Grappler vs Grappler+

I've read that the differences between the 74LS373 that is on many cards and the 74LS374 that is on others is negligible and that one can be exchanged for another on that card.

So based on that you could modify it...

Also, there are lots of Grappler+ clone cards out there - many of them direct clones. 

• Basic parallel printer cards send literal data to the printer.  With a basic parallel card it can be a challenge to print graphics from Applesoft BASIC because the regular printing routines only pass 7-bit ASCII to the printer, but you usually need to send 8-bit data to print pixels.  If you're running commercial software that prints graphics, a basic parallel card will work as well as any of the cards below.
• Grappler and clones (like Grafstar) include firmware to print screen dumps to an Epson dot matrix printer (or an Epson compatible printer).  This makes it easier to print graphics if you're writing a program in Applesoft BASIC, since you can simply print a command string like PR#1:PRINT CHR$(9);"G" instead of tediously converting the pixels into printer strings yourself.  Grappler offers limited benefit if you're using commercial graphics software that knows how to print graphics, like The Print Shop.
• Grappler+ adds DIP switches so you can configure it for printers other than Epson.  It also adds a hardware interrupt that stops your program to put a message on the screen if the printer stops for an error, which received negative reviews in the 1980s because the popup message corrupts the screen if you're running in 80-column mode.  On pre-enhanced-//e models, the Grappler+ can cause DOS 3.3 to corrupt disks if a printer error occurs.  In the 1984 book Apple Thesaurus, Aaron Filler wrote, "If you're using a Grappler+ with DOS you should be very careful and make sure your printer is online and ready to go whenever the Grappler+ tries to call it.  The Microtek Dumpling GX is an excellent alternative."  And yes, there's also a design flaw with the 74LS373 at U10 that can cause the Grappler+ to misread data from the bus on an Apple IIgs, so it prints the correct quantity of characters but not necessarily the correct contents of those characters.  There are two subsequent versions of the Grappler+ with fixes for the bugs, identifiable by a 74LS374 at board location U10 or U2, depending on which variant.  Also, there's a special version of Grappler+ that is only compatible with IDS and Prism printers -- it still has DIP switches, but the switches don't do anything.
• Buffered Grappler+ fixed the aforementioned bugs, and removed interrupts completely.  Its principal advantage is the buffer, which enables you to go on to other tasks while printing runs in the background.  A minor penalty is that most programs have no way to cancel printing once it's started, but that won't be a hardship when using it with NEWPRINT.
• ProGrappler adds more DIP switches, more settings, more brands of printers, and more firmware.  This could be exciting if you're writing new software, but you'd undoubtedly need the manual!

Personally, I wouldn't show any special preference for a Grappler over a basic parallel printer card.  And I would avoid the Grappler+ because it has pointless flaws that aren't compensated by any meaningful benefits, and it's a nuisance to distinguish which variants will work with which Apple models and which printers.

There's documentation in the IO category on Apple.org:

The Grappler Interface - manual for the original model

Orange Micro Grappler+ Operators Manual for Grappher+ (including IDS version)

[EDIT] Correction, according to the docs the original Grappler wasn't necessarily restricted to just Epson brand printers.  Rather, there were different versions of the card for different brands of printers.  If you didn't have the correct version to match your printer brand, it would still work like a basic printer interface but without any extended graphics features.

djc6 wrote:

Personally, I wouldn't show any special preference for a Grappler over a basic parallel printer card.

Thanks for the detailed response! I really tried, and I couldn't find this info anywhere. One reason I ended up buying a Grappler+ (U10 74LS373 varient) is that it seems to have the most documentation. Another card I had considered was the Q-Print II by Sequential

If you've got it working, then keep it working.

The GS issues were just luck-of-the-draw.  If you had a GS and Grappler+ that didn't play well together, it was just bad luck.  Our school office had a Grappler+ that inserted misspellings randomly into every print.  One or two randomly garbled characters every line.  I swapped-in a Grappler from an Apple //e in our computer lab, and they both worked fine.

The interrupts probably won't any cause harm on a GS because Apple improved the firmware over the original Apple ][ and un-enhanced //e.  Anything after the enhanced //e should be ok.  And, so long as NEWPRINT is always "ready" (not offline nor signalling an error) it won't even trigger those pop-up message that caused the problem.

