Toggle switch switching time...

[Fred Bloggs]

On Tuesday, May 30, 2023 at 6:34:28 PM UTC-4, Ed Lee wrote:

On Tuesday, May 30, 2023 at 2:50:28 PM UTC-7, John Larkin wrote:
On Tue, 30 May 2023 14:18:35 -0700, Don Y
blocked...@foo.invalid> wrote:

On 5/30/2023 6:13 AM, Don Y wrote:
Find a *documented* switching mechanism. If you\'re switching
*DC*, expect the switch to be heavily derated. And, expect to
have to factor in some arcing during the transition. You
may be better off with a mercury-wetted *relay*, depending on
your application.

Note that switches are not \"ideal\" even when in steady state
but in *transition*, are complex creatures. A matED contact
will \"bounce\" as it is opened and a matING contact will bounce
as it\'s closing. Ideally, you want to know the time when the
contact is \"in flight\" AND the voltage rating at that point
(assuming any arcing has subsided -- that\'s YOUR calculation
not the switch manufacturer\'s as it depends on your application).

I can\'t find (online) the more comprehensive reference that I have
locally but here\'s ganssle\'s take on switch bounce. It\'s one of
my go-to documents for folks who THINK they understand the
dynamics of switches. E.g., folks seem to think switches
don\'t bounce when *opening* -- oops! :
Sometimes they don\'t.

Note the observation of a *relay* driven repeatedly by an MCU attached
to a storage scope (hard to drive a *switch* thusly) -- emphasis mine:

\"When the relay OPENED it always had a max bounce time of 2.3 to 2.9 msec,
at speeds from 2.5 to 30 Hz. More variation appeared on contact CLOSURE:
at 2.5 Hz bounces never exceeded 410 µsec, which climbed to 1080 µsec at
30 Hz. Why? I have no idea. But it\'s clear there is some correlation
between fast actuations and more bounce.\"
A relay will twang, namely ring like a bell, for a long time after a
transition. And bells have very complex vibration patterns. The twang
affects future actuations.

Reed relays make terrible analog multiplexers, because the high-Q
mechanical ringing makes voltages in the mag field and takes a long
time to die out.
I don\'t care about bouncing and/or ringing, as long as they don\'t short out the two throws. Some switches has center disconnect or middle throw, but they are not heavy duty rating. I might have a separate disconnect switch in series with the function switch, perhaps in lock box door type. The function switch is hidden inside the box. When the user open the lock box, the circuit should be disconnected.

Pursuing the MOSFET idea further, you can use the bypass technique where the ON MOSFET parallels the switch contacts only during closing and opening, and thereafter remains OFF. Then you can get away with a cheap switch, even that Chinese junk in your first post.

You would do yourself a service if you could get rid of that 4pdt requirement.

 

[Ed Lee]

On Wednesday, May 31, 2023 at 8:14:57 AM UTC-7, Fred Bloggs wrote:

On Tuesday, May 30, 2023 at 6:34:28 PM UTC-4, Ed Lee wrote:
On Tuesday, May 30, 2023 at 2:50:28 PM UTC-7, John Larkin wrote:
On Tue, 30 May 2023 14:18:35 -0700, Don Y
blocked...@foo.invalid> wrote:

On 5/30/2023 6:13 AM, Don Y wrote:
Find a *documented* switching mechanism. If you\'re switching
*DC*, expect the switch to be heavily derated. And, expect to
have to factor in some arcing during the transition. You
may be better off with a mercury-wetted *relay*, depending on
your application.

Note that switches are not \"ideal\" even when in steady state
but in *transition*, are complex creatures. A matED contact
will \"bounce\" as it is opened and a matING contact will bounce
as it\'s closing. Ideally, you want to know the time when the
contact is \"in flight\" AND the voltage rating at that point
(assuming any arcing has subsided -- that\'s YOUR calculation
not the switch manufacturer\'s as it depends on your application).

I can\'t find (online) the more comprehensive reference that I have
locally but here\'s ganssle\'s take on switch bounce. It\'s one of
my go-to documents for folks who THINK they understand the
dynamics of switches. E.g., folks seem to think switches
don\'t bounce when *opening* -- oops! :
Sometimes they don\'t.

Note the observation of a *relay* driven repeatedly by an MCU attached
to a storage scope (hard to drive a *switch* thusly) -- emphasis mine:

\"When the relay OPENED it always had a max bounce time of 2.3 to 2.9 msec,
at speeds from 2.5 to 30 Hz. More variation appeared on contact CLOSURE:
at 2.5 Hz bounces never exceeded 410 µsec, which climbed to 1080 µsec at
30 Hz. Why? I have no idea. But it\'s clear there is some correlation
between fast actuations and more bounce.\"
A relay will twang, namely ring like a bell, for a long time after a
transition. And bells have very complex vibration patterns. The twang
affects future actuations.

Reed relays make terrible analog multiplexers, because the high-Q
mechanical ringing makes voltages in the mag field and takes a long
time to die out.
I don\'t care about bouncing and/or ringing, as long as they don\'t short out the two throws. Some switches has center disconnect or middle throw, but they are not heavy duty rating. I might have a separate disconnect switch in series with the function switch, perhaps in lock box door type. The function switch is hidden inside the box. When the user open the lock box, the circuit should be disconnected.
Your voltages and current are easily handled by power MOSFETs.

But I don\'t trust the failure modes of semis and/or electronics.

 

[Fred Bloggs]

On Wednesday, May 31, 2023 at 11:20:08 AM UTC-4, Ed Lee wrote:

On Wednesday, May 31, 2023 at 8:14:57 AM UTC-7, Fred Bloggs wrote:
On Tuesday, May 30, 2023 at 6:34:28 PM UTC-4, Ed Lee wrote:
On Tuesday, May 30, 2023 at 2:50:28 PM UTC-7, John Larkin wrote:
On Tue, 30 May 2023 14:18:35 -0700, Don Y
blocked...@foo.invalid> wrote:

On 5/30/2023 6:13 AM, Don Y wrote:
Find a *documented* switching mechanism. If you\'re switching
*DC*, expect the switch to be heavily derated. And, expect to
have to factor in some arcing during the transition. You
may be better off with a mercury-wetted *relay*, depending on
your application.

