HA-8-3 "2000" Board

[Terry Smedley]

I just finished building Norberto's latest board for the H8 - a sequel to the NOGDS HA-8-3 color graphics card with a boatload of updates to the original.  This is "retrographics" - it recreates faithfully what was possible in 1980.  If you're looking  for hi-def graphics on the H8 ..... this isn't it.  The board passes all the original NOGDS diagnostics for the video, sound, and joystick controllers.

Features that were added or updated:

- The VDP was upgraded from the TMS9918 (composite video) to the software-compatible TMS9928 (YPbPr component video) with MUCH improved video quality

- Dynamic VRAM was replaced with a single static RAM chip

- The AY-3-8910 audio output was split into stereo channels so sound effects can move spatially

- The fragile digital outputs of the AY-3-8910 (just one TTL load permitted) are now buffered so they can directly drive loads like LEDs with no external components

- The address space used by the board is "one jumper" selectable between the original HA-8-3 addresses (270Q-277Q) and an alternate set that does not interfere with the ports used in a three drive system (where the H67 occupies 274Q)

- Independent clock sources are provided for the video, sound, and optional AMD9511 APU so none of these are constrained by bus or CPU speed

- An I/O wait state generator permits many legacy cards to run at full 16MHz CPU speed

Norberto did a fantastic job on the board layout.  For a board of this complexity to be electrically perfect out of the gate is nothing short of miraculous.

A few of these boards are available now, and a few more are in the queue for production after COVID restrictions were lifted at the factory this week.

I've been experimenting with joystick options to connect to the HA-8-3 - this tiny one from Sparkfun uses XBee RF modules to connect wirelessly with the H8.  The PWM outputs of the XBee allow the receiver to connect to the HA-8-3 joystick inputs just like a wired controller - no software changes needed.

Terry

[Tom Wilson]

May I have one Terry?
Tom

--
You received this message because you are subscribed to the Google Groups "SEBHC" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sebhc+un...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/sebhc/5619797e-967d-4f47-a132-64a59bb2da3bn%40googlegroups.com.

-- 
Tom Wilson <><
Prineville Oregon

[Mark Garlanger]

I would be interested in one too.

To view this discussion on the web visit https://groups.google.com/d/msgid/sebhc/762efc4f-2453-50df-87ba-35e520deed8d%40gmail.com.

[Richard Davis Jr.]

Terry,

I am definitely intrested in getting a board. Send details on how to obtain one.

Rick Davis

[Norby]

Thanks Terry for all your help in testing such board. 👍

Sent from my iPhone

On Mar 29, 2022, at 1:06 PM, Richard Davis Jr. <rickdav...@gmail.com> wrote:



To view this discussion on the web visit https://groups.google.com/d/msgid/sebhc/CABvWWgZVmKP-1MVijePbVmukEeNRJ9L4%2BoL07g7_tCRsSbmpkw%40mail.gmail.com.

[Terry Smedley]

In the interest of full disclosure.....here is what is necessary to display the component video from the HA-8-3 card.

The TMS9928 outputs component video in the YPbPr format.  It is not RGB.  It requires either a component video display or a scan converter to convert the component signal to HDMI.  While it is possible to connect the TMS9928 directly to a modern component video display, the results are usually much less than spectacular.  Most component video displays that I have tried have either a) displayed the video with annoying "jail bar" vertical banding, or b) refused to display the signal at all.  The issue is that while the TMS9928 is outputting good component signals, the image size/resolution is 256x192:  something that modern displays don't understand well.  At least I haven't found a display that will display that signal well.  A Samsung TV displays it, but it has unacceptable vertical banding.  An Epson projector just gives up and says "no input".  

There's a whole community of "retrogamers" out there who have worked through these display issues with vintage gaming consoles (some of which use the TMS9928).  Their solution is a scan converter that takes the TMS9928 video and scales it to 480 or 720 lines on HDMI - something a modern display can deal with.  The most popular scan converter in the retrogaming community seems to be the "OSSC" - Open Source Scan Converter.  Glenn reports that he frequently sees the OSSC being used at vintage computer exhibitions.  This is open source hardware and firmware, and you can readily buy an OSSC in many places.  I purchased this one from Amazon:: https://www.amazon.com/dp/B083FG6PXH   RetroTINK makes similar scan converters, but with fewer tweaking options: https://www.retrotink.com/  The OSSC does a great job of converting the video.  It also encodes the stereo audio from the HA-8-3 onto the HDMI signal.  My projector and TV see a 1280x720 signal that looks quite good.  The AY-3-8910 sound effects sound great on a home theatre system.

The rub is that the OSSC is $140, the RetroTINK 2x-Pro is $130.

