I-top dac and software Help?

[carmangary]

LOL. You are describing analog and digital bevahior, not voltage. 5v = 5v = 5v = 5v.

An example; my lasers have analog inputs from 0-5v. But I can hook them to TTL outputs and they will blank just fine. My laser doesn't care whether the voltage is "analog" or "digital" because there is no such thing. It just wants something in the range of 0-5v and it will do what it is supposed to based on it.

I think you know this, though.

[buffo]

Ok... You're right. We're both on the same page then, arguing over semantics.

I agree it was a poor choice of words to say "analog voltage" (or ground)verses "digital voltage". I should have just said what I was thinking, which is the floating ground issue. (Well, that, and the fundamental differences between AC and DC. And though those terms actually refer to current, not voltage, it's still convenient to think in such terms.)

In my defence, I should point out that the grounds *are* labeled that way on the pinout diagram. (Analog ground and Digital ground). But it appears that they're all common. (I also confess that I never once checked them with a voltmeter to see if they were floating or not. My bad.)

And I also agree that a laser with analog blanking will work just fine with a TTL signal. (In fact, the opposite is also true to a point, though the laser won't turn on until the voltage gets really close to +5V)

Adam

[Pangolin]

Just one quick point. There CAN be (and often is) a difference between digital and analog grounds.

On the QM32, the digital ground is the ground plane, which is an inner layer within the QM32 board. The analog ground is just a really thick trace. Both of these meet ONLY AT ONE POINT on the ISA connector, otherwise they are isolated. All digital chips are referenced to the ground plane, which means that this ground may pick up high frequency digitally-excited noise (which you wouldn't want corrupting your analog signal). The big thick analog ground trace will not have this noise, since it is always separate from the digital circuits. The QM2000 has a digital ground plane and actually two separate analog ground planes!!

Note that you won't be able to measure a difference between these grounds with an ohm meter. It will appear as zero ohms. But there is a difference... you're going to have to trust me on this one

Also, even within complex all-analog systems such as servo amplifiers (and good audio equipment), there may be multiple grounds used internally. If memory serves, the original CB6580 amp from Cambridge had two or three internal grounds, each of which doing a separate job, and only meeting at one single place on the board.

If anyone is curious about this, you might want to google "single point gounding scheme" or "star grounding scheme"...

Best regards,

William Benner

[carmangary]

That's pretty interesting. I have never learned about ground schemes. However, even in these cases GND is still all tied together to provide a common reference. And if you measure voltage from an analog or digital device to either of the ground plane or trace you'll still get the same reading as opposed to connecting from + to gnd and + to - on an ILDA connector.

[The_Doctor]

Star grounding is a big thing in audio setups. It revolutionises the way people think about grounding and electrical safety, and it's easy to understand. Most of the apparent difficulties are caused by trying to work with existing ring ground systems that are still the way most places are wired.

Ideally, I like a single global ground, branching where needed, and balanced lines for signals, neither pole grounded, and a screen on a cable that is grounded at only one end. That way all mains powered equipment can be grounded for safety without causing noise loops. Considering how many cables default to screen grounded at both ends with no choice offered, the only way to get the ideal is NOT to trust the industry. If you want it, learn how it works and set it up. (Neutrik make very good connectors for low cost that can easily be adapted to this). If you start with good habits and arrangements it's very easy to build on that standard, and you can take a system into a new location with a much better chance that it will work right. This works on small boards, or whole rooms, so it always applies.

The main point is that all grounds have resistance. It might be too small to measure with a multimeter, but the moment a significant current flows through it (far bigger in all cases than the multimeter will be putting through it), any change in that current is also a change in voltage, a change in SIGNAL. These signals can go wherever the ground circuit goes, so the idea is to make it NOT a circuit. Keep circuits (loops) as small and as well defined and localised as possible. That way they don't talk to each other in ways you don't want or expect.