Continue to Site

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.

  • 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.

What is the best heat transfer method

Status
Not open for further replies.
Hi,

I am not sure what you are asking then. Current limiting with a resistor only works in certain situations, and if you have the current already being measured then there is a way to use that measurement to feed back to a circuit (perhaps another circuit altogether) to either shut down or just lower the current.

A typical method is to measure the current through the transistor itself and shut down if the current goes too high.


Me Neither, got any good suggestions about about how to handle the current at the end of travel when the current sky rockets due to the stall. I did use the LEM output once to trip a relay that just dropped all power to the motor until the input was zeroed. The current/present method seems to work by increasing the negative feed back as current increases, but it's so instantaneous, it seems to like it's messing with the wave form, but it is working, if I can keep the FETs alive, which I think I can now because of the advice I received in this thread.
Thanks all!!!
Jeff
 
Relays are quite slow, I would rather use a mosfet to disconnect the supply voltage, and when the overcurrent trips disconnect it for a few seconds. Even if the motor remains stalled turning the current on for somthing like 0.5ms with a period of a few seconds will keep the bridge mosfets nice and cool.
 
Me Neither, got any good suggestions about about how to handle the current at the end of travel when the current sky rockets due to the stall. I did use the LEM output once to trip a relay that just dropped all power to the motor until the input was zeroed. The current/present method seems to work by increasing the negative feed back as current increases, but it's so instantaneous, it seems to like it's messing with the wave form, but it is working, if I can keep the FETs alive, which I think I can now because of the advice I received in this thread.
Thanks all!!!
Jeff


Hi again,

Well, the typical way to do that is to add control to the device that is doing the switching in the first place. If something already turns the motor on (like a MOSFET) then that is the thing to target. Adding feedback to the control of that device can turn it off.

If that is not possible then you could use another MOSFET. The control for that could be a comparator and gate driver. If the current goes too high, turn off the MOSFET. Either keep it off until manually reset or wait a number of seconds to 'try' again. A relay may be too slow but it's hard to say depending on the dynamics of the entire circuit. Without seeing that i would say a MOSFET would be better because that would be faster than even the comparator.
 
Kinarfi,
I can give you a couple FAN7842 parts to play with if you'd like. They're in SO-8 SMT packages though.
Question for you, I been looking at the data sheet this and I'm thinking I can tie the hi and lo inputs together an then take HO to the top left of the bridge and LO to the bottom right of the bridge with one of the units and do the other side with another unit. This would eliminate my need for the lower driver setup I have and the LM555 oscillator and voltage booster. Is this correct?
Thanks,
Jeff
 
i havnt used that exact part but I have used an FAN73832 which probably has the same restrictions in that it doesn't have a charge pump for the high side driver it relys on the bootstrap diode. this means you cant just turn on the high fet and hope that it stays on.


as I pointed out before the hip4081 is ideal for what you want.

Your problem seems to be gate drive speed or level for the low side devices otherwise both top and bottom fets would die if was purely an I2R heating issue.


if you do decide to use a fast gate driver you may need to watch out for turn off spikes as your supply decoupling doesn't look very good
 
I usually just drive my Mosfets wih a 555 and Never have any Problems.
Even at Quite High Frequencies
.

How do you set up your 555 so it works as a FET driver. On Spice and on circuit board, a 10K from pin 8 to pin 6 gave a nice square wave out, until I hooked it up to a FET, then it looked like trash.

I have rebuilt my heat sink, it will be remote from the control box, has 2 separate halves and has a fan on it and the FETs are held against the sink with a spring clip and I used heat sink grease.
 
Question for you, I been looking at the data sheet this and I'm thinking I can tie the hi and lo inputs together an then take HO to the top left of the bridge and LO to the bottom right of the bridge with one of the units and do the other side with another unit. This would eliminate my need for the lower driver setup I have and the LM555 oscillator and voltage booster. Is this correct?
Thanks,
Jeff

Yes, you can tie the input to the drivers for opposite corners together, as they switch at the same time. One thing to be careful of is to make sure that one pair of mosfets is fully off before the next pair turns on. this is known as dead time. If one set is still even slightly on when the next starts to turn on will cause a spike of current straight straight through the left and/or right side pair. This is known as shoot through and is another cause of mosfet failures in bridge circuits.
 
I have been looking at IR2110, IR 2108, IR2104, IR2181, FAN7842 and other Half Bridge Gate Driver and playing with the IR models on spice and there doesn't seem to be a way to make 2 half bridge gate drivers into a full bridge gate driver because the hi and lo side outputs are 180 degrees out of phase, so I have come to the conclusion that richardb is correct and the HIP4080 type bridge gate driver is the way to go, now I need to further investigate the current limiting ability to see how I can use that feature.
 
Last edited:
Can you use two heat sinks? no my problem is not heat sink, it's heat transfer to heat sink, I'm using IRF3205 55V 110A 8.0mΩ FETS and running them at about 30 and maybe up to 60 at motor stall.

What he means is if you are using an H bridge configuration, two of the FETs are tied to the positive rail, and the other two are tied to the negative rail. If you pair them like that, then you don't need insulators, but your heatsinks will be at the rails.

Have you done the thermal calculations? Or are you just thinking you have a big enough heat sink?

The kits that Pommie suggested use shoulder washers to keep the metal screws centered in the hole and away from contacting the metal, thus insulating the heatsink from the device.
 
