Hi,

some of you will have seen this famous powered speaker since it has been around for about 18 years. Tens of thousands must have been sold and it spawned numerous clones. The original is better made than most others with a hefty 12 inch woofer and a 45mm diaphragm compression horn. The amp module is particularly solidly built with generous heatsinking and so no need for a fan.

Inside are two, discrete component power amps of 300W and 60W rating. The 300W amp has four power rails for high efficiency of plus and minus 80 & 40 volts.

Both power amps are DC rail fused and have VI limiting to protect output devices in case of driver failure. Both amps have power limiting too, which reduces the output level after short period of full drive to prevent overheating the woofer and horn driver's voice coils.

So they thought of everything - right? Nope.

Both amps are direct coupled to their respective loads with no protection in case of large DC offsets. Both amps use one half of the same dual op-amp as input gain stages - also direct coupled to the rest of the amp.

Failure of one of the 15V rails that supply the op-amp sends both amp's outputs hard over to one of the 40V rails. A few seconds later, the woofer's voice coil and the horn diaphragm are burnt and destroyed.

In the example on my bench, it looks like an intermittent solder joint in the feed to the -15V supply did the job. The amps test just fine.

One 10uF bi-polar cap would have saved the horn diaphragm and a simple triac crow bar would have saved the woofer. A couple of dollars worth of components.



.... Phil

The horn is DC coupled ? Why the hell would they do that ? Sometimes I wonder what kind of drugs these enginers are on.

I was on audiokarma the other day and someone mentioned a protect circuit that wa powered by the DC from the amp. Not hard to figure out how it works, it is a matter of finding a relay, normally closed of course, that will activate at a low enough voltage and current. Couple resistors going to a bridge and a filter should do it.

I commented about finding a suitable relay, someone replied that it has been done and I might be abe to find the link if anyone wants it.

Another thing was about the magnets on the sides of the relay, remember that thread ? you said that it was to draw the arc away from betweent the point so they would be less likely to weld themselves together due to the inductive CEMF.

I posted that a cap, like a 10 uF or whatever, across the cntacts should work just fine. The guy on AK said it might still pass some sound and I replied in the who cares ? you only care about that DC. If the amp is still putting out sound with a ****load of DC on the output, well isn't that just dandy. So what ?

My Phase Linear 400 Series 2 has no relay, no nothing. My buddy's Adcom GFA545-2 is the same. Both of these amps have plenty of current to smoke things. an Ampzilla is the same way. Depends on a fuse. **** really.

At any rate, with the right relay there could be a decent addon protector for this.

I also happen to like SCR crowbar protection. I like the SOUND of "crowbar protection" actually, and how it works. I designed a bridged car amp a loing time ago used that. Of course must short out the drive to the outputs when it is active, which it did. If any fusibles blow, then something is shorted for sure.

I said at the time "This baby is more than idiot proof, you hook up the speaker outputs to the car battery it will burn the wires off.

I like cool **** like that. The UL or CSA or whatever might not, but I do.

wrote:

The horn is DC coupled ? Why the hell would they do that ?

** Most are like that and a real lot have LM3886 amps driving the horn direct.


I was on audiokarma the other day and someone mentioned a protect circuit that wa powered by the DC from the amp. Not hard to figure out how it works, it is a matter of finding a relay, normally closed of course, that will activate at a low enough voltage and current. Couple resistors going to a bridge and a filter should do it.

** The relay MUST short the speaker to protect it and that means the contacts are normally open. Since the coil is not polarity sensitive, there is no need for bridge - just some RC filtering and a fuse on the amp side.


Another thing was about the magnets on the sides of the relay, remember that thread ? you said that it was to draw the arc away from betweent the point so they would be less likely to weld themselves together due to the inductive CEMF.

** Nope, it is to help break the arc that forms when contact open under heavy DC current and voltage. 40V and 5 amps are the limit for most small AC rated relays - adding a suitable magnet might increase that to 50V and 10 amps.


I posted that a cap, like a 10 uF or whatever, across the cntacts should work just fine.

** No it wont.


.... Phil

Phil Allison wrote:


In the example on my bench, it looks like an intermittent solder joint in the feed to the -15V supply did the job. The amps test just fine.

One 10uF bi-polar cap would have saved the horn diaphragm and a simple triac crow bar would have saved the woofer. A couple of dollars worth of components.


