New Saturn 128k compatible cards

100 posts / 0 new
Last post
Offline
Last seen: 1 month 1 hour ago
Joined: Apr 13 2006 - 22:28
Posts: 101
Actually, I just remembered

Actually, I just remembered that David has provided support for GAL20V8 devices which my programmer *does* support (I just fired up the programming software to check).

 

Probably the easiest (and cheapest) solution for me is to buy some of these GAL20V8's to use instead. Now to wait for more parts to come in... <grumble>

 

Offline
Last seen: 1 day 10 hours ago
Joined: Aug 4 2018 - 21:59
Posts: 123
I ended up dropping the 22V8,

I ended up dropping the 22V8, I'll have to check that it can do the timing updated code. So let me know if you need it and I'll look into it.

I can conform that my TL866 can no longer program lattice 22v10, one of the buggy software updates they have been releasing lately is to blame.
I now use me VS4800 to program both Atmel and Lattice GAL's

Offline
Last seen: 1 month 1 hour ago
Joined: Apr 13 2006 - 22:28
Posts: 101
I haven't ordered any new

I haven't ordered any new components yet, so maybe I'll just need to bite the bullet and get a VS4000 programmer (which seems to be a 40 pin variant of the VS4800). I can't see me needing to program anything > 40 pins. At least its pretty reasonably priced ~AUD$70.

Offline
Last seen: 2 days 10 hours ago
Joined: Jul 5 2018 - 09:44
Posts: 354
One problem is a lot of

One problem is a lot of people say that a lot of the Lattice marked chips on the market right now, especially those from "no telling" Chinese sellers are fakes.  My guess is that a lot of the complaints about programming failures with the Minipro TL866CS could be due to poor quality fake chips, or used chips being sold as new, remarked other chips being sold as 22V10, etc.

 

When I say it is getting harder and more expensive to find Lattice parts, I mean through "reputable" channels.  Those channels still have the Atmel ATF parts.

 

Offline
Last seen: 2 days 4 hours ago
Joined: Dec 19 2008 - 21:01
Posts: 282
There seem to be a lot of

There seem to be a lot of variants of the 22V10 and other PAL/GALs. If you choose the wrong type, it will not program, because they all seem to be different.  I was able to get some of the cheap Lattice 22V10D's from China and they program with my cheap G540 (if I go back and use Windows XP, anyway).

 

Offline
Last seen: 1 day 10 hours ago
Joined: Aug 4 2018 - 21:59
Posts: 123
The VS4000 and the VS4800 all

The VS4000 and the VS4800 use the same software. Not being able to use the software under Win10 x64 is annoying but it works fine with Win 7 x86 running in a Virtualbox virtual machine. As far as Lattice GAL's go, the same batch of chips that I got from ebay that used to program in the TL866 will now now longer program. They do however program fine in the VS4800, so it isnt a case of rebadged or clone chips that dont work.

Offline
Last seen: 2 days 10 hours ago
Joined: Jul 5 2018 - 09:44
Posts: 354
I don't use the supplied

I don't use the supplied software for the TL866CS, so I am curious to whether the Open Source software has the same issue with the Lattice GALs as the Windows software does.  It will be a while before I have time to try it though.

 

Offline
Last seen: 1 month 1 hour ago
Joined: Apr 13 2006 - 22:28
Posts: 101
So I did pull the trigger and

So I did pull the trigger and bought a VS4000 last night. Running under a VM is something I'm used to - the TOP853 I have works best running under an 32 bit XP VM.

