Placid-BP Bipolar Shunt Regulated Power Supply

[Goto]

Nic

I am running a placid BP, powering my Counterpoint boards (I am using B24).

Quite an upgrade over LCBPS

Mark

[avr300]

Raising another thought. A shunt doesn't like to drive capacitive loads, right ?

What about the 6 * 220uF on the IVY 2.0 ? Is it wise to remove these to make the shunt operate freely ?

[Russ White]

The placid will work fine into any practical capacitance as long as there is some ESL.

Still you could quite safely remove those caps.

I personally sometimes do remove large decoupling caps when using shunt regulators. This minimizes the possibility of having little parasitic tank ccts.

Cheers!
Russ

[avr300]

Thanks for a quick answer Russ.

[tailspn]

[QUOTE=Russ White;1989370]The placid will work fine into any practical capacitance as long as there is some ESL.

Is ESL the same as esr?

Thanks,
Tom

[BrianDonegan]

ESL = Equivalent Series Inductance

ESR = Equivalent Series Resistance

[luvdunhill]

TPA Guys!

Do you feel that the rails track enough for this to be used with a DC coupled amplifier, that doesn't have any sort of on board regulators (perhaps some of your designs meet this criteria?)

I've done a bit of testing, but haven't formed a solid opinion myself.

Here's one test run zoomed in across both rails (it's configured as a +-12VDC supply)



Here is some data over 5 hours of warm-up on a single rail:



Unfortunately, waiting two hours for the DC offset to stabilize is not practical, so the question is how well do things thermally track during warm-up.

So far it seems okay. Upon turning on the power supply, I see 0.055VDC difference in the rails. So, assuming this offset would be trimmed via the amplifier, we'll call this value 0. After 1 min. this was 0.046. After another min, this was 0.052. After 5 more minutes, 0.060. After 5 more minutes, 0.064. After 10 more minutes, 0.061. After a final 10 minutes, 0.061. So, during this time there was a +-9mV variance from the initial value.

The next plan is to build a simple circuit to do this math for me, so I can monitor it over a longer period of time.

Anyways, I do wish this was tighter, given the use in a DC coupled amplifier, so my question is, is there any suggestions on how this variance could be nullified somewhat? Any further tests that I could help you with?

[Russ White]

A small change in the rails over a long peroid of time should not effect a well designed amplifier, DC coupled or not.

The drift you are observing is a function of the tempco of the LED.

You could use a voltage reference with less tempco, but you would likely then have more noise. You could easily put just about any series VREF in there. Even a Zener if you please.

None of our amps would be adversely effected by the slight drift at all.

I would just use the LED VREF unless you have good reason not to.

Cheers!
Russ

[luvdunhill]

Originally Posted by Russ White

A small change in the rails over a long peroid of time should not effect a well designed amplifier, DC coupled or not.

I wouldn't call 300mV over 2 hours small (per the second graph) in the context I've presented. Also, I'm not sure what you mean by a well designed DC coupled amplifier? Let's assume a servo isn't used. If it uses a bipolar PSU and the rails drift apart by 10mV, this will show up as a DC offset. edit: I'm obviously more concerned about the rails drifting at different rates, then any absolute change over time.

Just trying to help!

[Russ White]

You would have to show me your amplifier circuit for me to understand. I have never used an amplifier (DC coupled or not) that would show any appreciable offset change with a 300mv rail difference.

In any case you have the solution. Use a VREF with very low tempco drift.

Cheers!
Russ