** I have now added a 20uF, 250VAC cap to the horn feed and an active relay board to the woofer amp. The NC contacts short the speaker while unpowered and open about 2 seconds after AC power is applied, the NO contacts then close and the amp is connected. If more than +/- 9VDC is detected the relay drops back to it's rest condition. There is a 7Amp fuse in series with the relay contacts in case of an arc developing to ground from the 80V rail.


.... Phil

Not what I would've done but sounds good.

Now how do you write tyhe bill ?

Really, when you modify something there is no price schedule. I now I only charged rtwenty bucks to wrap a flyback for an HK short in the CRT, but that was after a couple of basic charges that brought it up to a livable level..

Time and materials ? Alot of machne shop **** is judged that way. Not all, they have a couple different ways of doing it.

Here's a freaky one. How to figure out what to charge for a modification :

In the GTE Sylvania E 4X series chassis of color TVs there was a chip witht he part number of 15-43301. this was custom or made by them and only for themselves. I supported a countdown circuit for line sync and frequency.

Well a bunch of them even brand new werr coming out with no sync. You got the picture but it was floating out there, and nothing really fixed it. Well I hade a look at three of the pins and with the resistor, cap and whatever thee, determined that I could probably just put a transistor there to detect the sync. It worked.

Then I figured that this is what Sylvania had done,to simply put that transistor inside the package with the IC. I know people who lost a ****load of monet for not being able to get good ICs like this, if I had done this a few years earlier I could have made bucks and kept the landfills a bit cleaner.

wrote:

Not what I would've done but sounds good.

Now how do you write tyhe bill ?

Really, when you modify something there is no price schedule.


** Yep - always tricky to put a price on designing a custom modification that fixes a nasty problem the maker could not be bothered fixing. I simply had to fit the mods cos the dumb fault that fed full rail DC to the speaker voice coils just might return.

The owner has now seen the work and was happy and he has another 11 of them in constant use.

In this case, my bill is well less than the combined bills for re-coning the woofer and a new horn diaphragm - which I estimate will exceed A$400.



.... Phil

On Monday, December 8, 2014 8:09:52 PM UTC-5, wrote:
The horn is DC coupled ? Why the hell would they do that ?

I like cool **** like that. The UL or CSA or whatever might not, but I do..


Price I guess. I don't work on all that many music amplifiers, but I don't recall ever seeing one with speaker protection relays installed. In the case of home audio, they do provide protection because outboard speakers can be pricy (much more expensive than the amp), and folks wouldn't be thrilled to see their amp take out their speakers. In the case of a music amp, they're self contained so I guess manufacturers figure they can sell new drivers if they blow.

As for mods, I've done my share over the years. Technically, we're not supposed to make any modifications unless the manufacturer provides a procedure in a bulletin. From what I heard years ago, the manufacturer is supposed to submit the mod to U.L. (in the States) for approval before it's allowed, but I don't know if that's a fact.

"John-Del" wrote in message
...
On Monday, December 8, 2014 8:09:52 PM UTC-5, wrote:
The horn is DC coupled ? Why the hell would they do that ?

I like cool **** like that. The UL or CSA or whatever might not, but I
do.


Price I guess. I don't work on all that many music amplifiers, but I
don't recall ever seeing one with speaker protection relays installed.


Lots with semiconductor output stages have output relays, if for no other
reason, to stop the switch-on thump as the rails stabilize, but mainly for
speaker protection purposes. Some of the speaker protection circuits can be
very complex. One that comes to mind is that Ecler, that spawned the
original discussion about magnets on the relays.


In the case of home audio, they do provide protection because outboard
speakers can be pricy (much more expensive than the amp), and folks wouldn't
be thrilled to see their amp take out their speakers. In the case of a
music amp, they're self contained so I guess manufacturers figure they can
sell new drivers if they blow.


Only combos are self contained. There are just as many heads and slaves in
use, and these drive external speakers. Contrary to what you believe, the
prices of some of these drivers can be eye-watering ...

Arfa

"One that comes to mind is that Ecler, that spawned the
original discussion about magnets on the relays. "

There are possible other protection schemes when the power amp is married to the speaker. First of all the engineer should know all about what the speaker can take.

Right now I am on a Berhinger B115 with the woofer shutting down at a very low level. There is no relay but it has enough of a protection circuit that it knows the woofer voice coil is shorted. I find that pretty rare but most of the audio I have worked on over the years was consumer. (usually they are open) It is a class D amp, so any overcurrent detection is most likely in the PS. I have not checked.