Note that switches are not \"ideal\" even when in steady state
but in *transition*, are complex creatures. A matED contact
will \"bounce\" as it is opened and a matING contact will bounce
as it\'s closing. Ideally, you want to know the time when the
contact is \"in flight\" AND the voltage rating at that point
(assuming any arcing has subsided -- that\'s YOUR calculation
not the switch manufacturer\'s as it depends on your application)..

I can\'t find (online) the more comprehensive reference that I have
locally but here\'s ganssle\'s take on switch bounce. It\'s one of
my go-to documents for folks who THINK they understand the
dynamics of switches. E.g., folks seem to think switches
don\'t bounce when *opening* -- oops! :
Sometimes they don\'t.

Note the observation of a *relay* driven repeatedly by an MCU attached
to a storage scope (hard to drive a *switch* thusly) -- emphasis mine:

\"When the relay OPENED it always had a max bounce time of 2.3 to 2.9 msec,
at speeds from 2.5 to 30 Hz. More variation appeared on contact CLOSURE:
at 2.5 Hz bounces never exceeded 410 µsec, which climbed to 1080 µsec at
30 Hz. Why? I have no idea. But it\'s clear there is some correlation
between fast actuations and more bounce.\"
A relay will twang, namely ring like a bell, for a long time after a
transition. And bells have very complex vibration patterns. The twang
affects future actuations.

Reed relays make terrible analog multiplexers, because the high-Q
mechanical ringing makes voltages in the mag field and takes a long
time to die out.
I don\'t care about bouncing and/or ringing, as long as they don\'t short out the two throws. Some switches has center disconnect or middle throw, but they are not heavy duty rating. I might have a separate disconnect switch in series with the function switch, perhaps in lock box door type. The function switch is hidden inside the box. When the user open the lock box, the circuit should be disconnected.
Your voltages and current are easily handled by power MOSFETs.
But I don\'t trust the failure modes of semis and/or electronics.

You can use that failsafe technique where the one output actuates an NC relay that opens the other output. That way if the two outputs are on at the same time, they\'re both removed.

 

[Ed Lee]

On Wednesday, May 31, 2023 at 8:26:14 AM UTC-7, Fred Bloggs wrote:

On Wednesday, May 31, 2023 at 11:20:08 AM UTC-4, Ed Lee wrote:
On Wednesday, May 31, 2023 at 8:14:57 AM UTC-7, Fred Bloggs wrote:
On Tuesday, May 30, 2023 at 6:34:28 PM UTC-4, Ed Lee wrote:
On Tuesday, May 30, 2023 at 2:50:28 PM UTC-7, John Larkin wrote:
On Tue, 30 May 2023 14:18:35 -0700, Don Y
blocked...@foo.invalid> wrote:

On 5/30/2023 6:13 AM, Don Y wrote:
Find a *documented* switching mechanism. If you\'re switching
*DC*, expect the switch to be heavily derated. And, expect to
have to factor in some arcing during the transition. You
may be better off with a mercury-wetted *relay*, depending on
your application.

Note that switches are not \"ideal\" even when in steady state
but in *transition*, are complex creatures. A matED contact
will \"bounce\" as it is opened and a matING contact will bounce
as it\'s closing. Ideally, you want to know the time when the
contact is \"in flight\" AND the voltage rating at that point
(assuming any arcing has subsided -- that\'s YOUR calculation
not the switch manufacturer\'s as it depends on your application).

I can\'t find (online) the more comprehensive reference that I have
locally but here\'s ganssle\'s take on switch bounce. It\'s one of
my go-to documents for folks who THINK they understand the
dynamics of switches. E.g., folks seem to think switches
don\'t bounce when *opening* -- oops! :
Sometimes they don\'t.

Note the observation of a *relay* driven repeatedly by an MCU attached
to a storage scope (hard to drive a *switch* thusly) -- emphasis mine:

\"When the relay OPENED it always had a max bounce time of 2.3 to 2.9 msec,
at speeds from 2.5 to 30 Hz. More variation appeared on contact CLOSURE:
at 2.5 Hz bounces never exceeded 410 µsec, which climbed to 1080 µsec at
30 Hz. Why? I have no idea. But it\'s clear there is some correlation
between fast actuations and more bounce.\"
A relay will twang, namely ring like a bell, for a long time after a
transition. And bells have very complex vibration patterns. The twang
affects future actuations.

Reed relays make terrible analog multiplexers, because the high-Q
mechanical ringing makes voltages in the mag field and takes a long
time to die out.
I don\'t care about bouncing and/or ringing, as long as they don\'t short out the two throws. Some switches has center disconnect or middle throw, but they are not heavy duty rating. I might have a separate disconnect switch in series with the function switch, perhaps in lock box door type. The function switch is hidden inside the box. When the user open the lock box, the circuit should be disconnected.
Your voltages and current are easily handled by power MOSFETs.
But I don\'t trust the failure modes of semis and/or electronics.
You can use that failsafe technique where the one output actuates an NC relay that opens the other output. That way if the two outputs are on at the same time, they\'re both removed.

What if both outputs are off, with both NC relays and instant spark.

 

[Anthony William Sloman]

On Thursday, June 1, 2023 at 1:02:42 AM UTC+10, Ed Lee wrote:

On Wednesday, May 31, 2023 at 5:05:09 AM UTC-7, Anthony William Sloman wrote:
On Wednesday, May 31, 2023 at 7:50:28 AM UTC+10, John Larkin wrote:
On Tue, 30 May 2023 14:18:35 -0700, Don Y <blocked...@foo.invalid> wrote:
On 5/30/2023 6:13 AM, Don Y wrote:
snip
Reed relays make terrible analog multiplexers, because the high-Q
mechanical ringing makes voltages in the mag field and takes a long
time to die out.
The reeds are magnetic, and they are moving inside the actuating coil but they don\'t have any effect on the flux threading the actuating coil after they open or close.

