API 2520 Failure Rate

[Matt Syson]

Hi
It almost suggests there was a 'batch' problem for so many to fail. It would be interesting in an 'anal' kind of way to know WHY they failed.
'Theres always a reason for it' as some lyrics have it.
Matt S

[dandeurloo]

I think it is a very common problem. I just answered this in another Geakslutz thread as well. (API 55AB gone bad???)

Answer = Red Dots.

[drBill]

Originally Posted by dandeurloo

Answer = Red Dots.

Or GAR 2520's or 1731's....

[effitall]

Yeah...Dead 2520's aren't an uncommon occurrence unfortunately. I suspect it has to do with the potting. I had two genuine 2520's die on me...None of my "clone" 2520's have let me down. Also...The Melcor 1731 seems to be a fairly nice option if you want something a little "creamier" in sound to the 2520.

[Tim Abraham]

The Melcors sound like an interesting option. I actually have 4 red dots already. I did some fairly extensive testing with them when I first got them. I a/b'd them to the 2520's in my 2500, 5500, and 3124 by printing mixes through them (the 3124 with a Folcrom). In the end I preferred the 2520's everywhere except the 2500. (which felt tighter and a bit cleaner with the red dots

[jrp]

it would be fantastic if you could share some cips from those tests.

[Retinal]

Yep, I was the last victim of a dying 2520..

I have a bunch on API but this is the first that an op amp fails on me,
this eq was used but mint, really.. so I don't think it has anything to do with it
plus it was checked by VK before shipping.

After reading all these horror stories about dying 2520's I'm terrified,
especially because we can't really stack up on those due to API politics..
Now I find this really stupid, I understand - to some extent - why they do it
but dude, seriously.. either you make sure your stuff doesn't die on customers
or you make the spare parts easily available (and also not so expensive I'd add..)
I love API sound and as I stated in the other thread I don't want to replace
the ops with something different, whether is suppose to sound better or not,
I want that sound and I want it to be reliable.
I see it as a pretty normal request since API gear is not exactly cheap

[Tim Farrant]

I am guessing that the failures are due to thermal problems, the output transistors are running away and blowing. Perhaps if you have the gear in racks where it is getting really hot, this may not be helping the situation. This used to happen all the time to the AM10 opamp used in the Quad 8 console I maintained years ago. I understand also that the API 2520 is now surface mount and these parts are going to have less tolerance to high temperature than their big buddies.

[Retinal]

Originally Posted by Tim Farrant

I am guessing that the failures are due to thermal problems, the output transistors are running away and blowing. Perhaps if you have the gear in racks where it is getting really hot, this may not be helping the situation. This used to happen all the time to the AM10 opamp used in the Quad 8 console I maintained years ago. I understand also that the API 2520 is now surface mount and these parts are going to have less tolerance to high temperature than their big buddies.

Still, we all keep all the gear in racks :/
never had problems with the 19" units so far,
but I have my 550's in lunchboxes and those are not in racks
I never noticed lunchboxes to get particularly hot, def not as hot as 16/20 units
racks.

[Matt Syson]

Hi
For gear powered 24/7 it is surprising how even very small differences in 'airflow' can have a dramatic effect on component reliability. If a particular device is say 5 degrees warmer it can make a real difference to it failing. If you recorded data on such failures over hundreds of units, a pattern would emerge.
Matt S

[MicDaddy]

2/4 2520s failed out of a 3124+ unit that are still under warranty (intermittent, no audio, static/spitting) Just worried when the other 2 will fail, or if any of the 4 from another 3124+ with the same datecode will fail.

Built up some GAR1731 and GAR2520 to fit in until I can send off to API for replacements. Units are working great with these opamps.

Has anyone had any experience with API warranty service, turn times?

[Matt Syson]

Hi
Some early batches of SSM2017 mic amp chips had a strange failure mode where they would work correctly for hours or a few days, then the gain would drop to minimum but still work properly except no gain. This was eventually traced to a scratch on the photoplate used for the die manufacture which had a scratch on it giving a 'weak internal gain set resistor.
They 'fessed up' and replaced them (thousands!).
Matt S

[Retinal]

Would be nice to know from API is there was some sort of bad batch
or this is like.. how's always gonna be.
I love API like no other gear, but in my book reliability is as important
as the sound, actually even more, there's no point in having
great gear that can fail you any moment in the middle of a session.
I also wonder if the 1608 (since I'm planning on getting one next year)
has the same problem.. I assume it does since the op amps are the problem,
but damn, replacing op amps is gonna be a second job..

[Marcocet]

Originally Posted by Retinal

Would be nice to know from API is there was some sort of bad batch
or this is like.. how's always gonna be.
I love API like no other gear, but in my book reliability is as important
as the sound, actually even more, there's no point in having
great gear that can fail you any moment in the middle of a session.
I also wonder if the 1608 (since I'm planning on getting one next year)
has the same problem.. I assume it does since the op amps are the problem,
but damn, replacing op amps is gonna be a second job..

