Magnets are apparently the reason why this Omron relais can tolerate the DC load conditions and for the polarity as well. Besides that, I'd recommend measuring the amplifier's distortions with and without the chosen relais in series to the load to examine the possible effects of magnets and contacts.
Thanks for the link, it shows clearly that relays can fail shorted due to too high DC voltages. Bonsai posted something about it a couple of days ago and then removed it again.
I've had relay contacts weld on AC inrush... Turn it on once - never turn it off. OTOH, I've used relays rated for "up to" 75VDC to switch 400VDC - The trick is to oversize the relay 🙂 Using a 12V/40A automotive relay to switch 400VDC@500mA is child's play apparently - never had one fail in that service yet - YMMV.
that thread is worth reading
it is not a mech relay vs ss relay debate as such
it introduces many members to use of ssr
it is not a mech relay vs ss relay debate as such
it introduces many members to use of ssr
Everything can break down certainly when devices have higher voltages and higher power unlike the OP's 100W amplifier. Bonsais amplifier was a 250W one with 140V around. The relay was used way out of specifications which simply is a question of when things go wrong (not if they will go wrong). It were also 3 relays in parallel which I think did not help as they were all 3 used outside specifications.
Now just suppose you don't need so much power... when lowering rail voltages/output power to moderate levels (who needs 2 x 250W in a living room?) then riscs of defective relays because of high DC levels diminish at the same time. It is like buying a Porsche that drives 200 km/h within seconds and then wonder why you get speed tickets.
I work with high power and very high power devices on a daily basis. If one designs/builds high power devices then relays alone won't work out too well. Then all fault possibilities must be examined as results of a failure will practically always be destructive. Protection needs to be more complex then as those Joules want to go somewhere.
Now just suppose you don't need so much power... when lowering rail voltages/output power to moderate levels (who needs 2 x 250W in a living room?) then riscs of defective relays because of high DC levels diminish at the same time. It is like buying a Porsche that drives 200 km/h within seconds and then wonder why you get speed tickets.
I work with high power and very high power devices on a daily basis. If one designs/builds high power devices then relays alone won't work out too well. Then all fault possibilities must be examined as results of a failure will practically always be destructive. Protection needs to be more complex then as those Joules want to go somewhere.
Last edited:
Aha thank you. I think things get exaggerated big time just to prove the protection is a risk on itself whilst the goal is different. The situation is decribed by Bonsai in detail and parts were used outside specifications and in a non optimal configuration so here is your anecdotical evidence 🙂 For the average few tens of Watts amplifier and well chosen relays of course used within specifications it works out many years and protects stuff as it should in relatively simple, unobtrusive and reliable way. Just like the thread title suggests.
Feel free to introduce a competitive way of protecting loudspeakers that is just as simple, unobtrusive and reliable.
Feel free to introduce a competitive way of protecting loudspeakers that is just as simple, unobtrusive and reliable.
Last edited:
Since post #37 or so of this thread, the question has been whether relays always fail open or can also fail shorted when they have to switch off a DC voltage or current greater than specified in their datasheet. Bonsai's thread shows they can fail shorted; of course the voltage was some 60 % higher than for the 100 W in the title of the present thread (sqrt(250 W/100 W)), but I haven't seen anything proving that the problem is in the last 60 %.
Of course Bonsai used the relay way out of its specified range, but that was exactly what we were discussing: can you safely ignore relay DC switching specs when you only need the relay to switch off a higher than specified DC level once? Based on what I've read in the two threads, I'd say no.
Of course Bonsai used the relay way out of its specified range, but that was exactly what we were discussing: can you safely ignore relay DC switching specs when you only need the relay to switch off a higher than specified DC level once? Based on what I've read in the two threads, I'd say no.
Last edited:
If you like ESLs, then 250W amplifiers aren't oversized, as the sensitivity of most ESLs is surprisingly low; you don't need the power but the voltage swing.
Today it is not so much a problem to find a relay or industry contactor rated for high DC voltages and currents (used in vehicles) but the essential information about contact materials and/or minimal switching conditions is quite often not included in datasheets. So, you have to ask the manufacturers or distributors for the details and have to test the devices for unwanted audible (or measurable) effects. These contactors are rated for up to several hundreds of volts and amperes (DC); prices are high and the parts are quite large.
Edit: IIRC, in Bonsai's thread it was already mentioned, that a combination of SSR (rated for the load under error conditions) and EMR (optimized for the normal work conditions) might be the best
Today it is not so much a problem to find a relay or industry contactor rated for high DC voltages and currents (used in vehicles) but the essential information about contact materials and/or minimal switching conditions is quite often not included in datasheets. So, you have to ask the manufacturers or distributors for the details and have to test the devices for unwanted audible (or measurable) effects. These contactors are rated for up to several hundreds of volts and amperes (DC); prices are high and the parts are quite large.
Edit: IIRC, in Bonsai's thread it was already mentioned, that a combination of SSR (rated for the load under error conditions) and EMR (optimized for the normal work conditions) might be the best
Last edited:
The best .... for that power/voltages. Horses for courses.
