Authentic Reproduction ATR8000 Interest?

[+slx]

More drives here, but not cheap (250 USD).

[Gitbizy]

Count me in and I would love to help with this project as well.

[+Larry]

Thread appears to have gone cold.  Hope I'm wrong.

[leech]

This is sweet, always wanted one.

[Gitbizy]

I can certainly chip in on this and help with PCB design and procurement.  I need to know where the project is now.

[joeventura]

count me in, let me know what you need

[StartedWith400]

I created a mini ATR-8000 a couple of years ago. One problem I ran into was that the floppy disk controller chips are sometimes fakes. For a while there weren't many FD1797 available. I just looked and there are some in the US on eBay, but many seem to have the same labeling - a typical warning flag that it's a fake.

 

Also, the data separator needs a small ROM that used an old style chip (74S288 - 32x8). They require special programming voltages and timing. There are equivalent components but they all seem to need special programming. My various ROM programmers did not support them. So, that means I need to build another circuit just to program those ROMs.

 

I considered alternative FD controllers, but then I found out that the CP/M code would require changes to support an alternative FD controller.

 

Anyway, I have enough parts to make about 10, assuming all the FD controllers work.

 

Edited April 30, 2022 by StartedWith400
Some information was incorrect - fixed.

[+Larry]

Looks neat!  What functions does it support?  Quite a power supply section!  64k?

[venom4728a]

I would buy one too.

 

[StartedWith400]

3 hours ago, Larry said:

Looks neat!  What functions does it support?  Quite a power supply section!  64k?

It's an (almost) exact duplicate of the original circuit, just in miniature. Probably with all it's quirks too. It's essentially what the O.P. talked about doing.

 

I placed the components in the same relative locations as the original. The RAM chips, Z80, Z80-CTC, FD1797, and 74S288 are the original size parts. The rest are surface mount ICs. Since I couldn't make SIO jacks (and buying them is too expensive), I opted to use right angle DB-15 connectors. Then I will make cables that are SIO plug (that I 3D print) to DB-15. The picture shows the wires just direct soldered for testing. Also I chose to use DB-25 for the printer parallel port and serial port so that common, industry standard cables can be used. I also put a header for the floppy connection so that a industry standard floppy ribbon cable can be used. I also kept the card edge for compatibility with anyone who has older floppy cables and hardware.

 

The power supply is just an off-the-shelf AC to dual output DC (+/-12v and +5v). After I get it working, I could replace the true RS-232 with a TTL level port that could be used with modern FTDI USB to serial cables and then the -12V is not needed (unless you want to use 16K of RAM).

 

This has 64K of RAM on board. 16K x 1 chips are actually harder to find at a good price today than the 64K x 1 chips.

 

It should support all the functions of the original ATR8000 !

 

[StartedWith400]

Like many of you here, I always wanted to connect a hard drive to the Atari using the ATR 8000. Back in 1987, a college acquaintance found out that I had a 70MB full height 5.25 MFM hard drive and SCSI/SASI controller. It was something that I had been testing at a co-op job and it was pretty state of the art at the time.

 

He mentioned that he had an ATR8000 and wanted to try out the drive. So, since I couldn't use it for anything else, I visited his apartment and helped him connect it up electrically. I really didn't understand what I was looking at because I had never heard of an ATR8000 before that moment. Now that it was connected properly (SASI ID set correctly, termination correct, and power for the disk drive - it was a hungry beast), he very quickly ran the programs to format it and mount it as a disk volume....then he begged me to sell it to him. I said no because I had other plans for it.

 

I do remember that it was formatted with a 256 byte block format. Also, I don't remember multiple volumes, just a huge volume with many MB of space.

 

Today I understand what was going on. Apparently he must have had the Z-TEC 1000 SASI controller and the software to use it. I finally found some information on this board and would love to replicate one. Unfortunately they are quite rare and copying the circuit is made more difficult by the fact that it uses a PAL10L8.

