Gents,
just a question regarding a floppy controller from IBS sold as AP53.
Does anyone know the history behind while the raw pcb is marked as AP41 ?
Picture Source: https://www.facebook.com/photo/?fbid=3150692075074619&set=pcb.10162824251778677
Thanks...
Perhaps that is an internal part number for just the PCB?
That card is SUPER interesting though. Unlike a lot of MFM floppy controller cards it doesn't appear to use any chips that aren't fairly readily available with the exception of the 28L22N, which I think is a bipolar PROM. That could be replaced with another EPROM. It would be interesting for someone to clone this card in a modern buildable version.
Do you have the software to use it? Is there support for CP/M or reading other formats like 360K PC floppies or support for DOS 3.3 or ProDOS formatted floppies on higher density medai?
Fascinating!
This is a clone of the apple2 DISK II controller with a stub to drive alternatively PC drives but use them like DISK II drives minus ability to half and quarter step in GCR mode. No MFM here. Replacing the bipolar PROM with an EPROM would lead to amateur made contemporary DISK II controller cards with their huge incompatilities due to slow EPROMs ;)
So it controls PC style floppy drives but only in GCR and can't read PC format floppies? Still interesting. Can it control double sided and 80 track drives or just 40 track single sided 160k ?
There are quite a few contemporary designed floppy controllers which use EPROMs instead of BiPolar PROMs that work OK, but you do need to use fast enough EPROMs, the -450 versions of course probably won't work.
Now you started explaining things to me ;) How fast is the D2732C PROM on this card that already replaces one BPROM?
The 2732 on the card probably replaces the code used for the boot. The one that would be more speed critical is the one that is used for the Woz state machine.
I can't find a datasheet on that PROM, but the access time on typical UV erasable 2732 is between 200ns and 450ns. I've seen the 200ns parts used without any major issues on modern floppy controller designs for both the P5 and P6. I could certainly see possibilities of the 450ns parts being too slow for use with the P6 (state machine) but they'd probably work OK for the P5 (boot).
And do you think supposedly German engineers were that stupid not to replace the "more critical" BPROM, too? That's why I asked that question ;) The 250nS 2716/32 EPROMs were widely available in 1984 already.
Back in 1984 BiPolar PROMs were readily available and cheaper than EPROMs. It would make sense they would use that when it was cheaper. For the 2732 there weren't BiPolar PROMs that stored 4KB.
The datasheet for the TBP 24/28 series is here. The typical access time for the 28L22 (a low-power, slower part) is 45 ns.
Had the BPROMs been cheaper at the time of production of the pictured card why did the engineers of that card use a 2732 instead of the cheaper and more compact BPROM ;)? Obviously only a small amount of that 2732 memory is used in a DISK II controller clone. The rest is wasted.
Assuming that it only has a Disk ][ style 256 byte boot code in it, maybe. A lot of other disk controllers like the LiRON which support other than just Disk ][ drives have much larger firmwares. A bunch of vintage period disk controllers like the MicroSci, Franklin and Rana Systems ones use an EPROM also.
I'm guessing if this IBS card used a 2732 instead of a 2716 it was because they needed 4KB instead of 2KB. Do you know the IBS only uses 256 bytes? If they needed more than 256 bytes, most of the BiPolar PROMs were only 128 or 256 bytes so that wouldn't have worked for them and forced them to use something else. I think there might have been some BiPolar PROMs that were 512 bytes, but I don't think they were very common if there were. At least I can't recall anything for an Apple II using one back in the day.
The PROM I was looking for the datasheet on was the NEC D2732C which is I believe a write once version of the common 2732 EPROM. I'd suspect access times to be more typical of that style of chip which are usually 150ns or slower. 200 or 250 are pretty common.
Erphi like 2 .jpg
Erphi like 3.jpg
duplicate pics deleted
Thanks Wayne! Those look like different layouts of the same basic design.
Well their is some action ongoing at a German Forum to rebuild the Ehring FDC-4 Disk Controller which seems to be compatible with the mentioned AP53 card:
https://forum.classic-computing.de/forum/index.php?thread/36150-ehring-fdc-4-disk-controller-nachbau-tbp28l22-programmieren/&pageNo=1
The actual discussion is around the similar and/or same interface layout found at some clone cards. This was the reason to raise the above question, because we didn't find any background information yet, describing while the PCB is marked as "AP41" while the whole card itself was sold as "AP53".
Personally I don't believe that the pcb was "wrongly?!" marked this way, because other pcb cards from IBS are marked correctly.