Someone suggested mercury switches.

Ordered some to try out:
https://www.ebay.com/itm/195708179355?var=495574245862

I suggested mercury wetted reed switches, which aren\'t the same thing at all. Here is one example. Most suppliers of reed relays offer mercury wetted variants.

https://www.pickeringrelay.com/pdfs/10-50.pdf

The do have to be mounted vertically - or within 30 degree of the vertical - but they do have unique advantages

> So, how much current can I expect? The listing says 1 to 15A in general and 0.3A in specific.

Read the data sheet. Only the Pickering series 40 parts are mercury wetted.

--
Bill Sloman, Sydney

 

[Fred Bloggs]

On Wednesday, May 31, 2023 at 11:32:03 AM UTC-4, Ed Lee wrote:

On Wednesday, May 31, 2023 at 8:26:14 AM UTC-7, Fred Bloggs wrote:
On Wednesday, May 31, 2023 at 11:20:08 AM UTC-4, Ed Lee wrote:
On Wednesday, May 31, 2023 at 8:14:57 AM UTC-7, Fred Bloggs wrote:
On Tuesday, May 30, 2023 at 6:34:28 PM UTC-4, Ed Lee wrote:
On Tuesday, May 30, 2023 at 2:50:28 PM UTC-7, John Larkin wrote:
On Tue, 30 May 2023 14:18:35 -0700, Don Y
blocked...@foo.invalid> wrote:

On 5/30/2023 6:13 AM, Don Y wrote:
Find a *documented* switching mechanism. If you\'re switching
*DC*, expect the switch to be heavily derated. And, expect to
have to factor in some arcing during the transition. You
may be better off with a mercury-wetted *relay*, depending on
your application.

Note that switches are not \"ideal\" even when in steady state
but in *transition*, are complex creatures. A matED contact
will \"bounce\" as it is opened and a matING contact will bounce
as it\'s closing. Ideally, you want to know the time when the
contact is \"in flight\" AND the voltage rating at that point
(assuming any arcing has subsided -- that\'s YOUR calculation
not the switch manufacturer\'s as it depends on your application).

I can\'t find (online) the more comprehensive reference that I have
locally but here\'s ganssle\'s take on switch bounce. It\'s one of
my go-to documents for folks who THINK they understand the
dynamics of switches. E.g., folks seem to think switches
don\'t bounce when *opening* -- oops! :
Sometimes they don\'t.

Note the observation of a *relay* driven repeatedly by an MCU attached
to a storage scope (hard to drive a *switch* thusly) -- emphasis mine:

\"When the relay OPENED it always had a max bounce time of 2.3 to 2.9 msec,
at speeds from 2.5 to 30 Hz. More variation appeared on contact CLOSURE:
at 2.5 Hz bounces never exceeded 410 µsec, which climbed to 1080 µsec at
30 Hz. Why? I have no idea. But it\'s clear there is some correlation
between fast actuations and more bounce.\"
A relay will twang, namely ring like a bell, for a long time after a
transition. And bells have very complex vibration patterns. The twang
affects future actuations.

Reed relays make terrible analog multiplexers, because the high-Q
mechanical ringing makes voltages in the mag field and takes a long
time to die out.
I don\'t care about bouncing and/or ringing, as long as they don\'t short out the two throws. Some switches has center disconnect or middle throw, but they are not heavy duty rating. I might have a separate disconnect switch in series with the function switch, perhaps in lock box door type. The function switch is hidden inside the box. When the user open the lock box, the circuit should be disconnected.
Your voltages and current are easily handled by power MOSFETs.
But I don\'t trust the failure modes of semis and/or electronics.
You can use that failsafe technique where the one output actuates an NC relay that opens the other output. That way if the two outputs are on at the same time, they\'re both removed.
What if both outputs are off, with both NC relays and instant spark.

If your system can\'t take a 10-50ms blip like that then even a switch is risky. Think of something else.

 

[Ricky]

On Wednesday, May 31, 2023 at 11:20:08 AM UTC-4, Ed Lee wrote:

On Wednesday, May 31, 2023 at 8:14:57 AM UTC-7, Fred Bloggs wrote:
On Tuesday, May 30, 2023 at 6:34:28 PM UTC-4, Ed Lee wrote:
On Tuesday, May 30, 2023 at 2:50:28 PM UTC-7, John Larkin wrote:
On Tue, 30 May 2023 14:18:35 -0700, Don Y
blocked...@foo.invalid> wrote:

On 5/30/2023 6:13 AM, Don Y wrote:
Find a *documented* switching mechanism. If you\'re switching
*DC*, expect the switch to be heavily derated. And, expect to
have to factor in some arcing during the transition. You
may be better off with a mercury-wetted *relay*, depending on
your application.

Note that switches are not \"ideal\" even when in steady state
but in *transition*, are complex creatures. A matED contact
will \"bounce\" as it is opened and a matING contact will bounce
as it\'s closing. Ideally, you want to know the time when the
contact is \"in flight\" AND the voltage rating at that point
(assuming any arcing has subsided -- that\'s YOUR calculation
not the switch manufacturer\'s as it depends on your application)..

I can\'t find (online) the more comprehensive reference that I have
locally but here\'s ganssle\'s take on switch bounce. It\'s one of
my go-to documents for folks who THINK they understand the
dynamics of switches. E.g., folks seem to think switches
don\'t bounce when *opening* -- oops! :
Sometimes they don\'t.

Note the observation of a *relay* driven repeatedly by an MCU attached
to a storage scope (hard to drive a *switch* thusly) -- emphasis mine:

\"When the relay OPENED it always had a max bounce time of 2.3 to 2.9 msec,
at speeds from 2.5 to 30 Hz. More variation appeared on contact CLOSURE:
at 2.5 Hz bounces never exceeded 410 µsec, which climbed to 1080 µsec at
30 Hz. Why? I have no idea. But it\'s clear there is some correlation
between fast actuations and more bounce.\"
A relay will twang, namely ring like a bell, for a long time after a
transition. And bells have very complex vibration patterns. The twang
affects future actuations.

