• Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Diode recommendation for automotive pwm fan circuit?

Status
Not open for further replies.

atxz06

New Member
My research for the right diode selection has quickly exhausted my electrical expertise.

I'm wiring a pair of automotive (12V obviously) radiator fans (40-50A) through a Crydom D1D100 solid state relay controlled by Holley Dominator EFI via PWM up to 800Hz.

What would be a good diode choice to protect the relay from flyback and operate in this environment?

Appreciate someone pointing me in the right direction.

Thanks!
 

atxz06

New Member
Currently switching the positive side but could switch.

Snubber... are you asking if I know the direction should be negative to positive?

Thanks for direction...
 

crutschow

Well-Known Member
Most Helpful Member
A 50A minimum, 24V minimum bolt-down diode in series with a 0.22 Ohm 50W minimum, wire-wound, bolt-down resistor should do the job.
What's the purpose of the resistor? :confused:
All it will do is dissipate inductive energy that otherwise would help keep the fan running.
 

spec

Well-Known Member
Most Helpful Member
Currently switching the positive side but could switch.

Snubber... are you asking if I know the direction should be negative to positive?

Thanks for direction...
Below are the two switching arrangements and the related snubber circuits.

spec

2016_12_11_Iss1_ETO_FAN_SWITCH_V1.jpg
 

crutschow

Well-Known Member
Most Helpful Member
see post #5
That shows me the how, which I understood.
It doesn't explain anything about the why.
A resistor is normally used with a diode if you want to more rapidly stop the inductive current than a diode alone would.
In this case you have the opposite, you want to keep the fan current flowing when the PWM voltage goes low, so you don't want the resistor.

It similar to the inductor and free-wheeling diode in a switching regulator.
You want to keep the inductor current flowing efficiently when the transistor is off so you use a diode with as low a forward drop and resistance as practical.
 

spec

Well-Known Member
Most Helpful Member
That shows me the how, which I understood.
It doesn't explain anything about the why.
A resistor is normally used with a diode if you want to more rapidly stop the inductive current than a diode alone would.
In this case you have the opposite, you want to keep the fan current flowing when the PWM voltage goes low, so you don't want the resistor.

It similar to the inductor and free-wheeling diode in a switching regulator.
You want to keep the inductor current flowing efficiently when the transistor is off so you use a diode with as low a forward drop and resistance as practical.
Perhaps if you explained clearly and concisely what you meant it would help.

Also would you please post your circuit with recommended components for this application.

It is not exactly a big resistor and is intended to smooth the transients a touch and give the diode an easier time.

spec
 
Last edited:

spec

Well-Known Member
Most Helpful Member
Is there any concern with the diode linked being able to support the pwm frequency?
800Hz is not a particularly high switching frequency so an ordinary diode should be OK but you can use a Schottky diode if you are worried about switching speed.

The dissipation in a Schottky diode would also be less, although the diodes I have linked to have a fairly low forward voltage anyway.

spec
 
Last edited:

spec

Well-Known Member
Most Helpful Member
Thank you for your help!!!
No probs- hope it all turns out OK.

You may find that you have to put some decoupling capacitors across the supply lines near to the SS relay and/or fan.

And if you have an oscilloscope, take a look around the circuit and make sure there are no nasty high voltage spikes.

As CrutsChow points out, the circuit would be more efficient without the resistor, so you could try the circuit without the resistor and if there is little ringing and hash, leave the resistor out, but then i would be inclined to go for a 100A or over diode.

spec
 

crutschow

Well-Known Member
Most Helpful Member
Perhaps if you explained clearly and concisely what you meant it would help.

Also would you please post your circuit with recommended components for this application.

It is not exactly a big resistor and is intended to smooth the transients a touch and give the diode an easier time.
I thought I explained it reasonably clearly. Sorry if it wasn't.
My circuit would simply be the diode without the resistor.
I don't see how the resistor particularly smooths the transients or makes it easier for the diode (the diode will carry the same current with or without the resistor).
If there is any hash problem, then that would be better solved with a small RC snubber circuit.
The main thing the resistor will do when the SSR shuts off is increase the transient voltage from about 1V (a junction diode forward drop at 50A) to about 12V. That would tend to increase any radiated hash, not reduce it.
 

spec

Well-Known Member
Most Helpful Member
I thought I explained it reasonably clearly. Sorry if it wasn't.
My circuit would simply be the diode without the resistor.
I don't see how the resistor particularly smooths the transients or makes it easier for the diode (the diode will carry the same current with or without the resistor).
If there is any hash problem, then that would be better solved with a small RC snubber circuit.
The main thing the resistor will do when the SSR shuts off is increase the transient voltage from about 1V (a junction diode forward drop at 50A) to about 12V. That would tend to increase any radiated hash, not reduce it.
HI CC,

Thanks for the explanation.

Radiation can be caused by current and voltage.

I feel that the resistor will limit the current through the diode and give it some protection, di/dt for example, and will possibly reduce ringing, bearing in mind that in a practical circuit there will be stray capacitances and inductances.

Mind you, with the size of diode I found on eBay it will not be suffering too much.

Anyway, I don't think we are a mile apart in out views.

spec
 

crutschow

Well-Known Member
Most Helpful Member
I feel that the resistor will limit the current through the diode and give it some protection, di/dt for example
The maximum current and di/dt through the diode is unaffected by any resistance, being equal to the current through the inductor just before the switch is opened.
The resistance will cause that current to more rapidly decay (L/R time-constant).
High di/dt can be a concern with thyristors and MOSFETs but not for diodes.
 

spec

Well-Known Member
Most Helpful Member
The maximum current and di/dt through the diode is unaffected by any resistance, being equal to the current through the inductor just before the switch is opened.
Yes, but only with a pure inductor, and a motor is anything but a pure inductor.
The resistance will cause that current to more rapidly decay (L/R time-constant).
If you make R= zero as you suggest, surely the time constant is also zero.
High di/dt can be a concern with thyristors and MOSFETs but not for diodes.
I am afraid that is not true- all semiconductors have a maximum dI/dt and dV/dt limit, even though it is not always documented. It is not a case of just falsely triggering an SCR but damaging the semiconductor structure.

On a practical note, I have done a bit of messing about with automobile fan motors and they are truly nasty things to control.

spec
 

crutschow

Well-Known Member
Most Helpful Member
If you make R= zero as you suggest, surely the time constant is also zero.
You went the wrong way. ;)
Zero resistance makes the L/R time-constant infinity.
But, of course, it's not actually zero in practice, since the diode has resistance as well as a forward drop and any real inductor (except a superconducting one) also has some resistance associated with it.

A superconducting inductor with a shorted winding is an example of an infinite time-constant since any induced winding current will keep circulating indefinitely.
 
Status
Not open for further replies.

Latest threads

EE World Online Articles

Loading
Top