Is a New IRF540 the Same As an Old One?

In short, yes. No problem - go for it.

Thanks a lot.

I should have though a few hours ago to post the question here after a few hours looking. The daily digest reminded me.

Jim

No prob. Just make sure you order replacements with the same packaging as the old ones (almost certainly TO220, like this: . They also come in surface mount D2Pak, like this: ).

Yes, it's the top one: I had to unbolt big heatsinks before I could read the info on the unit.

Some have bevels on the front, but that makes no difference to me, if they do the same job.

Jim

The order for the gate / drain / source is also to be checked.

Please give two examples of IRF540s in TO220 packages with different pinouts.

Well, given my elementary question at the start, I took that as a nice way of saying 'Don't fit them the wrong way round'

Jim

Some IXYS MOSFETS and On Semi JFETs and Bipolar Transistors do come with pin assignment difference and are having suffix A, B, C or some other suffix or no suffix difference. However, for IRF MOSFET IRF540, I have not seen any one with pin assignment difference. These do come with different suffix for their package and also for their ON resistance difference. As some MOSFETs manufacture randomly change the pin assignment it is still worth looking at catalog.

Lots of fake MOSFETs now may also come that carry marking IRF540 but they may not be in any way worth using for replacement. Lots of such dirty parts come either from defective lots or from deliberately misleading markings by cheaters. Counterfeit parts have become a serious problem. Getting parts from reputed stockist or directly from manufacturer in sample quantity is worth. Sometime bad production lot from well known manufacturer is called back from market but not all parts come back so what happens for those do not come back?

I will say that look at catalog and also spend little time in testing few parts from the same lot to make sure you have what you want. I usually purchase MOSFETs in thousands from good source and if they are bad, then they go back to the supplier.

Well, I guess if I could do all that, guys, I wouldn't have asked the elementary question at the start.

The parts arrived, one with the N suffix. I had them fitted. I reinstalled the board. Hope turned to despair as the motor performed just as it did before the parts were swapped. I thought I might as well try the lead back on the right axis, BINGO. Eureka. Yahoo Voila. Slow return to home and pause.

Thanks you guys. It's all back together and the test cut at least seemed to work. If I wasn't so tired I would clean the machine up so it is spotless, and tell my assistant it is a new one in the morning.

Jim

Yippee, woo-hoo, hahaha .. phew.

Congratulations!

Well done mate - got there in spite of all the cr@p we were chucking at you.

Pub's still open .

Hi cnc jim;

Remember that you can not swap IRF540 with IRF9530. First one is N-channel type and second one P-channel type. Swat only with same number MOSFETs and same type packing. If these are different make then also look at ON resistance and gate voltage, gate charge for becoming ON. Normally same number means replacement types.

It is also worth to investigate, why these failed. Most likely that excessive load or over heating killed these. Also make sure heat sink is properly coupled for good heat transfer from MOSFETs.

HI Shyam. No I will only swap the same numbers, and I can get the same make, IRF, International rectifier, although they have been taken over by Vishay.

They are mounted back to back with a heatsink bolted between them.

They are 18 years old, but what I think caused the failure.. (if they are failed, I was told they can't be tested in the circuit, and to take them out and test them is more work than take them out and replace with known good ones) ..what I think caused the failure is:-

I am hoping they formed part of the homing device on a plotter. The carriage can be put manually in any position, even taken apart and put back on anywhere, and when switched on it will gently roll to one end, touch the side, and back off 5mm. There is no visible mechanical or optomechanical switch, therefore?? it must be resistance in the servo.

I got something in the way while it was plotting, which made the belt slip on that servo, and lose a few 'steps'. At the other end therefore it hit the side where no side should be, during a plot, and this i suspect has blown the components. Now it zips in a predictable pattern each time it is switched on. Each time it would crash into the side, but I swapped the feed to the other axis which just rolls rollers, which a; confirmed it was the signal not the servo and b; stopped it destroying the rest of the machine.

I really appreciate your advice and the help of everyone on the forum.

I did ask a local electronics person, but only a few words into my request he was sort of backing away with his hands up shaking his head. A couple of minutes of his time would have saved me hours of 'finding out'.

In his place, I would have had a look and made sure the person knew what to do next, but maybe that's why I'm having to spend hours repairing an old machine instead of just ordering a new one.

Jim

Pity you're so far away - I'd've happily dropped round for a look.

I know.

Thanks JohnDG

The one marked with N suffix have much lower ON resistance. As you have older versions, they may not have N suffix on numbers like IRF540N etc.

Usually MOSFET gate to channel resistance drops after they are blown by heat. This can be measured in the circuit itself.

You may have number of these in parallel and they get blown in chain one by one like attached crackers, much after one get finished.

You also have to see that heatsink is not isolated in these MOSFETs and is conne ted to Drain circuit of the MOSFET.

It is the back emf from inductive load or continuous current that kills these MOSFETs.

If current sense circuit is used then it switches to PWM control for safety. This is often design part of the control circuit. Some people avoid this and rely on failure and replacement.

I am using these MOSFETs for power switching I am seeing these with not same rating available. I have to get them tested before they are used.

None have the N suffix, and neither do the new ones, but I take your point that they may be N in nature, just not marked as such, but I have no way of finding the spec on the units I have.

In the manual:

Test transistor(s) w/ohm meter touching D (drain) & S (source) leads; should read 2.23 mega ohms, replace if reading is less.

They do seem to be in parallel, four per axis. The heatsink is linked to the motherboard but only for stability, it doesn't seem to be connected to anything else.

I guess if they are tested before fitting and give a reading as above, then they will do the job.

Thanks for the info

Jim

ON resistance is tested for gate voltage and load applied.

Normal IEF540 MOSFET may have 70-80 mill ohms ON resistance and N version 40-44 milli ohms 33A type.

As you have older version, they must be 70-80 milli ohms version. I think if you can get N version then they can conduct much high current and will not get heated easily.

N version will be available from Farnell and other sources.

If you have P and N channel working together as looks from the part number then worry more for P channel types as they have much higher ON resistance to 700-800 milli ohms.

If clamping Zener diode is dead then that will also damage MOSFETs. I use clamping Zener diode to limit the inductive induced voltage to 100V maximum permitted for these MOSFETs. You can over run to 124V for short time.