The AM Forum

THE AM BULLETIN BOARD => Technical Forum => Topic started by: kc2ifr on November 24, 2008, 05:12:26 PM



Title: electrolytic or oil caps
Post by: kc2ifr on November 24, 2008, 05:12:26 PM
If I have a power supply that currently has 4 electrolytic caps in series to come up with 40 mfd at 1000 volts, would it be ok for me to replace them with an oil cap rated at 40 mfd at 1500 volts?
I guess the question is this........any advantage to the oils or are the electrolytics a better choice.

Bill


Title: Re: electrolytic or oil caps
Post by: WA1HZK on November 24, 2008, 05:41:12 PM
Oils, they seldom fail and usually out live the "Old Buzzard" using them. I prefer them every time.
Keith
WA1HZK


Title: Re: electrolytic or oil caps
Post by: WA1GFZ on November 24, 2008, 06:29:10 PM
Many oil caps are self healing so if they flash over oil will flow back in and fill the void.


Title: Re: electrolytic or oil caps
Post by: KL7OF on November 24, 2008, 09:10:39 PM
Oil caps are my choice if I have one and if I have the room.....In my experience, they seldom fail, and if they do,    They sometimes heal themselves....


Title: Re: electrolytic or oil caps
Post by: KM1H on November 24, 2008, 10:08:22 PM
Oil caps have low ESR, less heat generated, and can handle higher ripple currents. Im using only 12 uF in a FWB SSB/CW 3500VDC supply and it only drops 250V at 1A key down.

OTOH some of the latest electrolytic designs are darn close and the amount of C that you can stuff in a small can is amazing. The slopbucket "serious QRO" bunch are using strings of up to 5600 uF @ 450V to obtain almost no voltage sag on voice peaks. This also requires special transformers with extremely low secondary DC resistance.

Carl
KM1H



Title: Re: electrolytic or oil caps
Post by: k4kyv on November 25, 2008, 01:04:03 AM
I have a couple of 400 mfd oil caps rated at 2500 volts each.  In parallel that would make 800 mfd @ 2500v, and in series, 200 mfd @ 5000 v.


Title: Re: electrolytic or oil caps
Post by: KM1H on November 25, 2008, 09:17:43 AM
I have a couple of 400 mfd oil caps rated at 2500 volts each.  In parallel that would make 800 mfd @ 2500v, and in series, 200 mfd @ 5000 v.

Thats some serious joules ::) but by the time you get enough strung together for the QRO voltages its going to require their own rack to hold them ;D. To some of those guys a 4CX5000A is barely entry level to the club.

Ive been staring at a 51.5 uF @ 9500V Aerovox for a decade or more but other than using it as a test cap for much lower voltages it just anchors a shelf.

Carl
KM1H


Title: Re: electrolytic or oil caps
Post by: WD5JKO on November 25, 2008, 09:21:28 AM
   There are many disadvantages of a series electrolytic string versus a single oil filled capacitor of similar ratings. These disadvantages are offset by several advantages:

* Lower cost
* much smaller in size for a given MFD, V rating
* Readily available


   The disadvantages include but not limited to:

* Reduced temperature range
* NOT self healing
* Finite operating lifespan

  We often ignore the limitations and design such that series strings of electrolytics fail. I think we can do a better job and use electrolytic series strings with confidence so long as we observe some simple design constraints:

* Keep the caps away from heat sources. One major issue is the equalizing resistor location. Move these resistors away from the capacitor. Choose low ESR capacitors to minimize internal heating from the ripple current.

* Keep steady state voltage across each capacitor no more than 2/3 the rating.

* Equalizing resistor value (bleed current) is important but does little to protect the capacitors during transient conditions of charge up (turn on the power switch) and sudden discharge (grounding hook or parasitic oscillation). Use the lowest value resistor that you can in a large series string, and mount away from capacitors to keep the heat off the capacitors.

* Use wirewound equalizing resistors

* Match the capacitors ( same mfd ) prior to selecting them. This will do more good than equalizing resistors during transient conditions. Matching techniques are another story since many C - meters will struggle when the capacitor is leaky (not formed).

* Add a back biased diode across each series capacitor. This diode will need to be stout in peak forward current rating. Maybe a 1N5408. The idea is the diode will prevent any cap in the series string from being reverse biased when the capacitor bank gets shorted out suddenly. These diodes are important to keep the stack alive after a big bang.

* Add a resistor in series with the stack that will limit the current into a short circuit. The value here is debatable, but maybe 5-10 X the sum of the capacitors ESR ( 10 ohms?).   This resistor will need to be stout such that it does not vaporize when the stack is shorted out.


I feel that if you follow these simple suggestions, that you will have a reliable HV capacitor filter using small, and inexpensive capacitors that are readily available.

Comments welcome!

Jim
WD5JKO


Title: Re: electrolytic or oil caps
Post by: WA1GFZ on November 25, 2008, 02:47:50 PM
A pair of CG 310s with 2 stage choke input 14 uf at 14 KV and 3 100s@3 KV  in series for output pretty stable at 1500 watts out.
AMfone - Dedicated to Amplitude Modulation on the Amateur Radio Bands