+/-12vdc, 5volt digital ground?

[argonrepublic]

I pulled a supply from an effects processor that is out of service.
It has +15VDC, -15VDC and +5VDC. It is exactly what I need for my synthesizer, except!!! This unit has a dedicated ground for the 5V. The synthesizer I am going to use shares 1 ground between +15, -15 and +5. I can not seem to get the voltages to work. Can I tie the grounds together or?

 

[wwenze]

Can't tell for sure from the photo, what's the measured resistance between the ±15V ground and 5V ground?

If they're separate, connect a high value resistor between them and measure the current. No current = tie them together.

 

[argonrepublic]

It worked, Thanks. No current across grounds. I like the idea of a designated digital ground. It seems to make sense. I won't dare try to make my modules use a separate 5V/ground. Would it in the bigger picture be beneficial to use independent grounds. One for logic, one for audio?

 

[wwenze]

I find this a good read:

http://www.analog.com/static/imported-files/tutorials/MT-031.pdf

 

[argonrepublic]

I just hooked up and found +14.8, -15.1, 5.04. Is the difference between + and - 15vdc ok for opamps? Is it possible that the offset is caused by the Digital ground?

 

[twest820]

A regulator or voltage reference's output voltage accuracy is usually in the 1 to 4% range---pull the part numbers and have a look at the datasheets to determine the expected range for this supply. You'll also need to account for the accuracy of the measurement equipment you're using---that'll specified in its data sheet as well if there's a guarantee, though calibration may be an issue---but it's probably in the 0.3% to 1% range.

If there's no current in the tie between the two grounds there'll be no offset between them. However, you'll have to be pretty specific about the circuit, layout, measurement technique, and so on for one to say anything meaningful about possible ground offsets in this case.

Once you've that sorted and have the ability to measure the common mode offset you can look at the circuits' CMRR to see how much error will occur. Generally, yes, it can be audible. Either directly as audio band noise or indirectly in the form of DC offset reducing a driver's linear throw.

 

[theAnonymous1]

It worked, Thanks. No current across grounds. I like the idea of a designated digital ground. It seems to make sense.

At the very least, you can use a ferrite bead on the ground link to help clean up any HF noise.