along with wiring one of the really big capacitors backwards and filling the room with snow.

It may be admitting defeat but with the large surface mount components like the 3v3 regulators I am going to try putting a tiny spot of a sticky glue (not superglue!) on the underside before I solder the little blighters.

say 'I') applied heat to the first pad, the moisture can flash steam and pop the SMT right off the board with a cloud of smoke and associated sound effect--instant heart attack.  Do you know a dropped solder gun can quickly melt a nice hole in the linoleum floor?  Do you also know a linoleum floor with burnt solder hole in it is a pilgrimage site for newly trained assemblers?

And family members making inquiries. First about the funny smell and then about the cost.

Alan

[Steve Cousins]

I've just worked out how to clone a public project on EasyEDA.

Search for the project and click the 'not so obvious' icon indicated below. This creates a copy of the project in your personal workspace. It also inherits the "shared" status which you may want to remove.

On Saturday, 19 October 2019 13:36:40 UTC+1, Francis Pierot wrote:

[Richard Lewis]

Try solder paste and a hot airgun, a ceramic hotplate or even a frying pan. The paste is quite sticky from the flux and will hold the parts down. 

[Richard Lewis]

Looking at the photo it reminds me of the hardest soldering job ever... It was getting all of the pins of the right angle header into the holes. 

[Francis Pierot]

It should, and the idea is so horrible that it deserves a try.

But I don't have spare ICs at the moment so I think I will wait ;-)

And there would be space conflicts with neighoring boards I suppose, unless of course all the other boards are soldered the wrong way out as well.

[Francis Pierot]

I just finished the part 1, should have read it before jumping into EasyEDA :-)

Thought you'd like to know I met two differences between EasyEDA (both offline and online) and what's in your tutorial. Probably changes in versions (I use v6.2.46 on Windows, dated 2019/10/10).

The first when I used Convert to PCB: it complained about the "screw terminal 1x2" having no package and refused to go further. I had used the one from EELib with the same name as in your tutorial. I found another component in library which had the required elements. (That is, a visible name in the "package" column in the Libraries dialog.)

The second when I hit Convert to PCB again, I didn't have any dialog for the PCB size. It immediately switched to the PCB editor, with a default (and quite minimal) size for board. I found no dialog box to edit the board size and position easily so I moved the sides and adjusted X and Y coordinates to make sure I had right angles.

I'm now starting to read Part 2, but it already starts a few levels above my knowledge. Which also means, I will learn interesting and useful things ;-)

Francis

[Tom Szolyga]

Hi Steve,

I tried your procedure to clone the SC126 and THE CLONING PROCESS WORKED PERFECTLY!   I have used EasyEDA for a long time, but was never able to get cloning to work well.  

Thank you for your research and explanation!

Tom 

[Francis Pierot]

News from this Sunday: I've jumped the gap - have to wait about 3 weeks for my sc129s and modified sc113 anyway - and put all ICs in their socket, and plugged a 5V source.

As I expected, my SC126 work s like a charm at first powerup! Linked to my PC/Teraterm via serial port A and the FTDI adapter from the kit, have both SCM and RomWBW boot fine. It's just amazing to see how SCM is so blasingly fast to boot, and RomWBW at least ten times slower, which at 18MHz is not really "slow": about one second. But scm boots in no time at all :-)

I have not checked on my MacBook Pro yet but if it works in Teraterm on Windows, I expect no issue on iTerm on my mac. (And so much prettier!)

What a dream machine compared to what we had in the 80's!

I have an issue with the pizero terminal though. I will talk about it on the existing topic in this group.

Francis

[Wayne Warthen]

On Sunday, October 20, 2019 at 3:17:26 PM UTC-7, Francis Pierot wrote:

As I expected, my SC126 work s like a charm at first powerup! Linked to my PC/Teraterm via serial port A and the FTDI adapter from the kit, have both SCM and RomWBW boot fine. It's just amazing to see how SCM is so blasingly fast to boot, and RomWBW at least ten times slower, which at 18MHz is not really "slow": about one second. But scm boots in no time at all :-)

RomWBW uses 1-2 seconds to measure the speed of the CPU.  That is actually the primary reason for the slow boot.  If you want, you can configure this step out of the boot process.

-Wayne  

[Francis Pierot]

Thanks for the info Wayne, that was intriguying me :-)

I certainly can live with a boot time less than a few seconds. I'm used to Linux, Raspbian, MacOS and Windows hefty boot times :-)

[David Reese]

Hello, All,

Got my SC126 Kit yesterday, 13 Jan 2020, after 16 days (U.S. Customs must have thought it was a "nook-ya-lur dee-vice" or something...).  

Inventory of parts completed with all present, and the small stuff sorted into pill bottles.  This will be fun!

I can see right away my Radio Shack pencil iron is not going to cut it, so I'll be buying a new iron for this build (something with a much finer tip).

I have a plan.  When the build is completed, I'll post again with pictures.

[Karl Albert Brokstad]

Hi David. 

SC126 is a good choice.

Good luck soldering!

Karl

[mark acierno]

Really having fun with  my SC126. I have a question about using the second serial port to access a WiFi modem. Are there patches that will allow terminal programs (and BYE) to access the second serial port? Would like to be able to use a WiFi modem for in and outbound communications. 

Thanks

-Mark 

[Blane Dabney]

My daughter and I soldered ours with a Weller WP35, with 0 issues.  Used MG chemicals flux paste and 60/40 solder. None of this lead-free mess :D

[David Reese]

On Wednesday, January 15, 2020 at 6:57:44 PM UTC-6, Blane Dabney wrote:

My daughter and I soldered ours with a Weller WP35, with 0 issues.  Used MG chemicals flux paste and 60/40 solder. None of this lead-free mess :D

Yeah... the Radio Shack iron I have is an Archer, catalog no. 64-2080, which I believe was made by Ungar.  (It's so old the internet seems to have no memory of it...)  The only tip I have for it is a 1/8" (3.5 mm) chisel point.  Definitely not fine enough for this kind of work, and I'm not taking a file to it.  I'm looking at a kit from Yome that costs less now than my present iron did new.  I have a roll of ECG/Sylvania rosin core 60/37/3 eutectic solder I'll be using, pretty fine gauge stuff (0.03", about 0.762 mm?), so none of that unleaded stuff for me, either - I'm driving.

I'm willing to wait a few more days until the right tools are in hand.