RCBus backplanes

[Steve Cousins]

I'm in the process of designing a number of backplanes that meet the full requirements for a generic RCBus backplane. Attached is a render of one of them. These use my established modular approach with a horizontal bus socket at the end to extend the backplane or connect a module.

There are a couple of things I can't decide:

I prefer my systems to be powered from a 12 volts supply. To this end I have a power supply module with a nice inefficient regulator that runs at a proper retro temperature (hot). This module also has a voltage supervisor that gives a clean reset, and some nice LEDs on the reset signal etc. To me this seems the correct solution for a modular system - or it least it does seem the right solution some days. However........

I know the 5v input via a barrel style connector on the backplane is a popular way to power these systems. 

It seems a shame to use an 80-pin bus socket just to connect a power supply. 

I have ended up with some CPU modules including a voltage supervisor to give a clean reset on systems that do not use the power supply module. 

I don't like using the same barrel connector for systems with different voltage inputs. And, yes, I have plugged a 12 volt supply into a 5 volt system!

I also have single board computers and motherboard style systems to consider.

There should be one "correct" way to do this. But I can't decide what that is.

I would be interested to hear what everyone else thinks about where and how to include power and reset in an RCBus system.

Steve

[Tadeusz Pycio]

I am in the process of designing an active backplane (5 slot + expansion), with MP1584 step-down converter, TL7705 supervisor chip and CH340E USB-Serial converter and mini USB connector. A project for my own needs, so I used SMD components. Also I will use a barrel connector 2.1/5.5 for the power supply and I intend to use it in a different colour (yellow, red?) to distinguish it from the 5V system.

[Dylan Hall]

I like the idea that the backplane is fed 5 volts by something, be that a barrel jack, screw terminal, or a dedicated power supply module. This keeps the backplane simple/compact.

If you want other voltages then you need extra bits like regulators, caps, etc so having the extra space a dedicated module provides feels like a good idea. I also have a couple of modules that need a lot of power (~750mA for a TIL311 display and the IDE module with a 2.5" hard drive bus powered). Under those circumstances I'm not sure a 7805 is enough, certainly not without a heatsink at least. Outsourcing voltage regulation to an external power supply solves this problem. 

In my recent backplane I went very non-retro and included a USB-C connector for power (https://www.adafruit.com/product/4907). I got sick of replacing my cheap power supplies when they start failing. My theory was it's easier to find a higher quality power supply that's USB-C (for example a laptop power supply). So far this is working nicely :)

I include a reset switch on my backplanes. The backplane tends to be solid where modules are wobbly and I don't like the idea that I'll carelessly press the reset switch one day and push one module into the next and short something. 

I've been including a voltage supervisor and clock on my CPU modules (I copied the design from you!). I justify this to myself by saying that the CPU module is responsible for providing signals on the bus where the other modules are consumers of signals. To put it another way, the CPU module provides the reset signal so it's responsible for ensuring it's quality, not a power supply card or backplane. I'm not sure this will hold up in a debate, but it's good enough for me :)

Dylan

--
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/cc844015-69d2-484e-a17b-672dfac61af5n%40googlegroups.com.

[Brad Hines]

I'm an outlier, but in case it's useful to see how some of us outliers live...  If this is useful to you, great.  If not, my feelings won't be hurt.

The short of this is that if one wants to entertain having a backplane that could accommodate a 8080A card, 12V is nice to have.  (The 8080A is pretty compatible from a signaling standpoint, not hard to interface.)

Since the 8080A requires 12V, I use a 12V supply, and then have added MP1584 modules to deliver 5V.  (This is for my legacy 8080A that I am interfacing to Z50Bus.)

In my current setup, I am providing power to my SC513 backplane via its screw terminals, from a MP1584 on my main motherboard.  I have three of the MP1584's in all, which is overkill, but there are legacy reasons having to do with the original layout of my system.  In any event, I am powering the SC513 off a single MP1584.  They're supposed to be good for 1.5A continuous.  Not sure if that's enough for folks.

This is ugly, and nobody would want to do it this way; I just want to present how one person solved the problem.  I found the screw terminals handy.  

As far as 5V, of course if you're making a motherboard you could integrate the parts on the motherboard rather than using prefab MP1584 modules.

Brad

--

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/cc844015-69d2-484e-a17b-672dfac61af5n%40googlegroups.com.

--

Brad Hines
Brade...@gmail.com

[Brad Hines]

Hi Steve,

Here's one other thought, not fully formed, but maybe it will inspire someone.

What if the system accepted a USB-C wall wart?  Those things are great because they are so versatile in what they can put out, and so powerful, and they take the regulator off-board.  I have tested with one of these eval units:

   https://www.digikey.com/en/products/detail/cypress-semiconductor-corp/CY4533/10238329

and it works as advertised - you can get 5V, 9V, 12V, 15V, or 20V by dialing the 5-position rotary switch.

It uses this chip, $2.51 in qty 1.  https://www.digikey.com/en/products/detail/infineon-technologies/CYPD3177-24LQXQ/10238328

The "not fully formed" part of this idea is, "so great, how would you actually use it in a system?"  You don't want the user to be able to blow up their system, and any system that wants 12V will also want 5V.

But maybe there is some possibility that makes sense.   do find the USB-C power solutions really slick.

Brad

On Fri, Mar 17, 2023 at 5:33 AM Steve Cousins <steve...@gmail.com> wrote:

--

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/cc844015-69d2-484e-a17b-672dfac61af5n%40googlegroups.com.