It is possible to use the composite video TMS9918 chip on this new HA-8-3 card.  Quick jumper settings select either the 9928 or 9918.  But composite video is, well, horrible!   Glenn has noted how the original HA-8-3 card, using the TMS9918 composite chip, outputs a video signal that's really not in the same league as what the retrogamers are exhibiting with their OSSC connected displays.

Other suggestions for displaying the TMS9928 component video signal are welcomed!

I just wanted potential builders to see the whole picture of using retrographics in a UHDTV world!

Terry

[norberto...@koyado.com]

Great information Terry! You just convinced me to buy from Amazon the OSSC scan converter.

 

Also, can you provide some pictures/instructions on how to build out the A/V cable hardness to interface properly to the OSSC scan?

 

Thanks,

Norberto

To view this discussion on the web visit https://groups.google.com/d/msgid/sebhc/2d7d6267-619c-40c9-9538-e3ec2d5a469cn%40googlegroups.com.

[Joseph Travis]

I would also like to add my name to the list for this board. Surpisingly, I kept the NEC Color Video Monitor I used with the H8 I had with the original boards. Right now it's on top of an Apple //e.

Thanks!

Joe Travis n6ypc

To view this discussion on the web visit https://groups.google.com/d/msgid/sebhc/003101d843fa%2463f3d9c0%242bdb8d40%24%40koyado.com.

[Royce Taft]

Hi all,

New here, I’ve only been lurking. 

I thought I’d chime in here to say if you’re interested in RGB output (15 KHz) from the TMS9928, there is an adapter board you can solder to the underside of the IC socket. 

https://tms-rgb.com/

I’ve used it on my TI 99/4A, and it works great. I connect it to an old RGB CRT monitor, but for those of you who prefer modern displays there’s really no reason to convert to RGB. The OSSC or retro rink should do a fine job. 

-Royce

On Mar 29, 2022, at 22:21, Terry Smedley <terry....@gmail.com> wrote:

In the interest of full disclosure.....here is what is necessary to display the component video from the HA-8-3 card.

To view this discussion on the web visit https://groups.google.com/d/msgid/sebhc/2d7d6267-619c-40c9-9538-e3ec2d5a469cn%40googlegroups.com.

[Terry Smedley]

Norberto:

Here is a document (also eMailed to you for posting to the website) that shows a couple ways to make the Audio/Video connections from the board to the OSSC.

Terry

[Terry Smedley]

You might be wondering why we didn't just use RCA connectors or a 1/8" minijack on the board itself to spare some of the cable grief.  We (that's the "Royal We", and it really means "Norberto") tried very hard to make jacks fit onto the board, but there just wasn't enough room.  The jacks themselves are reasonably thrifty of space, but you end up needing to leave a huge open space at the edge of the board to accommodate the length of the connectors and a reasonable cable bend to clear the H8 chassis lid on the way out.  I kept sending Norberto different jacks, and he kept showing me that there just wasn't enough room for them.  So we fell back to the good old pin header, much like NOGDS used on the original HA-8-3 board.

Terry

On Tuesday, March 29, 2022 at 10:52:53 PM UTC-7 Norby wrote:

[norberto.collado koyado.com]

Thanks, Terry, for the files! I think we should wait for the China pcb boards to arrive to do a final checkout. Then Todd's can take over this board for production.

On the different RCA jacks that you provided at least one of them landed onto the H8-Z5-4 cassette controller. Thank you for such HW.

Norberto 🙂

---

From: se...@googlegroups.com <se...@googlegroups.com> on behalf of Terry Smedley <terry....@gmail.com>
Sent: Wednesday, March 30, 2022 12:46 PM

To view this discussion on the web visit https://groups.google.com/d/msgid/sebhc/f3c3b957-07cf-4337-bfcf-2383bd58ee5an%40googlegroups.com.

[Norberto Collado]

I updated the website with latest files from Terry S. Link: http://koyado.com/heathkit/New-H8-Website/ha-8-3.html Thank you, Terry!

 

Also, during testing, Terry S. found that “ANY” device that adds wait states on the H8 bus will cause a “double” write pulse on some of my boards.

 

The following boards will add wait states to the H8 triggering this issue:

• Z-H8-CPU Trionyx
• Heath HA-8-3-1 GRAPHICS CARD
• SSI BOARD

 

Boards affected:

• Z80 V3.x CPU board
• Z80 V4.x CPU board
• H17 Combo board
• H37/H67 Combo board
• VDIP1 on Z80 V3.x & V4.x and H17 combo board

 

The following document provides information on how to update such boards. Link: http://koyado.com/heathkit/New-H8-Website/download/reworks_for_h17_h67_h37_z80-v3_z80_v4_boards%20(3).pdf

 

On the positive side, once the reworks are in place, you can run your system smoothly at 16MHz if needed. The process will be to boot the H8 at 2MHZ, and once you are in the OS, switch the speed to 16MHz. The Z67-IDE+ will run fine at such speed. Also, the stock H8-4 Serial board will run fine at 16MHz with the stock 8280 UARTS.