There's definitely was a bad batch of 2520, prob about four-five years ago. My 1608 was built in the first batch (serial 11) and had a whole pile of 2520's that went bad. Luckily API was extremely accommodating as far as sending spares and replacing ones that die. I imagine for a smallish company (I know they're big for pro audio, but that doesn't make them anywhere near big!) it would be a huge expense to do a complete recall, especially since most of them seem to be holding up fine. Easier to just replace them as they fail, which they will do, though I know nobody appreciates the downtime. If I had to guess I'd say that prob 10 or so have died on me over the last couple years. And considering the fact that the AC in our current space doesn't quite cut it and the desk is always hotter than it should be I don't think that's all so bad. But to each their own.

I've tried a lot of different 2520 sized opamps, and I must say that even though there were ones I'd prefer for certain uses, I would never pay to replace a functioning API 2520. I just haven't heard them beat THAT badly. But if API is trying to charge you for a non-warranty replacement I've had great experiences with the Gar2520 DIY kits, Avedis 1188s, Purple KDJ4s and I'm dying to hear the Scott Lieber red dots.

[George Necola]

doesn't help much, but there is a replacement OP amp ghosting trough the internet...

sorry, I forgot the manufacturer.

[ionian]

Could it be the potting? I heard that they pot their 2520s and I'm not entirely convinced that is the best course of action.

Anyway, the new Inward Connection VF600 amp blocks (used in their 500 series modules like the Brute) are supposed to be drop in replacements for 2520s as well.

Paul (Riffmachine) swapped them for the 2520s in his 5500 in this thread here with some samples.

API 5500 - 2520 vs. Inward Connections VF600 - Clips

Regards,
Frank

[Matt Syson]

Hi
Potting could be good or bad for a circuit but it takes the 'design' of the circuit to a whole new level when you have to consider thermal gradients in and among electronic circuit parasitic effects. You are then entering the specialised world of integrated circuit design although at a 'ham fisted' level in comparison to real ICs.
One would hope the boys and girls at API KNOW what the issue is, but you wouldn't expect them to tell anyone.
Could be something subtle like they changed to lead free solder or some surface mount parts a few years back and didn't remodel the thing properly.
Matt S

[DavidChampoux]

I needed to replace a lot of 2520's in my 1608... Most of them were from August or September 2008.

API always gave me VERY GOOD and QUICK service.

I think the 2520 are fragile to static. Beware of carpet, etc...

[nosebleedaudio]

Originally Posted by ionian

Could it be the potting? I heard that they pot their 2520s and I'm not entirely convinced that is the best course of action.

Anyway, the new Inward Connection VF600 amp blocks (used in their 500 series modules like the Brute) are supposed to be drop in replacements for 2520s as well.

Paul (Riffmachine) swapped them for the 2520s in his 5500 in this thread here with some samples.

API 5500 - 2520 vs. Inward Connections VF600 - Clips

Regards,
Frank

I disagree, Have seen many MUCH older 2520's (70's) that still worked great AND drew more current, around 15 to 18 mA...They are potted..

[Matt Syson]

Hi
Changing the internal design or layout of these devices or even the formulation of the 'potting' compound could contribute to failures. Altering even seemingly insignificant items can produce all manner of unforseen results. Politically correct solder and potting compound could be two, alternative manufacturer of the internal components another.
Matt S

[kaoss]

Originally Posted by Marcocet

There's definitely was a bad batch of 2520, prob about four-five years ago. My 1608 was built in the first batch (serial 11) and had a whole pile of 2520's that went bad. Luckily API was extremely accommodating as far as sending spares and replacing ones that die. I imagine for a smallish company (I know they're big for pro audio, but that doesn't make them anywhere near big!) it would be a huge expense to do a complete recall, especially since most of them seem to be holding up fine. Easier to just replace them as they fail, which they will do, though I know nobody appreciates the downtime. If I had to guess I'd say that prob 10 or so have died on me over the last couple years. And considering the fact that the AC in our current space doesn't quite cut it and the desk is always hotter than it should be I don't think that's all so bad. But to each their own.

I've tried a lot of different 2520 sized opamps, and I must say that even though there were ones I'd prefer for certain uses, I would never pay to replace a functioning API 2520. I just haven't heard them beat THAT badly. But if API is trying to charge you for a non-warranty replacement I've had great experiences with the Gar2520 DIY kits, Avedis 1188s, Purple KDJ4s and I'm dying to hear the Scott Lieber red dots.

Originally Posted by DavidChampoux

I needed to replace a lot of 2520's in my 1608... Most of them were from August or September 2008.

API always gave me VERY GOOD and QUICK service.

I think the 2520 are fragile to static. Beware of carpet, etc...

Ya, there definitely was a bad batch. I had to replace a lot of 2520s in my 1608 - some 20+ so far. It was almost always the ones marked Aug 2008.

Hopefully API have the issue under control now. Here's a link to their rep mentioning switching outside opamp manufacturer
API 2500 opamp malfunction (2520)

My board came with a 'service kit' which includes six spare 2520s and an extra channel strip, so this helped a lot when I had to ship a strip to my dealer for service a few times. The dealer (Economik) also changed ribbon cables in all my channel and echo strips.

The opamp and ribbon cable replacements were done fast and free of charge, and it's needless to say that both API and Economik are always very responsive and nice to deal with.

I love API sound and don't feel like I want to spend time trying third party opamps but Scott Lieber's website definitely looks pretty intriguing

[Radardoug]

+1 Jim! Don't you hate that bad design?http://static.gearslutz.com/board/im...lies/thumb.gif