 

The good news is that I still have the MFM disk drive (Vertex peripherals V170, later a Priam drive) and the SASI controller (a DTC peripherals controller). And they still work. I actually have two of them, plus many other 5.25 MFM hard disk drives. I used the V170 with an Adaptec RLL controller, as the V170 documentation said the platter was capable of the higher density RLL encoding. That gave me 90MB HD on my MAC 128. I tested it within the last few years. What a trip down memory lane that was.

 

So, the holy grail for me is the Z-TEC 1000 controller. If I get this mini ATR8000 working and someone with a Z-TEC 1000 controller is willing to let me try it out, I could try to duplicate it. Maybe a mini ATR8000 for them as well.

[+Larry]

Is the footprint for the SIO jack(s) present, so if a purchaser wanted to solder in actual SIO jacks, he could? (They are available new, and not terribly expensive.)  What is the purpose of the perf board, holding an eprom?  What are the apx dimensions? 

 

I'm interested if this happens.

[StartedWith400]

I have done another pass at the PCB and fixed some problems in the first spin. Here is a KiCAD rendering of the revised PCB:

[StartedWith400]

1 minute ago, Larry said:

Is the footprint for the SIO jack(s) present, so if a purchaser wanted to solder in actual SIO jacks, he could?  What is the purpose of the perf board, holding an eprom?  I'm interested if this happens.

No, there is no footprint for an SIO jack. It could be added in a revision of the PCB. I don't have a KiCAD model of one so that would have to be created. Why? Do you have a source of SIO jacks?

 

The perfboard is an interposer to fix a mistake I made in the first revision of the PCB. I got my wires crossed (literally). I socketed a 28pin ROM but forgot to adjust pin numbers from the original 24 pin part.

[+Larry]

Vintage Computer Center in CA. 

https://www.vintagecomputercenter.com

$6.99 ea. -- maybe less for a larger quantity?  (Or in my case, donor boards.)

[_The Doctor__]

serial port and printer port fingers need to protrude for actual use...

[reifsnyderb]

Here's my KiCad info on the SIO Connector:

 

 

The connector pins:

 

 

The layout on the PCB.  Note the silk screen lines coming up from the the two large "vias".  Those locate the edge of the PCB.  As I did this when learning to use KiCad to make my own footprints, I should probably change this a little so as to not put silkscreen on the board:

 

 

Finally, the KiCad file for the connector layout itself:

 

 

SIO_Port_Connector.kicad_mod

 

 

 

Edited May 1, 2022 by reifsnyderb

[+Larry]

But if there are two SIO connectors, they should be connected like a disk drive, shouldn't they?  Perhaps  @Dropcheck would have file(s) for the two port configuration.

[BillC]

4 hours ago, _The Doctor__ said:

serial port and printer port fingers need to protrude for actual use...

While more expensive than edge connection using DA15 and DB25 connectors would probably be a cost saving for the end user since standard PC serial/printer cables could be used.

Edited May 1, 2022 by BillC

[BillC]

3 hours ago, reifsnyderb said:

Here's my KiCad info on the SIO Connector:

 

 

The connector pins:

 

 

The layout on the PCB.  Note the silk screen lines coming up from the the two large "vias".  Those locate the edge of the PCB.  As I did this when learning to use KiCad to make my own footprints, I should probably change this a little so as to not put silkscreen on the board:

 

 

Finally, the KiCad file for the connector layout itself:

 

 

SIO_Port_Connector.kicad_mod 2.26 kB · 2 downloads

 

 

 

You may want to look at how Consumer implemented the SIO connection for the Sdrive NUXX, he included pads that allowed either a DA15/SIO connector to be installed. This is J1 at the right edge of the linked image, it takes the same amount of PCB area as for only the SIO connector.

https://nuxx.net/gallery/v/stuffivemade/sdrive_nuxx/design_documents/sdrive_nuxx_pcb_v1_1.png.html

[+mytek]

19 hours ago, BillC said:

You may want to look at how Consumer implemented the SIO connection for the Sdrive NUXX, he included pads that allowed either a DA15/SIO connector to be installed.