--

Brad Hines
Brade...@gmail.com

[Tom Storey]

You could maybe use a crowbar circuit to protect the 5V rail from over voltage, not just from someone plugging in a higher voltage supply, but also from power supplies that go faulty and happen to dump a higher voltage onto a lower voltage rail.

But youre adding a lot of additional components to what was perhaps meant to be a very simple board.

I guess you could also make it an optional add-on or module that can be plugged/soldered in somewhere.

To view this discussion on the web visit https://groups.google.com/d/msgid/retro-comp/CAABP5hu%2Bkb%2BoCb6HR2_QWsse3SoKLAOmYYg4iGTdFs9%3DchNuvQ%40mail.gmail.com.

[fixit9660]

My thoughts are:

Power should be supplied discreetly by a dedicated Power Supply. If there is more than one voltage required then use a keyed multi-pin plug, similar to the way PC power supplies are connected. This has several benefits;

1) You can't connect it the wrong way around.

2) You can't connect the polarity back to front.

3) You can't connect some supplies and not others.

4) You can't connect the wrong power supply.

5) The power supplies are connected "globally" across the system and in the correct sequence. I realise that having multiple power rails on the backplane requires heavier tracks and thus reduces the space for other tracks, but it means that all boards have power connected simultaneously, and so stand a good chance of powering-up together. You don't need to consider a group of independant boards all doing their own power regulation and production at different times.

6) You don't need a board/backplane socket for Voltage Supervision.

7) Having the power done by a dedicated PSU reduces complexity across the system by removing the function from all the individual boards.

I believe the backplane should be the least component-complicated part of the system, with any switches or indicators being "somewhere else" connected by keyed connectors.

When do you anticipate the first Production run of the new Backplane please?

[Steve Cousins]

Thanks for your thoughts.

My first two RCBus backplane PCB designs are in production at the moment. I should have them in a week or so. I expect them to work as planned so I shouldn't need to make any changes. I plan to offer them as kits and as PCBs in about two weeks. The designs will be in the public domain with the design files and Gerber files freely available.

One design (SC702) is a passive backplane section, while the other (SC701) includes a 5 volt input and a reset/supervisor chip. Both offer 6 vertical RCBus sockets and one horizontal RCBus socket.

Both of these designs have two pins connected with thicker tracks for possible future power use. I was tempted to make all the tracks thicker but the additional capacitance is something I'm not so sure about.

I'm also likely to offer versions with 12 vertical RCBus sockets.

I personally like to use a 12 volt supply so I also have a power supply module (SC703) in production. This has a 7805 regular allowing it to be powered from 8 to 15 volts. This module also includes circuitry to give a clean and fast rising reset signal - better than provided by the DS1233 supervisor on the SC701 backplane.

There is no reason we can't add a backplane section to the range with a keyed connector as suggested.

Steve

[fixit9660]

I look forward to the availability of the Production backplane kits. I was keen on the SC508 so something with a similar number of sockets would be interesting.

As I read the spec there are plenty of "User definable" tracks available. In the past I have used such for power tracks by beefing them up with additional wire soldered to the backplane.

The keyed power connector was only a passing thought. I think a separate backplane version to accommodate it might be going too far. The fewer variations to a standard/specification the better? The majority should decide on the spec and I'll go with their decision.

Andy

[Tadeusz Pycio]

The basic version is waiting to be tested.

[Steve Cousins]

My first RCBus backplane seems to be working as expected. More testing to do, but it is looking good. I'll take a better picture when the big bright ball of fire in the sky is next visible here.

[Mark T]

Will that be May or June :) 

[Steve Cousins]

The weather here in the UK is not as bad as is generally believed. Admittedly, I can't see the sun today and it is quite wet out there, but that's just normal everywhere, isn't it?

 :)

[Norberto Collado]

Hello,

I ordered the "SC709 RCBus-80pin Backplane Kit" and I was wondering is there a chassis that I could use.

Thanks,

Norberto Collado

[Steve Cousins]

Hi Norberto,

None that I know of.
Steve

[Norberto Collado]

OK! Also, I received the boards last night. They all look great and thank you!

[Tom Plano]

If want something very elegant (but with a price to match), Ive found this enclosure to be more then satisfactory. There is enough room for an rc2014 backplane pro which is wider and a tad shorter then yours, as well as 2x 5 1/4 floppy drives. If you dont want drives there are even more options from Hammond, is various styles. Just be prepared to do some hand drilling to get mounting holes. 

Tom

[Norberto Collado]

Thanks Tom and will check it out.

I completed the assembly of the  "SC709 RCBus-80pin Backplane Kit" and when I try to insert a board the pins do not aligned with the backplane female connector holes. It is very hard to insert any module. 

However I do not have any issues inserting same module at the right angle female connector at the end of the backplane.

Any suggestions or or any tricks already learned?  

Norby

[Steve Cousins]

Hi Norby,

Is the plastic housing of the module sockets in contact with the PCB along the whole length of the sockets?

Steve  

[Robert Porter]

I had the same problem - down to the pins fitting the bus extender socket but not the vertical.  I think the vertical sockets are just a hair off. Or the pins and the horizontal are both off. Been meaning to check against some other headers and sockets but haven't had time.  It's just a hair though so I was able to carefully work it in.  Had to do this both for the Z180 CPU module and the ROM/RAM module. 

