Internal drive select

[friijy2001]

Shouldn't the pull up resistors connect to 5V?

A slightly tweaked diagram showing it connected different with the resistor, does this look better?

image.jpg (1.05 MiB) Viewed 2956 times

[friijy2001]

Found this video on youtube

Although it is for an Amiga, it looks like the guy has it working with just a two way switch on pin 10 with no pull-up resistor.

So in theory I should be able to plug both drives in including the power and just put a switch on pin 10 to select the drive, I could put a switch on the motor too as this would not do any harm.

I'll order some parts and have a play

[tzok]

Drives are expected to have pull-up resistors on the Drive Select lines, so there's really no need for this resistor. Yet it is safer to put 4,7k pull-up resistors on the drive sides of the switch (1 resistor for each drive). They should not be switched.
Motor-On line is common for all 4 drives according to the Shugart standard (PC drives haves a separate motor-on line for each of 2 drives). It's up to the drive's logic whether all motors work together or the motor-on signal is internally gated with the drive select signal.

[exxos]

Just to clarify.. a MPF920..There are no pullups on the drive. So you need a pull up on it. Once drive select is "broken" with switch there isn't even ST connect anymore (but no pullup that side either)

Capture.PNG (1.21 MiB) Viewed 2921 times

DS0 is driven by a open collector inverter on the STF.. It goes to the keyboard, but I just measure and no pull up there either.

In fact the I may add resistor on future remake boards as its pretty bad not to have one. Plus with no pull up on the drive, as I said , the line needs pull up on it drive side. But in order to avoid floating pins, the ST side should really have one as well. Though that is just my preference in not rolling the dice with floating pins.

[tzok]

Controller IC in that drive has an internal pull-up of about 16kΩ. The Drive Select line on the host/controller side is supposed to be open collector. Atari is a little "noisy" so it is better to use additional pull-ups, but both options should work.

Code: Select all

SIMPLE
                                __________ PIN 10
PIN 10 ________________________/           FDD
ATARI                           __________ PIN 10
                                           GOTEK 

WITH PULL-UPS
                         4k7
                   +5V -/\/\/--------,
                                _____|_____ PIN 10
PIN 10 ________________________/            FDD
ATARI                           __________ PIN 10
                         4k7         |      GOTEK 
                   +5V -/\/\/--------'

[exxos]

Controller IC in that drive has an internal pull-up of about 16kΩ.

Where you find that info ?

[tzok]

I've measured it on the actual drive. But it is logical. The Specification says that the Drive Select line (among other lines) is an open collector type, so the drive has to deal with it.

One more thing - you don't need any adapters. All you need is a switch shorting and opening a jumper/shunt on the drive itself. If you remove DS0 jumper from the drive, it will never become active. All you need is a DPDT switch and 2 pairs of wires with 2-pin Dupont connectors, which you install on drives in place of the original DS0 jumpers.

Code: Select all

DS0   o–––––––––––––––––––––––,
GOTEK o––––––––––––––––––––––,|
                             ||
                         o|o–'|
DS0   o––––––––––––––––––o|o––'
FDD   o––––––––––––––––––o|o

[exxos]

I've measured it on the actual drive. But it is logical.

That is madness.. you cant do that, resistors in IC's are emulated with silicon, there are not actual resistors. Unless you have actual datasheet saying what the effective internal pull up is... Pullups are normally more like 1Mohm in IC's which are generally pretty ineffective.

[tzok]

Well it works without the external pull-up, so where is the problem? It was working like that for the last 30 years...
Shugart' specification says that the DS line is OC, so there should be no pull-ups on the host controller side.
The drive itself also doesn't have external pull-ups, but when powered, keeps this line high. Some other drives (including GOTEK) have discrete (resitor) pull-ups.

I've powered the drive, and measured the current sink into pin 10 - it is 1.6mA... so the "emulated" pull-up is closer to 3kΩ. I hope this clears your doubts.

[exxos]

Well it works without the external pull-up, so where is the problem? It was working like that for the last 30 years...

"works" is a very debatable topic.. A combustion engine badly blowing out smoke for 30 years can "work".. but should you run it like that ? This is again why I just want to define better standards and practices on here. Such things are debated for years on other forums, but this does nothing but confuse people and generally they end up with learning bad practices. I want better standards and user experience on this forum. Just do things properly, Atari took many shortcuts and look how all that went for the past 40+ years.

I've powered the drive, and measured the current sink into pin 10 - it is 1.6mA... so the "emulated" pull-up is closer to 3kΩ. I hope this clears your doubts.

If that is 3K then that would be fine, I am happy with that. But I would prefer to see proper specification on what the actual datasheet says, but these unlikely to be available anymore. A lot of electronics "works on paper" to specification, but in the real world, things are very much different. Internal pullups on IC's are generally pretty bad to the point I never even depend on them anymore. This has got worse over the years with low power devices.

But with so many drives out there, you cannot assume the same value anyway. Some could well be 1meg for all we know. When values start getting higher your back to "works on paper". Throw in a little noise or longer cables and the thing malfunctions. It is far easier just to use a pullup and just not worry about this stuff at all. Now I am aware of this , I am adding a pullup on figure H5 revisions. I just don't want to leave anything to chance. But thats is just me.