The defective VL works properly in the Timex FDD3000 disk drive.
In the original, WD1770 was used there.
Perhaps the faulty VL have a problem with stabilizing parameters when working with shorter access times ?
Maybe your FAFO works a little slower than the original FORMAT ?

I think I found the problem:
The Seek command don't reliable these faulty chips. Sometimes execute Restore instead of Seek (only one bit difference in command code!)
My FAFO only using Step In/Out commands, when Seek needed it is done by steps cycle.
EXDOS FORMAT use DISKIO calls which is use a general head positioning routine. It is don't care about if only one track step needed, always using the Seek command.

Probably that Timex drives also only use Step commands. There is disassembly of their ROM?

If my theory right then the faulty chips also will works with the Hungarian SpeccyDOS disk interface (for a ZX Spectrum), least with 4S ROM (it is developed for a old slow drives, where the 1772 step rates too fast. This ROM using Step cycles with a CPU delays for simulate the different step rates from 2 to 30ms).

Are you sure those 813 V L657S are working ?
Because I can to exchange them.

What happen finaly with your chips?

After some investigation now it is sure: the Seek command are the problematic in the fauilty VL1772 chips. Sometimes execute Restore instead of Seek.

To the DRVTEST (will be released soon) added a verification for a Seek commands, and report error if the head went to Track 0 when it is not expected. Then easy to see if the chip faulty.
At a manual seeks:
[ Guests cannot view attachments ]

During a test cycle:
[ Guests cannot view attachments ]

Test ended with tons of errors:
[ Guests cannot view attachments ]

Bruce built in a alternative seek mode to the program, which is use cycle of step command, it is called Stepseek mode.
In this mode the faulty chip works correctly:
[ Guests cannot view attachments ]

This mode will be built in to the EXDOS 3, then it is can work with these faulty chips.

Meanwhile I found another bad chip, got it from a Commodore 1571 drive (where probably no anyone using the MFM mode in the last 30 years ). It is have a 1986 week 33 datacode.
[ Guests cannot view attachments ]

Now I ordered cheap 5x pack WD1770 from a Aliexpress Let's see these are working or not

Finally, the seller replaced the faulty chips with other from 813V L657S series.
This series works fine. I checked all the pieces.

Great news!

For a change, IMO, they makes too hot. I did not check to what temperature.

Yes, it is a know the VL are more hot than WD.

It is possible that overheating will cause some errors, but this should be carefully checked.

But this is only make a problem when overclocking to 16MHz for using a HD disks.
For example there is Pertik Laci's Turbo EXDOS card, he are used a fan on the card. When VL1772 overclocked.
Few years ago it is replaced by WD1772, then fan not required even using a HD disks.
(The Turbo EXDOS 1.3 ROM left the last used clock on the chip. In a EXDOS 3 only at R/W operation used the turbo clock, then switch back to the normal 8MHz, for a less stress on the controller chip, and also solve the step rate problem.)

On a normal DD usage don't problem with VL. (with a good chips )
I read same experiences at Atari ST forums.

8921 date coded also looks good, tested two of them.
[ Guests cannot view attachments ]
[ Guests cannot view attachments ]

This version of VL1772 works best.

Nice young one Manufactured in 1992.
It is one piece find? Or still available in somewhere?

Now I ordered cheap 5x pack WD1770 from a Aliexpress Let's see these are working or not

Arrived
Extreme slow shipping, more than 1 month until the seller dispatched it... plus the usual shipping time from China.
But it was worth the wait!

These are really NOS chips , not a relabeled/refurbished/fake what I expected
All working normally.
[ Guests cannot view attachments ] [ Guests cannot view attachments ]
[ Guests cannot view attachments ] [ Guests cannot view attachments ]

I have made a new video.

A FlexiBridge PCB from Pear(Poland) has arrived to my hands and I can't resist to connect it to my Enterprise.

It is a buffered expansion card that allows to connect several expansions without problems. The original idea was to connect a 64 leads flat cable to it, and then, at the other tail connector, a train of expansions that would link between them to the cable in a row.
It is flexible also in its conception, because both classic or actual expansions can be connected to it. But only +9v are passed to the cable, so every expansion has to create its own +5v.

On the Mészáros' Bus Expander(Wilco's Bus Expander is based on it), a total of 27 signal lines where buffered with 2  74LS244 and 2 74LS367, while on the FlexiBridge from Pear the buffered signals are also 27, but made with 4 74LS245. I have to say that on the Enterprise every expansion has to provide buffers(74LS245) for that bus, so at the end a total of 35 lines are protected.

The name "Bridge" comes from Enterprise Computers, and it was an 1-slot-only bus expander to connect the EXDOS interface, already designed to be plugged on a never released original Bus Expander. On that Expansion Motherboard, 4 74LS244 would be used, and a fifth 74LS245 for the Data Bus direction, making a total of 40 protected lines.

Just for the test, I have made a "RigidBridge" with one of my spare Bus Expander PCBs, bypassing the buffer chips on it. The only precaution to connect the two boards is to jump one pin on the not used 4A/B contacts, because the FlexiBridge  uses only 64 leads to make a more standard flat cable, as advised on the original IS papers.

You can see on the video that the SymbiFace3 and the EXDOS+HxC work perfect connected to it.
Sorry for the clipping sound on the Webradio app, on my house the Wifi station is far from my "laboratory".



Thanks Pear for the information about your great FlexiBridge.

https://youtu.be/LU2Re95Zcwg