 

I will advise how the VS4000 goes when I finally get it in my (eagerly waiting) hands.  :-)

Offline
Last seen: 1 month 1 hour ago
Joined: Apr 13 2006 - 22:28
Posts: 101
OK, the VS4000+ *seems* to

OK, the VS4000+ *seems* to work fine for writing the Lattice GAL22V10D-15LPN that I have but I am getting weird results when testing my completed card:

  * won't boot when installed in slot 0 (in my Apple II+ or Apple II+ clone)

  * will boot when installed in slot 1 but when I run the RAMTEST128K on the Saturn 128 disks, I seem to get errors for every test (but I notice that the different LEDs are cycled through depening on the test, so it is doing *something*)

  * I've tried various versions of the GAL code - the original 1.0 version, the 1.1 version, the no DMA version, and the DMA fixed version and all seem to have similar problems

  * I've gone over all the soldering under magnification and everything looks ok, but I haven't tested continuity for every connection yet (although I have tested a number of them and they appear OK)

 

Is this likely to be a soldering issue or a bad GAL? Should I order some of the Atmel chips and if so would these fit the bill:

https://www.ebay.com.au/itm/182529820716

 

I notice that the V4000+ programmer software only lists the ATF22V10B and not the ATF22V10C variety,  but I don't know whether the C variety would work just as well?

 

It's the first time I've used a Saturn card, so if there is a better means of testing than running the RAMTEST128K program, please let me know.

 

Thanks!

macnoyd's picture
Offline
Last seen: 5 hours 54 min ago
Joined: Oct 15 2012 - 08:59
Posts: 487
My Experience ...

I'm using the Lattice chips with much success and the card works fine in Slot zero, but one thing I noticed is that the card seems to be sensitive to the 74LS74 chips.

It was by accident that I discovered the card works consistently better using SN74HC74N chips instead of the LS series.

I don't have an explanation for this but I noticed the card works with more Apple ][ variations when using these chips.  (HC version)

 

Programming the Lattice chips "reliably" requires that I use my old TOP3100 programmer under a 32-bit Windows XP environment.

For some reason, they never offered an upgrade to the old version programmer, but it becomes only a minor hassle to use in a virtual environment.  It seems to burn the Lattice GAL's just fine.

Overall, I've had good luck with using the Lattice chips.  Been using them successfully for all the "Reboot" projects.

Offline
Last seen: 1 day 10 hours ago
Joined: Aug 4 2018 - 21:59
Posts: 123
Use locksmith to test it.

Use locksmith to test it.

I have had reports from another person that ramtest128 failed but locksmith worked fine.

When you used the 1.1 GAL code did you also add the mod?

 

 

macnoyd's picture
Offline
Last seen: 5 hours 54 min ago
Joined: Oct 15 2012 - 08:59
Posts: 487
FYI the mod ...

My version of it... (note the 10K SMD resistor at lower left & the green jumper at lower center)

 

Offline
Last seen: 1 day 10 hours ago
Joined: Aug 4 2018 - 21:59
Posts: 123
It is unlikely to be the GAL.

It is unlikely to be the GAL. with the mod done use the 1.1 code the only other time I've seen a problem it has been the RAM chip.

Offline
Last seen: 1 month 1 hour ago
Joined: Apr 13 2006 - 22:28
Posts: 101
macnoyd wrote:My version of

macnoyd wrote:

My version of it... (note the 10K SMD resistor at lower left & the green jumper at lower center)

 

 

I've done the jumper wire, but I thought the resistor was only for Atmel GALs...

 

Anyway, looks like I've got more testing to do. I can always try it in my //e or //e clone to rule out anything to do with my II+ or II+ clone.

 

I think I have some 10K resistors for the MultiROM card (which I haven't assembled yet), so I can try this as well...

Offline
Last seen: 1 month 1 hour ago
Joined: Apr 13 2006 - 22:28
Posts: 101
macnoyd wrote:My version of

macnoyd wrote:

My version of it... (note the 10K SMD resistor at lower left & the green jumper at lower center)

 

Actually, what is the purpose of your 10K SMD resistor? Did this fix a particular issue you were having or...?

macnoyd's picture
Offline
Last seen: 5 hours 54 min ago
Joined: Oct 15 2012 - 08:59
Posts: 487
The only purpose of the 10K pull-up

The only purpose of the 10K pull-up resistor is for the Atmel GAL's, as David has indicated. 

I put the resistor there for design consistency.  It really does nothing for the Lattice chips.

It solved nothing for me, but the 74HC74 TTL chips did make a noticeable difference with

compatibility among different Apple II Platforms.  The 74LS74 version wouldn't work at all

in some versions of the Apple II.  Can't tell you why because I haven't looked into it.

