Illinois capacitors

[Helmholtz]

Form factor required the use of axial caps in many of our products.

As there are/were a few competitors as well, I estimate the annual worldwide high voltage ecap market to several hundred million pcs.

[nevetslab]

And, as many of us have been finding, Nichicon has left the market with Axial Lead Electrolytic Caps, and those that inventory Ruby, Panasonic, Vishay/Sprague and others (besides CDE/Illinois products) has left us with not a wide range of choices, bound to get worse, since our usage is a mere pittance of what Component mfgrs will tolerate to keep production of low-volume sales items.

[SoulFetish]

One word - Panasonic. (check the specs)

As far as electrolytics go, I can definitely get behind Panasonic
I ordered some some ultra low ESR (FM/FC series) caps to meet the requirements of a DC heater supply simulation for a project.
Very well constructed.

And we required 50.000 hours of service life with less than 10% failures.

50,000 hours our of an electrolytic?

These days the best high voltage ecaps are made by Japanese companies.

No love for your native F&T brand?

[Helmholtz]

50,000 hours our of an electrolytic?

50K hours at 60°C roughly corresponds to 3K hours at 105°.
We did real lifetime tests over years with hundreds of units for qualification (imagine the huge waste of energy).

No love for your native F&T brand?

I sometimes use them for vintage amp restoration. I consider them a niche product. No problems so far.
But I don't have long term lifetime statistics as with the caps we used in the company. And I don't consider them suitable for SMPS application.

Generally SMPS applications are much more demanding for ecaps than conventional power supplies because of high HF ripple currents.

[g1]

SMPSs used at 230V mains require high voltage ecaps. Especially where PFC (boost convertor) is required, you need 450V (sometimes 600V peak) rated caps.

10, 22, 47 uF? I see 100uF or more in SMPS?

[Helmholtz]

10, 22, 47 uF? I see 100uF or more in SMPS?

What do you mean?
Required capacitance depends on wattage.

[g1]

What do you mean?
Required capacitance depends on wattage.

I'm asking if it's common to find those values in SMPS. (10, 22, 47uF @ 450 or 500V). I don't work on much SMPS, especially for non-music related stuff.
If those kind of values are commonly used, then that is better for us. If not, maybe tube guitar amp makers should start designing with values that are more common to be mass produced for the consumer electronics industry.

[Chuck H]

I vote Nichicon. Though they don't offer axials for their higher performance product lines. I've been learning to work with radial caps on eyelet board builds.

[SoulFetish]

I'm asking if it's common to find those values in SMPS. (10, 22, 47uF @ 450 or 500V). I don't work on much SMPS, especially for non-music related stuff.
If those kind of values are commonly used, then that is better for us. If not, maybe tube guitar amp makers should start designing with values that are more common to be mass produced for the consumer electronics industry.

Manufacturers of SMPS require high performance, high capacitance filter capacitors with very low ESR, low inductance, and low reactance at HF. This has created a market for film capacitors with capacitance and voltage ratings not historically available. This is a really positive development for those of us in the niche market of sourcing parts for tube amp power supplies, because they are low profile, inexpensive, and readily available in values and voltage ratings found in almost every tube amp ever made.
So, you could surmize that the same would be true of electrolytic caps competing in the same market.

[Helmholtz]

I'm asking if it's common to find those values in SMPS. (10, 22, 47uF @ 450 or 500V).

That's the cap range we mostly used in the DC link of our products. Might be different with general purpose SMPS (flyback designs) that don't require active PFC and a 400V DC link.

[DFuller]

Honestly, I've seen more Illinois electrolytics go bad than any other brand. In the last year I've had 2 HRDs and two Peavey 6505+'s that needed a total recap of the HT supply. The 6505+'s were both at oldest a decade old and the caps were starting to go - not enough to hear hum, but enough for the amp to start misbehaving when a lot of current was demanded by the power amp (the plate supply caps were the worst offenders and lo and behold, scoping the plate pin showed a 10V p-p sawtooth!). As far as the HRDs, something about those amps seems to cause premature failures in their filter caps. I encountered the same exact electrolytics on DRRIs and TRRIs that are 2+ decades old and they worked just fine.