I have been looking at IR2110, IR 2108, IR2104, IR2181, FAN7842 and other Half Bridge Gate Driver and playing with the IR models on spice and there doesn't seem to be a way to make 2 half bridge gate drivers into a full bridge gate driver because the hi and lo side outputs are 180 degrees out of phase, so I have come to the conclusion that richardb is correct and the HIP4080 type bridge gate driver is the way to go, now I need to further investigate the current limiting ability to see how I can use that feature.

you invert the input signal to the second half bridge driver so that when A drives hi, B drives low, and vice versa.
 
Thanks, what I have done is order a HIP4081A to drive my FETs, redesigned my circuit using a NE555 for my oscillator and I have separated the FETs on to a two piece heat sink allowing the tabs to make direct contact with the heat sink. IMHO, my main problem was driving the FETs to slowly, but excessive heat sinking can't be a problem and if its just adequate, that's ok too. On a normal ride, the steering doesn't have to work very hard, but on occasion, I need all the power I can get up to the limits of my supply system with has a 20 amp circuit breaker in it, and if I find it to be the weakest point, I may up it to a 30 amp.
I am quite familiar with shoulder washers and found them to be a weak link also, even though they are glass filled nylon, they still melt and then the pressure that holds the tab of the FET to heat sink goes away and the heat transfer conduction is lost and the FET fries. Heat shrink tubing can handle the heat and circuit board pieces with holes drilled in them work well for isolating the screw from the tab while maintaining good pressure and heat transfer. But even better is the spring clip and the direct contact, photo later.
If any one is familiar with the HIP4081A and knows how the use the current limiting ability, please advise, I'm still studying it and it looks like a feed back to pins 8 &/or 9, but what kind and how?
https://www.intersil.com/content/dam/Intersil/documents/hip4/hip4081a.pdf
Any and all help is appreciated
Thanks
Jeff
 
I don't see that this part will current limit. Pins 8 & 9 are delay controls for the crossover switching. Look on page 13 of the data sheet, they use pots to adjust time delay. Also on that schematic are R30 and R31, which looks like they would do current limiting, but on the internal schematic they look like the low rail of the transistor drivers. The description doesn't say the part will do current limiting.

Edit: page 3 shows an external current control circuit.
 
OK, my mistake, I guess that it's sending a feed back to somewhere else that would control the input to the HIP4081A, so the HIP4081A is strictly driving the FETs according to the inputs on pins 5 & 6.
Thanks
 
if you are mounting the fets on the metal directly onlt the 2 top fets can be on the same heatsink. both bottom fets must be on there own sink.

if you just want to protect the fets there is a disable input on a single pin

do you have any experience with microcontrollers?

ps the highside input can be left high and just pulse the lowside input this may help your logic
 
I have been looking at IR2110, IR 2108, IR2104, IR2181, FAN7842 and other Half Bridge Gate Driver and playing with the IR models on spice and there doesn't seem to be a way to make 2 half bridge gate drivers into a full bridge gate driver because the hi and lo side outputs are 180 degrees out of phase, so I have come to the conclusion that richardb is correct and the HIP4080 type bridge gate driver is the way to go, now I need to further investigate the current limiting ability to see how I can use that feature.

On the FAN7842 at least, the high and low side drivers are are independent, so you can drive them with whatever timing you want to.
 
if you are mounting the fets on the metal directly onlt the 2 top fets can be on the same heatsink. both bottom fets must be on there own sink.
Had to think on this for a minute, then I realized that I was making a huge mistake by tying the hi side's source to positive, guess I have some rewiring to do, think I'll use up some of my mica insulators too, sure glad you said that!!!

if you just want to protect the fets there is a disable input on a single pin
It's my intentions to have a on-delay signal go to the disable pin for when I turn the key on.

do you have any experience with microcontrollers?
Nope, I used to work some with PLCs, but have stayed away from micros.

ps the highside input can be left high and just pulse the lowside input this may help your logic
Yes, that's the way I'm doing it.
Got any other suggestions, I'm expecting my HIP to show up tomorrow.
 
I just wondered about micros as they are ideal for this kind of thing

one suggestion is start with high gate resistors first, as when you turn off the fets with good drive you are likely to get big turnoff spikes if your supply isn't closely coupled to the rail capacitors
 
ps the highside input can be left high and just pulse the lowside input this may help your logic

If the high side fet is on when you pulse the low side fet, you will have a dead short through the two fets across your power supply rails.

Guaranteed big smoke!!!
 
I used the HIP4081 to drive some big old fets for some motors in a little car. As I recall I had 2 1/2 problems. :)

The first was I was testing it on a power supply rather than the batteries. When I reversed direction (without stopping) the generator action of the motors pumped up my power supply from 12 to about 25 volts and blew the HIPs.
Second. The maximum delay was not enough to keep from having some shoot thru. I didn't blow anything but the fets got hot. I had to play around with gate resistors bypassed in the off direction with a diode to fix that.
The 3rd I'm not so sure about, but I built it on a breadboard and even though I tried to keep everything short the source voltage went about 9 volts below ground, so I followed the directions in the app note. I'm not sure if this was a real problem because the things are so fast there is a lot of noise.
Here's the app note.
https://www.intersil.com/content/dam/Intersil/documents/an94/an9405.pdf
 
If the high side fet is on when you pulse the low side fet, you will have a dead short through the two fets across your power supply rails.

Guaranteed big smoke!!!

The driver prevents the 2 from being on at the same time, so it works ok.
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top