Offline
Last seen: 1 month 1 hour ago
Joined: Apr 13 2006 - 22:28
Posts: 101
macnoyd wrote:The 74LS74

macnoyd wrote:

The 74LS74 version wouldn't work at all

in some versions of the Apple II.  Can't tell you why because I haven't looked into it.

 

When I try loading up Locksmith 6.0 with the card installed, it boots to the Locksmith menu but then the computer locks up hard - even Ctrl-Reset is ignored and it requires a power recycle to reboot. Does this give any clues on which areas I should be concentrating on? I got my //e clone out and it behaves exactly the same way - even requiring a power recycle when booting into Locksmith. What slot signals could cause this behaviour?

 

My Apple II+'s lock up during boot with the card installed in slot 0 - did you have any of these kind of problems with the 74LS74?

 

 I've gone over the board and tested continuity on ~90% of the board (using the schematic) and eveything checks out OK so far. 

 

Maybe I should just bite the bullet and get some 74HC74's too and see if things improve, but to me it sounds like there's a dodgy connection somewhere.

 

Thanks for any insight you guys can provide. 

macnoyd's picture
Offline
Last seen: 5 hours 54 min ago
Joined: Oct 15 2012 - 08:59
Posts: 487
Polymorph69 wrote:My Apple II

Polymorph69 wrote:

My Apple II+'s lock up during boot with the card installed in slot 0 - did you have any of these kind of problems with the 74LS74?

 

Yes, I experience lock-up issues with the 74LS74, but not with the 74HC74.  As said earlier, it wouldn't work at all. Computer would randomly lock up, sometimes boot, acted unstable.

Offline
Last seen: 1 day 10 hours ago
Joined: Aug 4 2018 - 21:59
Posts: 123
The resistor is only required

The resistor is only required when using Atmel GAL's, it isnt needed for Lattice GALS.

 

The issue with the 74LS74's may be tied to a specific brand of chip. I've seen no issues at all with lockups with the TI chips I've been using on the preassembled cards.

The contunuity test is probably a waste of time, the factory that makes the boards does automated continuity and short circuit tests on every pcb before they ship.

 

What other cards do you have installed when it locks up?  Not all cards support DMA chaning and an empty slot in the chain could also cause the problem.

Offline
Last seen: 1 month 1 hour ago
Joined: Apr 13 2006 - 22:28
Posts: 101
David_M wrote:The resistor is

David_M wrote:

The resistor is only required when using Atmel GAL's, it isnt needed for Lattice GALS.

 

The issue with the 74LS74's may be tied to a specific brand of chip. I've seen no issues at all with lockups with the TI chips I've been using on the preassembled cards.

The contunuity test is probably a waste of time, the factory that makes the boards does automated continuity and short circuit tests on every pcb before they ship.

 

What other cards do you have installed when it locks up?  Not all cards support DMA chaning and an empty slot in the chain could also cause the problem.

 

I was mainly testing continuity to see if *I* stuffed up somewhere.  ;-)

 

When testing, the only other card in the II+'s is the disk ][ controller. //e clone is the same with the exception of the aux RAM card.

 

I'm using TI 74LS74's too, but it sounds like I might be best to try the 74HC74 and see if that helps.

 

Thanks guys.

 

Offline
Last seen: 1 day 10 hours ago
Joined: Aug 4 2018 - 21:59
Posts: 123
When it locks up which LED's

When it locks up which LED's are lit?
What are you trying to boot when it locks up?

Also you say it doesnt lock up when it is in slot 1.  Have you tried the locksmith RAM test with the card in slot 1?



 

Offline
Last seen: 1 month 1 hour ago
Joined: Apr 13 2006 - 22:28
Posts: 101
David_M wrote:When it locks

David_M wrote:

When it locks up which LED's are lit?
What are you trying to boot when it locks up?

Also you say it doesnt lock up when it is in slot 1.  Have you tried the locksmith RAM test with the card in slot 1?

