I ran across an ad for a used DC bench power supply titled "Racal Dana 9234 Precision Laboratory Dual DC Power Supply 15V/4A 60W". Judging by the listing, this unit was in service at a lab with professional users.
What caught my eye right away was a capacitor in parallel with the terminals (see attached photos). I zoomed in on the photo and it appears to be a 10uF 16V capacitor.
Is the purpose of this capacitor to reduce voltage ripple? If so, is connecting a capacitor in this manner considered a best/acceptable practice? I'm not sure if it's a risky hack or a good technique.
Also on topic, what is considered an acceptable limit for voltage ripple from a power supply? Max 2%? (I don't know if this unit is a linear or switched supply).
Is the purpose of this capacitor to reduce voltage ripple?
Unlikely. 10uF at 60Hz is not going to make much difference. It might make a difference to the hf noise present on the output.
But. In my opinion the vendor has no idea what they are selling. The capacitor is there for the purposes of taking a photo with something attached to the output terminals. Note that the sense terminal is not connected, also, the output is not turned on. It's a meaningless capacitor for the sake of a posed-photo!
These supplies are re-badged under the TTI (Thurlby Thandar Instruments) name.
Edit: I believe these are an early version of the TTi PL154.
Also on topic, what is considered an acceptable limit for voltage ripple from a power supply? Max 2%? (I don't know if this unit is a linear or switched
supply).
2% is high. I would expect tenths of a percent or less. It is a linear supply.
I don't think it means anything - the supply will have an internal capacitor across the output terminals anyway.
A supply has a control loop with a bandwidth and in an old supply, it could be 10kHz. This means the internal regulator cannot handle fast transient load changes well. Some capacitance across the output of a supply takes over at keeping the output voltage constant at time scales smaller then the regulator can cope. The higher the bandwidth of the supply control loop, the smaller the output capacitor you can get away with.
If this supply actually need that capacitor permanently to work, it just means that the internal output capacitor has probably gone dry and needs replacing. A cheap and easy thing to fix.
That cap isn't going to do much at all. Besides, it's leads are too straight to have been on that supply for long. The cap is going to get knocked around in general use - so, yes, it's a prop for the photo shoot.
I do that sometimes just to have a lead to clip onto. Had it been a better cap, there could have been a purpose.
Fair enough - but those leads are too neat.
There should already be a cap (or three) across the terminals inside the case. It's a very basic requirement for a power supply.
The internal capacitor(s) will be well chosen though, with capacitance and voltage suited to the job.
I sent a brief message to the seller inquiring about this capacitor in their ad listing photo. They replied they are a surplus consolidator and they do not use the equipment they receive and sell. And they did not place that capacitor on the psu, it came like that.
I don't think this is a prop for purposes of the photo. The leads look neat but the terminals can be connected by banana plugs or alligators with no contact to the leads. I would speculate this cap was used to filter out high freq transient noise as they took oscilloscope measurements on lower freq waveforms. Just a guess, and I might try it myself.
I do that sometimes just to have a lead to clip onto. Had it been a better cap, there could have been a purpose.
this!
I have a nice 1 uF red wima on one of my psu's. I needed to have leads to grab onto and for high freq noise, it lowered it a tiny bit. but it was 99.9% there just to have leads to clip onto
I don't think this is a prop for purposes of the photo.
That theory never made any sense. It' makes it look like there's something wrong with the power supply - hardly a good selling point.
I would speculate this cap was used to filter out high freq transient noise as they took oscilloscope measurements on lower freq waveforms. Just a guess, and I might try it myself.
It won't work, because:
a) There's already caps inside the supply.
b) Electrolytics are rubbish for high frequency.
There should already be a cap (or three) across the terminals inside the case. It's a very basic requirement for a power supply.
The internal capacitor(s) will be well chosen though, with capacitance and voltage suited to the job.
Unfortunately this isn't always the case. I have a cheap-ish Voltcraft (Conrad) PSU and it definitely needs an external capacitor for stability.
Yep, I can see why you might do this in special cases. If the smoothing circuit is poorly designed (mis-sizing of cap), or if the cap is drying out due to age (possibly the case for the photo in my original post of an older PSU), then adding an e-cap like this can make sense.
as I was cleaning up my place, I found another psu that had a cap across its output screw banana terms.
for me, it was 100% a case of 'I need something there that my alligators can grab onto; and I don't want to burn current, so a resistor is not going to do it'.
that was entirely the reason. I have caps and resistors within easy arms reach and I grabbed a cap that would do no harm, just so I could have stable leads to clip onto.
I'm pretty sure, this is all that there was to it, for the ebay item in the OP's post. just easy for grabbers.
Ha. I ordered this exact unit off ebay; serial numbers match.
It seems to be pretty uncommon but it's turned out to be a decent power supply, no noise at all on the output.
Ha. I ordered this exact unit off ebay; serial numbers match.
It seems to be pretty uncommon but it's turned out to be a decent power supply, no noise at all on the output.
Wow, it's a small world!
Did you measure for noise with the cap on or off the terminals?
Ha. I ordered this exact unit off ebay; serial numbers match.
It seems to be pretty uncommon but it's turned out to be a decent power supply, no noise at all on the output.
Wow, it's a small world!
Did you measure for noise with the cap on or off the terminals?
I measured with it on. I doubt it makes any difference, like others have mentioned it's probably just a convenient place to clip loads to. I'll measure it real quick just for curiosity's sake.
No difference at all. It has about ~5mVpp of noise with and without the cap on it.