What I did was start with the board 1/2 lined up (so the left 20 pins are in the right 20 sockets) and work it slowly into the board. Then keep adding another column of pins and do it again. Like the shampoo bottle says: Rinse and repeat.  When you get to the last couple columns I think you'll find it starts getting harder.  SLOWLY work the pins into the socket.  For me it's only over the full width of the bus does it become an issue.  If it was a 2x35 header and socket it would have been fine. 

I've attached a close up and this is after I'd gotten the pins worked in a couple times. 

Rob

--

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/62918614-6921-47fa-a645-b6bf7f317cc7n%40googlegroups.com.

[Steve Cousins]

Hi Rob

That photo makes the alignment error look quite significant. Good to know that you can resolve the problem with a little jiggling. I assume once you have forced them into alignment once it is easier the next time you need to plug them in.

I've just checked a few bus sockets from my current batch and they seem to line up okay with the two modules I tried. I guess there could be tolerance issues with some batches.

I've seen some bus sockets that are just slightly bowed which is why I asked earlier if the sockets are in contact with the PCB along the whole length. Any socket that is slightly bowed can be pressed against the PCB when soldering to straighten it.

Steve

[Mark Cohen]

I had the same issue with the header with my sc112 backplane that I received yesterday.

I had this with only 1 board and header. I didn't try to move it to a different header so I do not know if others had the issue.

I can tell you that the PCB was bowed and I did my best to make sure I put it in straight.

Thanks,

Mark

To view this discussion on the web visit https://groups.google.com/d/msgid/retro-comp/9adbb4b4-98d7-4835-a3cb-4431c8deed33n%40googlegroups.com.

[Steve Cousins]

Hi Mark,

I've just checked my stock of SC112 PCBs. I found three green PCBs with slight bowing. Pressing one corner down on a flat surface, the far corner is about the thickness of the PCB off the surface. Is that similar to what you experienced?

Your PCB is blue I believe. I didn't find any blue PCBs where I could see a bow in the board.

Steve

[Mark Cohen]

Yes. All correct. I saw the bowing visually while holding it in my hand and inserting headers.

I could not understand why one end was almost impossible to insert a board into the header but maybe this explains it.

Thanks,

Mark

To view this discussion on the web visit https://groups.google.com/d/msgid/retro-comp/da56ec6e-efb3-4335-b698-95565183775fn%40googlegroups.com.

[Steve Cousins]

I've checked my stock of double row (80 pin) bus sockets. Out of about 100 I measured, 7 are short and appear about the same as Rob's photo. The others all look spot on the correct length. Odd there didn't seem to be any that are between the correct length and the obviously short length. The short ones appeared to be in random positions in the trays they are supplied in but it could be there are batches of short ones in other trays. I guess I'm going to have to measure them all before putting them in kits.

I also checked a smaller number of single row sockets, single and double row angled sockets, single row and double row header pins. I didn't find any that appeared to have tolerance issues.

On Saturday 13 January 2024 at 17:15:39 UTC Robert Porter wrote:

[Norberto Collado]

Hello Steven!  Yes,  the sockets are completly flat. Here are some pictures.

[Norberto Collado]

All the sockets have the same problem. One of them is pretty bad to insert any module.

[Mark Cohen]

I don't know if it is related but the slot giving me a problem had a compact flash in it and I think it is corrupted now. I am getting panic error messages and it continues regardless of what slot I use now. It was working originally. I ordered an sc compact flash module and I will see if it makes a difference. It may not be related.

To view this discussion on the web visit https://groups.google.com/d/msgid/retro-comp/da3ef7ca-54a5-40cc-bb8a-d6cf6145cd52n%40googlegroups.com.

[Steve Cousins]

I've damaged a CF card by plugging it in when the system was powered up. Or at least that was when I first had a problem with it. I doubt the connector caused the problem.

[Steve Cousins]

If the pins visibly don't align with the all the sockets on your backplane then I've got a bigger problem than I thought. My measurements of about 100 sockets only found a few with the issue.  Your kit suggests there are clusters of such sockets in some trays of connectors. Sometimes offering kits is very unattractive.

I can send you a replacement kit if you wish.

I've experimented with one socket here that has similar alignment issues to the one in your photo. I found by partially inserting the pins and one end and pushing the module towards the misaligned socket holes the module would go in quite easily but it got tight as the module was fully inserted. It turns out I had one such socket on one of my recent backplanes. 

Steve

[Norberto Collado]

I'm planning to order a second backplane to figure out a solution if possible. No need to replace such Kit to avoid unnecessary charges at your end as you are far away from me. :) Thanks you for the offer.   Also, my own modules started to fail to work and the monitor was crashing all the time. It took a while to debug, but what I found is that the backplane /Reset signal was pulsing at a low voltage between .5 to 1V. I wonder if the Supervisor IC is bad. I installed such device based on the picture at your site. I added a 1K pull-up for the time being and everything is working fine. I ordered additional voltage supervisor IC's to replace current one unless it is backwards.

[Steve Cousins]

Regarding the reset issue, the only things I can think of to check are: 

1/ Is the voltage at the voltage supervisor above 4.75 volts?

2/ Is the capacitor C14 (1nF) correct? 

3/ Do you have anything unusual attached to the reset signal, such as another voltage supervisor?

I have seen systems repeatedly going into reset where the power supply is not quite able to support the load and the voltage drops below the voltage supervisor threshold. When measured off load the voltage might look fine but under full load it drops below the threshold. I've even seen systems repeatedly resetting when running RomWBW where they reset due to the extra load of all the diagnostic LEDs coming on during boot up.

