Precise voltage reference

[cj9]

Hi!

I'm using a -5v reference to shift a signal with a summing amplifier, currently with a voltage regulator. Is there anything that has a lot less noise (whatever capacitors I put on the regulator, there's still quite a bit of noise around -5V).

It's used as a signal which needs to be quite clean, so just wondering if anyone has any suggestions?

Thanks.

[IanB]

I doubt the regulator is making noise itself, presuming it is a linear regulator. More likely the noise is feeding through from the upstream supply. Perhaps you could try to create a cleaner supply?

[jahonen]

Try a buried zener reference (e.g. AD586) and put additional filtering around it, should be quite quiet.

Regards,
Janne

[samgab]

The TI REF5050 has good accuracy, low drift, and low noise:
http://www.ti.com/product/ref5050&lpos=Middle_Container&lid=Alternative_Devices

[cj9]

Thanks for the suggestions, I'll take a look at them

In reply to Ian, amazingly it is the regulator causing the problem (2v of noise). I have tried different batches from different sources and they all have ridiculous noise (-5v regulators from ST). I finally managed to reduce this with caps, but they arn't even the recommended ones on the datasheet (as those still had noise). Also tried on both PCB and breadboard. Directly replaced with 12v the signal is nice and clean (fed from a lab supply). Put the -5v reg's in and all of a sudden mega noise!

[IanB]

In reply to Ian, amazingly it is the regulator causing the problem (2v of noise). I have tried different batches from different sources and they all have ridiculous noise (-5v regulators from ST). I finally managed to reduce this with caps, but they arn't even the recommended ones on the datasheet (as those still had noise). Also tried on both PCB and breadboard. Directly replaced with 12v the signal is nice and clean (fed from a lab supply). Put the -5v reg's in and all of a sudden mega noise!

Ah, I seem to remember another thread about this. Do you have sufficient load on the regulator?

[amspire]

cj9, I wrote about this on your other thread on the same subject.

If you are not matching the 7905 specs for noise, then something major is wrong, and you are just not providing us any quality information. We have no schematics, no photos to show exactly how the 7905 is connected, along with the caps, no details of the power source, no details of the load. Something is wrong about your use of the IC's, but we don't have anything to work from. How can we help you if you just won't put the effort in?

These regulators have been used for decades without a problem, so what is different about your use of them?

Richard

[alm]

In reply to Ian, amazingly it is the regulator causing the problem (2v of noise). I have tried different batches from different sources and they all have ridiculous noise (-5v regulators from ST). I finally managed to reduce this with caps, but they arn't even the recommended ones on the datasheet (as those still had noise). Also tried on both PCB and breadboard. Directly replaced with 12v the signal is nice and clean (fed from a lab supply). Put the -5v reg's in and all of a sudden mega noise!

If you get 2 V noise from a 7905, no expensive low noise reference is going to help you, it might very well oscillate, too. First get the 7905 working according to specs with the schematic from the data sheet, then worry about improving on this.

[SeanB]

You definitely need to place more load on the regulator. 1K between output and ground will stop it definitely. Most negative ( and a lot of positive) regulators have a minimum load spec you need to supply or else they do strange things.

[cj9]

In response to those referencing my last post, I had taken your advice, which is why I am looking for voltage reference rather than regulators. As I had said, the circuit set up was as stated in the example in the datasheet. It wasn't worth providing pictures considering it was copied from the datasheet.
I just couldn't work out the problem because it was only happening with 5v regulators, the same circuit with other voltages was absolutely fine.

I'll try putting load on the regulator and see if it solves the problem too.

Thanks!