multi-section cap adaptor boards?

[nuke]

I'm working on an old amp with a 1.0" 3-section electrolytic. I've seen some circuit boards advertised that fit the twist-lock hole and retainer pattern and allow use of radial capacitors to replace the can.

Except, I can't seem to connect with anyone who can actually sell me one, or at ship them without prohibitive international postal rates (if they can sell me one).

Anyone got an open source board layout?

[lonote]

I am not familiar with the adapter boards you are looking for, but have encountered the same issue a few times & fabricated my own.

After a few cap can failures, I decided to convert my PR to radials & as it is a vintage amp, that meant no new holes in the chassis.

I was able to fabricate a shouldered plug that fit into the original punched-out cap can hole & using that as an anchor, fasten a small turret board to the inside for the radial replacement. No new holes & from the outside the plug looks pretty nondescript.

Easier to the show than to explain.

TDPR Cap2.jpeg

TDPR Cap4.jpeg

TDPR Cap5.jpeg

TDPR Cap7.jpeg

[R.G.]

I've never seen something like this, not that it doesn't exist. Making the twist-lock mounting will be the hard part. Frankly, I suspect that the advertised boards are a gimmick that didn't get made anymore. I could be wrong, but custom metalwork to do the twist-lock stuff is EXPENSIVE.

Better suggestions; as Ionote said, make your own. The turret thing is nice. I might approach it differently, using just a PCB and standoffs.

I'd measure how much PCB area could fit in the space that used to hold the original cap, and spot locations where I could put 3 or 4 standoffs in the chassis without drilling through things. Then I would make a drawing. I'd draw in the hole in the chassis for the original capacitor's wires, and circles the size of the caps to be used, and places for their leads to go through holes, and holes for the wires to the circuits. Even with bare copper clad and hand-drawn etch resist (like Sharpie or nail polish) and ferric chloride, this would take a couple of hours to finish.

A real PCB layout to gerber and fabrication at a PCB place would get you ten of the boards for ~ $30 in two weeks.

[solderhead]

At first glance I had thought this would be an easy problem to solve, as CE Manufacturing started reproducing cap cans a looooong time ago -- but I haven't even looked at multi-section cans since I adopted the snap-in caps and started making my own boards 30 years ago. I seem to recall CE M making cap cans that would drop into the holes, and be held in place by twist-locking tabs on the body of the cap cans using pliers. Would something like that be a decent solution or do I fail to understand @nuke's problem?

I really like the custom plug solution that @lonote came up with, but you really need to be able to turn metal to come up with a solution like that one, so I doubt many people could adopt that solution as the barrier to entry in machining your own shoulder plug is high. If there's enough restoration demand, maybe the market would support a cottage industry manufacturing plugs like that. I can't imagine that demand would be sufficient to justify volume production and stocking inventory unless somebody just wanted to make them as a hobbyist for hobbyists.

I have to wonder about vibration noise with a solution that mounts radial caps on their sides. I'm thinking that it's going to be necessary to place some sort of pad, or some RTV, between the board and the caps to eliminate the possibility of part buzzing.

I gave up looking at cap cans decades ago. The writing was on the wall that cap cans would continue to be unobtainium ... until CE M decided to start making repros. The most feasible long term solution at the time -- if you wanted to have premium-grade / low-ESR / ripple-tolerant caps with high voltage ratings -- appeared to be the adoption of snap-in caps, as they were available in values/ratings that I needed, and premium quality units were in large scale production back then (that is, until CRT TVs became extinct). I like to use snap-in caps because I've never had problems with them suffering from vibration noise. The leads on snap-in caps are designed to act as a spring to hold the cap to the board, and I've seen them working without problems on devices where someone failed to solder them in place. Wow. I like using snap-in caps on 1/16" boards because that board thickness is optimal for the performance of the springy leads to hold the cap in place firmly enough to avoid vibration noise. When I tried using 1/8" G10 boards I had problems with the snap-in function of the caps not working as well. 1/16" boards work better.

I continue to make my own boards out of G10 and PTP parts wiring. For one-off projects this seems to be the path of least effort and I always end up designing and fabricating a custom board. For me there's never been an opportunity to leverage the design into volume production so I've never bothered to scale the learning curve to draw my own Gerber files and have a batch of boards made.

The main problem I'm running into now is that it's getting hard to find snap-in caps with high voltage ratings. Ever since the world has gone to low voltage devices the demand for high voltage caps has really waned. I tried to restock CDE 381 series snap-in caps last night only to find that all of the values that I want for an amp PSU are no longer in production. So now it's looking like snap-in caps with voltage ratings for tube amps are becoming as rare as dodo bird eggs. Now it's beginning to look like adapting to radial caps will become a necessity. I guess the other option is to bend over and pay ridiculous money for axial leaded caps for niche applications, but those prices just seem predatory to me.

[solderhead]

CE-M has a web page that shows how their can caps are manufactured. Perhaps they could supply parts.

https://www.cemfg.com/can-capacitors

[lonote]

I had a couple of the CE cap cans go bad which precipitated the switch.

I had a JJ cap can in-stock & thought it would work, but it was too large a diameter to fit in the existing PR cap hole &, as already pointed out, things are pretty much going radial, so it seemed like the logical thing to do.

