Hi,
This is just development....once the software is done it will be very easy to update the Gotek drive. I will be selling adapter kits which include a PCB with sockets for the cables and stand offs to mount the Gotek into the sampler. The standard QD drive cable does need some minor modification at one end, and the Gotek casing needs some adjustment with a craft knife. So I plan to provide the drive, cable and adapter kit, along with a USB stick with the factory samples, so its “plug and play”.
I have checked the read support : I have connected 2 goteks together and made a QD file copy from one gotek to the other one : The CRCs in the destination gotek are fine.
one thing we didn't check : What is the read pulse width coming from the real QD drive ? Maybe the gotek pulse width are currently too short for the s10/s220.
AMSynths wrote: Wed Aug 28, 2019 9:37 am
I loaded the latest fw but verify error on save
I will try reading the files this evening and do a trace of the real QD read signal
Ok
First try to read the 3 above images.
Then if this doesn't work i think we should have a look around the write splice/(the WG "glitches") :
Can you redo this ? : viewtopic.php?f=2&t=773&start=120#p21092
With the WG signal and the Writedata signal ? I want to know what does the S10 at this place (before, under and after the WG "glitch"). These WG glitches appears to be placed right before the blocks headers. Pure speculation but maybe this is a part of the synchronisation to read the block since the MFM synchronisation "word" doesn't have any missing clock.
(Btw have a logic analyser able to capture this WG / Writedata flux ?)
AMSynths wrote: Tue Aug 27, 2019 10:14 pm
Hi,
This is just development....once the software is done it will be very easy to update the Gotek drive. I will be selling adapter kits which include a PCB with sockets for the cables and stand offs to mount the Gotek into the sampler. The standard QD drive cable does need some minor modification at one end, and the Gotek casing needs some adjustment with a craft knife. So I plan to provide the drive, cable and adapter kit, along with a USB stick with the factory samples, so its “plug and play”.
Latest fw loaded. 000 and 003 give I/O Error 3 - S-10 Has Broken Down, with 002 giving Illegal QD. There was no 001.
Save still gives Verify Error. I will do some traces of the real QD
The two glitches in the Write Gate appear to be consistent across all saves, now I have the scope at a high resolution.
Both glitches are around 50us in duration, and the first one does not interrupt the write data clocking but the second one does.
They appear around 220ms and 440ms after the write data is asserted
This is in fast save mode without any checking of the QD
The QD head moves back to the edge of the disk as soon as SAVE is pressed, write starts once the head is in position and continues to the end.
Then the read cycle commences with another pass of the disk surface, and then the verification error message
Let me know what further traces will be useful!
Looks like 4 data blocks then 1 short data block with the 2nd glitch in it, then 38 data blocks.
More tests with F1 and F2 saves shows the 1st glitch interrupts the write data stream as well. I was thinking that the early stages of the write data will be the 8kb of wave and performance parameters, followed by the sample data. But that doesn't make sense as the QD is only 64kB...
I'm looking forward to this!