Reed relays make terrible analog multiplexers, because the high-Q
mechanical ringing makes voltages in the mag field and takes a long
time to die out.
I don\'t care about bouncing and/or ringing, as long as they don\'t short out the two throws. Some switches has center disconnect or middle throw, but they are not heavy duty rating. I might have a separate disconnect switch in series with the function switch, perhaps in lock box door type. The function switch is hidden inside the box. When the user open the lock box, the circuit should be disconnected.
Your voltages and current are easily handled by power MOSFETs.
But I don\'t trust the failure modes of semis and/or electronics.

Good thing you don\'t drive an electric car then. You would live in constant fear.

--

Rick C.

- Get 1,000 miles of free Supercharging
- Tesla referral code - https://ts.la/richard11209

 

[Ed Lee]

On Wednesday, May 31, 2023 at 10:53:20 AM UTC-7, Ricky wrote:

On Wednesday, May 31, 2023 at 11:20:08 AM UTC-4, Ed Lee wrote:
On Wednesday, May 31, 2023 at 8:14:57 AM UTC-7, Fred Bloggs wrote:
On Tuesday, May 30, 2023 at 6:34:28 PM UTC-4, Ed Lee wrote:
On Tuesday, May 30, 2023 at 2:50:28 PM UTC-7, John Larkin wrote:
On Tue, 30 May 2023 14:18:35 -0700, Don Y
blocked...@foo.invalid> wrote:

On 5/30/2023 6:13 AM, Don Y wrote:
Find a *documented* switching mechanism. If you\'re switching
*DC*, expect the switch to be heavily derated. And, expect to
have to factor in some arcing during the transition. You
may be better off with a mercury-wetted *relay*, depending on
your application.

Note that switches are not \"ideal\" even when in steady state
but in *transition*, are complex creatures. A matED contact
will \"bounce\" as it is opened and a matING contact will bounce
as it\'s closing. Ideally, you want to know the time when the
contact is \"in flight\" AND the voltage rating at that point
(assuming any arcing has subsided -- that\'s YOUR calculation
not the switch manufacturer\'s as it depends on your application).

I can\'t find (online) the more comprehensive reference that I have
locally but here\'s ganssle\'s take on switch bounce. It\'s one of
my go-to documents for folks who THINK they understand the
dynamics of switches. E.g., folks seem to think switches
don\'t bounce when *opening* -- oops! :
Sometimes they don\'t.

Note the observation of a *relay* driven repeatedly by an MCU attached
to a storage scope (hard to drive a *switch* thusly) -- emphasis mine:

\"When the relay OPENED it always had a max bounce time of 2.3 to 2.9 msec,
at speeds from 2.5 to 30 Hz. More variation appeared on contact CLOSURE:
at 2.5 Hz bounces never exceeded 410 µsec, which climbed to 1080 µsec at
30 Hz. Why? I have no idea. But it\'s clear there is some correlation
between fast actuations and more bounce.\"
A relay will twang, namely ring like a bell, for a long time after a
transition. And bells have very complex vibration patterns. The twang
affects future actuations.

Reed relays make terrible analog multiplexers, because the high-Q
mechanical ringing makes voltages in the mag field and takes a long
time to die out.
I don\'t care about bouncing and/or ringing, as long as they don\'t short out the two throws. Some switches has center disconnect or middle throw, but they are not heavy duty rating. I might have a separate disconnect switch in series with the function switch, perhaps in lock box door type. The function switch is hidden inside the box. When the user open the lock box, the circuit should be disconnected.
Your voltages and current are easily handled by power MOSFETs.
But I don\'t trust the failure modes of semis and/or electronics.
Good thing you don\'t drive an electric car then. You would live in constant fear.

If they have NC relays shorting the battery, i won\'t even ride in it.

 

[Fred Bloggs]

On Wednesday, May 31, 2023 at 2:03:20 PM UTC-4, Ed Lee wrote:

On Wednesday, May 31, 2023 at 10:53:20 AM UTC-7, Ricky wrote:
On Wednesday, May 31, 2023 at 11:20:08 AM UTC-4, Ed Lee wrote:
On Wednesday, May 31, 2023 at 8:14:57 AM UTC-7, Fred Bloggs wrote:
On Tuesday, May 30, 2023 at 6:34:28 PM UTC-4, Ed Lee wrote:
On Tuesday, May 30, 2023 at 2:50:28 PM UTC-7, John Larkin wrote:
On Tue, 30 May 2023 14:18:35 -0700, Don Y
blocked...@foo.invalid> wrote:

On 5/30/2023 6:13 AM, Don Y wrote:
Find a *documented* switching mechanism. If you\'re switching
*DC*, expect the switch to be heavily derated. And, expect to
have to factor in some arcing during the transition. You
may be better off with a mercury-wetted *relay*, depending on
your application.

Note that switches are not \"ideal\" even when in steady state
but in *transition*, are complex creatures. A matED contact
will \"bounce\" as it is opened and a matING contact will bounce
as it\'s closing. Ideally, you want to know the time when the
contact is \"in flight\" AND the voltage rating at that point
(assuming any arcing has subsided -- that\'s YOUR calculation
not the switch manufacturer\'s as it depends on your application).

I can\'t find (online) the more comprehensive reference that I have
locally but here\'s ganssle\'s take on switch bounce. It\'s one of
my go-to documents for folks who THINK they understand the
dynamics of switches. E.g., folks seem to think switches
don\'t bounce when *opening* -- oops! :
Sometimes they don\'t.

Note the observation of a *relay* driven repeatedly by an MCU attached
to a storage scope (hard to drive a *switch* thusly) -- emphasis mine:

\"When the relay OPENED it always had a max bounce time of 2.3 to 2.9 msec,
at speeds from 2.5 to 30 Hz. More variation appeared on contact CLOSURE:
at 2.5 Hz bounces never exceeded 410 µsec, which climbed to 1080 µsec at
30 Hz. Why? I have no idea. But it\'s clear there is some correlation
between fast actuations and more bounce.\"
A relay will twang, namely ring like a bell, for a long time after a
transition. And bells have very complex vibration patterns. The twang
affects future actuations.