I also have a Yamaha EMX (something) powered mixer, a two channel wedge design. It actually uses a relay but that is the first I've seen of one at this place. They all seem to depend on fuses and the like. Like this decent Fender BXR or whatever, the dumbass put a 25 amp fuse in it when the fuse blew. One of the transistors blew its top clean off. Now it might become a tradein so I had to check the speaker, which we will use. It is good, but you know if the idiot had not turned it off soon enough it wouldn't be.

One of the problems with consumer audio is the sheer cheapness of it. They need a relay probably for the UL if the thing has any power, and if lower power need it so they don't have to replace woofers when their winpy ass output circuits inevitably fry out. I mean, somonabitch, I had a five channel Sony that backfed something in the protection and fried OTHER channels ! I forget now, but they forgot one ****ing diode. I forgot the exact failure mode because I WANT to forget the exact failure mode.

In a way I almost want to say that speakers should protect themselves, but that leaves the problem of dangerous wiring in case of a failure. The UL would not be happy with that. In fact I bet they would flunk my Phase Linear 400-2 these days. Don't want to add to the legal staff...

I remember working for the (ugh) rental place. They rented (SFS) Fisher systems wiht a CA-270 (discrete) amplifier. It had a relay but not DC protect circuit. You could see the traces on the board for it, as well as a commutatiing power supply for the outputs. All removed. I guess the bean counters and lawyers decided the risk was worth it. They did have some pretty hefty woofers in the some of the speakers, and depended on the five or six amp line fuse to blow I think for protection.

You know, on a manufacturig level, it probably would not cost all that much for those few transistors and ****. Really, the commutators that would save the outputs, a few diodes and **** to detect DC at the output ? This is not expensive. Not when you buy millions of each of the things, and the board is already etched, and even DRILLED for those components.

Ever since the Dodge brothers sued Henry Ford for making a too good product costing them dividends, manufacturers are forced to be so ****ing stingy, really. Few people realize that if a company has the wherewithall to move to China and is publicly traded, MUST do so if it will increase profits. This is not bull****, if they do not, they could be accused of a conflict of interest of worse.

And that really sucks and what has ****ed up this as well as many other industries.

And in the US, including medical. In fact I have a suspicion about that. In medical equipment, the reason a twevle cent transistor is fifty bucks is because that is whay pays the lawyers in this country.

wrote:


The UL would not be happy with that. In fact I bet they would flunk my Phase Linear 400-2 these days.



* I have a Phase 400 series 2 in the workshop, fully repaired but long ago abandoned by its owner.

FYI:

Those darn amps had to be the most DC failure prone of any model in common use - there were so many ways:

1. Output device failure followed by the opening of one DC rail fuse meant that BOTH channels went full rail DC, ie 78 volts.

Since the two DC fuses were common to both channels, a shorted output stage fed wrong polarity DC to the now open rail of the other channel and thence to the speaker via one of the four 1A catch diodes wired across the output transistor groups.

2. Resistor failu there were four resistors on the PCB that were certain fail open in time, the 1.4kohm 5W ones feeding the 15V zeners for the op-amp rails and the 7.5kohms 2W ones for the bootstrap circuit in each channel..

Failure of either 1.4kohms ( which ran stinking hot) sent the op-amps crazy and BOTH channels full rail DC while failure of a 7.5kohm did the same to one channel. Both parts suffered internal corrosion induced by
high DC voltage across them.

Output transistors were prone to fail because of the thermal pads under them - made of plain pink coloured silicone. These pads were thermal insulators and the TO3 devices ran up to 50 degress C above the temp they would if good old mica and white grease were used.

Oh, and there were no thermal cut outs fitted to the heatsinks as in all previous models.



..... Phil

"made of plain pink coloured silicone. "

Mine are not pink.

I agree about the thing being dangerous. the owner's manual says to put speaker fuses in, well why didn't they put in some sockets of they weren't going to do anything else ?

Hm, was speaker protection even out there back then ? that might be a subject of interest. Or not.

Yup, and I own one of those. I really like how it sounds, but it has the evil look of a speaker blower extraordinaire.