 

Answering your questions:

  1) no LEDs are lit when it locks up

  2) as I mentioned in a previous post, booting Locksmith 6.0 it gets to the main menu and locks up (requiring a power recycle)

  3) when installed in slot 0 in the Apple II+'s, it locks up part way through the boot process when booting the Saturn 128K disk (doesn't even get to prompt) . I haven't tried booting Locksmith with the card in slot 0. When installed in slot 1 in both the Apple II+'s and //e clone and booting Locksmith 6.0, it gets to the main menu and locks up straight away. In all cases when it locks up, a power recycle is required to reboot.

 

 

 

Offline
Last seen: 1 month 1 hour ago
Joined: Apr 13 2006 - 22:28
Posts: 101
I should have mentioned, when

I should have mentioned, when installed in slot 1 it boots the Saturn 128K utilties disk OK, but appears to fail all of the tests when running RAMTEST128K - but the different LEDs light up depending upon the test being run (so if I keep 'Continuing' through the tests I see each of the LEDs come on at some stage and in various combinations).

 

Offline
Last seen: 1 day 10 hours ago
Joined: Aug 4 2018 - 21:59
Posts: 123
RAMTEST128K  is not a good

RAMTEST128K  is not a good test, test it with locksmith it will give much more information.

Offline
Last seen: 3 months 2 days ago
Joined: May 20 2016 - 06:57
Posts: 46
Isn't there any space left in

Isn't there any space left in the 22V10 to accomodate these 6 D-triggers and to get rid of the external '74s? What are they used for (I am lazy enough to reverse the PCB pictures posted here)? Look at the pocket rocket. Just one 16L8 as far as I remember...Unfortunately I still don't have one...

Offline
Last seen: 1 month 1 hour ago
Joined: Apr 13 2006 - 22:28
Posts: 101
OK, I tried swapping out the

OK, I tried swapping out the 74LS74's for 74HC74's last night but it made absolutely no difference in my case.

 

I might wait for the extra parts to arrive for my 2nd board and see if swapping  the RAM makes any difference.

 

What confuses me is how the computer locks up hard after booting to the Locksmith main menu - I've hardly ever seen that on a //e. I would have thought bad RAM would just drop to the monitor or something, but the fact that it requires you to turn the power off and on to get any response from the machine leads me to think there's a bad connection somewhere. What could cause the machine to ignore Ctrl-Reset? I get no beep or anything...

 

My hunt for the problem continues...

macnoyd's picture
Offline
Last seen: 5 hours 54 min ago
Joined: Oct 15 2012 - 08:59
Posts: 487
Keep us posted

Keep us posted with your progress.

Offline
Last seen: 3 months 2 days ago
Joined: May 20 2016 - 06:57
Posts: 46
@Polymorph69, the INHibit

@Polymorph69, the *INHibit signal probably stays active when it shouldn't be or are you sure the diode is soldered properly (the orientation of its anode and cathode is correct). Is it a Shottky diode? Is it operational at all? I can't see from the pictures well but I am almost sure the *RESET signal does not go to the 22V10 and has no way to reset the ram card's *INH logic...There are still bugs with this design...

 

 

Offline
Last seen: 1 day 10 hours ago
Joined: Aug 4 2018 - 21:59
Posts: 123
No there isn't room in the

No there isn't room in the GAL to move the latches into it. If there was I would have done so.

The diode should be a 1N4148 it's voltage drop is low enough to work fine.

No the GAL does not need the RESET signal, nothing is latched in the GAL all outputs are a function of the inputs.

No it isnt a design bug.

macnoyd's picture
Offline
Last seen: 5 hours 54 min ago
Joined: Oct 15 2012 - 08:59
Posts: 487
Mine are all working ...

Mine are all working fine ... Not sure what the problem is.  I will post a photo later.  I will also post results of Saturn test and Locksmith tests. (later)

Offline
Last seen: 2 days 4 hours ago
Joined: Dec 19 2008 - 21:01
Posts: 282
FWIW, I built mine from David

FWIW, I built mine from David's boards and they both pass all the Saturn RAM tests in my II and II+. 

