USIfAC II:Convert a PC or USB stick to Amstrad HDD,access dsk's,and many more!

[ikonsgr]

Finally, after A LOT of trials & testing i manage to support the notorious C3 mode on a CPC 464 using ULIfAC:

You cannot view this attachment.

You cannot view this attachment.

In the end all it was needed was a minor modification of the board.
 The trick was to use a 16k block of the 512kb sram as a shadow video ram which is activated whenever the c3 mode is activated. Of course this reduces somehow the available extra RAM. In the end the only thing needed was to change common pin of the network resistor (that drives the 5 highest address bits of sram). These  previously used as pull down only, for low rom activation, but now it's connected to "rom enable" pin, thus it can be also used as pullup, for activating the highest 16k block of 512kb sram and use it a "shadow video ram"!
Anyway, i order a new batch of boards hopefully the last  and "official" one, and next time i mention ULIfAC would be in it's own thread! 

[GUNHED]

A well way to see if the RAM mode &C3 work is this: Start FutureOS and see if you can move the pointer. If it leaves a trace then mode &C3 isn't fully implemented.

Up to now only Revaldhino could do that for the CPC464. It would be awesome if the ULIfAC can do this too. 

[ikonsgr]

A well way to see if the RAM mode &C3 work is this: Start FutureOS and see if you can move the pointer. If it leaves a trace then mode &C3 isn't fully implemented.

Up to now only Revaldhino could do that for the CPC464. It would be awesome if the ULIfAC can do this too. 

 Unfortunately, Future os (as well as symbos) doesn't work right on CPC464. For some reason these OS  don't seem to like the masking of MREQ signal (in order to avoid write of the internal ram on the same address), although extra ram and dual mode (ram+rom board together) seems to work ok with everything else i've tried on cpc 464 (3 different mem diagnostic tests, dozens of games requiring 128k ram, low+upper roms along with 256k ram).
Also, i just tried batmanforever demo which i also think it needs c3mode, and it seems to run ok on CPC 464  (previously it couldn't pass the initial screen even when using cpc6128 fw and 256k extra ram but without C3 mode correctly implemented)

[GUNHED]

True this. FutureOS was initially made for CPC6128 and 6128plus (back the day when I started to develop it mode of the active CPC users went for these two machines. CPC664 was sold in small numbers anyway, and CPC464 was outdated somehow - that was the Zeitgeist of the 90ies. Around late 90ies to year 2000 on ebay they had lots of CPC6128 which went to the customer for around 10 Euros - However, it may have been different in other countries, guess in GB people loved the CPC464 more). 

And SymbOS, as far as I know it works with pretty much everything CPC. 

Anyway, my idea was just for testing &C3 mode (with proper relocation of the screen). But that can be done with other means too. 

Keep your great work going! I still love to work with my USIfACs, which sometime are used as CPC LAN. 

[d_kef]

A good test for C3 mode would be CP/M Plus.
It won't boot unless C3 mode works properly.
You will also need the 6128 ROM installed.

d_kef

[Prodatron]

Unfortunately, Future os (as well as symbos) doesn't work right on CPC464. For some reason these OS  don't seem to like the masking of MREQ signal (in order to avoid write of the internal ram on the same address)

SymbOS requires #C4-#C7, #C1 and #C2.
AFAIK some memory expansions don't work 100% in #C1 and/or #C2 mode on the 464. E.g. the SYMBiFACE II.
But e.g. X-Mem and SYMBiFACE 3 works fine with a 464.
It probably makes a difference, if a program is only reading from memory in #c1/#c2 mode or is writing as well.

[andycadley]

If the software cares about it, "classic" ram expansion should work on a Plus with activated ASIC as well. Somewhere here is a detailed thread about it.

No, a "classic" expansion doesn't work. You can avoid the particular use cases that cause conflicts but that isn't really the same thing. It's better to just accept that expansions like the Usifac just aren't really Plus compatible.

[Prodatron]

"Doesn't work" is a hard word. I am not sure, but wouldn't it already be ok, if the ASIC and the ram expansion isn't mapped in at the same time at #4000?
As Plus software should know about this, it could be a solution for many cases.

[andycadley]

"Doesn't work" is a hard word. I am not sure, but wouldn't it already be ok, if the ASIC and the ram expansion isn't mapped in at the same time at #4000?
As Plus software should know about this, it could be a solution for many cases.

Yeah, but that's working around the bug and it's actually difficult to do in practice because you find yourself really wanting to use the C1/C2/C3 paging arrangements (because you want all that extra RAM for sprite data) and they're not compatible. Trying to actually use 128K and not have banked RAM and the ASIC registers paged in just cripples the machine.

Worse than that, it confuses people who plan to buy the hardware. There is quite a lot of 128K Plus software that only works with RAM expansions that are Plus compatible, so it's better not to mislead people by suggesting these devices are.