Reed relays make terrible analog multiplexers, because the high-Q
mechanical ringing makes voltages in the mag field and takes a long
time to die out.
I don\'t care about bouncing and/or ringing, as long as they don\'t short out the two throws. Some switches has center disconnect or middle throw, but they are not heavy duty rating. I might have a separate disconnect switch in series with the function switch, perhaps in lock box door type. The function switch is hidden inside the box. When the user open the lock box, the circuit should be disconnected.
Your voltages and current are easily handled by power MOSFETs.
But I don\'t trust the failure modes of semis and/or electronics.
Good thing you don\'t drive an electric car then. You would live in constant fear.
If they have NC relays shorting the battery, i won\'t even ride in it.

The short would require a welded contact. If I understand you correctly your switch is a mux selecting one of two batteries to be put on the pole?

 

[Ed Lee]

On Wednesday, May 31, 2023 at 11:11:05 AM UTC-7, Fred Bloggs wrote:

On Wednesday, May 31, 2023 at 2:03:20 PM UTC-4, Ed Lee wrote:
On Wednesday, May 31, 2023 at 10:53:20 AM UTC-7, Ricky wrote:
On Wednesday, May 31, 2023 at 11:20:08 AM UTC-4, Ed Lee wrote:
On Wednesday, May 31, 2023 at 8:14:57 AM UTC-7, Fred Bloggs wrote:
On Tuesday, May 30, 2023 at 6:34:28 PM UTC-4, Ed Lee wrote:
On Tuesday, May 30, 2023 at 2:50:28 PM UTC-7, John Larkin wrote:
On Tue, 30 May 2023 14:18:35 -0700, Don Y
blocked...@foo.invalid> wrote:

On 5/30/2023 6:13 AM, Don Y wrote:
Find a *documented* switching mechanism. If you\'re switching
*DC*, expect the switch to be heavily derated. And, expect to
have to factor in some arcing during the transition. You
may be better off with a mercury-wetted *relay*, depending on
your application.

Note that switches are not \"ideal\" even when in steady state
but in *transition*, are complex creatures. A matED contact
will \"bounce\" as it is opened and a matING contact will bounce
as it\'s closing. Ideally, you want to know the time when the
contact is \"in flight\" AND the voltage rating at that point
(assuming any arcing has subsided -- that\'s YOUR calculation
not the switch manufacturer\'s as it depends on your application).

I can\'t find (online) the more comprehensive reference that I have
locally but here\'s ganssle\'s take on switch bounce. It\'s one of
my go-to documents for folks who THINK they understand the
dynamics of switches. E.g., folks seem to think switches
don\'t bounce when *opening* -- oops! :
Sometimes they don\'t.

Note the observation of a *relay* driven repeatedly by an MCU attached
to a storage scope (hard to drive a *switch* thusly) -- emphasis mine:

\"When the relay OPENED it always had a max bounce time of 2..3 to 2.9 msec,
at speeds from 2.5 to 30 Hz. More variation appeared on contact CLOSURE:
at 2.5 Hz bounces never exceeded 410 µsec, which climbed to 1080 µsec at
30 Hz. Why? I have no idea. But it\'s clear there is some correlation
between fast actuations and more bounce.\"
A relay will twang, namely ring like a bell, for a long time after a
transition. And bells have very complex vibration patterns. The twang
affects future actuations.

Reed relays make terrible analog multiplexers, because the high-Q
mechanical ringing makes voltages in the mag field and takes a long
time to die out.
I don\'t care about bouncing and/or ringing, as long as they don\'t short out the two throws. Some switches has center disconnect or middle throw, but they are not heavy duty rating. I might have a separate disconnect switch in series with the function switch, perhaps in lock box door type.. The function switch is hidden inside the box. When the user open the lock box, the circuit should be disconnected.
Your voltages and current are easily handled by power MOSFETs.
But I don\'t trust the failure modes of semis and/or electronics.
Good thing you don\'t drive an electric car then. You would live in constant fear.
If they have NC relays shorting the battery, i won\'t even ride in it.
The short would require a welded contact. If I understand you correctly your switch is a mux selecting one of two batteries to be put on the pole?

What if power is loss to both relays?

 

[Ricky]

On Wednesday, May 31, 2023 at 2:03:20 PM UTC-4, Ed Lee wrote:

On Wednesday, May 31, 2023 at 10:53:20 AM UTC-7, Ricky wrote:
On Wednesday, May 31, 2023 at 11:20:08 AM UTC-4, Ed Lee wrote:
On Wednesday, May 31, 2023 at 8:14:57 AM UTC-7, Fred Bloggs wrote:
On Tuesday, May 30, 2023 at 6:34:28 PM UTC-4, Ed Lee wrote:
On Tuesday, May 30, 2023 at 2:50:28 PM UTC-7, John Larkin wrote:
On Tue, 30 May 2023 14:18:35 -0700, Don Y
blocked...@foo.invalid> wrote:

On 5/30/2023 6:13 AM, Don Y wrote:
Find a *documented* switching mechanism. If you\'re switching
*DC*, expect the switch to be heavily derated. And, expect to
have to factor in some arcing during the transition. You
may be better off with a mercury-wetted *relay*, depending on
your application.

Note that switches are not \"ideal\" even when in steady state
but in *transition*, are complex creatures. A matED contact
will \"bounce\" as it is opened and a matING contact will bounce
as it\'s closing. Ideally, you want to know the time when the
contact is \"in flight\" AND the voltage rating at that point
(assuming any arcing has subsided -- that\'s YOUR calculation
not the switch manufacturer\'s as it depends on your application).

I can\'t find (online) the more comprehensive reference that I have
locally but here\'s ganssle\'s take on switch bounce. It\'s one of
my go-to documents for folks who THINK they understand the
dynamics of switches. E.g., folks seem to think switches
don\'t bounce when *opening* -- oops! :
Sometimes they don\'t.

