New Capacitors Lower Than Expected Capacitance

[AllTheGearNoIdea]

I am rebuilding the power supply section of my spectrum analysers after it exploded last week. I am taking the opportunity to replace the main smoothing capacitor 470uf and 250V. I have received today the new capacitor from RS component EPCOS B43544-A9477-M  470uF 400V

I was just doing a quick confidence check regarding ESR and capacitance and found they are measuring only 400uF. I know electrolytic have a typical wide tolerance but this appears very low.  Is there any kind of  burn in on new capacitors?   

I’m not sure if the power supply would be classed as a switching topology, but its rectifying the mains then chopping up the dc to supply a high frequency  conventional transformer.

Any idea regarding low capacitor value

Regards Chris

[SiliconWizard]

Looks like it's perfectly within tolerance.
It's a +/-20% capacitor. Min value is 376 µF. You'll be hard pressed to get better tolerance with this kind of capacitor.
The circuit design should have taken this into consideration. What was the original capacitor reference?

[AllTheGearNoIdea]

Yes your right regards tolerance I still just thought they would be more closely matched. I guess I thought would go on the plus side.

The Original Part was  Philips 470uf 250V  2222 052 43471

I think they listed the ESR as 300mOhm, whereas I have measured them all at better than 90mOhm. All 4 measurement around 500uF.  I suspect the old capacitors are good but thought it worth changing them after 30 years etc.

Thanks for the advice I guess your correct but maybe I have been lucky in the past. I don’t normally check capacitor before installation.

[SiliconWizard]

If you want to make sure the cap has a higher value, there are some electrolytic caps of this value/voltage range the tolerance of which is tilted towards higher values: commonly +30%/-10% instead of the more common +/-20%. Much less choice of values though.

I doubt that it will make any difference in your case though.

[AllTheGearNoIdea]

That’s interesting I had no idea regarding the tolerance bias. I have installed them and checked all the power rails and everything is within expected limits, so I guess I’m worrying over nothing. Many thanks for the reply and advice.

Best regards Chris.

[Zero999]

If it has a tolerance of +/-20%, then there's equal chance of it being above, as below 470μF. The fact the ESR is well below the specification, is good and is the most important thing.

[wraper]

In addition to all said above, capacitance is specified for measurements with sine wave @ 100Hz. Anything else will give different result. Measuring at higher frequencies normally will show lower capacitance.