Offline
Last seen: 3 months 2 days ago
Joined: May 20 2016 - 06:57
Posts: 46
The Si 1N4148 gives 0.6V

The Si 1N4148 makes 0.6V voltage drop under average current which also depends on specific computer and its peripherals connected. Add to this the logical 0 voltage of a TTL/GAL of 0.4V and you are shifting logical 0 to 1V! Could work in some computers - in others not.  Better use open collector ot tri-state output.  But you are using a registered GAL device without getting use of its latches...David, here is a puzzle for you...How could the AE engineers manage to design the 16K card with only one "mere"  PAL 16L8?

 

https://mirrors.apple2.org.za/Apple%20II%20Documentation%20Project/Interface%20Cards/Language%20Cards/AE%2016k%20Pocket%20Rocket%20RAM%20Card/Photos/AE%2016k%20Pocket%20Rocket%20RAM%20Card%20-%20Front.jpg

You need 3 more bits/flip-flops for the 8 banks (the 8 language cards of which  saturn card consists)...why did you use 6 D-triggers then? I am still not convinced there is no room in REGISTERED GAL 22V10. Will analyze this later...More comments and ideas to come...

Offline
Last seen: 1 month 1 hour ago
Joined: Apr 13 2006 - 22:28
Posts: 101
Casablanca wrote:@Polymorph69

Casablanca wrote:

@Polymorph69, the *INHibit signal probably stays active when it shouldn't be or are you sure the diode is soldered properly (the orientation of its anode and cathode is correct). Is it a Shottky diode? Is it operational at all? I can't see from the pictures well but I am almost sure the *RESET signal does not go to the 22V10 and has no way to reset the ram card's *INH logic...There are still bugs with this design...

 

OK, I might check things around the INHibit signal as you suggest.

 

I used a 1N4148 as per David's design. My diode does look quite a bit smaller than that shown in David's pics of his board however. I'll take a photo of my board tonight and post here to see if you can spot anything that looks suspicious. I'm pretty confident that the diode is soldered in the correct orientation though.

 

I have to confess to be a software guy by trade (hardware is just a hobby for me) so I might well miss the fact that I ordered/was delivered an incorrect part or something.

 

Thanks for the help guys.

Offline
Last seen: 3 months 2 days ago
Joined: May 20 2016 - 06:57
Posts: 46
You can replace it with any

You can replace it with any Shottky diode if you have one handy, but if you don't for now you can wait a bit to see if the author of the project will come with a fix. I started to remember the language card by checking its schemtics and noticed that instead of using RESET signal it forms its own power-up signal to reset its internal state.  This maybe was useful if someone wanted to use  more persistent "shadow ROM" image or to hook the 6502 reset vector. Is there a schematics diagram of the original saturn card around?

macnoyd's picture
Offline
Last seen: 5 hours 54 min ago
Joined: Oct 15 2012 - 08:59
Posts: 487
Front Pic

Not as clean as the back pic, but certainly good enough ...

macnoyd's picture
Offline
Last seen: 5 hours 54 min ago
Joined: Oct 15 2012 - 08:59
Posts: 487
One important note worth mentioning...

When I ordered the Memory IC shown in the photo, nearly half of them were BAD.

One was DOA, and 5 failed at the exact same memory address.  ! !  Be aware of that.

I ordered from another vendor and bought KM681000CLP-71 RAM memory IC's and

all worked well. (all good)  Keep that in mind if you have doubts of the RAM you're using.

The only other possibility is the GAL, provided you programmed it properly with the v1.1 code.

One last note:  The resistors shown on the board were to accomodate the LED brightness.

The red LED was a high brightness/low power device and lit at the same brightness as the

3 green LED's using a 10K resistor.  Ther green LED's weren't so efficient. :/

Hope this all helps!

 

Offline
Last seen: 1 month 1 hour ago
Joined: Apr 13 2006 - 22:28
Posts: 101
Here's the front of my board:

Here's the front of my board:

 

And the back: 

 

macnoyd's picture
Offline
Last seen: 5 hours 54 min ago
Joined: Oct 15 2012 - 08:59
Posts: 487
Something to try ...

