Is it just me or have any of you wondered if replacing an axial electrolytic with a radial might not be 100% ideal?
Some years back a very knowledgeable technician commented on how some people like to mount components on long leads....way off the board. I'd never thought about it before, in fact I've done it myself thinking that the part, usually a hot running resistor, would run cooler.
As it turns out, those nice long leads act as antennae and will pick up any RF that happens to be in the vicinity. Also, although they will discolour the board, resistors apparently run cooler closer to the board because the board and traces act as a heatsink.
Back to the caps. What worries me with radials replacing axials is the rather long leads one inevitably ends up with. This is even more the case with vintage equipment as the new component will be a fraction of the size of the original meaning more lead is needed to get to the solder pad.
Am I being paranoid? I've done it in the past and always got away with it. But after reading that most interesting thread on Tiger amps, inherently unstable at the best of times....
Perhaps a good compromise would be to mount the radial with the 'hot' lead as close to the pad as possible and make the ground the long one?
What say the experts?
Some years back a very knowledgeable technician commented on how some people like to mount components on long leads....way off the board. I'd never thought about it before, in fact I've done it myself thinking that the part, usually a hot running resistor, would run cooler.
As it turns out, those nice long leads act as antennae and will pick up any RF that happens to be in the vicinity. Also, although they will discolour the board, resistors apparently run cooler closer to the board because the board and traces act as a heatsink.
Back to the caps. What worries me with radials replacing axials is the rather long leads one inevitably ends up with. This is even more the case with vintage equipment as the new component will be a fraction of the size of the original meaning more lead is needed to get to the solder pad.
Am I being paranoid? I've done it in the past and always got away with it. But after reading that most interesting thread on Tiger amps, inherently unstable at the best of times....
Perhaps a good compromise would be to mount the radial with the 'hot' lead as close to the pad as possible and make the ground the long one?
What say the experts?
Radial capacitors have both leads on the bottom. Axial capacitors have one lead at each end.
In general component lead lengths should be minimized for best results, especially at high impedance nodes.
Internal heat is a big cause of capacitor failure. Screw terminals connected to a metal bus bar can help with this.
Many components of various types do rely on the pcb foil to dissipate some heat.
Last edited:
Thanks for clarifying that rayma, I just assumed everyone would know the difference between an axial and a radial capacitor. Perhaps I should have given an example like this - outgoing cap has an effective lead pitch of 40mm, the new replacement radial has a lead pitch of 5mm. No matter how you snug the new cap up against the board, there is no way of getting around 35mm of exposed lead on at least one leg.
All comments are welcome. I only have thirty odd years of hands-on experience at the bench, so compared to some of you I am a novice. My Achilles Heel is that I have no formal electronics training, which is why I come here to learn from people more knowledgeable than myself. I'm also have no problem conceding I may be wrong - it would be nice to know why though.
It sucks for old gear, but that's the way it is. I'm very happy with radials, more space effective and even shorter leads. Even their size to capacitance/voltage rating seems better. 20 years ago axials were already on the exit.
Believe it or not, HV axial capacitors are alive and well, made for guitar amps by CDE for example.
Yeah, but admittedly that still a little niche. Radials have become very wide spread up to 100V, after that they quickly become unobtainable.
Also, a guy I know involved in industrial electronics told me he prefers axial because they have a lower profile when mounted making the boards more mechanically robust. We had this conversation about 25 years ago so perhaps things have changed, I can ask. But in (shall I say domestic solid state) audio you rarely see them.
Yes, there are low profile electrolytics that are about as wide as they are tall, not so good for ripple current, though.
Sometimes they have an extra lead or two to mechanically stabilize them.
Frankly rayma, I feel completely overwhelmed whenever I start thinking along those lines. I never got into embedded design - every now and again I start to research something and when I realise how little I know...well I just feel plain stupid. I thought I knew a little about caps until I read the Wiki on them last night. Then I realised I know nothing at all 🙁
Well, once you realize that you need to learn, you're halfway there. There are various levels of understanding.
Just look around for material that seems to be on the right level, and go on from there. What's your background?
If you'll send me your email, I can send you some ebooks.
Last edited:
Thanks rayma, that's very kind of you. I heard someone once say they'd reached a stage of 'paralysis of analysis'....kind of feel like that sometimes. Will send you a PM.
You just go step by step, each time building new knowledge on what you've already learned.
don't forget to insulate those long leads. I've used both and not had any issues but it's good that you are thinking deeply on the subject
Legs acting as antennae to pick up RF signals? Even if this were the case, it wouldn't be a problem as RF isn't audible, and the amplifier cannot amplify it, due to lack of bandwidth.
- Status
- Not open for further replies.