[ikonsgr]

Worse than that, it confuses people who plan to buy the hardware. There is quite a lot of 128K Plus software that only works with RAM expansions that are Plus compatible, so it's better not to mislead people by suggesting these devices are.

According to leexus RAM expansion of ULIfAC works ok with plus machines, so maybe it's a bit unfair to suggest that these devices are NOT plus compatible   

[andycadley]

Worse than that, it confuses people who plan to buy the hardware. There is quite a lot of 128K Plus software that only works with RAM expansions that are Plus compatible, so it's better not to mislead people by suggesting these devices are.

According to leexus RAM expansion of ULIfAC works ok with plus machines, so maybe it's a bit unfair to suggest that these devices are NOT plus compatible   

The hardware has to be built to monitor the ASIC unlock sequence and fundamentally understand how RMR2 works or it just will not work properly. It's not something that can just happen without it being built to do so.

[ikonsgr]

Well, let me state a few things about PLUS machines:
- I NEVER owned an Amstrad PLUS so i never test any of my boards with it.
- I NEVER stated that any of my board is "PLUS certified", on the contrary, i always notify anyone told me that he intended to use board with an Amstrad PLUS, that i can't guarantee it will work properly.

BUT, what i DO know is that many people who have tried USIfAC II over the years (and lately ram/rom board and ULIfAC too) reported that for the "casual" needs of the VAST MAJORITY of Amstrad CPC users (which of course is to load and play games  ) , board CAN be used with Amstrad PLUS without problems.
 Having said that, i think that it would be wrong (and unfair) to declare board "PLUS incompatible" as it would be wrong and misleading to declare it "PLUS certified" too.
Bottom line:
- If you want to load fast and easy most games on Amstrad PLUS, USIfAC II/ULIfAC is a very nice and cheap solution!
- If you want to use any "plus specific" software (most probable custom OS like symbos/futureos or "plus only" games) ,then you should look for some other hardware which is "plus certified"

[Cwiiis]

Are there people using the usifac2 on the Plus with no problems whatsoever...? For me, it only works with the original Burning Rubber/BASIC cart, only with no other expansions and I still don't get perfectly stable behaviour. This seems to tally with what I've seen from other Plus owners...

I think the Usifac 2 is great and amazing value for money, but I'm not sure I'd ever recommend someone with a Plus machine get one. That's totally fine though, there isn't exactly a huge swathe of Plus owners, and until one of us lends you their machine, I don't think there's anything you could be reasonably expected to do about it!

[ikonsgr]

I just took a quick search on ebay for Amstrad plus machines.... 400+euros for this and.... ~1500euros for a PLUS with monitor   
Is this HUGE overpricing  for PLUS machines, something new (you can find a VAST SUPERIOR amiga 1200 for much less, as A1200 is something like Amstrad plus compared to Amiga 500 and amstrad cpc respectively....) , or always happened because of...."collectors value" due to  rarity? 
Don't get me wrong, but i don't think it's worth to invest so much money,(and of course extra time and effort) just to make USIfAC II/ULIfAC "plus certified".
After all ,how many are the PLUS owners that actually USE their machines (and not having it in a closet,a shelf or a showcase just for demonstration....)? Afaik, Amstrad PLUS sales were a few tens of thousands at most, were Amstrac CPC series, sold in millions! So would it be valid to suggest that actual Amstrad PLUS "active" owners today, would be ~1-3% of total Amstrad CPC users?

[andycadley]

The creation of the C4CPC (which at the time was dismissed by many as too niche) made the GX and Plus machines a lot more desirable. Especially as the wiki collection of CPR conversions shot up.

They're rare enough that they don't show up on eBay much these days, but I'm not convinced anyone is actually buying them at ridiculous prices like £1500 though.

[eto]

My gut feeling is, that although the only sold in the tens of thousands, a much higher percentage of Plus' survived and were not scrapped. So they are still more rare than a normal CPC but not as rare as one would expect from the original 1:100 ratio.  And: enthusiasts are more likely to have a Plus - and those enthusiasts are also more likely to buy expansions - sometimes just for the sake of owning the expansion. So over all, maybe the "market" for a Plus compatible expansions is not that small. 

I think there is a general issue with the Usifac II as we saw a few comments that it doesn't work with some Plus', with some 464s, if I am not mistaken someone even had issues with a 6128. As it's similar behavior in all cases I could imagine it's related to the same root cause, maybe signal timing/ROMDIS. So maybe(!) fixing it the 464s would also fix it for the Plus'. I could e.g. send my 464 to Greece for investigation. If it helps, great! If not, it was at least worth trying it. (I just want to get it back some day ;-) )

[eto]

im actually trying to get my hands on another 464 from mate atm, just to test the usifac on, and if it works, then do some oscilloscope probing over the two boards to try to see if theres a difference...

I wondered if the ROMDIS could sometimes be slightly too slow. If the ROM (in the 464 or the Cartridge) would put data on the bus in parallel to the Usifac that could explain the behavior we see. But it's just a wild guess as I don't have the equipment to test it. 