I have a couple non-grappler parallel printer cards I will likely never use, at least one that I have an extra of is this one:

https://mirrors.apple2.org.za/Apple%20II%20Documentation%20Project/Interface%20Cards/Parallel/Practical%20Peripherals%20PRinterface%20Card/

I may even have the docs for one of those. If that looks at all helpful let me know. 

djc6 wrote:

djc6 wrote:

Despite the card being the reportedly non-IIGS compatible version, works fine with the NEWPRINT.

 

So this unexpected.. the U10 74LS373 Grappler+ card works fine text/graphics when using Apple II software, but I'm getting "lines" of missing data throughout the printouts when using GS/OS applications l

It's worth a try.  Chances are the ROM image won't matter because I'd guess GS/OS probably isn't using it.  It's probably writng directly to the card.  The 74LS373/374 maybe makes a timing diffference.

softwarejanitor wrote:

It's worth a try.  Chances are the ROM image won't matter because I'd guess GS/OS probably isn't using it.  It's probably writng directly to the card.  The 74LS373/374 maybe makes a timing diffference.

Plainly, it's a functional difference...and not merely timing.  Consider what these parts actually are:

• 74LS373 Octal Transparent Latch: Enters transparent (pass-thru) mode when pin 11 is high, latches data when pin 11 goes low.
• 74LS374 Octal D Flip-Flop Bus Interface: Captures data when pin 11 transitions from low-to-high.

Choosing the 74LS373 was a design blunder, probably a side effect of engineers testing parts they already had on hand.  There's no functional reason to put any 74LS373 chips on the Grappler+, let alone two.  U10 was wired so that it latched data too late, hence the GS timing bug, while U11 is wired so that it cannot ever function as a latch.  The obvious explanation is that their engineers defaulted to the LS373 because it was a part they were already using in previous designs.

Illustrating with Apple's timing diagram, peripherals should latch data on the leading edge of the select signal.  Instead, the Grappler+ puts its LS373 into transparent mode until the trailing edge.  It might capture the correct data.

Grappler timing.png

Complicating matters even more, the Grappler+ is wired so that it detects the trailing edge when any of four signals changes: DEVICE SELECT, R/W, A0, or A1

S.Elliott wrote:

peripherals should latch data on the leading edge of the select signal. 

 

Are you sure about that? The leading edge just means that the address bus has settled and we can now read/write data. But the data bus is now changing and may not be correct until just before the trailing edge. That's when you want to latch data.

Great information on the Grappler and Grappler+.

I've had a Grappler C in my parts box for years, because I maintain a working Apple //c, but I have never had the need to use it; not having a parallel printer.

I would love to hear your input about this flavor of Grappler and/if there is any utility for me keeping it all these years?  

Thank you!

jeffmazur wrote:

S.Elliott wrote:

peripherals should latch data on the leading edge of the select signal. 

 

Are you sure about that? The leading edge just means that the address bus has settled and we can now read/write data. But the data bus is now changing and may not be correct until just before the trailing edge.

Oopsie, no, I'm not sure about the right way to do it.  Just sure that Grappler+ definitely did it the wrong way.

The Grappler+ implementation latches data after the end of ϕ0, while Apple's reference manual says cards need to read the data strictly-before that.  But my 1981 edition of the Reference Manual is slightly ambiguous:

• text on p. 90 says data is valid 300 nsec after the end of ϕ1 (ie: 200 nsec before the end of ϕ0),
• Figure 11 says data is valid 300 nsec before the end of ϕ0 (ie: 200 nsec after the end of ϕ1).

Despite that superficial inconsistency in the Reference Manual, both descriptions could be correct: the data could be valid at both of those times, and a card should conform to at least one of them.  Grappler+ didn't.

MichaelLAX wrote:

Great information on the Grappler and Grappler+.

I've had a Grappler C in my parts box for years, because I maintain a working Apple //c, but I have never had the need to use it; not having a parallel printer.

I would love to hear your input about this flavor of Grappler and/if there is any utility for me keeping it all these years?  

Thank you!

That is a different beast than what is being discussed. It is essentially a serial to parallel converter. It doesn't have to worry about timings (well, it cares about serial port settings) or how it reacts to the A2 expansion bus. It just converts serial data to parallel. I have a similar device made by some mom and pop for my //c as we had a Prowriter Jr growing up. Unfortunately you can't get ribbons for it any more. So I got an IW2.

But about a year or so ago I also picked up a couple Commodore MPS 1250 printers that are parallel and can mimic an Epson. I pulled out The Kameleon (the name of my interface) and hooked it up. It worked with print shop. 

