can cap question

[surfsup]

Having never used older style cancaps, this cap has the four ground tabs on the cap. Should i isolate the can from the chassis and run the ground to a point?

Also this is new out of the box. The solder tabs are all charred at the bottom (hope you can see that from the picf). Is that normal? Just got it from AES.

Hmmm edit, cant seem to attach a pic from android....

[surfsup]

pic deleted...no longer needed

[Structo]

I think what you are seeing on those lugs is the spot welds.

The mounting lugs provide the ground.

[Firestorm]

What amp? Is this the only filter(s)?

[surfsup]

Its for a restoration of a heathkit ag10 signal generator. One of three can caps...

[Firestorm]

The first filter in that one looks like two caps in series. If those are cans, you have to insulate one and be careful how the negative tabs are connected. Don't think I'd worry about insulating any of the rest.

[surfsup]

I got a couple JJs for the other two, and now that you mention it, I think there is only one ground on those cans. I might have to parllel the two sectios and put another axial cap in series. I'll have to check when I get home.

[surfsup]

Firestorm and all, a quick question about the can cap common ground. I have a JJ that I received to replace teh 50/50 original seen in the photo on the right (leftmost can cap)...

The JJ has three tabs. two (+) and one (-). I assume all these caps have a commond ground (including the old originals in the AG10). My question is looking at the wiring, if the (-) are connected internally the AG10 wiring is wrong and not according to the schematic. The PT secondary is connected to one cap (+) and the other half of the cap to the diode. The ground goes to the other diode. But...the ground of the first 50u should go to the (+) of the other cap. It doesn't...what am I missing? Again, I'm not familiar with cancaps, let alone old ones...TIA

[surfsup]

Could someone verify for me if this is some sort of voltage doubler circuit? I can't find this rectifier online anywhere...and the wiring seems off. I've been trying to figure out how this circuit would work but I dn't get it.

I am thinking of just using a rectifier I am more familiar with since I'm replacing stuff anyway, but it seems this is a voltage doubler but I'm not 100%.

EDIT okay I think this is definitely a doubler...on the positive cycle, D1 allows C1 to charge. On the negative cycle, C2 charges due to D2. The B+ is going to be the sum of C1 + C2 ... correct?

[surfsup]

Taking a very hard look at the cap cans, there is some writing that is rubbed away that looks like the cap can (original in photo) actually has teh sections wired in series. I'm 99% convinced. The cap shows type "FP" I can't google any info on this type. Looking at AES, their FP type caps seem normal (common ground for all sections) so not sure where to even find a cap to replace this.

Here's what I'm thiinking, wire the cap can I got in parallel, then put another regular electrolytic in series. Do the two have to be equal? if I parallel the 50/50 I will get 100u, so I'll need another 100u for the series cap or can I slap a 47u in there to complete the circuit? I would assume the difference in capacitance will just bea ripple/time constant issue that will be filtered out down the road.

Or I can use just one 50u side, with a separate cap, then run a 10R to the next side as an extra filter...

[martin manning]

Yes, this is a voltage doubler. The multi-cap in the picture must be a special, with the caps series connected instead of having the typical common ground (like Firestorm says above). The lead going directly from the PT to the cap can is connected to the +/- common, the case is the negative end of the stack, and the other terminal is the positive end. I've never seen this before, but that is the way it is wired. Are there any markings on the can the describe the internal arrangement?

[Firestorm]

Very strange indeed. I've never heard of a cap with an internal series connection (and frankly, something in the back of mind wants to believe it isn't actually even possible given the way they're made, but who knows?) If you can read all the markings on the side of the can (or sleeve, if it's in cardboard) there might be a clue. And don't overlook the possibility that this is an error committed sometime between the factory and now. The fuse looks like it was added.

[surfsup]

The cap can is definitely a series cap can. Leave it to me to discover something new. unsolderable turrets, conductive G10, series can caps...I should be used to this by now.

So does it matter if I use a 100u and a 47u in series for the doubler like I posted earlier (below)?

Here's what I'm thinking, wire the cap can I got in parallel, then put another regular electrolytic in series. Do the two have to be equal? if I parallel the 50/50 I will get 100u, so I'll need another 100u for the series cap or can I slap a 47u in there to complete the circuit? I would assume the difference in capacitance will just be a ripple/time constant issue that will be filtered out down the road.

[martin manning]

I think what I'd do is either untwist the tabs (carefully) on the 50-50 series can to save the phenolic mounting plate, then cover the center hole, or make a new one. Mount it back where it was with a 3-lug terminal strip (center ground) under one of the mounting screws. Then get a couple of radial 47u 200V caps and mount them on top of the phenolic insulator plate with the leads passing through and soldered to the terminal strip. The radials will be small and cheap, and the chassis hole will be covered.

Unequal caps in the doubler will give a funny ripple voltage, something between a half and a full-wave rectifier, but it might still work just fine.

[surfsup]

I might ditch the first can cap all together and just wire two axials in and leave the hole. It will be inside the box anyway so who cares. But, i'm on the road until next sunday so...but thanks to all who responded.

Firestorm, the fuse is on the original schemo, it may not b original but looks like it is, and is in the right spot of the circuit.