 

This happened because I copied the write latch from the H89 Serial port for writes (my bad), instead of using the write latch from the H8-4 schematics. The rework follows the H8-4 circuit. Any new board carries the H8-4 write latch or similar circuit to avoid such issue.

 

Thanks,

Norberto

[Glenn Roberts]

Great info. Will the affected boards still work fine (after the rework) even without one of the wait state generating boards installed? (I.e. should it just be standard practice to just do the rework? Or if/when we do another production run just make the change?)

Sent from my iPad

On Mar 31, 2022, at 7:33 AM, Norberto Collado <norberto...@koyado.com> wrote:



--

You received this message because you are subscribed to the Google Groups "SEBHC" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sebhc+un...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/sebhc/EC202446-D904-4B18-9837-A0F00EB5D85F%40koyado.com.

[Terry Smedley]

Glenn:

The (very minor) rework allows the boards to work with or without wait states so it can be done as standard practice.  Later boards include this method of handling the latched write pulses.

Terry

You received this message because you are subscribed to a topic in the Google Groups "SEBHC" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/sebhc/XvDMrnbll7o/unsubscribe.
To unsubscribe from this group and all its topics, send an email to sebhc+un...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/sebhc/DA383A67-797B-42FC-8E4B-9932B51BCB1A%40gmail.com.

[Norberto Collado]

I will update the Gerber files in the background with a note, so that when Todd’s supply is depleted, he can order new boards with the changes. I will update only the Z80 V4.x and H17/H37/H67 combo boards.

 

Thanks,

Norberto

 

From: "se...@googlegroups.com" <se...@googlegroups.com> on behalf of Terry Smedley <te...@tsmedley.net>
Reply-To: "se...@googlegroups.com" <se...@googlegroups.com>
Date: Thursday, March 31, 2022 at 12:08 AM
To: "se...@googlegroups.com" <se...@googlegroups.com>
Subject: Re: [sebhc] HA-8-3 "2000" Board

 

Glenn:

To view this discussion on the web visit https://groups.google.com/d/msgid/sebhc/CALt%2BXpPpo026e6-5DrvO0XCPWUzWsF8hXnaybKu3y%2B53WoxRZg%40mail.gmail.com.

[norberto...@koyado.com]

I received an HA-8-3 “2000” board along with the hard to find devices from Terry S. Also, I want to thank Terry for his guidance and support as I went thru the assembly of such board. I just completed the assembly of the board per schematics and it worked great. I took the time to solder and inspect every component and ensure a successful power-on tests. The HA83DIAG application is a great tool to test the board functionality. I’m happy on how this board turn out, super great! It is beautiful! The only pending circuit to test out will be with the OSSC controller and joystick.

 

As Terry S. is planning to design a remote joystick, in the meantime I will like to get a vintage joystick from eBay. Any recommendations? I want to play eventually Les’s “Warlords” game.

 

HA-8-3-1 picture attached!

 

Thank you, Terry!

 

Norberto

 

From: se...@googlegroups.com <se...@googlegroups.com> On Behalf Of Terry Smedley

[Glenn Roberts]

Great! We had some discussion here about joysticks last fall while I was preparing an HA-8-3 demo for VCF (I ended up not doing any joystick in the demo). I think Les had something that was working OK.

 The NOGDS schematic has a specification for the potentiometer joystick (10K I think??) so unless the circuitry has changed I would start there.

Sent from my iPad

On Apr 8, 2022, at 6:41 AM, norberto...@koyado.com wrote:



--

You received this message because you are subscribed to the Google Groups "SEBHC" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sebhc+un...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/sebhc/022401d84b02%24c9086ce0%245b1946a0%24%40koyado.com.

[norberto...@koyado.com]

Thanks Glenn for the information. I found the discussion and to play with it, I will follow Terry suggestion on possible joystick.

 

I also used the thumbsticks from Adafruit with good results (but my human thumb control is not well developed): https://www.adafruit.com/product/512  

 

It s a low cost joystick just to play with it. I will have my son print a case for it along with the fire switch. Also I will add an LED as well. I just need something simple to play with the analog circuit and to play Les’s game.

 

Thanks,

Norberto

 

From: se...@googlegroups.com <se...@googlegroups.com> On Behalf Of Glenn Roberts
Sent: Thursday, April 7, 2022 10:21 PM
To: se...@googlegroups.com
Subject: Re: [sebhc] HA-8-3 "2000" Board

 

Great! We had some discussion here about joysticks last fall while I was preparing an HA-8-3 demo for VCF (I ended up not doing any joystick in the demo). I think Les had something that was working OK.