The voltage supervisor has a very distinctive shape and should be fitted to match the shape outlined by the silkscreen.

A good way to check the functionality of the reset signal and the voltage supervisor is to isolate the supervisor from the reset signal with jumper JP23.

[Norberto Collado]

The input voltage to the backplane is 5.15 Volts with current load. The +5V AC adapter is a 3A supply. Yes, C14 is correct. I do not have any other Supervisor as I rely on the backplane circuitry. 
The Supervisor shape matches the silkscreen. If I take out the jumper JP23, the /RESET voltage goes to +5V. Either the supervisor or C14 is pulling down the voltage. For now a 1K pullup with JP23 installed fixed teh issue. Once I get the parts I ordered, I will provide feedback on which comaponent is causing the issue.  

[Norberto Collado]

Steve, also I did order another backplane last night, but I was not able to find the new spacer that you have on the Flock board to add to the order. I need about 10-pcs for my new boards. Send me the link to order them. 

[Norberto Collado]

Steve, my bad!  I took out all the boards out of the backplane and the /RESET line on the backplane worked fine. I found the module that was causing the isue. I had a 1K resistor driving the base of a transitor and was causing the /RESET line to be less than a 1V.  I changed the 1K to a 10K and that fixed my last issue and thanks for the fedback as number 3, kept me thinking.

"3/ Do you have anything unusual attached to the reset signal"

[Steve Cousins]

No problem. Always good to find a solid reason for a problem.

[Steve Cousins]

Re: "I was not able to find the new spacer that you have on the Flock board to add to the order"

The flock board is not one of my products. See here:

https://www.tindie.com/products/weird/flock-v2-rcbus-module-kit/

My SC700 series modules are supplied with simple nylon, hex shaped, spacers. These are the same spacers and screws that are supplied as legs or mountings for SC700 series backplanes. I can throw in some spacers and screws with your backplane order. They are not as nice as the one you asked about but they do keep the modules safely spaced.

Steve

[Mark Cohen]

I have been using these. 

https://www.amazon.com/dp/B0BXT71WJK?psc=1&ref=ppx_yo2ov_dt_b_product_details

the 10mm with nylon screw in back and sometimes in front depending.

To view this discussion on the web visit https://groups.google.com/d/msgid/retro-comp/d59650ea-5f64-419a-a58e-c72f80cf3bb4n%40googlegroups.com.

[Norberto Collado]

Steve, 

Thank you for building out the SC709 12-slot backplane as it allowed me to eventually developed an idea I had for a long time. I created a pdf file listing all the modules required to developed such idea. Now I have a working prototype and refer to attached file for more infrmation.

Thank you very much!

Norby

[Steve Cousins]

Hi Norby

Wow, that's a huge set of modules. Are you going to publish the full design details?

Steve

[Mark T]

Hi Norby,

Thats a very interesting collection. I’d like to see more details on each module, do you have a webpage or github?

Mark

[Norberto Collado]

Steven, how I publish the full design details? Is there a procedure? Yes a lot of modules. The core of the system is just 5 modules and the rest are just optional I/O boards. The only 3 missing are; Floppy Soft Sector Module, Floppy Hard sector Module, and SASI Hard Drive Module.

Mark, I'm planning to add such information to the website as I'm not familiar with github. Link:    H8/H89 PCB's (koyado.com)

[Steve Cousins]

Hi Norby,

You could post PDF files of the schematics here. The schematics will answer many of our questions.

If you share the Gerber files then others can get PCBs manufactured that are identical to your PCBs and can thus duplicate your journey.

Depending on the software you used to design the PCBs you could post or share the PCB design files allowing others to create modified versions of your PCBs.

If you are happy to share then all of the above would be great. I publish my files through my website, together with links to my projects on the online design system I use (EasyEDA). Some people publish using GitHub. Others use Hackaday.

Steve

[Norberto Collado]

Once I'm done testing and cleanning up any mistakes, i will post the files. I'm using KiCad 7.010.

Norby 

[J.B. Langston]

I will take a little credit for bringing Norby into our little community ;).  I know him from the SEBHC mailing list. A little background: the Heathkit H89 was the first computer I ever used around age 7, and is part of the reason for my fondness for the Z80.  Unfortunately, my dad sold it long before I became interested in retro computers, but I have always wanted to have one. I finally found one this summer and bought it, but unfortunately it was not a happy reunion. I tried valiantly to get it running again, but the computer is quite a basket case, especially the video board.  Norby and the other guys on the list were a great help in offering advice. But in the end, working on high voltage stuff stresses me out and trying to fix the computer was bringing me more frustration than joy, so for the time being I have put it in my closet and focused my updating z80ctrl project.  I was telling Norby that I designed some boards for the RCBus ecosystem and he was interested in doing his own boards to recreate a Heathkit-like system on the RCBus, and now he's done it! Congrats Norby!

[Norberto Collado]

Thanks J.B. for all your suport and congratulations on your  z80ctrl project!  :)

[J.B. Langston]

Well, let's save the congrats until I actually have working boards, lol.  According to JLCPCB my parts should arrive Feb 6, so I have until then to finish tweaking my designs and test out the breadboard prototype.  At this point I really need to close KiCad and do some software. That is my goal this week, now that the boards are pretty much finished.

