advertisement


Elcap / Alcap capacitors in a 80's crossover

Are the 3MFD and 4MFD (bottom left of picture in original post) in parallel?

Sadly not - the circuit is more complex in that area - they are neither in parallel nor in series - one of the inductors sits between the two.

never understood MFD, is this American?

I was just writing it as the values are printed on the component! I agree that technically it's wrong, but I guess in those days they didn't find it so easy to print the mu symbol?
 
Thanks, but I've gone with Jantzen NP electrolytics. I spent way too much time messing about figuring out combinations of caps in series, only to realise that it's easier (and cheaper) to do what a.palfreyman suggested, and look at combinations in parallel first!

I would have placed an order with Falcon, but I got to the checkout and discovered that they have a £15+VAT minimum order, and their minimum shipping charge is £5.40. If I can ship a few capacitors for less than £3 then I don't see why I should be paying them getting on for double that.

So I went with HifiCollective - no silly minimum charge and reasonable postage costs too.
 
Going high cap' is totally impossible. Think about it logically - to go high cap' you'd have to increase electrode surface areas in proportion. It is likely a consequence of high(er) leakage current as that screws the measurement method.

The capacitance is not just dependent on the actual surface area but also on the dielectric and electrolyte though and back in the '60's and 70's it was not unusual for tolerance on general purpose electrolytics to be -50/+85%! I can well imagine the capacitance drifting high as well as low. Maybe over time chemical action of the electrolyte or impurities in the aluminium or electrolyte can cause some further etching and forming of the aluminium foil leading to increased effective surface area? Of course in the case of bipolar caps used for crossovers these tolerances are useless and such caps were normally +/- 20% tolerance when new.
 
The capacitance is not just dependent on the actual surface area but also on the dielectric and electrolyte though and back in the '60's and 70's it was not unusual for tolerance on general purpose electrolytics to be -50/+85%! I can well imagine the capacitance drifting high as well as low. Maybe over time chemical action of the electrolyte or impurities in the aluminium or electrolyte can cause some further etching and forming of the aluminium foil leading to increased effective surface area? Of course in the case of bipolar caps used for crossovers these tolerances are useless and such caps were normally +/- 20% tolerance when new.

Pretty much all very true, but being a Chemist/Chemical process engineer by qualification and inclination, I will stick with my statement that it is totally impossible, or at least to any extent that could be detected compared to a spec' and/or without before and after measurements.
For instance, to re-etch a foil and hopefully increase surface area and hence increase cap', the dielectric would have to be removed, the etching take place and then the dielectric reformed. With no dielectric, it would be a resistor, not a capacitor. I'd bet a lot against that happening, especially in a non-polar as it would be far more complicated than that..

Non-polar electrolytic caps are normally two polar caps with their anodes soldered together, so become extremely poor in terms of volumetric efficiency as you obviously halve the cap' when doing so, though hardly a hassle in something like a cross-over. (An anode absolutely must not become a cathode, but a cathode can perform the duty of an anode).
 
Pretty much all very true, but being a Chemist/Chemical process engineer by qualification and inclination, I will stick with my statement that it is totally impossible, or at least to any extent that could be detected compared to a spec' and/or without before and after measurements.
For instance, to re-etch a foil and hopefully increase surface area and hence increase cap', the dielectric would have to be removed, the etching take place and then the dielectric reformed. With no dielectric, it would be a resistor, not a capacitor. I'd bet a lot against that happening, especially in a non-polar as it would be far more complicated than that..

Non-polar electrolytic caps are normally two polar caps with their anodes soldered together, so become extremely poor in terms of volumetric efficiency as you obviously halve the cap' when doing so, though hardly a hassle in something like a cross-over. (An anode absolutely must not become a cathode, but a cathode can perform the duty of an anode).

Hmm... I'm not so sure...

