How old is too old for an electrolytic cap

[Mike_J]

Does anyone have a method of reviving some Sprague Atoms that I have probably had in my cap bin for around ten years?  These things cost a small fortune and it be would nice to be able to use them if they can be revived.

Thanks
Mike

[HotBluePlates]

Get a 100kΩ resistor and solder insulated alligator clips on either end. Recharge/re-form the cap by connecting the - to ground in an amp, and use your 100kΩ resistor to connect the + to a positive d.c. voltage in the amp. When done with reforming (and with the power source off), move the far end of the 100kΩ resistor from positive voltage to 0v to discharge the cap.

Better yet, put all of this in a box. Use something like plastic capped banana jacks to hold the cap leads. Have your 100kΩ resistor run from the red banana jack to a SPDT switch. Have a + d.c. voltage feed one side of the switch, and 0v (and black banana jack) feed the other side of the switch. Be careful of exposed leads and verify with a meter the cap is fully discharged before handling.

The positive voltage reforms the insulating oxide layer inside the cap, while the resistor limits cap leakage current to a safe value while reforming occurs. A 100kΩ resistor limits charging current to a few milliamps (4.5mA for 450vdc, 1mA for 100vdc). You can make the resistor larger, but reform and discharge time will be somewhat longer.

You don't need a variac for forming caps: the suggestion of the variac is a hope that reduced input voltage will reduce any cap leakage current to a safe value while forming. A resistor makes sure the leakage is limited regardless.

You may wish to even add meter jacks to your box to directly measure leakage current (when the cap is fully formed, leakage current will drop to probably under a milliamp even without the resistor) to know when the cap is reformed, or if it can't be reformed and needs to be tossed.

[Mike_J]

Get a 100kΩ resistor and solder insulated alligator clips on either end. Recharge/re-form the cap by connecting the - to ground in an amp, and use your 100kΩ resistor to connect the + to a positive d.c. voltage in the amp. When done with reforming (and with the power source off), move the far end of the 100kΩ resistor from positive voltage to 0v to discharge the cap.

Better yet, put all of this in a box. Use something like plastic capped banana jacks to hold the cap leads. Have your 100kΩ resistor run from the red banana jack to a SPDT switch. Have a + d.c. voltage feed one side of the switch, and 0v (and black banana jack) feed the other side of the switch. Be careful of exposed leads and verify with a meter the cap is fully discharged before handling.

The positive voltage reforms the insulating oxide layer inside the cap, while the resistor limits cap leakage current to a safe value while reforming occurs. A 100kΩ resistor limits charging current to a few milliamps (4.5mA for 450vdc, 1mA for 100vdc). You can make the resistor larger, but reform and discharge time will be somewhat longer.

You don't need a variac for forming caps: the suggestion of the variac is a hope that reduced input voltage will reduce any cap leakage current to a safe value while forming. A resistor makes sure the leakage is limited regardless.

You may wish to even add meter jacks to your box to directly measure leakage current (when the cap is fully formed, leakage current will drop to probably under a milliamp even without the resistor) to know when the cap is reformed, or if it can't be reformed and needs to be tossed.

HBP

Thank you for the response.  Is there a minimum wattage value for the 100K resistor?  How do I measure the leakage current to determine if the cap is reformed or needs to be tossed?

I did a little schematic of what I think is a way to do what you are saying.  I would use turrets and some G-10 to make a board.  I  would solder each cap onto the board and unsolder it when it is done.  I would attach the board to probably another piece of G-10 with some standoffs.  Should make it safe.  Do you think this would work?

I am envisioning pulling a chassis out of one of my amps and attaching the positive side to my HT point and the negative side to ground.  Is this what you had in mind?

Thanks
Mike

[sluckey]

Is this what you had in mind?

No. The resistor should be in series with the cap. Half watt is sufficient, but 1 watt is sturdier. I wouldn't bother with soldering. Gator clip leads is all you need.

[HotBluePlates]

Is there a minimum wattage value for the 100K resistor?

Teach a man to fish... Learn to use Ohm's Law and the Equation for Power.

Ohm's Law: Current = Voltage/Resistance
Power = Voltage * Current

Use algebra to rearrange the Ohm's Law equation to solve for the quantity you don't already know, and substitute Ohm's Law into the Equation for Power to reduce the number of steps.

Power = Voltage * Current
= Voltage * (Voltage/Resistance)   (from Ohm's Law)
= Voltage²/Resistance

If you have a supply voltage of 500vdc and a 100kΩ resistor, then resistor dissipation is no more than Voltage²/Resistance = 500²/100k = 2.5w. Double that so the resistor doesn't cook, and you'd get a 5w resistor (which are hard to find in a 100kΩ value).

The resistor wattage rating is entirely dependent on the dc volts you use to form the cap. if you only use 200vdc, the 100kΩ only needs to be 1w (do the math & see what you come up with).

How do I measure the leakage current ...?

Say you have everything connected. Disconnect the 100kΩ from your power source. Connect the red lead of your meter (set to measure current) to the power source, and the black lead to the 100kΩ resistor. Now the meter is in series with the current you're trying to measure. When you first apply power, the current will spike to a high value, but it quickly settles down to a low value (how quickly depends on the resistor).