You received this message because you are subscribed to a topic in the Google Groups "retro-comp" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/retro-comp/vz1Qpshl8zY/unsubscribe.
To unsubscribe from this group and all its topics, send an email to retro-comp+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/retro-comp/96eaf56e-99d1-4a15-80c3-59c0eee52509n%40googlegroups.com.

[Tadeusz Pycio]

I have made my Backplane 4 available on GitHub

[J.B. Langston]

O hai, look what just arrived. 

On Thu, Jan 25, 2024 at 7:39 AM Tadeusz Pycio <ta...@wp.pl> wrote:

I have made my Backplane 4 available on GitHub

--

You received this message because you are subscribed to a topic in the Google Groups "retro-comp" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/retro-comp/vz1Qpshl8zY/unsubscribe.
To unsubscribe from this group and all its topics, send an email to retro-comp+...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/retro-comp/c949be3e-b927-41b4-a6ea-a19ed9910511n%40googlegroups.com.

[Ronny Ribeiro]

I know that handwriting in those plastic bags 😂

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/CAMwROAxKe8Q9C2cAeK6MfeazvqW5UCKwsYq4WVLq-THTcNCawQ%40mail.gmail.com.

[J.B. Langston]

The SC701, SC723, and both SC724s have all been assembled.  So far I have only tested the SC723 with my current z80ctrl board and z80/ram/rtc boards.  I get a bus timeout with the z80ctrl, which I believe is caused by the DS1233 on the new backplane, since it doesn't happen when that is disabled.  I could probably change the z80ctrl to have a longer reset timeout to make it work with the reset supervisor.  I also have an intermittent issue where the SD card doesn't properly initialize on the z80ctrl when the system is turned on. I believe this may be down to the 100 uf cap on the SC723 since the SC112 I was previously using didn't have any cap on board, though I had added a 10uf across the screw terminals.  I may try adding a bleeder resistor across the screw terminals of the SC723 and see if this solves the issue.  I haven't tested the SC701 yet, but I am expecting it will behave similiarly.

[J.B. Langston]

A bit more experimentation later, I think I have solved the mystery. When only the z80ctrl and CPU are installed in any of the three buses (SC112, SC701, and SC723), the SD card has trouble initializing on every other power on.  I believe the serial adapter is parasitically powering the z80ctrl enough that the SD card never fully shuts down. If there is enough load on the bus, then this does not occur. I found that installing any board containing an NMOS chip (TMS9918, SN76489, or YM2149) loads the bus enough to prevent the problem, so this is why I hadn't noticed it before. Likewise, installing a 1.5K resistor between ground and VCC prevents the problem. Installing a resistor in the screw terminals does not fix the problem because the screw terminals are connected to the power supply before the switch; it has to be put across the power pins on the bus.  So I don't think there's any issue with either of the new backplanes, just an issue with the z80ctrl that I hadn't noticed before. On the current revision of z80ctrl, I had installed a 1K resistor in the TX line to prevent this from happening but apparently that's not sufficient if the bus only lightly loaded. I may need to use a resistor > 1K on the TX line on my new z80ctrl revision to prevent parasitic powering.

[Norberto Collado]

On my first backplane order, I received this red board. How I used it?

[Steve Cousins]

Hi Norby,

That board (SC713) is just a freebie a I put in with some orders. Think of it as a quick reference card for the RCBus pin-out. 

Documentation:

https://smallcomputercentral.com/sc713-rcbus-tool/

Steve

[Harry Speer]

Re: TI TL7705A - versus -  Analog Devices / Maxim Integrated DS1233-5+

Wednesday, April 17th, 2024 @ 10:46 PM 

An open question to those of you who have a Lot of experience in design and construction of boards,  including,  but not limited to,  Steve Cousins  and  Tadeusz Pycio  but also to any of you who have done this a Lot. 

Hello to  Steve Cousins,  Tadeusz Pycio,  and others,   I am beginning a project,  a mc68k SBC,  part of which uses a Power Supervisor Chip for Power-On Reset.   I cannot decide between the  1. TI TL7705A  - versus the -  2.  Analog Devices / Maxim Integrated DS1233-5+  ---  and I do not have experience with either of them.   Can any of you give me some feedback on  Which you would prefer to use and Why ??   Or a recommendation of anything related ??   I would very much like to use one of them on this prototype board,   and,  in addition,  I plan to put in a debounced RESET push-button momentary switch.    I'm using an mc68k Family CPU Chip  so these /RESET signals will be buffered by Open Collector Drivers with Pull-Ups.   /RESET and /HALT being both Inputs to the 68k CPU  but them also being outputs from the CPU some of the time.   I've seen the debounced push-button momentary  feeding into a 555 timer as a one-shot which looks to me like a very good idea also.   But I lack the experience that might help me to decide between these  2  "Power Managers" options.   As I will be starting with Wire-Wrap,  I am currently leaning towards the  TI TL7705A  chip as it mounts into an 8-Pins DIP package,  well-suited to Wire-Wrap Sockets.  

Please excuse the Very Large Bolded font but my eyesight is very poor.   

So Thanks in Advance,  

Harry S. Speer 

On Friday, March 17, 2023 at 6:06:20 AM UTC-7 Tadeusz Pycio wrote:

I am in the process of designing an active backplane (5 slot + expansion), with MP1584 step-down converter, TL7705 supervisor chip and CH340E USB-Serial converter and mini USB connector. A project for my own needs, so I used SMD components. Also I will use a barrel connector 2.1/5.5 for the power supply and I intend to use it in a different colour (yellow, red?) to distinguish it from the 5V system.