@softwarejanitor: to be honest... I don't have any software nor any additional information, yet.
It's entirely possible it was just a mistake or even that they originally intended it to by AP41 and then changed their mind later.
I hope you are able to find the software and docs for this card or one of the compatible clones of it.
I already told you it was a DISK II controller with no special software for it. Presently PC floppy drives cost more than DISK II drives and with such controller PC drives would not be able to half and quarter step so there is no advantage in using such controller. I think I own a couple of these controllers. Knowng the humble abilities of the German cloner I doubt the result would be rational but JLCPCB and similar allow such projects to exist nowadays.
The first pic from post #13 seems to be an ERPHI-clone, the second one is an original Ehring-FDC4.
The pic in post #1 is an IBS-AP53, which is basicly an FDC4 clone.
For what reason ever, the circuit board of the AP53 was marked as AP41!
For all of theese a special patch software is needed, to access higher capacity and double sided operation. This may be the same for the FDC4 and AP53, they are nearly equal, but the software for the ERPHI is definitely different.
The FDC4 is able to drive 2 shugart or 2 APPLE drives, but not at the same time. As far as i know, a mix is possible, but it's not allowed to connect an APPLE and a shugart drive as drive #1/2 at the same time
Regards
How higher will the capacity be, and can you post that software or link to such software for non-erphi controlers? Not talking about 40 tracks general DOS 3.3 patches.
The zip contains an ADTPro disk image, the disk is not bootable, load a standard DOS 3.3 disk and change disks, or run PATCH165 from another drive.
Patches are available for DOS3.3 , UCSD , CP/M.
Info for the original FDC4:
(Hope i remember them correct ...!?)
SW1: On/Off - Drive 1: Shugart/APPLE
SW2: On/Off - Drive 2: Shugart/APPLE
SW3: not used
SW4: On - automatic doublesided operation
The Shugart bus supports drives with
2 x 40 tracks (360k)
2 x 80 tracks (720k)
Note that not all doublesided 5,25" drives are working with the APPLE, and they are hard to find today. Unfortunately i don't have a list of working drives. ;-(
The Shugart bus supports drives with
2 x 40 tracks (320k)
2 x 80 tracks (640k)
Some more info can be found here:
https://www.applefritter.com/appleii-box/H085_AppleIIFDC4.htm
Good information from the late speedyG. My German is weak but the article appears like it kind of explains why the EPROM is that big. It apparently does have more than just the 256 byte boot code most clone controllers usually copy more or less directly from Apple's Disk ][ Controller Card.
The 320k and 640k floppy options with patched OSes and other OS support is really what makes this card interesting. If it were only a limited Disk ][ level controller that worked with 35/40 track single sided Shugard bus drives then it wouldn't be all that useful.
Of course these days all floppies are more or less for nostalgia, but that's OK.
I'm wondering if this card would work with something like a GoTek to do 320k and 640k images? That's really probably more of a fun-to-do thing than practical, but do much retro stuff is that anyway.
It is big because it was cheaper then a PROM. Look at the 8 address lines connected to the EPROM in the schematic in the article -- only 256 bytes of it were used.
transwarp2 wrote:
...Knowng the humble abilities of the German cloner I doubt the result would be rational but JLCPCB and similar allow such projects to exist nowadays.
Dear transwarp2,
Thank you for the compliment. Building my first smart home with an Apple II and an FDC-4 in the early 1980s paved the way for my professional career. Thanks to it, I can support my family without any worries, and I am delighted that almost 40 years later, so many people are still working on this topic — for example, on this discussion board.
Bildschirmfoto 2025-06-12 um 18.38.53.png
Having a circuit board made used to be a very laborious and expensive process. There was usually a minimum order of 100 pieces. That's why I opted for the wire-wrapping technique. The photo shows the Smart Home relay output stages being switched by the Apple II.
Bildschirmfoto 2025-06-12 um 18.40.29.png
Today, conditions are like paradise: It's no longer unusual to buy a 3D printer or laser cutter in a supermarket, for example, or to have high-quality circuit boards produced cheaply by JLC. The freely accessible Asimov FTP collection, which contains information on Apple II computers from around the world, provides unlimited opportunities to supplement your knowledge of Apple II computers. Hopefully it will still be around in 50 years' time for future generations.
In any case, I'm really looking forward to revisiting this topic, rebuilding the long-forgotten hardware for fun and archiving it in a freely accessible digital format. You are welcome to help with this. Don't get frustrated!
Best regards, Joerg