A perfect cap charges to the full value of the supply voltage you're using, then drops to a measured current of zero. Electrolytic caps aren't perfect, and usually have some residual leakage current. I wouldn't be surprised by 1mA reading on a new cap with the setup I described. I would view any cap passing 10mA after charging as a candidate for the trash can. The better the cap, the less current after it fully charges.

I would use turrets and some G-10 to make a board.  I  would solder each cap onto the board and unsolder it when it is done.  I would attach the board to probably another piece of G-10 with some standoffs.  Should make it safe.

Don't solder. Do as Sluckey said and use insulated gator clips, or the banana jack binding posts I suggested (below). Don't forget that if you use a 300vdc supply, the cap will become a 300vdc piranha waiting to bite you. It would probably be good to have the whole item in a plastic box, use the binding posts, have a set of tip jacks to connect the meter, keep the resistor inside the box, have a switch to allow you to discharge the cap without touching it, and extra jacks to allow you to verify the cap voltage is down to 0v before removing from the binding posts. add a 2-conductor cable from inside the box to insulated gator clips/connector of your choice to connect to your source of d.c. voltage.

I'd draw it out for you, but I'm on a new computer with no way to draw. If the above isn't so clear to you to be ridiculously-obvious, don't build/do it because the cap will be a major shock hazard and you'll only make a silly mistake once (I've been bit by 450vdc, and you don't want to feel that).

[Mike_J]

Is this what you had in mind?

No. The resistor should be in series with the cap. Half watt is sufficient, but 1 watt is sturdier. I wouldn't bother with soldering. Gator clip leads is all you need.

Thanks Sluckey

I am going to have to draw it out until I get it right but I have an amp to build.  For the time being I will buy new caps.  I only need the 100uf and one 20uf cap so I will order.  After the amp is finished I will figure out how to reform and test my old electrolytics.

Thanks
Mike

[Mike_J]

Is there a minimum wattage value for the 100K resistor?

Teach a man to fish... Learn to use Ohm's Law and the Equation for Power.

Ohm's Law: Current = Voltage/Resistance
Power = Voltage * Current

Use algebra to rearrange the Ohm's Law equation to solve for the quantity you don't already know, and substitute Ohm's Law into the Equation for Power to reduce the number of steps.

Power = Voltage * Current
= Voltage * (Voltage/Resistance)   (from Ohm's Law)
= Voltage²/Resistance

If you have a supply voltage of 500vdc and a 100kΩ resistor, then resistor dissipation is no more than Voltage²/Resistance = 500²/100k = 2.5w. Double that so the resistor doesn't cook, and you'd get a 5w resistor (which are hard to find in a 100kΩ value).

The resistor wattage rating is entirely dependent on the dc volts you use to form the cap. if you only use 200vdc, the 100kΩ only needs to be 1w (do the math & see what you come up with).

How do I measure the leakage current ...?

Say you have everything connected. Disconnect the 100kΩ from your power source. Connect the red lead of your meter (set to measure current) to the power source, and the black lead to the 100kΩ resistor. Now the meter is in series with the current you're trying to measure. When you first apply power, the current will spike to a high value, but it quickly settles down to a low value (how quickly depends on the resistor).

A perfect cap charges to the full value of the supply voltage you're using, then drops to a measured current of zero. Electrolytic caps aren't perfect, and usually have some residual leakage current. I wouldn't be surprised by 1mA reading on a new cap with the setup I described. I would view any cap passing 10mA after charging as a candidate for the trash can. The better the cap, the less current after it fully charges.

I would use turrets and some G-10 to make a board.  I  would solder each cap onto the board and unsolder it when it is done.  I would attach the board to probably another piece of G-10 with some standoffs.  Should make it safe.

Don't solder. Do as Sluckey said and use insulated gator clips, or the banana jack binding posts I suggested (below). Don't forget that if you use a 300vdc supply, the cap will become a 300vdc piranha waiting to bite you. It would probably be good to have the whole item in a plastic box, use the binding posts, have a set of tip jacks to connect the meter, keep the resistor inside the box, have a switch to allow you to discharge the cap without touching it, and extra jacks to allow you to verify the cap voltage is down to 0v before removing from the binding posts. add a 2-conductor cable from inside the box to insulated gator clips/connector of your choice to connect to your source of d.c. voltage.

I'd draw it out for you, but I'm on a new computer with no way to draw. If the above isn't so clear to you to be ridiculously-obvious, don't build/do it because the cap will be a major shock hazard and you'll only make a silly mistake once (I've been bit by 450vdc, and you don't want to feel that).

Thanks HBP

If I am going to be using the cap as my HT or screen supply cap I would think I would want to reform the cap at an approximation of the voltage rating at which it will be used.  Why wouldn't I use a one watt 470K resistor and form it at the 500K voltage at which it will be used?  I know you said it will take longer to form but how much longer?

Unfortunately, I am a pretty visual person so I will have to wait until I finish my current build to do this because I have not fully visualized what you are saying.  For the time being I will buy new caps but will return to this when my build is completed.