Note the observation of a *relay* driven repeatedly by an MCU attached
to a storage scope (hard to drive a *switch* thusly) -- emphasis mine:

\"When the relay OPENED it always had a max bounce time of 2.3 to 2.9 msec,
at speeds from 2.5 to 30 Hz. More variation appeared on contact CLOSURE:
at 2.5 Hz bounces never exceeded 410 µsec, which climbed to 1080 µsec at
30 Hz. Why? I have no idea. But it\'s clear there is some correlation
between fast actuations and more bounce.\"
A relay will twang, namely ring like a bell, for a long time after a
transition. And bells have very complex vibration patterns. The twang
affects future actuations.

Reed relays make terrible analog multiplexers, because the high-Q
mechanical ringing makes voltages in the mag field and takes a long
time to die out.
I don\'t care about bouncing and/or ringing, as long as they don\'t short out the two throws. Some switches has center disconnect or middle throw, but they are not heavy duty rating. I might have a separate disconnect switch in series with the function switch, perhaps in lock box door type. The function switch is hidden inside the box. When the user open the lock box, the circuit should be disconnected.
Your voltages and current are easily handled by power MOSFETs.
But I don\'t trust the failure modes of semis and/or electronics.
Good thing you don\'t drive an electric car then. You would live in constant fear.
If they have NC relays shorting the battery, i won\'t even ride in it.

What are you talking about???

--

Rick C.

+ Get 1,000 miles of free Supercharging
+ Tesla referral code - https://ts.la/richard11209

 

[Ed Lee]

On Wednesday, May 31, 2023 at 12:49:02 PM UTC-7, Ricky wrote:

On Wednesday, May 31, 2023 at 2:03:20 PM UTC-4, Ed Lee wrote:
On Wednesday, May 31, 2023 at 10:53:20 AM UTC-7, Ricky wrote:
On Wednesday, May 31, 2023 at 11:20:08 AM UTC-4, Ed Lee wrote:
On Wednesday, May 31, 2023 at 8:14:57 AM UTC-7, Fred Bloggs wrote:
On Tuesday, May 30, 2023 at 6:34:28 PM UTC-4, Ed Lee wrote:
On Tuesday, May 30, 2023 at 2:50:28 PM UTC-7, John Larkin wrote:
On Tue, 30 May 2023 14:18:35 -0700, Don Y
blocked...@foo.invalid> wrote:

On 5/30/2023 6:13 AM, Don Y wrote:
Find a *documented* switching mechanism. If you\'re switching
*DC*, expect the switch to be heavily derated. And, expect to
have to factor in some arcing during the transition. You
may be better off with a mercury-wetted *relay*, depending on
your application.

Note that switches are not \"ideal\" even when in steady state
but in *transition*, are complex creatures. A matED contact
will \"bounce\" as it is opened and a matING contact will bounce
as it\'s closing. Ideally, you want to know the time when the
contact is \"in flight\" AND the voltage rating at that point
(assuming any arcing has subsided -- that\'s YOUR calculation
not the switch manufacturer\'s as it depends on your application).

I can\'t find (online) the more comprehensive reference that I have
locally but here\'s ganssle\'s take on switch bounce. It\'s one of
my go-to documents for folks who THINK they understand the
dynamics of switches. E.g., folks seem to think switches
don\'t bounce when *opening* -- oops! :
Sometimes they don\'t.

Note the observation of a *relay* driven repeatedly by an MCU attached
to a storage scope (hard to drive a *switch* thusly) -- emphasis mine:

\"When the relay OPENED it always had a max bounce time of 2..3 to 2.9 msec,
at speeds from 2.5 to 30 Hz. More variation appeared on contact CLOSURE:
at 2.5 Hz bounces never exceeded 410 µsec, which climbed to 1080 µsec at
30 Hz. Why? I have no idea. But it\'s clear there is some correlation
between fast actuations and more bounce.\"
A relay will twang, namely ring like a bell, for a long time after a
transition. And bells have very complex vibration patterns. The twang
affects future actuations.

Reed relays make terrible analog multiplexers, because the high-Q
mechanical ringing makes voltages in the mag field and takes a long
time to die out.
I don\'t care about bouncing and/or ringing, as long as they don\'t short out the two throws. Some switches has center disconnect or middle throw, but they are not heavy duty rating. I might have a separate disconnect switch in series with the function switch, perhaps in lock box door type.. The function switch is hidden inside the box. When the user open the lock box, the circuit should be disconnected.
Your voltages and current are easily handled by power MOSFETs.
But I don\'t trust the failure modes of semis and/or electronics.
Good thing you don\'t drive an electric car then. You would live in constant fear.
If they have NC relays shorting the battery, i won\'t even ride in it.
What are you talking about???

He is suggesting using NC relays to replace toggle switches, which I use for mechanical isolations.

 

[Don Y]

On 5/31/2023 8:02 AM, Ed Lee wrote:

Someone suggested mercury switches.

Ordered some to try out:
https://www.ebay.com/itm/195708179355?var=495574245862

No. Mercury *wetted* switches/relays/contactors. The switch contacts
are coated witha bit of mercury that allows the connection to be
maintained AS the switch would otherwise be bouncing; it \"snaps\"
open/closed more cleanly.

<https://www.ebay.com/itm/155576690745>
<https://www.ebay.com/itm/134585850699>
<https://www.ebay.com/itm/155535895796>

Selecting the proper \"switch\" will depend on your application and
the faults you are likely to encounter *or* that must be guarded
against.

You may have to augment the \"switch\" with other circuitry to
guard against things like welded contacts -- if such an event
can \"cause a problem/loss\".

Or, may choose to use different technologies to address
different aspects of your particular problem.

[E.g., UPSs have a mechanical shorting bar that can be removed
to isolate the battery; far more reliable than the relays inside!]

> So, how much current can I expect? The listing says 1 to 15A in general and 0.3A in specific.