I did the same on the 1088XEL and XLD boards, maintaining compatibility with Consumer's implementation. Also because of size constraints, the 576NUC+ only has the DA-15 jack for it's external SIO connection.

 

 

 

 

I believe the idea for the DA-15 alternative and the standard that came from it began in 2006 by Raster.

 

 

[cwilbar]

On 4/30/2022 at 1:09 PM, StartedWith400 said:

Like many of you here, I always wanted to connect a hard drive to the Atari using the ATR 8000. Back in 1987, a college acquaintance found out that I had a 70MB full height 5.25 MFM hard drive and SCSI/SASI controller. It was something that I had been testing at a co-op job and it was pretty state of the art at the time.

 

He mentioned that he had an ATR8000 and wanted to try out the drive. So, since I couldn't use it for anything else, I visited his apartment and helped him connect it up electrically. I really didn't understand what I was looking at because I had never heard of an ATR8000 before that moment. Now that it was connected properly (SASI ID set correctly, termination correct, and power for the disk drive - it was a hungry beast), he very quickly ran the programs to format it and mount it as a disk volume....then he begged me to sell it to him. I said no because I had other plans for it.

 

I do remember that it was formatted with a 256 byte block format. Also, I don't remember multiple volumes, just a huge volume with many MB of space.

 

Today I understand what was going on. Apparently he must have had the Z-TEC 1000 SASI controller and the software to use it. I finally found some information on this board and would love to replicate one. Unfortunately they are quite rare and copying the circuit is made more difficult by the fact that it uses a PAL10L8.

 

The good news is that I still have the MFM disk drive (Vertex peripherals V170, later a Priam drive) and the SASI controller (a DTC peripherals controller). And they still work. I actually have two of them, plus many other 5.25 MFM hard disk drives. I used the V170 with an Adaptec RLL controller, as the V170 documentation said the platter was capable of the higher density RLL encoding. That gave me 90MB HD on my MAC 128. I tested it within the last few years. What a trip down memory lane that was.

 

So, the holy grail for me is the Z-TEC 1000 controller. If I get this mini ATR8000 working and someone with a Z-TEC 1000 controller is willing to let me try it out, I could try to duplicate it. Maybe a mini ATR8000 for them as well.

I have a Z-tec 1000, and I know of at least one other forum member that has one as well.  I've tried two times over the years to make this work, but the one thing I have NOT been able to figure out/find is how to format (low level) the hard drive.  I would think this would be a program within CP/M.  And there are 2 or 3 low level formatting programs on a CP/M disk that I have (that I did get from the same person as the ATR8000 so many years ago).... but none of them seem to recognize the controller.  I don't know if I'm doing something wrong, or if the Z-tec is bad, etc.  I tried it with my Xebec bridge controller, I tried it with an ACB4000A bridge controller.

 

My next step was to look into using Linux or possibly a Sun workstation to do a low level format (I know that Sun supported ACB4000A in SunOS 4.X for sure).... just don't know if you can make it do 256 byte sectors.  The other idea is using Linux.  I also wanted to try some early SCSI drives which are supposed to be 'compatible' with SASI.  I think you have to do some SCSI commands (hopefully only once) to put this drive in that mode.  Then format it to 256 byte sectors.  I have some early CDC/Imprimis/Seagate (depending on when they were purchased) drives that support this.  Plus I was going to try some older smaller Maxtor, Quantum, and Seagate drives.

 

I haven't gotten back to this though.

 

I haven't found much of any documentation on the Z-tec.  One document someone sent me, and one document I have.  The former appears to have been preparred by someone from a user group, and I don't know who created the other one.  Neither talk about low level formatting the drive.  I have yet to find anyone who knows enough or has docs to cover this part of the magic 'sauce'.

 

 

[StartedWith400]

Thanks for those KiCAD SIO parts. I like the idea of both options on the PCB.

 

To _The Doctor__: "serial port and printer port fingers need to protrude for actual use..." - I intentionally made these flush and put contacts on both sides of the PCB. Then I pressed and soldered DB-25 solder cup connectors so that one row is on each side of the PCB. The end result is that the DB-25 shell protrudes to the proper distance for a case mount. See the actual photo of my PCBA, post #82.

