Solenoid driver

[earckens]

Some years ago an interesting discussion went on about a small unit to drive solenoids for model railroad turnouts: https://www.electro-tech-online.com/threads/capacitor-circuit-and-current-flow.125521/

Now I would like to continue/restart/update this discussion although with some extra input.
Again, given is a solenoid with approx. 10mH inductance that has to be powered during a short (300ms to max 1sec) with a DC pulse of between 16V and 25V in order to get the solenoid to drive a model railroad turnout.
Requirements:
1. as fast charge as possible
2. no charge when load is applied
3. as high discharge current as possible
4. protection against back-emf (although this is not such a problem when reviewing the discussion of 2012)
Here is a simple explanation of the concept: https://www.rebelhosts.com/tt/te/p16.htm

My questions are:
1. one circuit (train2.gif) uses a power 2N3055, the other (Cap_Dis_Circuit.gif) a Darlington BD679. Although I would rather use a higher power Darlington TIP120 I wonder what the performance difference is between the 2N3055 and the Darlington. the purchase price of a Darlington is about half that of the 2N3055
2. circuit "CDblock" uses full bridge rectification: is there any advantage re. one of the requirements over the previous two circuits?
3. is circuit "CDblocksplit" in an advantage of "CDblock" re. charging time when both simultaneously discharge over 2 solenoids?
4. Circuit "CDSCR" uses SCR's to do the switching operation of the current but what interests me is the use of an opamp as voltage regulator:
a. is there an operational advantage in this?
b. does any of the above requirements get better resolved in this circuit?

Thank you for your valuable inputs, it is very appreciated!
Erik

 

[chemelec]

Using a 555 driving a Mosfet would be much better.
When triggered, it just give one pulse to the solenoid, than shuts Totally Off .

 

[spec]

Hy earckens,

You asked me to provide a bit more information in response to your opening post of this thread; wow, quite a wide ranging scope.

But right from the start, chemelec in post #2 has given the optimum method for generating a pulse to smartly and safely drive the point switching solenoid on a model railway.

I will respond to your questions in your opening thread with individual posts.

spec

 

[spec]

Issue 4 of 2016_10_04
Question 1: difference between following transistors:

(1.1) 2N3055: https://www.onsemi.com/pub_link/Collateral/2N3055-D.PDF
Metal can, NPN, high power, bipolar junction transistor (BJT): rugged, cheap, ubiquitous. The original consumer high power transistor. The 2N3055 can handle much more current (15A) and power (115W) than the other two transistors, but being only a single transistor only has a current gain (hFE) of around 50, dropping badly to 15 at 10A.

versus

(1.2) BD679: https://www.st.com/content/ccc/reso...df/jcr:content/translations/en.CD00000939.pdf
Plastic case, NPN, BJT medium power Darlington: cheap. popular. A Darlington transistor is essentially two transistors in one case to give double the hFE of a single transistor. In this case the hFE is 750 at 1.5A. Also included in the case are bias resistors and, normally, as in this case, a snubbing diode to catch voltage over swing with an inductive load.

versus

(1.3) TIP120: https://www.onsemi.com/pub_link/Collateral/TIP120-D.PDF
Plastic case, NPN, BJT medium power Darlington: cheap, ubiquitous. The TIP 120 is simply better than the BD679 in all respects and can be safely used as a substitute for the BD679 in all normal applications. hFE= 1,000 at 3A.

Darlington transistors, and to a large degree medium power and high power BJTs, have been completely eclipsed by MOSFETs in most circuits these days. High end audio power amplifiers still use BJTs though.

spec

 

[spec]

Issue 0 of 2016_10_04
Question 2: half wave versus full wave rectification.

