eti 480
Hello i have been talking to my lecturer about the ETI 480,
He has had experiences with this amp in the past,and has tested them and found that they oscillate at about 180khz.
Probably this is what happened to your amplifier.
Often this is caused by different suppliers of the transistors from different manufacturers. The gains varied dramaticaly from batch to batch.
The mismatch of gains led to instability.
Ideally the frequency response should be rolled off above 30khz
by fitting capacitors in the feedback path or using a new set of transistors with lower overall gain.
Can anybody add some more info to this?
Hello i have been talking to my lecturer about the ETI 480,
He has had experiences with this amp in the past,and has tested them and found that they oscillate at about 180khz.
Probably this is what happened to your amplifier.
Often this is caused by different suppliers of the transistors from different manufacturers. The gains varied dramaticaly from batch to batch.
The mismatch of gains led to instability.
Ideally the frequency response should be rolled off above 30khz
by fitting capacitors in the feedback path or using a new set of transistors with lower overall gain.
Can anybody add some more info to this?
haha ive posted so many times
http://www.alphalink.com.au/~cambie/
get the eti480 first pdf file-
i scanned it all-it says-
''notice that the designers have ommited an inductor on the output,and seem to ignore its absence'' or somthing like that.
its needed for non oscillation i think.(they didnt expect the later transistors to be different ?)
the eti 466 has more stable features.
and i think im nearly doen posting lots in this thread..but i dont want to be an eti 480 casualty.
http://www.alphalink.com.au/~cambie/
get the eti480 first pdf file-
i scanned it all-it says-
''notice that the designers have ommited an inductor on the output,and seem to ignore its absence'' or somthing like that.
its needed for non oscillation i think.(they didnt expect the later transistors to be different ?)
the eti 466 has more stable features.
and i think im nearly doen posting lots in this thread..but i dont want to be an eti 480 casualty.
These Are An Ok Amplifier, Even Good.
The original magazine article said "unconditionally stable" I think.
I never had instability problems with the ones I bulit decades ago, but I did improve earthing and power feed arrangements.
From memory the ground needs to be moved close down to the output stage - this improves stability.
I also added smaller caps across pcb electros.
I fitted a pair of such amps into a dead Nikko amplifier, and with twin turbo 5" cooling on top, it cranked and ran a million parties flat out till much too late.
I used Jap flat-pack output transistors, and I did not have any reliability problems whatsoever.
The unusual output stage is a good arrangement imo.
Eric.
The original magazine article said "unconditionally stable" I think.
I never had instability problems with the ones I bulit decades ago, but I did improve earthing and power feed arrangements.
From memory the ground needs to be moved close down to the output stage - this improves stability.
I also added smaller caps across pcb electros.
I fitted a pair of such amps into a dead Nikko amplifier, and with twin turbo 5" cooling on top, it cranked and ran a million parties flat out till much too late.
I used Jap flat-pack output transistors, and I did not have any reliability problems whatsoever.
The unusual output stage is a good arrangement imo.
Eric.
Hi everyone, hope I'm not intruding.
I've not personally worked on the Eti480, but looking at the schematics, the most likely problem is instability.
This amplifier has 3 gain stages, the last gain stage is in the driver stage. Designs with more than 2 gain stages are notoriously difficult to stabilise. Tell-tale signs that the designer was trying to contain oscillations are quite evident. I would suggest that a reliable speaker protection be used as marginal instability is very unpredictable.
Regards
I've not personally worked on the Eti480, but looking at the schematics, the most likely problem is instability.
This amplifier has 3 gain stages, the last gain stage is in the driver stage. Designs with more than 2 gain stages are notoriously difficult to stabilise. Tell-tale signs that the designer was trying to contain oscillations are quite evident. I would suggest that a reliable speaker protection be used as marginal instability is very unpredictable.
Regards
speaker protection
Hi mikee12345
I'm afraid putting a 47 ohms in series won't do, neither would a series fuse.
A common method is to use a fast responding circuit with a relay. If any harmful dc is sensed, it will disconnect the speakers until the dc is removed.
You can check out my site for this:
http://www.ampslab.com/dcpro.htm
It is simple, fast and reliable. It has been tested on amplifiers up to 800Watts/rms. It has even saved some very costly high-end speakers from certain destruction.
If you like to veroboard this circuit, just send me a mail. I will furnish you with the parts list so you can test it out.
Regards
Hi mikee12345
I'm afraid putting a 47 ohms in series won't do, neither would a series fuse.
A common method is to use a fast responding circuit with a relay. If any harmful dc is sensed, it will disconnect the speakers until the dc is removed.
You can check out my site for this:
http://www.ampslab.com/dcpro.htm
It is simple, fast and reliable. It has been tested on amplifiers up to 800Watts/rms. It has even saved some very costly high-end speakers from certain destruction.
If you like to veroboard this circuit, just send me a mail. I will furnish you with the parts list so you can test it out.
