There is quite a bit about why it doesn't work here :

http://eab.abime.net/showthread.php?...id=4&styleid=1

So it is about DMA bandwidth 500 kpbs is the max for paula.

I don't really understand.

When it comes to a usb flooppy emulator, there is no obstacles concerning drive's rpm, hall sensors,timing of motor step etc.
I kind of "prove" this my self, by testing the 68000/14Mhz hack on my A500. Using a real drive didn't work because of the double speed of "E" signal which drives the I/O CIA chips, that affect all the electro mechanical signals of the floppy. But when i tried using gotek, it worked fine!
So what i'm asking is, if the way to achieve double speed on a usb floppy emulator ONLY (and not a real drive where much more issues emerge), is changing some settings of the gotek firmware, or there is an "unbeaten" limit of the max speed paula can encode/decode mfm data stream!

afaik the reason why FD HD drives, on amigas, only transfered with half speed (500kbps) (as compared to the native speeds of a HD drive) was because the chips couldn't transfer data at those higher transfer rates (1000kbps).

So this might be the answer to your question.
This can ofc be that even with a higher clockrate the transferrates remain the same (this is just speculation).

Sounds like extra speed ain't ever gonna be an option. Well at least it works, that's the important thing. THough yes, it would be great if this was as fast as similiar drives for the Atari 8-bit line of computers. My SDrive Nuxx is blazing fast! :)

Hi Ikonsgr, et al.
With the 14MHz hack, the timing loops on the track stepping for the floppy are running twice as fast. Violates the minimum track stepping pulse width requirement of a "real" floppy drive.

Gotek effectively has a track to track access time of ZERO ms, so doesn't have the same requirements.

Check your A500 schematic.
Paula chip is clocked by CCK and CCKQ (staggered 3.5MHz clocks) , and *not* by the 'E' clock.
This is why it still works with the 14MHz hack active.

It also means it you are stuck with the 250kbps transfer rate (or 125kbps in the so called faulty chips).
At the moment, that means DD (880kb) disks only.

I would imagine it is possible to make the firmware in the Gotek boards emulate the half speed transfer rate of the Chinon FZ-357a drives, to give HD floppy compatibility.

I guess we will know for sure when H.M. releases his source code :)

Cheers,
Red

hopefully source code will be released soon, was just looking at the schematics and it would be nice to add more buttons and add code support for them :)

Has anyone tried Soccer Kid ECS? I've tried all ADFs from TOSEC pack and it's "no go". First disk (boot) loads fine (tried Fairlight, Ministry ATX versions), then it asks for first disk (second of four), I press the button on gotek, it changes but game doesn't seem to notice I changed the disk, it doesn't even try to read it (floppy led doesn't blink).

I've tried ADFs on WinUAE and they work fine. I thought it may be because my A500 is not quite like standard one. It's Rev8 MoBo that I modded to A500+, added memory chips to empty spots and also inserted standard 512kB memory expansion, so it has 1.5 MB chip ram, also I swapped CPUs so it has 68010 instead of 68000. I configured WinUAE like this but game loaded fine so I'm not sure if it's gotek or Amiga configuration. Kickstart is 1.3.

I guess it could be cause in WinUAE the 68000 and 68010 has same timings ...
in a real world the 68010 runs a little faster, so 68010 is probably still not emulated perfect !

p.s.
In practice, the overall speed gain compared to the 68000 at the same frequency was less than 10%. (according to wiki)

... but in emulation this is still not true ?! (as shown on pic. below)

You pressed fire after changing the disk?

Yup, fire, mouse button, enter, space, escape everything I could think of :). In WinUAE it automatically continues when it recognizes disk change.

Producing sound from Gotek on track stepping
 

I have another schematic for a buzzer for those people that want sound from their goteks, it has a very good sound level and is similar to the one that Robinsonb5 posted earlier but uses fewer components...

http://kipper2k.com/amiga/gotek/gotekbuzzer.jpg

1 x BC547 or similar bipolar transistor
1 x 1k resistor

and the buzzer (i salvaged this one off an old motherboard... http://www.jameco.com/Jameco/Products/ProdDS/138740.pdf

Nice! However, be aware that a lot of external drive enclosures don't bother to connect the +12v pin, so check that if you want to use a Gotek drive in an external case.

Interesting - is the storage time of the BC547 not a problem there? (The step signal's nominally high, and goes low very briefly when the head steps.)

Nice feature. I will test it at weekend.

This particular buzzer is rated at 5v so you could actually power it from the 5v rail, The reason i tried the BC547 is because i had it handy, it switches on/off fast enough to allow the buzzer to get a good pulse of power to beep. I figure that attaching it to the 12v rail is not overdriving it too much considering the on time

I did try the flat piezo buzzer with this ciruit and it didn't want to play. there are lots of magnetic type buzzers listed here... http://www.digikey.ca/product-search...=0&pageSize=25 I am ordering a couple of different ones to test them out


For those people who are intested in how this circuit works, check out this link...

http://www.allaboutcircuits.com/vol_3/chpt_4/2.html

I'd agree, except the duration of the step pulse is the *off* time. (Step is an active low signal.)

Yeah that's what I found - and I think the reason is the storage time in the BJT, exacerbated by the tiny current through a piezo transducer and its capacitive nature.

Mine has kick1.3. I'll try to write these ADFs to floppies and try with original chinon to see if problem is in configuration or in gotek.

Funny thing is that I tried FTL+trainers version first, it didn't work. Then I tried Ministry version and that one worked, but only first time. After that I couldn't get any version to recognize floppy no2.