But it's a novelty now that I have an IW2. Which answers the other part of your question. It has no real use unless you have a parallel printing device. 

djc6 wrote:

So, bad news.  I socketed U10 - the Grappler+ continues to function with the original 74LS373 in the socket, but the NEWPRINT doesn't work when 74LS374 installed in the socket.

Based on @jeffmazur's previous comment, they undoubtedly did more than just swap an LS374 for LS373.  The trailing edge of ϕ0 was certainly too late, but blindly switching to LS374 would just make it too early.

Apple's text and diagrams suggest the idea latching time would be 300 nsec away from one of the edges of ϕ0.  In other words, sometime near the falling edge of Q3.

Your closeup view reveals two vias were added to the later Grappler+ near the edge connector (yellow circle) and the LE latch-enable output from the NAND gate was removed (blue circle).  The yellow circle is just above the location of the Q3 and 7M bus signals on the reverse side of the card, pin 37 on the edge connector.  Since the original Grappler+ schematic didn't show any slot connections in that area, the sudden appearance of these two vias suggests they added another signal in that area.

Grappler with two extra vias.jpg

My guess would be that they added a connection to Q3 (pin 37) and routed it to U4, and inserted the spare fourth NAND gate into the circuit to combine the latch-enable with Q3 to generate a clock input for the LS374.  (The schematic for the original board is on page 23 of the Grappler + Operators Manual.)

Stick with something basic, like this $15 Apple Parallel interface on eBay.  [EDIT: watch out, though, it might not include a cable!]

From my experience as a student-employee in the 1980s, users didn't really benefit from the so-called enhanced features of boards like Grappler, Grafstar, and Dumpling-GX, etc.  Occasionally there was a placebo effect, like users who chose the Grappler-equipped computer because they wanted to print graphics, not realizing that it didn't have any effect whatsoever on software like The Print Shop.

[edit] Whatever you choose, try to find a card that includes its own cable.  Most cards had a simple header to connect the cable, but there wour four incompatible styles: 26-pin (Orange and clones), 20-pin (Apple and clones), 14-pin (Apricorn), 34-pin (Microtek).  Connectors of the same size were usually interoperable, but not interchangeable with those of a different size.

djc6 wrote:

I've noticed many cards are missing their parellel cable, I do see a listing for new Grappler compatible cables being made.    That said, NEWPRINT includes a 26 pin ribbon cable, so I guess it expects to be plugged into Orange Micro cards and clones only:

Oopsie.  I hadn't seen a picture of NEWPRINT until now, so all this time I erroneously pictured it using a Centronics-36 connector just like a printer.  Nope, it's a 34-pin header like you'd find in a 70's-era  S-100 machine.  How retro!

Newprint80.jpg

The Centronics connector was the great equalizer: no matter what size connector an interface used (26, 20, 14, or 34), the Centronics-36 always connected it to the printer.  The signals themselves were mostly interchangeable, so you didn't worry about the size of the interface card's connector unless you were chaining it through another card, like The Bufferboard or Bufferlink.  The occasional handshaking issues could be resolved with a DIP switch or bodging the SEL wire to RDY/BUSY or something like that.

djc6 wrote:

So, bad news.  I socketed U10 - the Grappler+ continues to function with the original 74LS373 in the socket, but the NEWPRINT doesn't work when 74LS374 installed in the socket.  NEWPRINT manual states: 

Grappler+ Dip Switch Settings:

• "older" Grappler+ with 74LS373 chip in U10 need to have 1=Off

• "newer" Grappler+ card with 74LS37

There are a few differences...  the newer card does not have Q1, R2 or C7 but does have C8.  I also see some other minor traces here and there that look different.  It would be interesting to see schematics of both cards.

That looks like probably a 1 to 1 clone.  So yeah, probably same issue.

Happened to see this card on ebay:

s-l1600a.jpg

s-l1600b.jpg

s-l1600c.jpg

It's an older version of the card, which normally has a 74LS373. But this one has a 74LS374 on it, with bodges added, possibly all from the factory. The modifications make it plain what changes are needed to to use a '374 instead.

Cut the traces going to U3 pins 11, 12, and 13. This frees up a NOR gate, at the expense of removing the ability of the card to drive interrupts.

Cut the trace going from U4 pin 11 to U10 pin 11, the clock input of the '374.

Wire the first input of the NOR gate, U3 pin 11, to Q3 (slot pin 37).

Wire the second input of the NOR gate, U3 pin 12, to U4 pin 8 (or 12) - this is the line that goes low when the Grappler receives the "output" command by writing to $C080,y. Previously this was just being inverted by that NAND gate on U4.