[Chuck H]

Honestly, I've seen more Illinois electrolytics go bad than any other brand. In the last year I've had 2 HRDs and two Peavey 6505+'s that needed a total recap of the HT supply. The 6505+'s were both at oldest a decade old and the caps were starting to go

I think this is actually supportive of any counter arguments.?. The HRD's and 6505's are probably the most popular production amps on earth. If several out of millions fail that's not a surprise. Certainly there's a bean counter somewhere that knows the exact failure rate of these caps in these amps and he/she has determined them to be acceptable.

But I agree with you actually. Ergo I worded the above response with "acceptable". Certainly there are better caps available. I think that the savings on the caps must outweigh the warranty period failure rate expense. Of course the warranty period isn't the only thing considered either. No major brand wants a bad reputation. So whatever pissing and moaning from repair techs or in the know DIYers and players must be small enough compared to accolades that it too is considered acceptable.

This is just the nature of production amps now. Though it wasn't always that way for all brands it was for most even historically. And that must be how we ended up here. The cheaper built amps eventually demonstrated a business model that forced the market.

In this light IC caps are just fine, right? They are industry standard. We shouldn't expect better (and I just threw up a little in my mouth).

[Enzo]

Whenever we have this sort of discussion, we never have the context..OK< just what IS the failure rate? My Cousin Vinnie was on again tonight. Central to the trial was this (paraphrased) exchange. So are you telling me the most popular size of the most popular model of the most popular brand of tire was on the defendant's car? Well...yeah...But TWO metallic mint green cars with the same tire?

People only bring their amps to me when they fail. No one brings an amp in to say it works fine.

[DFuller]

I think this is actually supportive of any counter arguments.?. The HRD's and 6505's are probably the most popular production amps on earth. If several out of millions fail that's not a surprise. Certainly there's a bean counter somewhere that knows the exact failure rate of these caps in these amps and he/she has determined them to be acceptable.

But I agree with you actually. Ergo I worded the above response with "acceptable". Certainly there are better caps available. I think that the savings on the caps must outweigh the warranty period failure rate expense. Of course the warranty period isn't the only thing considered either. No major brand wants a bad reputation. So whatever pissing and moaning from repair techs or in the know DIYers and players must be small enough compared to accolades that it too is considered acceptable.

This is just the nature of production amps now. Though it wasn't always that way for all brands it was for most even historically. And that must be how we ended up here. The cheaper built amps eventually demonstrated a business model that forced the market.

In this light IC caps are just fine, right? They are industry standard. We shouldn't expect better (and I just threw up a little in my mouth).

You know what's funny, though? I owned an older 5150 II from like 98 that was bone stock when I got it last year, and it had ChemiCons for the bulk filters! That amp had no issues at all with the caps.

Also, I did some looking, those IC LBA series caps that Peavey uses are sub-$3 in quantity (Mouser shows 2.99 per in packs of 2500, so probably in the $1.50 region per in the massive quantities Peavey would be buying). They're 3000hr at 85C/350V parts though which isn't bad, but 2000hr 105C/400V Nichicon LGG series are $3.54 each in quantities of 1,000. So, assuming pricing keeps dropping, I doubt they'd be substantially different in price.

But yeah, they don't care. Why would they? They're good enough to get out of warranty most of the time, and they save on BOM costs. They've gotta have profit margins somewhere on these....

w/r/t Fender, it's bizarre that I only see the consistent TTA series cap failure on Hot Rods. They use the same parts in the Blues Jrs, PRRIs/DRRIs/TRRIs, and so on, but I don't see those go bad with anywhere near the regularity that I do HRDs. But, Enzo's right as usual. We only see it from the perspective of broken amps.

[g1]

w/r/t Fender, it's bizarre that I only see the consistent TTA series cap failure on Hot Rods. They use the same parts in the Blues Jrs, PRRIs/DRRIs/TRRIs, and so on, but I don't see those go bad with anywhere near the regularity that I do HRDs. But, Enzo's right as usual. We only see it from the perspective of broken amps.

What failure mode are we talking about? I could be mistaken, but seem to recall Chuck talking about the number of phase inversions on the preamp supply node, and thus the HR series having the issue of howling when the preamp filters (decouplers) stop decoupling, where other amps don't exhibit the same symptom and it goes unnoticed?
Chuck H ?