It has done in my tweeters and tore the rubber right of my one 500 watt rated car subwoofer. I mean it ripped the rubber right off the aluminum. Believe it or not, at lower levels it stil sounds OK, but the fact still remains I need a pair of woofers. I am due to get them in the next couple days but then I still got the tweeter problem. Either way, for regular house speakers, **** like the 400-2 is danggerous. there is more power out there but it is usually protected.

some of the best sounding amps have nothing. Those old GFAs. the fuvck made them ? Oh, Adcom. And the Ampzilla, though I did not like the fact that the outputs are arrainged in series. However they DID have an actual speaker fuse. I am not sure of what else, and you kow what ? I am not sure what kind of distortion people seek. Sound is available thatis so perfect now that all this old **** would be discarded. WOULD BE.

As long as they pay as they say...

What do youse people down under and in the English grip call a garage ? I don't mean here you get a car fixed, I mena where you go out to smoke and drink becaus the olady doesn't like it.

I mean, I cannot predict this. TYou call the hood of a car the bonnet. You ccall the trunk a boot. What am I to do here ? And don't gimme this **** about l;earning English, there is noone here to teach me.

"1. Output device failure followed by the opening of one DC rail fuse meant that BOTH channels went full rail DC, ie 78 volts. "

Only ONE of them, I think the negative. and it was not full current. you owuld read that on a meter or whatever, but it was not putting out all the current it can.

Yes, they are bad in that respect. That is why I used MJ15024s in mine.

wrote:

"1. Output device failure followed by the opening of one DC rail fuse meant that BOTH channels went full rail DC, ie 78 volts. "

Only ONE of them, I think the negative.


** If one output transistor fails short ( ie suffers punch through ) - then it immediately causes similar failure of at least one other transistor on the opposite side. Then you have as hard short, rail to rail, that makes at least one DC fuse blow.

Yes, they are bad in that respect. That is why I used MJ15024s in mine.

** Most PL400-2s had MJ15024s, later versions were "fully complimentary" with 024s and 025s.

Maybe very early PL400-2s had PL909s - not sure.


.... Phil

On Mon, 15 Dec 2014 23:36:42 -0800 (PST), wrote:

"made of plain pink coloured silicone. "

Mine are not pink.

I agree about the thing being dangerous. the owner's manual says to put speaker fuses in, well why didn't they put in some sockets of they weren't going to do anything else ?

Hm, was speaker protection even out there back then ? that might be a subject of interest. Or not.

Yup, and I own one of those. I really like how it sounds, but it has the evil look of a speaker blower extraordinaire.

It has done in my tweeters and tore the rubber right of my one 500 watt rated car subwoofer. I mean it ripped the rubber right off the aluminum. Believe it or not, at lower levels it stil sounds OK, but the fact still remains I need a pair of woofers. I am due to get them in the next couple days but then I still got the tweeter problem. Either way, for regular house speakers, **** like the 400-2 is danggerous. there is more power out there but it is usually protected.

some of the best sounding amps have nothing. Those old GFAs. the fuvck made them ? Oh, Adcom. And the Ampzilla, though I did not like the fact that the outputs are arrainged in series. However they DID have an actual speaker fuse. I am not sure of what else, and you kow what ? I am not sure what kind of distortion people seek. Sound is available thatis so perfect now that all this old **** would be discarded. WOULD BE.

As long as they pay as they say...

What do youse people down under and in the English grip call a garage ? I don't mean here you get a car fixed, I mena where you go out to smoke and drink becaus the olady doesn't like it.

I mean, I cannot predict this. TYou call the hood of a car the bonnet. You ccall the trunk a boot. What am I to do here ? And don't gimme this **** about l;earning English, there is noone here to teach me.

Speaker protection was fairly common back then. The HK Citation 16's
protection circuit was almost bullet proof. The physical build of the
amp was also impressive. Our speaker company was broken into and the
thief grabbed one of the test amps and drove away with the police in
hot pursuit. The passenger door flew open and the amp rolled down the
hill. We powered up the amp the next day and it worked perfectly.

On Wed, 17 Dec 2014 13:42:17 -0600, chuck wrote:

Speaker protection was fairly common back then. The HK Citation 16's
protection circuit was almost bullet proof. The physical build of the
amp was also impressive. Our speaker company was broken into and the
thief grabbed one of the test amps and drove away with the police in hot
pursuit. The passenger door flew open and the amp rolled down the hill.
We powered up the amp the next day and it worked perfectly.

So did you hire the crook as a stress-tester?

Mike.