Newsgroups: sci.electronics
Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!wuarchive!uunet!medar!jbertoia
From: jbertoia@medar.com (Jeffrey A. Bertoia)
Subject: Re: solid state relays
Organization: Medar, Inc.     Farmington Hills, MI
Date: Fri, 17 May 1991 17:12:03 GMT
Message-ID: <1991May17.171203.29529@medar.com>
Summary: They're really pretty good.
References: <12500@qisoff.phx.mcd.mot.com> <1991May14.135524.25014@wsrcc.com> <1991May15.132910.1396@specialix.co.uk>

In article <1991May15.132910.1396@specialix.co.uk> stevem@specialix.co.uk (Steven Murray) writes:
>wolfgang@wsrcc.com (Wolfgang S. Rupprecht) writes:
>
>>hbg6@citek.mcdphx.mot.com writes:
>>> I am building a controller for the evaporative cooler on my house. The
>>> blower motor has two speeds, selected by providing power to either
>>> the HIGH terminal, or the LOW terminal.
>

>>> I really want to use solid state relays to avoid the cost and 
>>> complexity of contactors and the associated drivers.
>
Good move!

>>You will also want to use a good snubber network and a zero crossing
>>detector (for turn on).  This will reduce RF bursts that will lead to
>>radio interference and also an early destruction of the SCR's.
 
>Expanding on this a bit, driving motors with SCR's or solid state relays,
>especially anything more than 'one motor, one relay' can be painful.  You
>can tend to get problems with inductive spikes and back EMF - zero crossing
>switch-on is fine, but that doesn't control the switch off.  Snubbers are
>great, VDR's aren't such a bad idea either.  Aim to protect the SSR's with
>VDR's rather than stop interference from the load.

Expanding yet further and refuting some.

In the following text I use SCR/SSR relatively interchangably.

The zero crossing detector for turn on will definately reduce the RFI generated
but will have little effect on the overall life of the SCR/SSR as long as it
has been properly sized for the application.  Note: In sizing you must consider
not only the steady state but also the surge ratings.

Medar's business is resistance welding controls which generally involves phase
controlling a very large transformer.  More often than not we swicth SCR's on
at times other than the voltage zero-crossing.  We rarely 'lose' SCRs due to
current or voltage transients.  A far more common failure mode is improper
cooling.

Now a snubber, will help extend the life of an SCR in it's off state (i.e. it
will help protect the SCR for overvoltage transients in it's non-conducting
mode).  It will also assist in turn-on if the gate drive circuit is
inadequate,  however if I remember, the original post regarded SSRs which
have an integral drive circuit which is properly sized.  The problem with
snubbers in these applications is that the off state over voltage
protection is marginal due to the capacitors inability to store the energy.

An additional consideration in using a snubber is that 'some' voltage will
be present on the load side of the snubber at _all_ times do to the leakage
through the snubber.  This may or may not be tolerable.

If overvoltage is the main concern a MOV (Metal Oxide Varistor) will do the
job nicely.  A MOV will 'break-over' at a specified voltage quencing the
spike.  Note: Size the MOV correctly otherwise things WILL get exciting...
Lots of noise and smoke will follow.

As I remember the original post... There was also a question about what if
the SSR fails on.  One of the frequent failure modes in SCRs is to fail on,
but the is not necessarily a problem.  The circuit can be properly fused to
stop any potential problems.  Most motors have sigificant thermal mass and
can tolerate short over current conditions.  For the very cautious a small
protection circuit using (small) interlocking relays should be no problem
to design thus preventing both SSRs from coming on simultaniously.  This
exercise is left to the reader.

jeff
-- 
Jeffrey A. Bertoia				Medar, Inc.
...!uunet!medar!jbertoia			38700 Grand River Ave.
jbertoia@medar.com				Farmington Hills, MI 48335