 

[Ed Lee]

On Wednesday, May 31, 2023 at 3:25:18 PM UTC-7, Don Y wrote:

On 5/31/2023 8:02 AM, Ed Lee wrote:
Someone suggested mercury switches.

Ordered some to try out:
https://www.ebay.com/itm/195708179355?var=495574245862
No. Mercury *wetted* switches/relays/contactors. The switch contacts
are coated witha bit of mercury that allows the connection to be
maintained AS the switch would otherwise be bouncing; it \"snaps\"
open/closed more cleanly.

https://www.ebay.com/itm/155576690745
https://www.ebay.com/itm/134585850699
https://www.ebay.com/itm/155535895796

Selecting the proper \"switch\" will depend on your application and
the faults you are likely to encounter *or* that must be guarded
against.

You may have to augment the \"switch\" with other circuitry to
guard against things like welded contacts -- if such an event
can \"cause a problem/loss\".

Or, may choose to use different technologies to address
different aspects of your particular problem.

[E.g., UPSs have a mechanical shorting bar that can be removed
to isolate the battery; far more reliable than the relays inside!]

Yes, i am thinking about using mercury switches for lock box door. When the user flip up the door to access the function switch, it should disconnect the power source first.

Also got couple of ex-soviet military surplus heavy duty 16P2T switches. But probably hard to find any more. It can switch 8 independent power sources upto 3A.

https://www.ebay.com/itm/353292464636

Ex-soviet is good for something.

 

[Ricky]

On Wednesday, May 31, 2023 at 4:39:09 PM UTC-4, Ed Lee wrote:

On Wednesday, May 31, 2023 at 12:49:02 PM UTC-7, Ricky wrote:
On Wednesday, May 31, 2023 at 2:03:20 PM UTC-4, Ed Lee wrote:
On Wednesday, May 31, 2023 at 10:53:20 AM UTC-7, Ricky wrote:
On Wednesday, May 31, 2023 at 11:20:08 AM UTC-4, Ed Lee wrote:
On Wednesday, May 31, 2023 at 8:14:57 AM UTC-7, Fred Bloggs wrote:
On Tuesday, May 30, 2023 at 6:34:28 PM UTC-4, Ed Lee wrote:
On Tuesday, May 30, 2023 at 2:50:28 PM UTC-7, John Larkin wrote:
On Tue, 30 May 2023 14:18:35 -0700, Don Y
blocked...@foo.invalid> wrote:

On 5/30/2023 6:13 AM, Don Y wrote:
Find a *documented* switching mechanism. If you\'re switching
*DC*, expect the switch to be heavily derated. And, expect to
have to factor in some arcing during the transition. You
may be better off with a mercury-wetted *relay*, depending on
your application.

Note that switches are not \"ideal\" even when in steady state
but in *transition*, are complex creatures. A matED contact
will \"bounce\" as it is opened and a matING contact will bounce
as it\'s closing. Ideally, you want to know the time when the
contact is \"in flight\" AND the voltage rating at that point
(assuming any arcing has subsided -- that\'s YOUR calculation
not the switch manufacturer\'s as it depends on your application).

I can\'t find (online) the more comprehensive reference that I have
locally but here\'s ganssle\'s take on switch bounce. It\'s one of
my go-to documents for folks who THINK they understand the
dynamics of switches. E.g., folks seem to think switches
don\'t bounce when *opening* -- oops! :
Sometimes they don\'t.

Note the observation of a *relay* driven repeatedly by an MCU attached
to a storage scope (hard to drive a *switch* thusly) -- emphasis mine:

\"When the relay OPENED it always had a max bounce time of 2.3 to 2.9 msec,
at speeds from 2.5 to 30 Hz. More variation appeared on contact CLOSURE:
at 2.5 Hz bounces never exceeded 410 µsec, which climbed to 1080 µsec at
30 Hz. Why? I have no idea. But it\'s clear there is some correlation
between fast actuations and more bounce.\"
A relay will twang, namely ring like a bell, for a long time after a
transition. And bells have very complex vibration patterns. The twang
affects future actuations.

Reed relays make terrible analog multiplexers, because the high-Q
mechanical ringing makes voltages in the mag field and takes a long
time to die out.
I don\'t care about bouncing and/or ringing, as long as they don\'t short out the two throws. Some switches has center disconnect or middle throw, but they are not heavy duty rating. I might have a separate disconnect switch in series with the function switch, perhaps in lock box door type. The function switch is hidden inside the box. When the user open the lock box, the circuit should be disconnected.
Your voltages and current are easily handled by power MOSFETs.
But I don\'t trust the failure modes of semis and/or electronics.
Good thing you don\'t drive an electric car then. You would live in constant fear.
If they have NC relays shorting the battery, i won\'t even ride in it.
What are you talking about???
He is suggesting using NC relays to replace toggle switches, which I use for mechanical isolations.

I don\'t know why I expected an explanation, but, I\'ll try again.

What does that have to do with shorting the battery???

--

Rick C.

-- Get 1,000 miles of free Supercharging
-- Tesla referral code - https://ts.la/richard11209

 

[Don Y]

On 5/31/2023 4:41 PM, Ed Lee wrote:

On Wednesday, May 31, 2023 at 3:25:18 PM UTC-7, Don Y wrote:
On 5/31/2023 8:02 AM, Ed Lee wrote:
Someone suggested mercury switches.

Ordered some to try out:
https://www.ebay.com/itm/195708179355?var=495574245862
No. Mercury *wetted* switches/relays/contactors. The switch contacts
are coated witha bit of mercury that allows the connection to be
maintained AS the switch would otherwise be bouncing; it \"snaps\"
open/closed more cleanly.

https://www.ebay.com/itm/155576690745
https://www.ebay.com/itm/134585850699
https://www.ebay.com/itm/155535895796

Selecting the proper \"switch\" will depend on your application and
the faults you are likely to encounter *or* that must be guarded
against.

You may have to augment the \"switch\" with other circuitry to
guard against things like welded contacts -- if such an event
can \"cause a problem/loss\".