Having access to a lathe (& a just-right scrap of aluminum) did make that job easier, though I started down that path initially planning to use a few layers of plastic sheet I had on hand. As long as you can sandwich the chassis with some material to secure the cap board, you are good to go. I would think something similar could be created with G10 or even fender washers.

I have used radials on their side for nearly all recent projects & I do goop them down after they are dry fit first; then goop, then solder. They seem quite secure after the adhesive has set-up. I don't have thousands of hours on any off them, but no failures yet.

[maxkracht]

I default to replacing cans with radials most of the time. Theres often room to fit them in closer to where they are needed, potentially improving the ground scheme. Depending on what it is, drilling a hole to mount a terminal strip or just soldering one to the chassis or some other convenient location shouldn't wreck the value/aesthetics and will make the next guys job much easier when they need to be replaced. Dot of silicone underneath is a good idea.

[solderhead]

I have no first-hand experience with the CD-M can caps, so I can't comment about them, other than to say I'm glad to see that somebody whose business originated in the antique radio sector is manufacturing reproduction cap cans in the USA using the old Mallory equipment ... but one has to wonder if manufacturing variance or reliability problems is why Mallory retired that design/manufacturing equipment in the first place, and whether it's possible in the modern era to manufacture a high quality part by modern standards on the antiquated manufacturing equipment. With modern era manufacturing and parts reliability we've come to expect a lot of our passive components. What used to be the vintage norm in cap failure rates is considered an unacceptable rate of failure for modern electronics. The result is that our expectations of reliability are much higher today than they used to be. The old folks among us will remember when the TV / Radio Repairman had to come to the house periodically to fix our appliances because condenser failures were common.

If you're in a situation where aesthetics really matter and you absolutely have to have a period-correct looking part for something like an antique radio restoration, maybe it's going to be a necessary to accept a higher failure rate that's on par with Old World historical norms. To avoid that I've gone so far as to gut the cap can and to shoehorn modern caps into the can.

I just realized that my abbreviation of Cornell-Dublier Electronics as CDE, and of CE Manufacturing as CE M might be a source for confusion to someone who doesn't already know the suppliers. CDE has long been a premier supplier of caps.

I would think that G10 would be very hard to use in this sort of application. G10 is very hard, difficult to cut, and wears tooling quickly. IIRC it is actually harder than aluminum. But then I'm no machinist.

@lonote, what are you using as a standoff between the turret board and the plug? From the pics it appears that the plug is flush with the chassis but the board is sitting on some sort of standoff that would have to be a wider diameter than the hole, and made of a non-conductive insulator.

[jabguit]

CE-M has a web page that shows how their can caps are manufactured. Perhaps they could supply parts.

https://www.cemfg.com/can-capacitors

very cool

[martin manning]

Another option: https://hayseedhamfest.com

[lonote]

@lonote, what are you using as a standoff between the turret board and the plug? From the pics it appears that the plug is flush with the chassis but the board is sitting on some sort of standoff that would have to be a wider diameter than the hole, and made of a non-conductive insulator.

Good catch, there is indeed a lower insulating board that is tight to the chassis creating the sandwich with the plug. I smeared a little goop under the sub board to keep anything from spinning since the plug didn't have any means of indexing.

I left that image out as I didn't want to upload too many pics. Here is the set before installing.

TDPR Cap1.jpeg

Side note, I have seen many installations of radials on their side with the leads all the same length & eyelets/turrets all in a long row. I have found that staggering the length of the leads allows 2 parallel rows of terminals, making a ground buss a straight shot & even with tight cap spacing, the dropping resistors link + to +. It can make a nice tidy package. I didn't implement it in either of these instances, but that layout also facilitates giving the reservoir cap it's own dedicated ground if needed.

TD RevAmp PSU board.jpg

[R.G.]

One other option is the arts-and-crafts approach. Drill out the center of the existing capacitor, then pick and pull out the stuffings, leaving an empty shell. Insert same-capacitance but much smaller modern radial caps into the shell and glue them into place. It's tricky and leaves possible thermal issues, but works. I've done it when desperate to preserve looks over function. At one point, I tried to buy empty shells from one of the remanufacturers.

[nuke]

Here are some links to such things:

Granger:

https://grangeramp.com/product/multi-se ... tor-board/

https://www.proaudioe.com/3-in-1-cg/

The idea is three or four radial caps mounted to a board with the same footprint as a can cap, retained by the two screws that hold the twist-lock adapter plate in place. The lugs are simply press-fit lugs from Keystone - off the shelf hardware.

Here's an example image:

s-l1600.png

[nuke]

A 3-section 20-10-10 @ 450v is easily replace by 3-radial capacitors even in the 1-inch format.

The 4-section type is 1-3/8" diameter and easily accommodates 4 such caps, and there are "wide" 4-sections.

I managed to resolve my own immediate need, remembering that I'd ordered a 3-section, 1-inch can cap for a Magnatone that I never got around to using and found it in my parts inventory. I think it is about 10 years old, so I just let it reform for 24 hours for general principles.

Unless you have a "hot" negative on something, the circuit board and radial capacitor board would be a great solution.

[nuke]

Another option: https://hayseedhamfest.com

Yeah, that's a good source too.