Wire the output of the NOR gate, U3 pin 13, to U10 pin 11, so the '374 now has the proper clock input.

bradleyb wrote:

...This frees up a NOR gate, at the expense of removing the ability of the card to drive interrupts.

For what it's worth, Orange Micro removed that capability from the firmware some time in the mid-80s because it was causing problems and didn't offer any useful benefit.  Clones of the Grappler+ simply omitted the interrupt connection to prevent the problem from occurring.

Here's a 1984 discussion of "The Grappler Interrupt Problem" from Apple Thesaurus.  He's absolutely right: the two Grappler+ cards in our computer lab frequently caused users to lose data because it erased disks and crashed programs like Bank Street Writer and Magic Window if a printer error occurred.

Grappler interrupt problem.jpg

Bonus blunder: the Grappler+ circuit doesn't gate interrupts through the DMA IN-DMA OUT daisy chain, so the card doesn't comply with Apple's prescribed method of preventing interrupt conflicts.

So, removing the interrupt connections could only improve the Grappler's usefulness.

This is a much appreciated discussion on the Grappler parallel printer card.  I have the earlier Grappler+ with the soldered 74LS373 I acquired several years ago but never used it.

It is nice to know the downsides/upsides of the Grappler+ version I have.

I've always used the Perisoft PrinterLink parallel card with my Epson LX-810 in both my Apple II Rev 0 and Apple IIe's with no problems.  (and yes, it was 3-4 years before I had printing capability with my Rev 0.)

20230917_125804.jpg

Insane that a card this universal had so many flaws. Searching for more images, it seems like they made a dozen or more different revisions - did they ever get it right I wonder?

bradleyb wrote:

Insane that a card this universal had so many flaws. Searching for more images, it seems like they made a dozen or more different revisions - did they ever get it right I wonder?

In general they worked well enough for what most people wanted.  If you had one of the common dot matrix printers that was supported like the Epsons the Grappler worked just like the simple Apple Parallel Printer Card, with the addition of the built in graphcs dump software on the ROM.  In an Apple ][+ or a //e they usually didn't have any issues.  It was some clones and the IIgs that generally caused the flaws to manifest themselves.  I don't think they sold a lot of them to IIgs users because most of those people bought serial printers, mostly Imagewriter IIs.  However, some people who upgraded to a IIgs from a ][+ or //e might have not wanted to buy a new printer...  and that's when they would run into the problems people have been talking about.

The interrupt bug, people I guess just quickly learned to make sure they didn't have the printer off-line...

I never had one of the fancy printer cards back in the day, although I've got a couple of Grappler+ cards now, I have swapped out the EPROM for the LIRONGP image and I'm not actually using them as a printer card.

softwarejanitor wrote:

bradleyb wrote:

Insane that a card this universal had so many flaws. Searching for more images, it seems like they made a dozen or more different revisions - did they ever get it right I wonder?

In general they worked well enough for what most people wanted.  If you had one of the common dot matrix printers that was supported like the Epsons t

The Liron firmware is my main interest in these as well. Is LIRONGP the same thing as the SoftSP DIY from MFA2 or is that a separate project?

Here's what I've come up with for that, a Grappler+ clone anyone can build: https://github.com/btb/GraphicPlus

GraphicPlus.jpg

It should now be equivalent to the bug-fixed later versions.

If you only want to use it for SoftSP, you can eliminate about half the components.

I'm also working on a standalone, slot-independent version (essentially just like the real Liron) but that's going sorta slow.

To be honest I am not sure exactly whether LIRONGP and DIY SoftSP are the same project or separate.  The .BIN file that I have is LIRONGP.bin  The person on the FB Apple II Enthusiasts group that is working on that is Petar Puskarich from the DFW area.  After searching on that group and reading a bunch of posts I think that Petar started with the DIY SoftSP code but has tweaked it and has several different versions for other cards and bug fixes.

Ah yes, thanks for the link. I hadn't seen a few of those versions.

FYI v6 is out now at http://www.mfa2lab.com. Also yet another version for a simpler printer card: http://www.mfa2lab.com/diysp-v1-for-graphicard/

Could be design or layout flaws or manufacturing faults. I notice that the pins reworked with wires in the two pictures are not 100% the same, which points towards the latter.

Looks like they could have done a better job of it.  Also looks like two different people did those boards.  I've seen a lot better looking point to point wiring.

I would guess Orange Micro couldn't afford to scrap boards at the time those were made.  Interesting piece of history there.