[Tadeusz Pycio]

Hi Harry,

I am in favour of separating the RESET button from the RESET line. My choice of TL7705 is related to its availability and price, but it requires external components to operate. My favourites in DIP8 chassis are the MAX1232 or the MAX705-708/813 family, which do not require external components for the RESET signal to operate and have optional additional functions (watchdog, supply voltage monitoring), but they were hard to come by when I was designing my backplane (I don't know how it is now with their availability).

@All: The backplane project I mentioned is temporarily on hold due to the unavailability of the CH340E or CH340X, which met my expectations. (I wanted to outsmart the system by sourcing a Chinese product, which I don't think they counterfeit anymore, instead of buying Chinese FTDI counterfeits ;-) ).

[Steve Cousins]

Hi Harry,

I've not used the TL7705A so I'm not well placed to make a comparison.

The DS1233 is readily available from the big distributors and does the job of generating a reset signal and supervising the supply voltage. It will debounce a reset button for you but the resulting rise time is rather slow. On some of my dedicated power modules I use a schmitt triggered buffer to improve the rise time and an open collector stage connecting the cleaned up reset to the bus. There is definitely room for improvement but I find it adequate.

As a kit supplier I don't want to stock more different components then I have to so I've stuck with the DS1233. Also, it is important to me that the component is readily available from more than one major distributor.

Steve

[Bill Shen]

Harry,

I use MCP130 for all of my projects.  It’s only advantage is cheap, about 1/5th of DS1233.

What 68k project are you working on?

  Bill

[Harry S. Speer]

   

   8:37 AM, Thursday, April 18th, 2024

   Re: [retro-comp] New 68008 SBC for the S-100 / S100 Bus 

    

   Bill -

   I am beginning a new SBC for the S-100 / S100 Bus.   We've had the

   JAIR 8080 SBC from Canadian Josh Bensadon, (sp?)  See his board here:

   http://www.s100computers.com/My%20System%20Pages/8080%20CPU%20Board/8080%20CPU%20Board.htm

    and we've also had the Z80 SBC from John Monahan (sp?) here:

   http://www.s100computers.com/My%20System%20Pages/SBC%20Z80%20Board/SBC%20Z80%20CPU%20Board.htm

   And John also recently added the Z180 SBC S100 Board based on a design by Steve Cousins here:

   http://www.s100computers.com/My%20System%20Pages/Z180%20SBC/Z180%20SBC.htm 

   

   And these are all great options for the whole 8080 / Z80 enthusiasts,  but I've long been

   interested in owning a 68000-based computer.   In College, about 1981 to 1983,  (about 40 

   years ago when I  was in my early 20's),  I was lucky enough to be able to  work on a 

   68000-based SAGE IV Computer at my T.E.S.C. College in  Olympia,  Washington, USA.   

   The SAGE IV was in their College Computer lab.   The  SAGE IV info from wikipedia: 

   https://en.wikipedia.org/wiki/Sage_Computer_Technology    

   and from   http://www.oldcomputers.net/sage-iv.html   

  

   The SAGE IV was one of 2 computers in the lab running the UCSD Pascal P-System.  The  

   second computer was an Apple II.   I was able to setup a 2-way communication between the 2 

   computers over an RS-232 cable.  This allowed the transfer of Pascal Program Listings 

   between the two machines.  The 6502-based Apple II needed to link with some Assembly 

   Language Code whereas the 68000 didn't even blink.  Zoom  -->  Done !   The 68000 SAGE 

   was like one of my best dreams come true !!   :-)   I've Always wanted one of my own ever

   since then.   :-)  

   

   The biggest reason for using the 8-bits data bus of the 68008 for an S100 SBC is half the RAM 

   chips and half the EPROMS.  And the 68008 CPU itself is also a physically smaller DIP chip than 

   the 16-bits HUGE 68000 DIP.  Looks like an Aircraft Carrier !!    My hope is to gain enough empty 

   board space on the S100 prototype card to allow squeezing in either a 68B50 UART (24 pins)  or  

   a  (40-pins)  DUART  like the MC68681 chip or it's equivalent chip from other companies.  Use this

   for Console IO aka Console:  and a Second UART could be used for other things like  an

   Uplink / Downlink  to / from a  "Host"  PC.  Transfer programs and data.

  

   The 68000 S100 CPU card we have right now just does  NOT  have enough room for a UART. 

   I'm hoping that the 68008 - with a physically smaller CPU and half as many Memory chips - will

   leave me enough room on the S100 prototype board to install either -  1.  A single 24-pins DIP 68B50 UART 

   or   2.  A larger 40-pins DIP DUART like the MC68681 (or equivalent)  40-pins DIP.  Kind of Shoe-Horn it in there.

Clearly there will also need to  be room for an oscillator for the baud rate clock as well.  For RAM, use the

Alliance Memory AS6C4008-55PCN Static RAM  512k x 8-bit. And an EPROM or EEPROM as needed up to

512k x 8-bit. Or Flash. To gain more address lines, I am considering mounting the 52-Pins PLCC 68008 CPU

into an adapter to provide 52-pins DIP to use in a WireWrap socket. Getting 4 more address lines will help

this card to be more useful for all the other cards in the S100 system. And provide more memory-mapped IO.   

