So I've been fooling around with my fpga on a peripheral card doing reads and writes to the bus, trying to get the timing just right, but it's such a pain to power off the apple ii, remove the peripheral card, plug in the usb cord to the fpga, flash the new vhdl update, unplug usb cord, put the card back in the slot and power up the apple II again, ad nauseum. To simplify things I decided to use an external usb power with an on/off switch and plug the usb cord into that and leave the cord plugged into the fpga while it's in the apple ii slot. I have to make sure to have either the usb port or the apple ii off when the other's on, or else there'll be a lot of smoke.

As I finished flashing an update to the fpga card, I happened to glance at my oscillosope which was connected to my fpga and I saw it generating a wave form, a 1mhz signal that can only be generated if phase0 was actively feeding it, but the apple II was off! I was scratching my head... Surely the motherboard clock can't still be working with the apple ii powered off, can it? So I unplugged the power cord from the apple ii and... the waveform was still there! At that moment I realized that the USB powering the fpga was also feeding power to the apple ii via the 5v line, and that must be powering the clock. It must also be powering the ROM (and maybe RAM?) because I was able to read from it just fine.

I wanted to ask you guys out there in the know... is this okay or am I at risk of damaging something? Assuming it's okay, can I rely on debugging it this way with the apple ii power off, or will the timing be different with the apple II on thus making any debugging while off pointless?

Many thanks in advance.

Anthony

On 8/31/2022 6:08 PM, Anthony Ortiz wrote:
> So I've been fooling around with my fpga on a peripheral card doing reads and writes to the bus, trying to get the timing just right, but it's such a pain to power off the apple ii, remove the peripheral card, plug in the usb cord to the fpga, flash the new vhdl update, unplug usb cord, put the card back in the slot and power up the apple II again, ad nauseum. To simplify things I decided to use an external usb power with an on/off switch and plug the usb cord into that and leave the cord plugged into the fpga while it's in the apple ii slot. I have to make sure to have either the usb port or the apple ii off when the other's on, or else there'll be a lot of smoke.
>
> As I finished flashing an update to the fpga card, I happened to glance at my oscillosope which was connected to my fpga and I saw it generating a wave form, a 1mhz signal that can only be generated if phase0 was actively feeding it, but the apple II was off! I was scratching my head... Surely the motherboard clock can't still be working with the apple ii powered off, can it? So I unplugged the power cord from the apple ii and... the waveform was still there! At that moment I realized that the USB powering the fpga was also feeding power to the apple ii via the 5v line, and that must be powering the clock. It must also be powering the ROM (and maybe RAM?) because I was able to read from it just fine.
>
> I wanted to ask you guys out there in the know... is this okay or am I at risk of damaging something? Assuming it's okay, can I rely on debugging it this way with the apple ii power off, or will the timing be different with the apple II on thus making any debugging while off pointless?
>
> Many thanks in advance.
>
> Anthony

I don't know if it is okay or not but it got me curious so I got out my
old scope and tried to replicate what you are doing only with my Carte
Blanche II card.

I first connected my probe to phase0 pin of the Apple IIgs slot and to
ground on the slot. I then removed my USB connection to the Carte
Blanche II.

Powered up the IIgs and got a wave with some ringing but less than what
I see in your pictures (my scope is an old analog scope).

Turned off the IIgs. No signal on scope.
Plugged in USB. Still no signal on scope.
Tried programming the Carte Blanche II. It won't program without the
IIgs power on and still no signal on scope.

For what it is worth. I leave my USB connected all the time and program
the FPGA or Flash the non-volatile memory over and over without turning
anything on/off.

Is your USB connection to a USB chip on your board or is it a USB to
JTAG cable?

Charlie

Anthony Ortiz <anthonypaulo@gmail.com> wrote:
> So I've been fooling around with my fpga on a peripheral card doing reads
> and writes to the bus, trying to get the timing just right, but it's such
> a pain to power off the apple ii, remove the peripheral card, plug in the
> usb cord to the fpga, flash the new vhdl update, unplug usb cord, put the
> card back in the slot and power up the apple II again, ad nauseum. To
> simplify things I decided to use an external usb power with an on/off
> switch and plug the usb cord into that and leave the cord plugged into
> the fpga while it's in the apple ii slot. I have to make sure to have
> either the usb port or the apple ii off when the other's on, or else
> there'll be a lot of smoke.
>
> As I finished flashing an update to the fpga card, I happened to glance
> at my oscillosope which was connected to my fpga and I saw it generating
> a wave form, a 1mhz signal that can only be generated if phase0 was
> actively feeding it, but the apple II was off! I was scratching my
> head... Surely the motherboard clock can't still be working with the
> apple ii powered off, can it? So I unplugged the power cord from the
> apple ii and... the waveform was still there! At that moment I realized
> that the USB powering the fpga was also feeding power to the apple ii via
> the 5v line, and that must be powering the clock. It must also be
> powering the ROM (and maybe RAM?) because I was able to read from it just fine.
>
> I wanted to ask you guys out there in the know... is this okay or am I at
> risk of damaging something? Assuming it's okay, can I rely on debugging
> it this way with the apple ii power off, or will the timing be different
> with the apple II on thus making any debugging while off pointless?
>
> Many thanks in advance.
>
> Anthony
>

It can be dangerous to connect two power supplies together without
isolation precautions.

Each supply feeding the card can be isolated by connecting a diode in
series with the +5v lines on the card.

To minimize voltage drop, Schottky diodes should be used, connected like an
OR gate, to prevent backfeeding one supply from the other.

More complex FET-based approaches can be used to minimize voltage drops,
but diodes should be sufficient to allow you to operate safely.

--
-michael - NadaNet 3.1 and AppleCrate II: http://michaeljmahon.com

> Is your USB connection to a USB chip on your board or is it a USB to
> JTAG cable?
>
> Charlie

It's a JTAG cable that supplies power to the fpga. It seems the Carte Blanche isn't powered this way.

> It can be dangerous to connect two power supplies together without
> isolation precautions.
>
> Each supply feeding the card can be isolated by connecting a diode in
> series with the +5v lines on the card.
>
> To minimize voltage drop, Schottky diodes should be used, connected like an
> OR gate, to prevent backfeeding one supply from the other.
>
> More complex FET-based approaches can be used to minimize voltage drops,
> but diodes should be sufficient to allow you to operate safely.
>
> --
> -michael - NadaNet 3.1 and AppleCrate II: http://michaeljmahon.com

Thanks Michael, I'll keep that in mind!

On 9/2/2022 9:50 PM, Anthony Ortiz wrote:
>> Is your USB connection to a USB chip on your board or is it a USB to
>> JTAG cable?
>>
>> Charlie
>
> It's a JTAG cable that supplies power to the fpga. It seems the Carte Blanche isn't powered this way.

Right, my Carte Blanche II is powered by the Apple.

Possibly you could have your peripheral board work that way (with
appropriate isolation as Michael J. Mahon mentioned). I can't imagine
it being much fun having to take out the board every time you have to
program it. Also you can program the FPGA directly for a quick check of
a change in code without overwriting the the flash memory chip.

Charlie