To view this discussion on the web visit https://groups.google.com/d/msgid/sebhc/BCCED489-DCBF-4E73-9C31-57F14EF9B808%40gmail.com.

[Joseph Travis]

Atari / Commodore style joysticks are readily available and inexpensive.

How do we go about ordering the board? Is this thru Norberto's website?

Thanks!

Joe Travis n6ypc

To view this discussion on the web visit https://groups.google.com/d/msgid/sebhc/023201d84b13%247f027940%247d076bc0%24%40koyado.com.

[Terry Smedley]

We are waiting .... waiting .... waiting for a final test set of boards from OurPCB in Shenzhen.  Their factory was closed for about two weeks for Covid lockdown, and they are having a difficult time getting caught up.  While doing the initial builds (with domestic boards), we found a couple of silkscreen labeling issues which were corrected on the OurPCB board order.   When the OurPCB boards are checked out, they will be released for production through Todd.  Anyone who wants one should have no difficulty in obtaining one.

This board is on Norberto's web site:  http://koyado.com/heathkit/New-H8-Website/ha-8-3.html

Terry

[Terry Smedley]

It's my understanding that Commodore joysticks were arcade-style switches rather than analog devices.  I think some other similar vintage game consoles (Coleco?) also used switch joysticks?  I have no firsthand knowledge of these devices - it's just what I glean from researching them on the web.  The switching devices are electrically compatible with the HA-8-3, but most of the software for that card that I have seen expects an analog input.   

Glenn and Les researched this a while ago (look for "Paddle Controller" in the forum discussions) and Les found something with joystick pots, but I believe they were very high resistance (1 Meg) and resulted in a pretty noisy signal.  The target resistance for the HA-8-3 would be 10K.

Another wrinkle is that the HA-8-3 uses a +10v analog reference voltage.  Depending upon how a controller is wired, that might be an issue.  For example, I can imagine a controller using a common 5v supply for switches, pots, and LEDs or other electronics.  While 5v is available on the HA-8-3 header, using that as the reference voltage for X-Y pots would probably require a software change since the analog inputs would become half-range.  Depending upon how such a controller was wired, using the +10v reference voltage to power it might work OK....or might end up damaging either the controller or digital inputs on the HA-8-3.

10K "thumbsticks" are available everywhere, but as Norberto pointed out, they can be very tough to control precisely.  The ones I experimented with (from Adafruit and Sparkfun) were really more like switches than analog inputs.  I found it difficult to modulate the thumbstick to an intermediate position - it wanted to go right to the limit when moved away from center.  I used larger 10K devices (with a stick more suitable for aging hands) and they were much easier to control.

Terry

On Friday, April 8, 2022 at 5:16:41 AM UTC-7 Joe Travis N6YPC wrote:

[Glenn Roberts]

I think as the boards become available (and built) we will have to experiment to find what works best. Will be good if people share their experiences here.

Sent from my iPad

On Apr 8, 2022, at 6:06 PM, Terry Smedley <terry....@gmail.com> wrote:

It's my understanding that Commodore joysticks were arcade-style switches rather than analog devices.  I think some other similar vintage game consoles (Coleco?) also used switch joysticks?  I have no firsthand knowledge of these devices - it's just what I glean from researching them on the web.  The switching devices are electrically compatible with the HA-8-3, but most of the software for that card that I have seen expects an analog input.   

To view this discussion on the web visit https://groups.google.com/d/msgid/sebhc/566a256e-5372-428a-931f-bd15ee13a933n%40googlegroups.com.

[Joseph Travis]

You are correct Terry. Commonly available analog joysticks were used on Apple and PC compatible computers although, the joysticks are not compatible.

The Apple used a capacitor in the joystick circuit to create an RC network and a software timing loop to determine the joystick position.

Atari and Commodore used analog for paddles, digital for joysticks.

I've written programs for digital and analog joysticks. Digital is easier as the analog usually requires some form of software filtering to help smooth the readings.

The SNES controller interface is pretty cool, using a 12 bit shift register. It's a 5 wire interface: +5V, Gnd, Clock, Latch and Data. I've written programs for this too.

I guess the question's are: What does the currently available software expect?  What will future software expect?

I use all of those mentioned, yet I'm not married to any one of them. I don't know if I actually provided any answers here or if I'm just rambling on. My blood sugar is off and I gotta go.

Regards,

Joe Travis n6ypc

To view this discussion on the web visit https://groups.google.com/d/msgid/sebhc/566a256e-5372-428a-931f-bd15ee13a933n%40googlegroups.com.