   And Lastly, credit to whom it is due,  Australian Damian Wildie  has been VERY helpful to me so far

   and I am Truly Deeply Grateful.  :-)    Not everyone is willing to converse with a "newbie" like myself.  

  

   Harry S. Speer

   cc: self

   MCP130 - or - DS1233 - or - TI TL7705A

============================================  

On 4/18/24 04:36, Bill Shen wrote:

Harry,

I use MCP130 for all of my projects.  It’s only advantage is cheap, about 1/5th of DS1233.

What 68k project are you working on?

  Bill

==============================================

On Thursday, April 18, 2024 at 12:15:46 AM UTC-6 h.s....@gmail.com wrote:

Re: TI TL7705A - versus -  Analog Devices / Maxim Integrated DS1233-5+

Wednesday, April 17th, 2024 @ 10:46 PM 

An open question to those of you who have a Lot of experience in design and construction of boards,  including, 

but not limited to,  Steve Cousins  and  Tadeusz Pycio  but also to any of you who have done this a Lot. 

Hello to  Steve Cousins,  Tadeusz Pycio,  and others,   I am beginning a project,  a mc68k SBC,  part of which uses

a Power Supervisor Chip for Power-On Reset.   I cannot decide between the  1. TI TL7705A  - versus the -  2.  Analog

Devices / Maxim Integrated DS1233-5+  ---  and I do not have experience with either of them.   Can any of you give

me some feedback on  Which you would prefer to use and Why ??   Or a recommendation of anything related ??  

I would very much like to use one of them on this prototype board,   and,  in addition,  I plan to put in a debounced

RESET push-button momentary switch.    I'm using an mc68k Family CPU Chip  so these /RESET signals will be buffered

by Open Collector Drivers with Pull-Ups.   /RESET and /HALT being both Inputs to the 68k CPU  but them also being

outputs from the CPU some of the time.   I've seen the debounced push-button momentary  feeding into a 555 timer

as a one-shot which looks to me like a very good idea also.   But I lack the experience that might help me to decide

between these  2  "Power Managers" options.   As I will be starting with Wire-Wrap,  I am currently leaning towards

the  TI TL7705A  chip as it mounts into an 8-Pins DIP package,  well-suited to Wire-Wrap Sockets.  

Please excuse the Very Large Bolded font but my eyesight is very poor.   

So Thanks in Advance,  

Harry S. Speer

=============================================

On Friday, March 17, 2023 at 6:06:20 AM UTC-7 Tadeusz Pycio wrote:

I am in the process of designing an active backplane (5 slot + expansion), with MP1584 step-down converter, TL7705

supervisor chip and CH340E USB-Serial converter and mini USB connector. A project for my own needs, so I used SMD

components. Also I will use a barrel connector 2.1/5.5 for the power supply and I intend to use it in a different colour

(yellow, red?) to distinguish it from the 5V system.

=============================================

--
You received this message because you are subscribed to a topic in the Google Groups "retro-comp" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/retro-comp/vz1Qpshl8zY/unsubscribe.
To unsubscribe from this group and all its topics, send an email to retro-comp+...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/retro-comp/41539cbe-cc03-4e1b-ab9f-938183fe15dfn%40googlegroups.com.

[Greg Holdren]

Harry,

Like Bill, I have used the MCP130 in the past.

The Rosco_m68k uses a 555.

https://github.com/rosco-m68k/rosco_m68k

Greg

[Shawn Reed]

Harry,

Why stress about board realestate, adopt the RCBus. ;) I also would love a new 68k option so imagine there must be interest out there.

-Shawn

[Bill McMullen]

Although I haven't used them, one thing I spotted in the TL7705A datasheet is a slow rise time (max 3.5 uS) which may require the use of a Schmitt buffer in addition to the other external components.

My preference is for a simple fixed voltage supervisor that can directly handle a reset switch input.  I quit using the DS1233 due to increasing prices.  While I have used the MCP130, I primarily use the STM1813 (5V) or STM1818 (3.3V) due to it's shorter reset pulse (200 mS max vs 700 mS) and fast rise/fall times (<= 5 nS).  The negative of the STM1813 is that it's only available in a SOT23-3 surface mount package and it's prices have also increased.

[Dylan Hall]

I've recently used an ADM707 in a project that needs both RESET and /RESET signals.

I started with the DS1233-5+ as they've worked really well for me in the past. I ran the /RESET signal into a CPLD and produced RESET from it.

However, because of the slow rise time and the CPLD not having schmitt trigger inputs the RESET signal oscillated horribly. 

I chose the ADM707 because it drives its outputs (e.g. it's not open collector). The only downside with the ADM707 is the drivers aren't very strong, e.g.

Given I like an LED on the reset signal I added an 2N7000 for the LED, e.g.

Thanks,

Dylan

--

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.