Or, may choose to use different technologies to address
different aspects of your particular problem.

[E.g., UPSs have a mechanical shorting bar that can be removed
to isolate the battery; far more reliable than the relays inside!]

Yes, i am thinking about using mercury switches for lock box door. When the user flip up the door to access the function switch, it should disconnect the power source first.

(again, without knowing your application...) You can treat the \"door\"
as an armature and have it\'s opening *break* the contacts formed
by simply mating connectors (the traveling connector on the door
with the stationary one on the box)

Also got couple of ex-soviet military surplus heavy duty 16P2T switches. But probably hard to find any more. It can switch 8 independent power sources upto 3A.

https://www.ebay.com/itm/353292464636

Ex-soviet is good for something.

What are you trying to *do*? (I have assumed, in mu comments, this was
something EVbattery related)

 

[Don Y]

On 5/31/2023 5:43 PM, Don Y wrote:

On 5/31/2023 4:41 PM, Ed Lee wrote:
On Wednesday, May 31, 2023 at 3:25:18 PM UTC-7, Don Y wrote:
On 5/31/2023 8:02 AM, Ed Lee wrote:
Someone suggested mercury switches.

Ordered some to try out:
https://www.ebay.com/itm/195708179355?var=495574245862
No. Mercury *wetted* switches/relays/contactors. The switch contacts
are coated witha bit of mercury that allows the connection to be
maintained AS the switch would otherwise be bouncing; it \"snaps\"
open/closed more cleanly.

https://www.ebay.com/itm/155576690745
https://www.ebay.com/itm/134585850699
https://www.ebay.com/itm/155535895796

Selecting the proper \"switch\" will depend on your application and
the faults you are likely to encounter *or* that must be guarded
against.

You may have to augment the \"switch\" with other circuitry to
guard against things like welded contacts -- if such an event
can \"cause a problem/loss\".

Or, may choose to use different technologies to address
different aspects of your particular problem.

[E.g., UPSs have a mechanical shorting bar that can be removed
to isolate the battery; far more reliable than the relays inside!]

Yes, i am thinking about using mercury switches for lock box door.  When the
user flip up the door to access the function switch, it should disconnect the
power source first.

(again, without knowing your application...) You can treat the \"door\"
as an armature and have it\'s opening *break* the contacts formed
by simply mating connectors (the traveling connector on the door
with the stationary one on the box)

Think of the power cords that used to be part of the back panel on
old (CRT/tube) TVs... removing the panel (for service access)
effectively disconnected power to the set.

Also got couple of ex-soviet military surplus heavy duty 16P2T switches.  But
probably hard to find any more.  It can switch 8 independent power sources
upto 3A.

https://www.ebay.com/itm/353292464636

Ex-soviet is good for something.

What are you trying to *do*?  (I have assumed, in mu comments, this was
something EVbattery related)

 

[Ed Lee]

On Wednesday, May 31, 2023 at 5:43:46 PM UTC-7, Don Y wrote:

On 5/31/2023 4:41 PM, Ed Lee wrote:
On Wednesday, May 31, 2023 at 3:25:18 PM UTC-7, Don Y wrote:
On 5/31/2023 8:02 AM, Ed Lee wrote:
Someone suggested mercury switches.

Ordered some to try out:
https://www.ebay.com/itm/195708179355?var=495574245862
No. Mercury *wetted* switches/relays/contactors. The switch contacts
are coated witha bit of mercury that allows the connection to be
maintained AS the switch would otherwise be bouncing; it \"snaps\"
open/closed more cleanly.

https://www.ebay.com/itm/155576690745
https://www.ebay.com/itm/134585850699
https://www.ebay.com/itm/155535895796

Selecting the proper \"switch\" will depend on your application and
the faults you are likely to encounter *or* that must be guarded
against.

You may have to augment the \"switch\" with other circuitry to
guard against things like welded contacts -- if such an event
can \"cause a problem/loss\".

Or, may choose to use different technologies to address
different aspects of your particular problem.

[E.g., UPSs have a mechanical shorting bar that can be removed
to isolate the battery; far more reliable than the relays inside!]

Yes, i am thinking about using mercury switches for lock box door. When the user flip up the door to access the function switch, it should disconnect the power source first.
(again, without knowing your application...) You can treat the \"door\"
as an armature and have it\'s opening *break* the contacts formed
by simply mating connectors (the traveling connector on the door
with the stationary one on the box)
Also got couple of ex-soviet military surplus heavy duty 16P2T switches.. But probably hard to find any more. It can switch 8 independent power sources upto 3A.

https://www.ebay.com/itm/353292464636

Ex-soviet is good for something.
What are you trying to *do*? (I have assumed, in mu comments, this was
something EVbattery related)

Switch 8 48V cells into 384V.

 

[Don Y]

On 5/31/2023 6:56 PM, Ed Lee wrote:

What are you trying to *do*? (I have assumed, in mu comments, this was
something EVbattery related)

Switch 8 48V cells into 384V.

Saying so LITTLE -- yet wanting such a switch -- suggests you
want to treat the 8 cells as *individually* \"0V referenced\"
(perhaps for a 48V charger?) and then STACK them in series
(throwing a switch) to develop the 384VDC.

I.e., why aren\'t you just wiring them in series permanently?

 

[Ed Lee]

On Wednesday, May 31, 2023 at 7:26:19 PM UTC-7, Don Y wrote:

On 5/31/2023 6:56 PM, Ed Lee wrote:
What are you trying to *do*? (I have assumed, in mu comments, this was
something EVbattery related)

Switch 8 48V cells into 384V.
Saying so LITTLE -- yet wanting such a switch -- suggests you
want to treat the 8 cells as *individually* \"0V referenced\"
(perhaps for a 48V charger?) and then STACK them in series
(throwing a switch) to develop the 384VDC.

I.e., why aren\'t you just wiring them in series permanently?

They are in different capacities and state of charge, and total voltage vary between 384V and 408V. Changing voltage allows them to discharge deeper. It\'s deeper cycle batteries.