[ikonsgr]

I think there is a general issue with the Usifac II as we saw a few comments that it doesn't work with some Plus', with some 464s, if I am not mistaken someone even had issues with a 6128. As it's similar behavior in all cases I could imagine it's related to the same root cause, maybe signal timing/ROMDIS. So maybe(!) fixing it the 464s would also fix it for the Plus'. I could e.g. send my 464 to Greece for investigation. If it helps, great! If not, it was at least worth trying it. (I just want to get it back some day ;-) )

Indeed, and about 50% of all messages noting issues with USIfAC II in this thread, are from....WacKEDmaN! 
Funny thing, the dozens of messages regarding all "weird and exotic" problems, in the end was due to a bad power connector as WacKEDmaN himself finally noted!
In any case, i've sold ~350 boards so far, and all the individual cases regarding problems are counted on fingers of one hand. USIfAC II may be a very nice companion for the "old timer" Amstrad CPC, but it seems that some Amstrad cpc's have evolved over the years to... old grumpy ladies instead! 

[eto]

thing is ive tried original 464 rom...and the 6128 rom.. and both are doing the same thing!...
ive even changed EVERY chip except for the AY-3!

@eto do you have the schneider 464? im wondering if the schneiders extra filtering (all them toroidal coils) is the reason it works good for some but not others.... (but then why would it work on other amstrad branded 464s that dont have all that filtering the schneiders have!)

ill look into ROMDIS when i get my hands on the other 464 to test.. but id probably need a logic analyser to make sure the signal is timed right...(ill have to make one with an arduino or something!)

I also tried different ROMs. They don't make a difference. Which also makes sense, if it's related to the signal and not to the ROM. Or maybe the signal is not strong enough in some cases? Not sure if that could make sense.

Yes I do have Schneider 464s. And Amstrad Plus 6128.

[Prodatron]

USIfAC II USB mass storage driver added to SymbOS!

You cannot view this attachment.

Currently in "slow mode", and writing is still missing, but it already works great!
I am very happy to have this very important hardware now included as well!

[ikonsgr]

that power connector fixed it for like a day...then it was back to doing stupid shit....eg.. now i can ONLY use DSKs and |464 mode, i cannot use anything direct coz as soon as i enable |USB or |EN and try to reboot... it locks up before the rom line..... BUT the |464 reboot works... SMH
im really starting to get pissed off with this whole thing....
i feel ive been totally ripped off....coz there NO real help from anyone, including you the developer...
THERE IS AN ISSUE AND IM NOT THE ONLY ONE WITH IT.....
@eto well that rules out that idea!

I'm really sorry that you have all these problems,but i really can't understand what i, (or anyone else) can do to help (apart from suggesting a few hints or tips, which i already did many times).
Now,apart from you,there were two,maybe three other reporting similar issues.
On the other hand,me and more than 300+ other people, don't seem to encounter ANY issue at all!
So, contrary to your claim that there might be some "general" systemic issue with the board,the fact that more than 99% of people have NO issues, most likely suggests that RARELY, issues with SPECIFIC Amstrad cpc machines might occur, something quite difficult to diagnose from distance, and FAR from claiming there is an issue with the board itself...   

[ikonsgr]

USIfAC II USB mass storage driver added to SymbOS!

It was about time....   
Thank you very much for your time & effort to add USIfAC II support to symbos, much appreciated!

[ikonsgr]

@WacKEDmaN, at some time in the future (probably later this year),i will order a new batch of USIfAC II boards. These will have some design improvements (fine tuning of some traces, use wider traces), which hopefully will resolve the issues with some "grumpy" CPC boards 

[ikonsgr]

I've been looking at Amstrad CPC expansion port and i saw there is an "EXP" signal that can be used by a device to signal it's presence.
  Since i never used that signal  (e.g. activated by driving it to gnd), with any of my boards, does anyone knows if there is any practical use of this signal (for example program that might use it to detect the presence of an external device)?
 And mostly, if it is used by Amstrad CPC itself, and when a device is connected to expansion port, but EXP is not activated, this can affect the behavior of Amstrad.

[andycadley]

I've been looking at Amstrad CPC expansion port and i saw there is an "EXP" signal that can be used by a device to signal it's presence.
  Since i never used that signal  (e.g. activated by driving it to gnd), with any of my boards, does anyone knows if there is any practical use of this signal (for example program that might use it to detect the presence of an external device)?
 And mostly, if it is used by Amstrad CPC itself, and when a device is connected to expansion port, but EXP is not activated, this can affect the behavior of Amstrad.

from what i understand it isnt called anywhere in the 464s firmware.. its purely used as a GPIO...

Not sure about the old school CPC, but it is used on the Plus. From the Arnold V documentation:

"On the Plus, it depends on /EXP. If /EXP is low, page 1 will be visible at &c000-&ffff, otherwise page 3 will be visible and CPM will be auto booted."