Maker Pro
Maker Pro

Coil driving

N

Nuby

Jan 1, 1970
0
What's wrong with this picture?

TTL control lines with small current sinking capability (say 20ma)
need to activate electromagnetic coils (50VDC, 300ma). Each control
line gates an SCR which runs Vcc into a coil. The group of coils has
a common disable in the form of a big relay that cuts the common
connection to ground.
 
J

JeffM

Jan 1, 1970
0
Nuby said:
What's wrong with this picture?
TTL control lines with small current sinking capability (say 20ma)
need to activate electromagnetic coils (50VDC, 300ma). Each control
line gates an SCR which runs Vcc into a coil. The group of coils has
a common disable in the form of a big relay that cuts the common
connection to ground.

Are you planning to protect the semiconductors from inductive
kickback?
http://groups.google.com/groups/sea...diode+OR+kick-back.diode+OR+suppression.diode
 
What's wrong with this picture?

TTL control lines with small current sinking capability (say 20ma)
need to activate electromagnetic coils (50VDC, 300ma). Each control
line gates an SCR which runs Vcc into a coil. The group of coils has
a common disable in the form of a big relay that cuts the common
connection to ground.

You're missing an interface driver. Perhaps a common ULN2003 would
help. If you're running DC with SCRs, it will turn on one time and
latch. Is that the goal? Lifting the ground as a master reset will
complicate the TTL interface. Can you interrupt the positive instead?

What are you really trying to do?

GG
 
D

default

Jan 1, 1970
0
What's wrong with this picture?

TTL control lines with small current sinking capability (say 20ma)
need to activate electromagnetic coils (50VDC, 300ma). Each control
line gates an SCR which runs Vcc into a coil. The group of coils has
a common disable in the form of a big relay that cuts the common
connection to ground.

Maybe nothing. Assuming you have the correct polarity to gate the
SCR's into conduction.

A snubber across the coils isn't a bad idea, but SRR's tend to be
pretty rugged. I'd probably just go for it, and see if anything
smokes.
 
J

Jamie

Jan 1, 1970
0
Nuby said:
What's wrong with this picture?

TTL control lines with small current sinking capability (say 20ma)
need to activate electromagnetic coils (50VDC, 300ma). Each control
line gates an SCR which runs Vcc into a coil. The group of coils has
a common disable in the form of a big relay that cuts the common
connection to ground.
You should be using opto-isolators that will drive the SCR etc.
You must insure you use the proper relay coil for this. With
SCR's, it would be a DC coil. with an TRIAC , it would be an AC coil.

This only assumes that you're using AC as the load source.
if DC is going to be the source. The main relay you speak of must
not be an SCR. it has to be something like a real relay, GTO or
FET switch other wise, you'll get latch up until power is removed.


Actually, you can get SSR's (Solid State Relays) that can be driven
from a TTL signal.
 
E

Eeyore

Jan 1, 1970
0
Nuby said:
What's wrong with this picture?

TTL control lines with small current sinking capability (say 20ma)
need to activate electromagnetic coils (50VDC, 300ma). Each control
line gates an SCR which runs Vcc into a coil. The group of coils has
a common disable in the form of a big relay that cuts the common
connection to ground.

In overall concept it's ok but there are some practical considerations.

a. Do you really mean actual TTL ? TTL has annoying limits on output voltage
swing and 20mA may also be an inadequate trigger current for the thyristors.

b. The common disable would make more sense in the 50V DC supply positive input.

c. As others have noted, you need protection from inductive kickback when the
loads switch off.

Graham
 
E

Eeyore

Jan 1, 1970
0
Jamie said:
You should be using opto-isolators that will drive the SCR etc.

Why ?

You must insure you use the proper relay coil for this. With
SCR's, it would be a DC coil. with an TRIAC , it would be an AC coil.

Ne never mentioned AC, why confuse the issue ?

Graham
 
N

Nuby

Jan 1, 1970
0
You're missing an interface driver. Perhaps a common ULN2003 would
help. If you're running DC with SCRs, it will turn on one time and
latch. Is that the goal?

Not the end goal, but an acceptable method.
Lifting the ground as a master reset will
complicate the TTL interface.

I don't see the connection/problem. Can you explain?
Can you interrupt the positive instead?

I can, but having a common connection to Vcc and distinct/controlled
connections to ground appears slightly perverse. I can afford to
ground the frame, which simplifies the harness, but will not connect
the frame to Vcc.
What are you really trying to do?

Control 100 miniature/high-density coils as cost-effectively as
possible.
 
N

Nuby

Jan 1, 1970
0
You should be using opto-isolators that will drive the SCR etc

Um, have you priced such isolators recently? I'm looking for a
circuit as simple as possible.
..
You must insure you use the proper relay coil for this. With
SCR's, it would be a DC coil. with an TRIAC , it would be an AC coil.

