The AM Forum

A few recent threads toss around the topic of replacing electrolytic caps with larger values. For a better margin of safety, to improve bass response, and so on.

The voltage rating makes sense. But when it comes to boosting the size (µF), what are the guidelines to follow? When I recapped the audio deck in my transmitter, voltages were upped considerably (100-200v) and the size or value, slightly (.25-100 µF, if that much).

Previous reading on the topic revealed that going with too large of a cap would impact charging/recovery time with potentially bad results. Can't recall specifics.

What's a good rule of thumb, and how much is too much?

http://www.amwindow.org/tech/pdf/geps1.pdf
http://www.amwindow.org/tech/pdf/geps2.pdf

Hi Todd,

I don't think that there are any universal guidelines but I can throw out a few ideas:

1.  Apparently the tolerance range for many electrolytics was quite wide and I have a couple of documents that refer to the typical tolerance as being within -40%/+100% of the marked value.  I think this indicates that in general the values were not too critical if you stay fairly close to the original center value. 

2.  In the typical audio chain, increasing the electrolytic values (cathode bypass caps anyway) has the same general impact as increasing the value of the coupling caps and that is of course increased low frequency response.  Whether that is a good thing or not I am not entering that debate.  However, whenever you change something from the original design intents there will be the anticipated major effect and also the (perhaps anticipated) side effects.  Hitting the modulator section and  power supply with increased demands may have a decidedly negative impact on the life of some components.  One of my favorite pharmaceutical commercials from a few years ago was a product that promised great benefits but one of the potential side effects (listed in small print) was "sudden unexpected death".  If you set up a nice subsonic oscillation then your mod tranny may experience the same undesired side effect.

3.  The input capacitors in a capacitor input PS filter are fairly critical as increasing the value here will greatly increase the output voltage, particularly under light current demands.  The output caps in a choke input (or pi/multi-pi) filter can be increased safely without causing soaring output assuming that there is sufficient current draw at all times to insure that the critical inductance requirement is met.  In general, power supplies using chokes didn't really skimp on filtering anyway.

4.  Obviously from the above I like to keep my rigs fairly stock so I am freely admitting my "bias".  I certainly wouldn't be concerned about replacing a 40 uf with a 47 uf but I would want to make sure I understand all of the circuit parameters before I put a 125 uf in place of a 40 uf.

It will be interesting to see the responses to your query and thanks for posting it!

Rodger WQ9E  (now back to plumbing in the whirlpool tub...)

This is something that I have never given much thought to, I have pretty much always done a "shoot from the hip" on selecting filter caps, and I'm usually somewhere in the ballpark.

This can be much more critical on capacitor input filters than choke input filters. With cap input filters, the regulation goes out the window and the inrush current can get to the point that things go boom. With the choke input filter the choke definately has a cushioning effect. With really oversized filter caps you can reduce the ripple but you will need some kind of step start of ramp up device to protect the transformer and rectifier from the inrush current.

But on the other hand, caps will have a "stiffening" effect on the voltage under rapidly changing loads (voice peaks).

The reason some choke input supplies get away with small filter caps is the amount of inductance in the chokes. Larger chokes mean less capacitance is needed to remove the ac ripple.

I dont like to go smaller than original, but the +100% limit is probably a good stopping point to minimize the extra inrush current.

                                                      the Slab Bacon

I'm going to sort of vary a little with Frank AHE (aka Slabo Baconovsky), or maybe "amplify" the issues...  :D ;D ::) :P

In some cases there is a "tuning" in the choke input filter section, wherein increasing or deacreasing the value of the cap following the choke causes more ripple, so there may be an optimal value for the cap.

As far as the inrush current, it can or might be an issue, but I'd rather have the extra capacitance there, if possible. There are a number of clever ways to mitigate excessive inrush current, if they are needed. The simplest is a relay with a fixed time delay, voltage/current sensing, or manual switching of the relay that places a suitable resistance in series with the primary of the transformer, limiting the initial current, then jumping it out.

                   _-_-Wobbly Bear Two Giant Chocolate Resistors   

As previously noted the tolerance for electrolytics is pisspoor.  I've measured brand new ones on a HP digital bridge as generally being about +10 to +25%.  I wouldn't hesitate replacing, say, a 16 uf in a R-649 with 22 uFs  ;)

Get as close as you can with the standard values and you'll be fine.

<snip>
"In some cases there is a "tuning" in the choke input filter section, wherein increasing or deacreasing the value of the cap following the choke causes more ripple, so there may be an optimal value for the cap."
<snip>

Bear,
          This is true, but not often the case. Usually a good clue if you are dealing with a "tuned" filter circuit is that you will find a small value cap across (in parallel with) the filter choke. However you dont find this too often.

As a general rule when it comes to filter caps, I also agree with the "more is better" phylosophy as long as the inrush current doesnt create problems.

                                                                                           the Slab Bacon