Hello, I'm looking for a heavy duty relay that can break 35 to 50A at 45VDC. It's relatively easy to find relays that will break that much current for an AC circuit, but without a 0 crossing DC relays are much more difficult. Any help you can provide in finding such a beast would be great!!!
Thanks
Scott
Thanks
Scott
Large amp relays are common for automotive, so you might want to see if you can find any with a high enough voltage rating. Otherwise, check surplus places and if all else fails maybe try paralleling smaller relays i guess. Thats all I can think of.
-Chris
-Chris
I suggest that you tell us (me) a little bit more about what you are going to use this relay for. 50A DC is a beast! If the relay should break this current a couple of times you can use a smaller relay but if you want long durability (millions of times) you must choose more carefully.
I work with relays and also sell them so I know that people often choose the wrong types and forget about proper contact protection. Switching DC is VERY difficult if we talk mechanical contacts. What about a bauta transistor?
I work with relays and also sell them so I know that people often choose the wrong types and forget about proper contact protection. Switching DC is VERY difficult if we talk mechanical contacts. What about a bauta transistor?
More details...
Sure, I can give you some more details. I may just end up going with a MOSFET switch which I can use as a soft start. I am trying to provide a measure of DC offset protection for a very high power amp. If the amp goes to the rails for any reason 80VDC (It's a bridged amp) could end up accross the speaker. I don't want to use a big relay at the output because heavy duty relays can add quite a bit of noise and/or distortion at low signal levels relative to their maximum. Not good for audio use.
So, I would like to insert the relay in series with each supply rail. When the offset detector kicks off/on (depending on the logic you'd like to go by) I want to power down the amp, separating it from the bulk capacitance.
Does that help?
Scott
Sure, I can give you some more details. I may just end up going with a MOSFET switch which I can use as a soft start. I am trying to provide a measure of DC offset protection for a very high power amp. If the amp goes to the rails for any reason 80VDC (It's a bridged amp) could end up accross the speaker. I don't want to use a big relay at the output because heavy duty relays can add quite a bit of noise and/or distortion at low signal levels relative to their maximum. Not good for audio use.
So, I would like to insert the relay in series with each supply rail. When the offset detector kicks off/on (depending on the logic you'd like to go by) I want to power down the amp, separating it from the bulk capacitance.
Does that help?
Scott
why not just install fuses between amp and filter caps and switch to the ground ? with a hefty switch transistor probably, and a small current surge limiting resistor .
edit: probably not so small current surge resistor 🙂
/rickard
edit: probably not so small current surge resistor 🙂
/rickard
rikkitikkitavi said:
I thought about that. I would do it with very heavy duty SCRs, it's cheap, quick and dirty. However, I will have ~100,000uF of bulk capacitors in the power supply circuit. The short circuit current of that capacitor bank is a frightening prospect. BIG current. Big.
Very big. Lots even. 😉
Scott
I don't want to use a big relay at the output because heavy duty relays can add quite a bit of noise and/or distortion at low signal levels relative to their maximum. Not good for audio use.
I'm interested in any data you have on this. My plan is to use
a DPDT speaker output relay, contacts are rated for 20A @ 28V,
parallel the contacts for 40A and contact redundency.
Magnetcraft W92S11D12 series is the only thing I was able
to find.
My plan is to use a "rail detector" circuit in conjunction with
the standard generic DC offset detector for extra protection.
When there is a rail voltage imbalance (threshold determined by
user), it triggers a fault. Some people have told me that
sometimes only one rail fuse blows, not both, causing DC at the
output in which case my rail detector will respond first before the
DC detector (at least I hope so... LOL). I haven't done extensive
tests on reaction time.
I'm interested in any data you have on this. My plan is to use
a DPDT speaker output relay, contacts are rated for 20A @ 28V,
parallel the contacts for 40A and contact redundency.
Magnetcraft W92S11D12 series is the only thing I was able
to find.
My plan is to use a "rail detector" circuit in conjunction with
the standard generic DC offset detector for extra protection.
When there is a rail voltage imbalance (threshold determined by
user), it triggers a fault. Some people have told me that
sometimes only one rail fuse blows, not both, causing DC at the
output in which case my rail detector will respond first before the
DC detector (at least I hope so... LOL). I haven't done extensive
tests on reaction time.
thylantyr said:
Take a look at the minimum current capacity of the part. This is often listed as a specification. I've seen relays in a situation where 10% of the time they needed to pass/break 5ADC and 90% of the time they handled a small analog voltage. Very often we would get returns on them because over a very short time they would "go bad" and distort the small analog signal. The MFG wouldn't take them back because we were using currents BELOW what they were rated for, not rated for the application.
So, I don't want to put a relay in the output circuit.
Scott
I noticed the relay I chose has a rating of 500mA @ 12v
as minimum load.
Hypothetical.
If the relay is "bad" and won't allow a small current to
pass from the amplifier output stage to the speaker,
will you not have the same problem if the same bad relay
is connected to the rail voltage ? at low current draw,
won't the relay not allow the rail power supply current
to flow to the amplfier?
Another wierd idea.
Parallel the monster relay contacts to a small
signal relay contacts, both relays engage at the same
time. One allows tiny current to pass, at high current,
current takes the least path of resistance, the bigger
relay. ??
as minimum load.
Hypothetical.