[Harry S. Speer]

    

    17:12,  5:12 PM,  Thursday,  April 18th, 2024

    Re: SBC Sources of RESET - Momentary Push-Button -and- Power-Supervisor Power-On /RESET 

   

    Tadeusz -

    Regarding 2 separate sources of /RESET - Momentary Push-Button -and- Power-Supervisor 

    chip Power-On /RESET - in this case, the outputs of each circuit is intended to be an Open-Collector

    driver like 7407, so it occurred to me each output could be there and the pair of them driving the

    same pins on the CPU.  Wire OR'ed ??   Not sure about that name but either one or both of them 

    together could be triggering the same pair of pins on the CPU - /RESET and /HALT - the key

    point here being that I want a nice clean fresh start (so to speak) when the CPU Card powers up,

    and does so automatically, AND I also want a good old "Big Red Switch" to "Kick it in the Head" 

    when it gets lost due to "wandering away".  ;-)   Perhaps this is excessive or overkill or whatever 

    but I really do want a nice clean RESET for the 68000 or 68008 CPU Card.  The S100 SMB board

    from S100Computers.Com here: 

            

    http://www.s100computers.com/My%20System%20Pages/SMB%20Board/S100%20Bus%20SMB.htm

           
    By stuffing the S100 SMB Board into the Backplane along with his 68000 CPU Card, Damian Wildie was able 

   to remove the /RESET mod he had had to make for his 68000 CPU Card.  The Schematic for the SMB board shows 

   the U29 LM555N on the SMB board and I would prefer it to be on the CPU  card itself.   My whole goal with the 

   SBC is the same as John Monahan addressed on his website S100 website - Make this a "Beginner's Board"

   Or call it "The First Card for Newbies".   ;-)    It Contains CPU, RAM, EPROM, and Serial IO for  " Console: " and,   

    if there is still any "free" available board space after that, maybe add a CF Card or an SD Card as a "pretend" 

    A: Boot Drive for CPM-68k.  Right now, we have the 68000 CPU card, which requires a separate Serial IO Card,

    and we have the 68030 Card, which likewise is dependent on resources only available from other cards elsewhere 

    on the Bus.  The goal is for the Newbie to only have to debug ONE Card and once She / He gets it going right,

   THEN they can use that card as a stable base to work from.  We already had this for the 8080 and Z80 and Z180 

    But this is the "Beginners' Choice" that didn't exist.  So naturally I want to build one.  However, in parallel with the

   SBC,  I also still want to work on bringing up the 68000 CPU Card we have now,  to make sure that I have "Problem

   Solved" by providing the same kind of "stable reset" that the SMB provides.  And I have NO Source for the

   /RESET on my Motherboard / Backplane right now.   I need something to connect that front-panel RESET switch

   to.   Don't get me wrong - the SMB Board is a great board which provides a TON of great resources, but, For Right Now,

   I just want and need a nice solid stable source for the S100 /RESET for the Backplane.  The theory is that the 68000

   CPU card and the 68030 CPU Card will "pick this up" off of the /RESET on the Backplane.  Is that a Correct Assumption ???

  

   It Seems like there's an echo in here.   ;-)   Please forgive me repeating things over and over again.  It's a "birth

   defect". 

     

   My Very Deepest Thanks to all of you who responded with info about your various choices for Reset, for Power

   Supervisors, etc.  That Gives me a LOT to work with.  I am VERY Grateful !!!  I think I am still leaning towards a 

   DIP Package chip however that may change when I get a prototype circuit to display on an O'scope.  :-)  

   

   Thanks Again,

   Harry

            

   P.S.  Tadeusz - Could you please expand on what you meant by  " I am in favour of separating the 

                             RESET button from the RESET line " ???   I want to make sure that I understand your meaning.

  

  cc: Retro-Comp Google Group and S100 Google Group  and self

  

===================================================================

--
You received this message because you are subscribed to a topic in the Google Groups "retro-comp" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/retro-comp/vz1Qpshl8zY/unsubscribe.
To unsubscribe from this group and all its topics, send an email to retro-comp+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/retro-comp/e1317b71-77fd-425a-8d87-6be2c07551b8n%40googlegroups.com.

[Tom Storey]

The TL7705 will debounce a button connected to its RESIN input, and the DS1233 will debounce a button connected to its output also, so there is no need for extra circuitry like a 555 in either case.

As others have mentioned, the TL7705 requires additional components (a couple of resistors and capacitors) which means more board space, but has the advantage that you can configure how long you want the reset pulse to be using an external resistor and capacitor if it matters, whereas the DS1233 is fixed by an internal timer.

The TL7705 also has reset outputs of both polarities which could be useful, although in my experience I determined that it is better to use one polarity and generate the other polarity yourself as I noticed a significant timing difference between the two as generated by the TL7705 itself.

The DS1233 has the advantage of being very small in comparison.

For supervisors that dont have an open collector/drain output themselves and cant monitor and debounce a button like the DS1233, you can connect a button on the VCC input of the supervisor after a resistor to pull its power supply to ground instead. This should achieve much the same outcome.

Also remember that to initiate a hard reset on a 68k the RESET and HALT signals must both be pulled low at the same time. This needs two separate open collector/drain drivers (or perhaps transistors) connected together on their inputs, but driving two separate nets on their outputs - one for each signal. This then allows the CPU to generate the RESET signal independently of the supervisor, and that can form the reset signal to all other peripherals to allow for hard and soft resets.

--

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/28fe5144-719b-4d5f-8faf-02ff64201502n%40googlegroups.com.

[Tadeusz Pycio]

Hi Harry,

When I talked about separating the RESET button from the RESET line, I meant that if I was adding a reset supervisor to the circuit, it would be a good idea for it to also take care of such a button so that I wouldn't have to think about a circuit that would suppress contact vibrations and be compatible with the voltage supervisor circuit. Another point, the TL7705 was a choice out of necessity and was due to the lack of availability of other chips. My intention was to use the MAX707/708, but at the time there was a problem with its availability. It is worthwhile for you to take an interest in this solution. My choice was based on the fact that I have too many RCBus computers to use chips that are only available in single pieces.