Parts parameters should always be rated for the load under error conditions optimized for the normal work conditions. 100% agreed. Too bad they are sometimes selected for low price, or not chosen at all i.e. no protection. The latter is a design error which seems cheap but it presents real operation costs. It just takes a while.
Parts parameters should always be rated for the load under error conditions optimized for the normal work conditions. 100% agreed. Too bad they are sometimes selected for low price, or not chosen at all i.e. no protection. The latter is a design error which seems cheap but it presents real operation costs. It just takes a while.
Last edited:
The ESL-63 and its derivatives would just short the amplifier to protect themselves...
Then again, they are fairly sensitive, 86 dB at 2.83 V extrapolated to 1 metre. I used them with a 2 times 20 W into 8 ohm amplifier for years without clipping the amplifier (which has an LED that turns yellow for about 1 second when it clips a few microseconds).
Maybe you did not realise that but I did 🙂 Written communication so no non verbal communication... always difficult.
Beware... we might end up having a sensible discussion about anything.
Beware... we might end up having a sensible discussion about anything.
@MarcelvdG,
using Quads quite often, I can relate to your description, but I think the sensitivity is more around 82 - 83 dB (2,82V/1m). I've combined older Quads with 20W class-A designs, but it surely depends on the room, listening distance, and style of music.
Other ESLs (like some Audiostatics models for example) sensitivity was even lower and playing contemporary orchestral music in a larger room proved that even a ~64 V swing of the amplifier wasn't enough to prevent clipping during certain pieces (of course short time clipping).
@jean-paul,
from a manufacturer's point of view, it obviously depends; bulletproof protection circuitry is costly and will never be needed by most users most of the time (I know that's the caveat).
But, even for a 100W amplifier, it seems that the goal is hardly achievable with just a single relay, but I haven't tested these said newer DC relays so far.
using Quads quite often, I can relate to your description, but I think the sensitivity is more around 82 - 83 dB (2,82V/1m). I've combined older Quads with 20W class-A designs, but it surely depends on the room, listening distance, and style of music.
Other ESLs (like some Audiostatics models for example) sensitivity was even lower and playing contemporary orchestral music in a larger room proved that even a ~64 V swing of the amplifier wasn't enough to prevent clipping during certain pieces (of course short time clipping).
@jean-paul,
from a manufacturer's point of view, it obviously depends; bulletproof protection circuitry is costly and will never be needed by most users most of the time (I know that's the caveat).
But, even for a 100W amplifier, it seems that the goal is hardly achievable with just a single relay, but I haven't tested these said newer DC relays so far.
For the ESL-63, Quad recommended 50 W to 100 W into 8 ohm amplifiers and warned against anything above 150 W into 8 ohm, see https://www.manualslib.com/manual/407328/Quad-Esl-63.html?page=5#manual (As far as I know, the protection kicks in sooner when the loudspeaker is off, so you should never turn off the loudspeaker while it is playing loud.)
The official sensitivity figure from Quad is 1.5 ubar/V referred to 1 m, which is equivalent to 86.54 dB with respect to 20 uPa at 2.83 V referred to 1 m, see https://www.manualslib.com/manual/407328/Quad-Esl-63.html?page=8#manual No idea how accurate or inaccurate it may be.
The official sensitivity figure from Quad is 1.5 ubar/V referred to 1 m, which is equivalent to 86.54 dB with respect to 20 uPa at 2.83 V referred to 1 m, see https://www.manualslib.com/manual/407328/Quad-Esl-63.html?page=8#manual No idea how accurate or inaccurate it may be.
Last edited:
Off topic: I once built an amplifier that could produce 2 kV peak differential output voltage to directly drive an electrostatic loudspeaker made by a then colleague of mine. The supply voltage was 2.66 kV and the supply's short circuit current was well above what is needed to kill someone, making it by far the most dangerous circuit I ever built.
It sounded very good (in an informal uncontrolled test), but as soon as you turned up the volume a bit higher than for background music, the clipping LED would start blinking...
It sounded very good (in an informal uncontrolled test), but as soon as you turned up the volume a bit higher than for background music, the clipping LED would start blinking...
Then familiar tools like these should be used in a home environment to be able to work safely. Insane. Never got used to those ESL's, many of those I heard were arcing when the weather was moist. Cats did not seem to like them either 🙂
https://www.horstmanngmbh.com/fileadmin/products/Earthing_and_short-circuiting_devices.pdf
https://www.horstmanngmbh.com/fileadmin/products/Earthing_and_short-circuiting_devices.pdf
Attachments
Out of the eight cats who lived here while there were ESL-63s standing here, none complained about their sound quality, one urinated against a bass panel, but only once in many years, and one used one of the ESL-63s for climbing and sometimes jumped off, causing the loudspeaker to fall - which is the reason why there are no ESL-63s here anymore. I only heard the loudspeakers arcing after the urinating incident.