Bipolars are in fact apparently not made as you say, and as I'd assumed, from two caps in series inside one case but by a different process which does IIRC have something to do with a cathode being able to become an anode but combined with some special process for 'tother terminal.. sorry can't remember details and it ain't important enough for me to route through piles of WW mags to find the article by Cyril Bateman, who was once a capacitor designer for IIRC Erie capacitors. I recall thinking "well I never!" at his explanation.... Electrically of course it is two caps back to back but physically it is kind of "one capacitor who's construction makes it behave like two in series back to back"... again IIRC! I ain't a chemist... and I don't make capacitors... I'm just an end user partially recalling an article on it I read a year or two back....

This bit may interest some though... that apparently in some respects bipolars beat the specs of normal electrolytics in terms of distortion and ESR... seems unlikely but so said the electrolytic cap designer...
 
That is how they were made by STC/AVX in Paignton - I stood and watched them being made.

If I stumble across the article I'll add some detail... I vaguely recall mention of "true bipolars" as opposed to ones which are physically two caps in one case...
 
ESR is an oddball thing, not subject to obvious logic. The obvious one being that small case/anode sizes have high ESR - my logic would have said low..

Don't quote me - it was over 20 years ago, but I have a very vague recollection that non-polar tantalums were sold to some customers for their low ESR.
 
HAST I cannot remember the meaning of but DHSS is damp heat

In semiconductors HAST is 'Highly Accelerated Stress Test' and uses Temperature, Voltage and Humidity @ Pressure...so a big autoclave. Not sure if the same thing for caps?
 
In semiconductors HAST is 'Highly Accelerated Stress Test' and uses Temperature, Voltage and Humidity @ Pressure...so a big autoclave. Not sure if the same thing for caps?

Thanks for that.
The test-house in Paignton was NPL approved and was actually a secure area as a result, but I did have one look around, and autoclaves were not something that I noticed. Whatever the specifics for the test were/are, they did not include humidity - hence my conclusion about migratory ionic species within one of the resins. HAST was/is, for tantalum cap's, done without moisture above what is in the air naturally, indeed, I have an extremely vague recollection that warm and humid conditions all but shut the Test House down as it was not air-conditioned, so if I am right, there wil be an upper limit for HAST in terms of dew point and that would likely be in the region 10-15C, as an educated guess.
 
See link:
https://passive-components.eu/capacitors-electrolytic-capacitors/
Some useful info, but only includes a cursory note of back-to-back NP elcaps

I only know about tantalum wet and dry electrolytics, and quite a few details for those in the link, are wrong, but not so far out as to be entirely misleading. Given that all the big three manufacturers produce very similar products, in terms of electrical properties and also things like case size, suggests that the STC/AVX/Kyocera products that I am familiar with are very, very similiar to the others in how they are made. There are or were also only 2 major suppliers of tantalum powders suitable for cap' manufacture, and they offered standard ranges.

One of the most intersting bits about powder manufacture is the shape of the particles - they are minute flat sheets for the great majority of end uses. Now that is one very, very clever bit of chemical engineering.
 
You need to poke 'em with a sharp stick and tell 'em to get there act together. ;)
.......One of the most intersting bits about powder manufacture is the shape of the particles - they are minute flat sheets for the great majority of end uses. Now that is one very, very clever bit of chemical engineering.
Ants with anvils!
 
An SEM of a modern tantalum, low volt anode looks like a porous, random stack of thick roofing slates with all the slates of very similar size/surface area. Without uniformity in particle surface area, you could not predict final cap' of the finished component. VERY clever.
 
HiFiCollective were quick - the parts arrived this morning.

Back down to the mundane:

Is it fair to assume that there are no special rules about mounting several capacitors together (where I'll have to put two in parallel)? It seems unlikely that there will be at audio frequencies. So no reason not to just stack them - no need to keep leads especially short or the same length?
 
No, not that I'm aware of. This isn't like feedback caps in an amp or similar where loop area is important. You can always drill judiciously placed holes in the circuit board and pass tie-wraps through to make things nice and stable. You don't want something 'falling over' with the possibility of shorts or broken wires.
 
Thanks - there's a lot of track area on the PCB so I'll just drill the holes out so that 2 leads will go through one hole, and I'll glue them down to the boards and/or to each other (if stacked) - the originals were glued down from the factory.
 


advertisement


Back
Top