In general full wave rectification offers the following advantages:
(2.1) charges the reservoir capacitor twice as fast
(2.2) halves the ripple voltage
(2.3) halves the ripple current
(2.4) doubles the ripple voltage frequency, which means easier ripple voltage filtering
(2.5) balances the current taken from the mains transformer/mains supply

spec

 

[spec]

Issue 1 of 2016_10_04
Question 3: split charging

I may be missing something specific to model railways, but as far as I can see there is no advantage with this circuit that could not be achieved by simply doubling up on the positive part of the circuit.

spec

 

[spec]

Issue 1 of 2016_10_04
Question 4: use of opamp

The opamp does not provide any real benefit; it does better define certain circuit parameters. But the base drive to the BJT is limited due to the opamp limited current sink capability.

spec

 

[earckens]

Using a 555 driving a Mosfet would be much better.
When triggered, it just give one pulse to the solenoid, than shuts Totally Off .

Hi chemelec, I think I do not understand your response. In which circuit drawing would you use a 555 and MOSFET?
Thank you,
Erik

 

[earckens]

Issue 1 of 2016_10_04
Question 3: split charging

I may be missing something specific to model railways, but as far as I can see there is no advantage with this circuit that could not be achieved by simply doubling up on the positive part of the circuit.

spec

spec, my question 3. is not really relevant I realise now because the charge current is dependent on the transformer current supply rating, not on the charging transistor. And I just now see the rest of your answer, I concur. Thank you!

 

[earckens]

That answers about all of it pending explanation of the use of a 555 with a MOSFET (chemelec post #2). And then my further question: MOSFET or SCR?

 

[spec]

spec, my question 3. is not really relevant I realise now because the charge current is dependent on the transformer current supply rating, not on the charging transistor. And I just now see the rest of your answer, I concur. Thank you!

Yes, you are correct that the charge current is limited by the input power.

In general there are many ways to implement a particular function: one designer will favor one approach and another designer will favor another approach and, quite often there is little to chose between the two approaches. Also, tradition and common practice seems to have a big influence on the design approach used, as in this area where discharging a capacitor seems to be favored.

chemelec is proposing a new approach which will required a completely new circuit.

I will do a schematic showing chemelec's approach (update: no need, chemelec has posted a circuit).

spec

 

[spec]

That answers about all of it pending explanation of the use of a 555 with a MOSFET (chemelec post #2). And then my further question: MOSFET or SCR?

MOSFET by a mile for this application.

spec

 

[chemelec]

possibly something like this:
(500Ms on time)
The 10k And/Or the 47uF cap can be changed to Change the On Time.

 

[alec_t]

With that circuit you need a pull-up resistor (e.g. 10k) from the trigger switch to the 12V rail.

 

[chemelec]

Only Pin 2 Requires a Pull Up, unless you need Quick Repetition.
The cap and Switch Pull pin 2 Low, for just a Split Second to trigger the 555.
The cap doesn't hold a charge for very long.

 

[chemelec]

Hi chemelec, I think I do not understand your response. In which circuit drawing would you use a 555 and MOSFET?
Thank you,
Erik

Not With them, But Instead of those Other Circuits.

 

[spec]

possibly something like this:
(500Ms on time)
The 10k And/Or the 47uF cap can be changed to Change the On Time.
View attachment 101487

In addition to your circuit's elegance: accuracy, simplicity, and low cost, there is another advantage: the NMOSFETs intrinsic diode from drain to source will catch negative voltage swings (kickback) from the solenoid coil.

spec

 

[spec]

Hi earckens,

You would not need one of chemelec's circuits for each solenoid, as is shown with the other circuits. Instead, you can simply have an N way switch to select any one solenoid to operate from from N solenoids.

spec

 

[earckens]

MOSFET by a mile for this application.

spec

Why? Excuse my ignorance, I had my studies finished 30 years ago

 

[earckens]

possibly something like this:
(500Ms on time)
The 10k And/Or the 47uF cap can be changed to Change the On Time.
View attachment 101487

Wow, great! Thanks a million.

Just what do you suggest to use for the MOSFET? I happen to have IRL520N and IRLB8721 in stock here..