H-bridge vs relays?

HR19 · Feb 9, 2021

So, what would be the advantages of using an H-bridge vs using just a simple forward driver and then relays for reversing a motor? An H-bridge seems somewhat complex, but using a couple of NO/NC relays seems simple (I'm sure the relays would have slightly longer response times, but generally probably not enough to matter).

Nigel Goodwin · Feb 9, 2021

With an H-bridge you may be able to apply braking to the motor, if that's something you need.

Personally, for something like a Robot Wars machine, I'd probably use a single-ended driver and a DPDT relay (no need to use two relays).

schmitt trigger · Feb 9, 2021

If you employ a pluggable relay, they are very easy to replace in case of an overload damages it.

And they will resist more abuse in the first place.

HR19 · Feb 9, 2021

Nigel Goodwin said:
With an H-bridge you may be able to apply braking to the motor, if that's something you need.

Ah, OK, that makes sense.

schmitt trigger said:
If you employ a pluggable relay, they are very easy to replace in case of an overload damages it.

I've had a hard time finding pluggable relays for the right voltage at a good amperage rating.

KeepItSimpleStupid · Feb 9, 2021

HR19 said:
I've had a hard time finding pluggable relays for the right voltage at a good amperage rating.

Automotie relays (40A contacts, 12VDC). You can dual sockets basically pre-wired.
Most are missing the coil diode.

Two SPDT relays allow:
1. dynamic braking (motor windings shorted)
2. limit switches at the relay coil contact rating
3. No power draw when at the limit.
4. Positive or negative logic
5. truth table
relay A on, relay B on - brake
relay A on, relay B off - brake
Relay A on, relay B off - direction A
relay B on. relay A off - Direction B
6. Make the relays DPDT and you can get status: at limit A or at limit B
7. Add a bridge rectifier, a small capacitor, and some optoisolators, you can get a "Motor running" signal.

it's the power window circuit,

HR19 · Feb 9, 2021

KeepItSimpleStupid said:
Automotie relays (40A contacts, 12VDC). You can dual sockets basically pre-wired.
Most are missing the coil diode.

Two SPDT relays allow:
1. dynamic braking (motor windings shorted)
2. limit switches at the relay coil contact rating
3. No power draw when at the limit.
4. Positive or negative logic
5. truth table
relay A on, relay B on - brake
relay A on, relay B off - brake
Relay A on, relay B off - direction A
relay B on. relay A off - Direction B
6. Make the relays DPDT and you can get status: at limit A or at limit B
7. Add a bridge rectifier, a small capacitor, and some optoisolators, you can get a "Motor running" signal.

it's the power window circuit,

Ooh, I like that plan!

KeepItSimpleStupid · Feb 9, 2021

Oops: https://www.sonicelectronix.com/item-8400-Bosch-RS2-Dual-Relay-Socket.html

here's one: https://vetco.net/products/dual-40a-automotive-relay-socket

anyway check out www.parts-express.com use bosch relay or automotive relay.

The dual socket had everything nearly wired.

The relay terminals are supposed to face downward.

AnalogKid · Feb 10, 2021

HR19 said:
I've had a hard time finding pluggable relays for the right voltage at a good amperage rating.

And those values are ... ???

Diver300 · Feb 10, 2021

I suggest that you put resistors in series with the freewheel diodes if you are using relays anywhere near their rating, especially with motors as the loads. Aim for around 10 times the resistance of the coil.

That will limit the spike voltage to around 10 times the supply voltage. Obviously the transistors that drive the coils need to be rated for that, but the peak value is quite well known. Without the resistors it could be much higher and would depend on all sorts of factors.

The reason is to make sure that the relay turns off quickly. A freewheel diode without a resistor can mean that the contacts open slowly, which can lead to excessive arcing of the contacts.

You don't actually need the freewheel diode if there's a resistor in parallel with the coil. The advantage of no diode is that it's one less thing to worry about, and it doesn't matter which way round the resistor / relay is connected. The disadvantage of no diode is 10% extra power used when on.

HR19 · Feb 10, 2021

AnalogKid said:
And those values are ... ???

48V,40A

HR19 · Feb 10, 2021

Diver300 said:
I suggest that you put resistors in series with the freewheel diodes if you are using relays anywhere near their rating, especially with motors as the loads. Aim for around 10 times the resistance of the coil.

That will limit the spike voltage to around 10 times the supply voltage. Obviously the transistors that drive the coils need to be rated for that, but the peak value is quite well known. Without the resistors it could be much higher and would depend on all sorts of factors.

The reason is to make sure that the relay turns off quickly. A freewheel diode without a resistor can mean that the contacts open slowly, which can lead to excessive arcing of the contacts.