I know it will cost you $6.50 to try, but try replacing your RAM chip with this one...

https://www.ebay.com/itm/223266861788/

I see nothing wrong with your build and you've got the 74HC74's present.

Beyond that, there's only one component left.

Offline
Last seen: 3 months 2 days ago
Joined: May 20 2016 - 06:57
Posts: 46
Did you test the failing

Did you test the failing SRAMs in this RAM card or in a programmer?

macnoyd's picture
Offline
Last seen: 5 hours 54 min ago
Joined: Oct 15 2012 - 08:59
Posts: 487
I tested in this RAM Card

I tested all Memory in this RAM Card and (in the end) used the same Card to test ALL the RAM chips.

I built up 12 of these cards and found consistency with the IC's as noted but the RAM itself was inconsistent among those with the same part number.

Keep in mind I bought the cheapest RAM I could find from a Chineese supplier.  Basically a crap-shoot with regard to quality & reliability.

Offline
Last seen: 3 months 2 days ago
Joined: May 20 2016 - 06:57
Posts: 46
It would be nice if you test

It would be nice if you test those failing in this RAM card  SRAMs in another device, many (old?) programmers can test these chips...The price is often not related to quality.

Offline
Last seen: 1 day 10 hours ago
Joined: Aug 4 2018 - 21:59
Posts: 123
@Polymorph69  If you use

@Polymorph69  If you use locksmith to test the card as I suggested it will show you more detail on what is failing.

Also knowing what LED's are active when it locks up will help diagnose the problem.

I used the same brand of chip in most of the cards as you and found the failure rate to be around one in ten chips faulty.

Offline
Last seen: 3 months 2 days ago
Joined: May 20 2016 - 06:57
Posts: 46
I checked the photos of the

I checked the photos of the original SATURN card (I still cannot find mine) and it does not use DMA at all...What is the DMA in the David's card used for?

Offline
Last seen: 1 month 1 hour ago
Joined: Apr 13 2006 - 22:28
Posts: 101
David_M wrote:@Polymorph69 

David_M wrote:

@Polymorph69  If you use locksmith to test the card as I suggested it will show you more detail on what is failing.

 

As i mentioned in previous posts, when booting Locksmith 6.0 it gets to the main menu at which point the computer locks up and requires a power recycle (Ctrl-Reset is ignored). Because of this,  I am unable to test in Locksmith.

Offline
Last seen: 1 day 10 hours ago
Joined: Aug 4 2018 - 21:59
Posts: 123
Polymorph69 wrote:David_M

Polymorph69 wrote:

David_M wrote:

@Polymorph69  If you use locksmith to test the card as I suggested it will show you more detail on what is failing.

 

As i mentioned in previous posts, when booting Locksmith 6.0 it gets to the main menu at which point the computer locks up and requires a power recycle (Ctrl-Reset is ignored). Because of this,  I am unable to test in Locksmith.

When booting the saturn disk, what LED's are lit when it fails?

 

Also try locksmith 7, I have found at least one dsk image of locksmith 6.0 that has the problem of freezing at the menu in some machines.

Offline
Last seen: 3 months 2 days ago
Joined: May 20 2016 - 06:57
Posts: 46
Locksmith 7 is a vintage fake

Locksmith 7 is a vintage fake ;) You can compare the  disk images  and find out that only one character is different -- the version number that is output to the screen :) 

Offline
Last seen: 3 months 2 days ago
Joined: May 20 2016 - 06:57
Posts: 46
.
macnoyd's picture
Offline
Last seen: 5 hours 54 min ago
Joined: Oct 15 2012 - 08:59
Posts: 487
Awesome!

Awesome job!  Please make next batch using ENIG plate process rather than HASL process.

It's worth the difference in price.  Everyone likes the gold plated look over tin/lead.

 

Congrat's!

Offline
Last seen: 3 months 2 days ago
Joined: May 20 2016 - 06:57
Posts: 46
Thank you. I will consider

Thank you. I will consider your advice when ordering next batch of PCBs.  There are several other things that need polishing...My SCSI cards never had "golden" edge connector pads  too though.  

 

 

Pages

Log in or register to post comments