Regards
thanks mcp
its hard to find good information-all too often a google search wil find stupid things - ie not what one wants :-D
my eti480 status-1 assembled ,untested.
oscillation- can be mreasured and fixed.
Speaker safety-good
so im all set!
i noticed that the guy from 'almighty subwoofer' who has a 2226h JBL with sand filled panels,he runs only 20watts!!!!
i think 150watt rms would be fine :-D (97db/1watt)
im gona buy some earmuffs aswell
i thought a resistor wouldnt stop it unless it was a higher resistance than the speaker itself so most of the Vdrop would be across it?(which wudnt be good for heat and audio
its hard to find good information-all too often a google search wil find stupid things - ie not what one wants :-D
my eti480 status-1 assembled ,untested.
oscillation- can be mreasured and fixed.
Speaker safety-good
so im all set!
i noticed that the guy from 'almighty subwoofer' who has a 2226h JBL with sand filled panels,he runs only 20watts!!!!
i think 150watt rms would be fine :-D (97db/1watt)
im gona buy some earmuffs aswell
i thought a resistor wouldnt stop it
unless it was a higher resistance than the speaker itself so most of the Vdrop would be across it?(which wudnt be good for heat and audio 
I remember seeing a speaker protector similar to that and it had a changeover contact in the relay instead of only a normally open one. The idea was that the amplifier is connected to the moving contact, the speaker to the normally open one, and ground to the normally closed one. When a DC fault occurs, the amplifier output getsshorted to earth the same time as the speaker gets disconnected. This stops any relay contact arcing (arcing means current is not fully interrupted but still flowing through the speaker coilmcp said:
and crossover inductance) and also blows the amplifier fuse straight away. Easy to implement to an existing speaker protector too./Circlotron - has already demonstrated conclusively he knows nothing about speaker protection.
Maybe it will blow the amp instead of the fuse when you happen to trigger the DC protection with "real" LF content. I don't think it is necessary to do measures against relay arching since the energy stored in the voice-coil, even of drivers like a 2226, isn't that large either (W=I ^2 * L /2).
Take into account that the arching situation should occur between never and just a few times during an amplifier's lifetime.
Regards
Charles
Protection Ala Deluxe...
The Accuphase E-202 Amlifier used output protection relay changeover contacts in a clever way that I have not seen elsewhere.
The speaker was connected to the moving contact, the amplifier to the NO terminal, and a transformer winding via a series resistor connected to the NC contact.
With the contacts in the NC rest position, a small current was fed through the speaker to ground, and the voltage across the speaker was sensed, or maybe the current through the series resistor was sensed - I forget.
If the detected load impedence is too low or shorted, the protection relay will not operate and connect the speaker.
Many amplifiers sense amplifier output current, and will drop out the relay if the speaker lines are shorted or too low impedence.
Both of these functions together would make a really good protection stage.
Ultrasonic sensing (>20 kHz) would be an additional worthwhile function for complete protection.
Another thought is Rod Elliots 'Sound Imparment Monitor' to trigger the output relay to dropout also.
Just some ponderings....
Eric.
The Accuphase E-202 Amlifier used output protection relay changeover contacts in a clever way that I have not seen elsewhere.
The speaker was connected to the moving contact, the amplifier to the NO terminal, and a transformer winding via a series resistor connected to the NC contact.
With the contacts in the NC rest position, a small current was fed through the speaker to ground, and the voltage across the speaker was sensed, or maybe the current through the series resistor was sensed - I forget.
If the detected load impedence is too low or shorted, the protection relay will not operate and connect the speaker.
Many amplifiers sense amplifier output current, and will drop out the relay if the speaker lines are shorted or too low impedence.
Both of these functions together would make a really good protection stage.
Ultrasonic sensing (>20 kHz) would be an additional worthwhile function for complete protection.
Another thought is Rod Elliots 'Sound Imparment Monitor' to trigger the output relay to dropout also.
Just some ponderings....
Eric.
reaction time
There are quite a few ways to protect loudspeakers in the event of amp failure. Like Circlotron mentioned earlier, some designs short the outputs to ground, while others measure low impedance loads as well (ref Eric). Some uses opamps while others opto-couplers or triacs.
Whatever the designs are, the most important aspect is that it must work when called upon to. By that, I mean not just disconnecting the speakers but minimising the damage to the voice coils. If the reaction time is too slow, it can either fry the voice coils altogether, or weaken it substantially.
Regards
There are quite a few ways to protect loudspeakers in the event of amp failure. Like Circlotron mentioned earlier, some designs short the outputs to ground, while others measure low impedance loads as well (ref Eric). Some uses opamps while others opto-couplers or triacs.
Whatever the designs are, the most important aspect is that it must work when called upon to. By that, I mean not just disconnecting the speakers but minimising the damage to the voice coils. If the reaction time is too slow, it can either fry the voice coils altogether, or weaken it substantially.
Regards
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.