You don't actually need the freewheel diode if there's a resistor in parallel with the coil. The advantage of no diode is that it's one less thing to worry about, and it doesn't matter which way round the resistor / relay is connected. The disadvantage of no diode is 10% extra power used when on.

Freewheel diode? Is that the reverse-biased diode I put in parallel with the motor to help it slow down and not bust the electronics?

AnalogKid · Feb 10, 2021

HR19 said:
48V,40A

Way to bury the lead ...

I think 100 V, 80 A power MOSFETs will cost much less than contactors rated for 80 A ***DC***. Plus, the drive power is trivial compared to the coil current. Plus the long-term reliability will be much better.

Digi-Key power MOSFET, 80 A min., 100 V min. start ar $1.21 each in ones. At 1.75 each, power dissipation at 40 A is less than 0.5 W per transistor, and only two are on at a time. Basically, the cost is so much less than a DPDT 60-80 A contactor that you can overdesign the hell out of it and not come close to cost parity.

Note: At these power levels, the NC contacts of a SPDT or DPDT relay often have a lower current rating then the NO contacts, because of the contact pressure provided by the return spring pushing the armature is lower than the pressure caused by the coil pulling the armature.

ak

Diver300 · Feb 10, 2021

HR19 said:
Freewheel diode? Is that the reverse-biased diode I put in parallel with the motor to help it slow down and not bust the electronics?

I was meaning the reverse biased diodes in parallel with the relay coil.

It's a good idea to have some suppressor device in parallel with the relay coil, so that when the transistor that feeds the relay turns off, the magnetic energy in the relay coil has somewhere to go. Without anything, there can be a large voltage spike which can damage the transistor.

A diode is very effective at reducing the voltage, which will only be about 1 V, the forward voltage of the diode, when the transistor turns off. However, with only 1 V across the relay coil, the current in the relay coil will only reduce slowly. If the current is only reducing slowly, the relay may only open slowly, which can cause more arcing of the contacts.

HR19 · Feb 10, 2021

AnalogKid said:
Way to bury the lead ...

I think 100 V, 80 A power MOSFETs will cost much less than contactors rated for 80 A ***DC***. Plus, the drive power is trivial compared to the coil current. Plus the long-term reliability will be much better.

Digi-Key power MOSFET, 80 A min., 100 V min. start ar $1.21 each in ones. At 1.75 each, power dissipation at 40 A is less than 0.5 W per transistor, and only two are on at a time. Basically, the cost is so much less than a DPDT 60-80 A contactor that you can overdesign the hell out of it and not come close to cost parity.

Note: At these power levels, the NC contacts of a SPDT or DPDT relay often have a lower current rating then the NO contacts, because of the contact pressure provided by the return spring pushing the armature is lower than the pressure caused by the coil pulling the armature.

ak

Maybe the answer is just buying a motor controller from Amazon or something and then figuring out how to wire it up to a micro controller.

Diver300 said:
I was meaning the reverse biased diodes in parallel with the relay coil.

It's a good idea to have some suppressor device in parallel with the relay coil, so that when the transistor that feeds the relay turns off, the magnetic energy in the relay coil has somewhere to go. Without anything, there can be a large voltage spike which can damage the transistor.

A diode is very effective at reducing the voltage, which will only be about 1 V, the forward voltage of the diode, when the transistor turns off. However, with only 1 V across the relay coil, the current in the relay coil will only reduce slowly. If the current is only reducing slowly, the relay may only open slowly, which can cause more arcing of the contacts.

Ah, did not realize that's a thing, but it makes sense.

HR19 · Feb 11, 2021

You know what, what I said earlier about 48V and 40A is actually the wrong project, I confused myself. 18V (so probably 24V rated relay), and probably not more than 20A, maybe peaks of 30 but probably just 20A continuous would be enough.

H-bridge vs relays?

HR19

Member

Nigel Goodwin

Super Moderator

schmitt trigger

Well-Known Member

HR19

Member

KeepItSimpleStupid

Well-Known Member

HR19

Member

KeepItSimpleStupid

Well-Known Member

AnalogKid

Well-Known Member

Diver300

Well-Known Member

HR19

Member

HR19

Member

AnalogKid

Well-Known Member

Diver300

Well-Known Member

HR19

Member

HR19

Member

Similar threads

H-bridge vs relays?

Member

Super Moderator

Well-Known Member

Member

Well-Known Member

Member

Well-Known Member

Well-Known Member

Well-Known Member

Member

Member

Well-Known Member

Well-Known Member

Member

Member

Similar threads

Privacy & Transparency

Privacy & Transparency