 

To cwilbar: I have a lot of knowledge and documentation on SCSI/SASI. I even created my own Atari 800 joystick based SCSI "host adapter" back in the 1980s. I was trying to write a driver for it but that was way over my head back then. I was able to successfully send format commands and read write data from Atari Basic. I connected that same drive and controller combination to my Mac Plus "back in the day" using the Ephriam Vishniac's driver (where you used ResEdit to enter the proper command bytes for your controller). Ow, I think I just hurt my brain remembering that old stuff. Anyway, my point is that SCSI controllers usually take only a few bytes to perform a whole drive format. The tricky part, as you alluded to, is the low level format.

 

I do remember providing the controller commands the the kid that connected my SCSI drive to the ATR8000. He definitely was using some program that prompted for the command bytes. I wish I had paid more attention - I can't be certain that he was using a Z-TEC 1000.

 

My first step is to get my circuit working. As I don't have, and never have had, an ATR8000, I am not completely certain how to validate the completed project. Power is good, Z80 clock looks clean, M1 jiggles for a little while then stops. I think it is going off into the weeds. I need to hook up my new Rigol logic analyzer and debug the beast.

 

Does anyone know the answer to this: If set up the ATR8000 for terminal mode, use the special standalone terminal cable (SIO connector to RS-232), connect to a terminal (9600, O, 7, 1), and turn on the ATR8000 without the disk drive attached, what happens?

[cwilbar]

3 hours ago, StartedWith400 said:

Thanks for those KiCAD SIO parts. I like the idea of both options on the PCB.

 

To _The Doctor__: "serial port and printer port fingers need to protrude for actual use..." - I intentionally made these flush and put contacts on both sides of the PCB. Then I pressed and soldered DB-25 solder cup connectors so that one row is on each side of the PCB. The end result is that the DB-25 shell protrudes to the proper distance for a case mount. See the actual photo of my PCBA, post #82.

 

To cwilbar: I have a lot of knowledge and documentation on SCSI/SASI. I even created my own Atari 800 joystick based SCSI "host adapter" back in the 1980s. I was trying to write a driver for it but that was way over my head back then. I was able to successfully send format commands and read write data from Atari Basic. I connected that same drive and controller combination to my Mac Plus "back in the day" using the Ephriam Vishniac's driver (where you used ResEdit to enter the proper command bytes for your controller). Ow, I think I just hurt my brain remembering that old stuff. Anyway, my point is that SCSI controllers usually take only a few bytes to perform a whole drive format. The tricky part, as you alluded to, is the low level format.

 

I do remember providing the controller commands the the kid that connected my SCSI drive to the ATR8000. He definitely was using some program that prompted for the command bytes. I wish I had paid more attention - I can't be certain that he was using a Z-TEC 1000.

 

My first step is to get my circuit working. As I don't have, and never have had, an ATR8000, I am not completely certain how to validate the completed project. Power is good, Z80 clock looks clean, M1 jiggles for a little while then stops. I think it is going off into the weeds. I need to hook up my new Rigol logic analyzer and debug the beast.

 

Does anyone know the answer to this: If set up the ATR8000 for terminal mode, use the special standalone terminal cable (SIO connector to RS-232), connect to a terminal (9600, O, 7, 1), and turn on the ATR8000 without the disk drive attached, what happens?

While I haven't done that, you should get a monitor prompt where you could then boot (I believe).

 

I've only used the sio to the computer, booted atrmon to get the prompt from the ATR.

 

A 'b' command then would boot CP/M (provided you have the disk and an attached floppy drive).

 

 

[StartedWith400]

On 5/2/2022 at 9:47 AM, mytek said:

I did the same on the 1088XEL and XLD boards, maintaining compatibility with Consumer's implementation. Also because of size constraints, the 576NUC+ only has the DA-15 jack for it's external SIO connection.

 

 

 

Is there a freely available KiCAD symbol for this dual connector?