Sorry you got bit.  Now I know why you always have 83 safety steps in everything you recommend.

Thanks
Mike

[SoundmasterG]

I wanted to do the same thing several years back with some Sprague 20uF 600V caps that had sat around here for 12 years. I have a Sprague Tel-ohmike which you can use to reform caps since the device limits the current to the caps....so I used that and hit the caps with the full 600V...it took about 12 hours for each cap to reform....and even though they function ok and don't leak much, I don't trust them enough to use in customer's amps. They are such a large cap that I can't use them in many of my own projects either....so they still sit, though I do hit them with voltage from time to time, hoping I will have a use for them some day. I also did the same thing with some LCR 50uF 50UF 525V cans.....Those are a little easier to use but I haven't used them in anything yet either.....


Greg

[Mike_J]

I wanted to do the same thing several years back with some Sprague 20uF 600V caps that had sat around here for 12 years. I have a Sprague Tel-ohmike which you can use to reform caps since the device limits the current to the caps....so I used that and hit the caps with the full 600V...it took about 12 hours for each cap to reform....and even though they function ok and don't leak much, I don't trust them enough to use in customer's amps. They are such a large cap that I can't use them in many of my own projects either....so they still sit, though I do hit them with voltage from time to time, hoping I will have a use for them some day. I also did the same thing with some LCR 50uF 50UF 525V cans.....Those are a little easier to use but I haven't used them in anything yet either.....


Greg

Greg

I have three or four of the 20uF @ 600 volts and quite a few at 450 or 500 volts and some at 16, 40 and 100uF.  When added up at new prices it adds up to quite a bit of money.  I now only buy exactly what I need for the build I am making. 

Thanks
Mike

[SoundmasterG]

I wanted to do the same thing several years back with some Sprague 20uF 600V caps that had sat around here for 12 years. I have a Sprague Tel-ohmike which you can use to reform caps since the device limits the current to the caps....so I used that and hit the caps with the full 600V...it took about 12 hours for each cap to reform....and even though they function ok and don't leak much, I don't trust them enough to use in customer's amps. They are such a large cap that I can't use them in many of my own projects either....so they still sit, though I do hit them with voltage from time to time, hoping I will have a use for them some day. I also did the same thing with some LCR 50uF 50UF 525V cans.....Those are a little easier to use but I haven't used them in anything yet either.....


Greg

Greg

I have three or four of the 20uF @ 600 volts and quite a few at 450 or 500 volts and some at 16, 40 and 100uF.  When added up at new prices it adds up to quite a bit of money.  I now only buy exactly what I need for the build I am making. 

Thanks
Mike

Yes I agree completely Mike. I got those years ago before I knew what I was going to need in planned builds. I only get electrolytics now that I will use on repairs or builds and I order them on demand rather than keeping stuff in stock.


Greg

[HotBluePlates]

Sorry you got bit.  Now I know why you always have 83 safety steps in everything you recommend.

I used to be in the Navy. They had a saying: "Navy regulations are written in blood," because the vast majority of rules about how you must do something are the result of investigations into accidents and fatalities.

The shop I worked in had a safety poster hanging up. I was a photo of a person's hand with a piece of gauze pad beside it. On the gauze was what appeared to be a severed finger. Looking at the hand, the middle finger was stripped down to the bone beginning right about where you'd wear a ring (oddly, the bones of the finger were still in place). The mangled ring was also in the photo.

The poster was meant to reinforce the Navy's rule to remove ALL jewelry when working on any machinery (or any electrical/electronic equipment). The subject of the photo apparently thought the power was locked out on the motor he was repairing, and also didn't remove his ring. I just hope he was seriously sedated when they were snapping the photo.

You can find the poster's image in an internet search, but I'll spare everybody and not actually link it here.

[sluckey]

I couldn't even wear metal rim glasses at Keesler or Robins AFB.

[PRR]

> hope he was seriously sedated when they were snapping the photo.

Such injuries sometimes hurt less than you'd think. I remember seeing thumb-bone, saying "That's not good, let's go to the ER", but no real pain.

> oddly, the bones of the finger were still in place

Tough little ligaments hold the bones together. Whatever he did, it ripped the soft stuff but didn't get a grip on the harder parts. (A ring would do that, it is much larger than the joints.)

Thanks for the post. I'm installing a furnace. I'm getting a little crazy, and tired, and not being careful enough about drilling and sharp edges. I'm mentally posting your image-description next to the job.

I did do one smart thing. The old furnace had to come out of the cellar. In my youth I woulda humped it up myself. But this time I hired two husky young guys to hump the furnace and the dozen concrete blocks it was on. They think they were over-paid, I figure it is a lot cheaper than hurting myself. (And over-pay is OK, because next month I want to move the 300 pound oil-tank out.)

[Willabe]

I did do one smart thing. The old furnace had to come out of the cellar. In my youth I woulda humped it up myself. But this time I hired two husky young guys to hump the furnace and the dozen concrete blocks it was on. They think they were over-paid, I figure it is a lot cheaper than hurting myself. (And over-pay is OK, because next month I want to move the 300 pound oil-tank out.)

Yep, smart.


             Brad