If the relay is "bad" and won't allow a small current to
pass from the amplifier output stage to the speaker,
will you not have the same problem if the same bad relay
is connected to the rail voltage ? at low current draw,
won't the relay not allow the rail power supply current
to flow to the amplfier?
Another wierd idea.
Parallel the monster relay contacts to a small
signal relay contacts, both relays engage at the same
time. One allows tiny current to pass, at high current,
current takes the least path of resistance, the bigger
relay. ??
I asked Richard Clark who has done extensive tests with
speaker relays used for ABX. In a nutshell, the contacts
degrade over time and a good solution seems be to
use 4 contacts all in parallel with blowout magnets, he said
it's audibly flawless.
The large magnets (user added if needed be) are placed
near the contacts to steer the arcs away. Over time
you see the crud on the inside of the relay plastic housing.
The fastest contact (of a 4 contact configuration) becomes the
sacrifical lamb for arcing and takes most of the beating preserving
the other contacts.
I shall seek a 4PDT for my project, perhaps one with 10A
per contact all in parallel.
Final thoughts. I've seen many high end home audio
and car audio amplifiers get away with use a "standard"
30A relay SPDT (forgot part number) for their speaker
relay designs, using 4 contacts should satisfy even the most
critical audiophiles.
Over time, replace the $10 relay. No big deal. Protect
your speaker investment.
speaker relays used for ABX. In a nutshell, the contacts
degrade over time and a good solution seems be to
use 4 contacts all in parallel with blowout magnets, he said
it's audibly flawless.
The large magnets (user added if needed be) are placed
near the contacts to steer the arcs away. Over time
you see the crud on the inside of the relay plastic housing.
The fastest contact (of a 4 contact configuration) becomes the
sacrifical lamb for arcing and takes most of the beating preserving
the other contacts.
I shall seek a 4PDT for my project, perhaps one with 10A
per contact all in parallel.
Final thoughts. I've seen many high end home audio
and car audio amplifiers get away with use a "standard"
30A relay SPDT (forgot part number) for their speaker
relay designs, using 4 contacts should satisfy even the most
critical audiophiles.
Over time, replace the $10 relay. No big deal. Protect
your speaker investment.
thylantyr said:I noticed the relay I chose has a rating of 500mA @ 12v
as minimum load.
Hypothetical.
If the relay is "bad" and won't allow a small current to
pass from the amplifier output stage to the speaker,
will you not have the same problem if the same bad relay
is connected to the rail voltage ? at low current draw,
won't the relay not allow the rail power supply current
to flow to the amplfier?
Another wierd idea.
Parallel the monster relay contacts to a small
signal relay contacts, both relays engage at the same
time. One allows tiny current to pass, at high current,
current takes the least path of resistance, the bigger
relay. ??
Not usually. If the relay is in the supply lines there is usually enough bias current to ensure that the minimum current for the part is exceded, thereby a good contact is made. Now if the bias current isn't enough, then you can always add a few 10watt bleeder resistors to make sure it is.
I have an issue with paralleling relays when they are used for breaking signals in a fault situation. In the ABX application he's trying to create a high quality signal relay using cheaper relays. That's fine. What I want to do is make sure that if the amp output stage fails and the amp goes to one rail or the other, or in my case if an instability causes the bridged output to head to either rail the relay can break the current. Well, what is the major failure mode of a relay? A short circuit. The sacrificial lamb that you speak of will actually weld itself shut, the other relays can be as open as they want. The fault still has not been cleared.
The more I've looked at this...the more I'm going to use BIG mosfets as switches in the supply rails to open the circuit. I can control them causing them to open over a 100ms period after excess (1.4VDC) is detectd on the output. The dissapation during this time for the mosfets may be high but it's a VERY low duty cycle so minimal heatsinking will be required.
Scott
thylantyr said:
As I have noted in the past, Amplimo of the Netherlands have specially dveloped speaker relays. (www.amplimo.nl, ask for LRZ). They consist of a wolfram (tungsten) main contact rated for 100 Amps. This is shunted by a gold plated contact to short the residual resistance of the main contact. Closure order is tungsten first, then low level. Opening in reverse order. Sets you back about 8 euros. I used them over 10 years now without a hitch.
Jan Didden
thylantyr said:
Ahh, see my post above. I will have very very large currents. Before that fuse has time to blow, I'm worried about what might happen to my nice PCB.
Oh well..
Scott
janneman said:
That sounds interesting. I'll have to give it a try. It looks like Plitron sells them in North america but I can't get that page to come up on their site. Thanks for the information!
Scott
thylantyr said:
This is a design for personal use. That said, I do design speakers professionaly, and I design Medical and Telecom equipment to make a living. A lot of my professional circuits have operated as high as 2.4GHz for mW type signals and 1GHz and less for signal levels as high as 10-25 watts. As a result, I see the value of a well layed out PCB and use them whenever possible.
Scott
Received an email from plitron,
the amplimo relay is not stocked,
2-3 weeks..
I'm gonna get me some to try out.
the amplimo relay is not stocked,
2-3 weeks..
I'm gonna get me some to try out.
thylantyr said:
I talked to them today. While on their site I noticed their medically rated isolation transformers...which I'd never seen before. I think I may end up getting some of those for some work I'm doing right now. At the same time I asked about the relays as well...they told me 2-5 weeks. I'm probably going to order 5-10 or so.
Scott
- Status
- Not open for further replies.