Hunh? None of that makes sense to me. What relay?
This only assumes that you're using AC as the load source.

What load are you talking about?
if DC is going to be the source. The main relay you speak of must
not be an SCR. it has to be something like a real relay, GTO or
FET switch other wise, you'll get latch up until power is removed.

The only purpose of the big relay is to interrupt the latch effect.
Actually, you can get SSR's (Solid State Relays) that can be driven
from a TTL signal.

Yes, but they are neither simple nor cheap. Why do I need such a
complex component?
 
N

Nuby

Jan 1, 1970
0
In overall concept it's ok but there are some practical considerations.

a. Do you really mean actual TTL ?

I think so. Typical outputs would drive 10-20 TTL loads. The
actual control chip is a serial-to-parallel shift register.
TTL has annoying limits on output voltage
swing and 20mA may also be an inadequate trigger current for the thyristors.

Interesting point.
b. The common disable would make more sense in the 50V DC supply positive input.
OK.


c. As others have noted, you need protection from inductive kickback when the
loads switch off.

How can I determine how much of a problem this would be? The energy
in each active coils is going to be pretty small.
 
E

Eeyore

Jan 1, 1970
0
Nuby said:
I think so. Typical outputs would drive 10-20 TTL loads. The
actual control chip is a serial-to-parallel shift register.

What's this chip's part number ?

Interesting point.

Depends of course on the SCR you use. Did you have a specific one in mind ?

How can I determine how much of a problem this would be? The energy
in each active coils is going to be pretty small.

A simple inverse parallel diode such as a 1N400x across each relay coil should be
fine.

Graham
 
E

Eeyore

Jan 1, 1970
0
Nuby said:
Um, have you priced such isolators recently? I'm looking for a
circuit as simple as possible.
.

Hunh? None of that makes sense to me. What relay?


What load are you talking about?


The only purpose of the big relay is to interrupt the latch effect.


Yes, but they are neither simple nor cheap. Why do I need such a
complex component?

Never mind Jamie, he talks a lot of nonsense, including posting things that are
100% wrong. Every group has one. In this case he merely posted mainly random
data.

Graham
 
N

Nuby

Jan 1, 1970
0
What's this chip's part number ?

We have not chosen a specifc one yet. An example would be the
MM74HC164 family.
Depends of course on the SCR you use. Did you have a specific one in mind ?

Again we don't have a specific target yet. Any sensitive-gate SCR
that can handle the (light) duty cycle would work for us.

[...]
A simple inverse parallel diode such as a 1N400x across each relay coil should be
fine.

OK, but I still don't know whether the protection is necessary.
 
E

Eeyore

Jan 1, 1970
0
Nuby said:
We have not chosen a specifc one yet. An example would be the
MM74HC164 family.

Ah, fine. That's not TTL, it's HCMOS. Much better.

Again we don't have a specific target yet. Any sensitive-gate SCR
that can handle the (light) duty cycle would work for us.

This would appear to do the job.
http://uk.farnell.com/jsp/Semicondu...DUCTOR/2N5064G/displayProduct.jsp?sku=9556540

It's nice and cheap too.
OK, but I still don't know whether the protection is necessary.

It would be very silly not to fit it.

Graham
 
J

Jamie

Jan 1, 1970
0
Eeyore said:
Nuby wrote:




Never mind Jamie, he talks a lot of nonsense, including posting things that are
100% wrong. Every group has one. In this case he merely posted mainly random
^^^^^^^^^^^^^^ One should practice what one preaches.
data.

Graham
You're such an anus Mr. Ham...

When some one gives incomplete information. It can lead to to many
roads. Until that poster makes it clear as to what they're trying to do,
which apparently has taken place now. People that have been down many
avenues can come up with so many scenarios that it would make some one
like you get all confused. Sorry for your disablity in this area.

I guess that is what most would call "separating the men from the
boys" when it comes to problem solving which you seem to be lacking.

Have a good day sucker.
 
J

Jamie

Jan 1, 1970
0
Nuby said:
What's this chip's part number ?


We have not chosen a specifc one yet. An example would be the
MM74HC164 family.

Depends of course on the SCR you use. Did you have a specific one in mind ?


Again we don't have a specific target yet. Any sensitive-gate SCR
that can handle the (light) duty cycle would work for us.

[...]

A simple inverse parallel diode such as a 1N400x across each relay coil should be
fine.


OK, but I still don't know whether the protection is necessary.
yes, it's required on the coils, other wise, you'll take out the SCR
at some point, and may also take out the TTL device.

The other problem is the (time on) gate signal. The coil is inductive
which means you'll get low current in the SCR at the initial signal.
If the Gate isn't kept on for the minimum time required that gets the
SCR => it's holding current. It will not latch. This can be solved with
a resistor across the coil if it becomes a problem.

To know that, you would have to know the inductive value of the coil
and perform some calculations for the time frame your signal will be on.


Have a good day.
 
Top