Hello all AAC members, I'm new to the forum and would like to ask what some might see as a dumb question but here goes, my concerns are if I can replace a FQPF5n80 mosfets with a FQPF6n80 mosfets? Viewing the datasheets on both I found slight differences for example the 5n80 features 2.8a,800v,Rds(on)=2.6 ohms @Vgs=10v. The 6n80 featured 3.3a,800v, Rds.(on)=1.95 ohms @Vgs=10v. My question is will these small differences matter the 5n80 was a part originally on a SMPS and I want to replace it with the 6n80 I'm thinking the 5 tenths of an amp is not much or the 11 one hundredth's of an ohm am I right or wrong need help before putting part in and destroying another component.

Thanks for all of you folks help in advance

 

The 6N80 is a slightly better part

 

Go for it.

 

All things being equal, it's a good choice, but all things aren't equal.

The 5N needs 25 nanocoulonbs to charge the gate and the 6N needs 31. That's a difference of 1.24 to 1.
IF the driver can supply that current, that fast, you have no problems.
If not, that FET is going to get extra warm...maybe too extra warm.
Proceed, but watch for overheating.

 

All things being equal, it's a good choice, but all things aren't equal.

The 5N needs 25 nanocoulonbs to charge the gate and the 6N needs 31. That's a difference of 1.24 to 1.
IF the driver can supply that current, that fast, you have no problems.
If not, that FET is going to get extra warm...maybe too extra warm.
Proceed, but watch for overheating.

Based on this info and thanks for it maybe I should despite the Moderator MrChips and Ronv being in favor of it being a" a slightly better part" part maybe for this application I should stick with what the board was made with. Thanks guys for the quick response

 

I did not suggest that one part was better than the other.
I simply assumed that one part was available and not the other and that was your reason for making the switch.

 

We can't see the schematic from here, and most FET drivers are far more capable than necessary, but you never know in these days of computer aided design where obsessive penny pinchers can accurately remove every bit of excess quality from a design. Try it, but don't assume.

 

Hello all AAC members, I'm new to the forum and would like to ask what some might see as a dumb question but here goes, my concerns are if I can replace a FQPF5n80 mosfets with a FQPF6n80 mosfets? Viewing the datasheets on both I found slight differences for example the 5n80 features 2.8a,800v,Rds(on)=2.6 ohms @Vgs=10v. The 6n80 featured 3.3a,800v, Rds.(on)=1.95 ohms @Vgs=10v. My question is will these small differences matter the 5n80 was a part originally on a SMPS and I want to replace it with the 6n80 I'm thinking the 5 tenths of an amp is not much or the 11 one hundredth's of an ohm am I right or wrong need help before putting part in and destroying another component.

Thanks for all of you folks help in advance

It depends on what parameters are important for a particular application. Many applications are not too picky about small changes in parameters -- that's one of the hallmarks of a good design -- so if you replace a component with something that is close (preferably "better" in the specs that matter the most), then it will probably work.

But "better" has to be viewed in the context of what is important for that application. I killed an IC design once because we moved to a smaller process and, due to schedule and budget constraints, I didn't do detailed FET-level simulations of the logic since we were keeping the device dimensions exactly the same (and the costumer was made aware of that and the risk involved, but they needed prototype chips in hand by a certain date to meet their contract obligations). But the process was enough faster that the logic failed to meet set-up and hold time requirements and so didn't function at all. Strangely, that wasn't a huge problem because the prototypes were delivered on time and that was what mattered contractually and the contract (their contract with their customer) allowed for a possible respin of the design, which we did (for free since our normal guarantee is that the design will be functional or we will do the redesign (but not the refab) for free although performance against spec is not guaranteed) and the second run worked beautifully. But it was still a bitter pill and a valuable lesson for me.

If you CAN use the original part, then you generally should. If you can't, for whatever reason, then you try to find one that is as close as possible, particularly if you don't know which parameters the design is most sensitive to, and then you test the result as well as you can.

 

I don't think there is a case where the old one will perform better than the new one.
The new one has 25% lower on resistance. So assuming it spends more time on than switching it will be better than the slower one.

 

I don't think there is a case where the old one will perform better than the new one.
The new one has 25% lower on resistance. So assuming it spends more time on than switching it will be better than the slower one.

I suspect that, for this application, that is probably true. But in general it isn't guaranteed to be that way. For instance, a design may not work if the on resistance is too low or if the transistor switches too quickly. Then there are plenty of other parameters that might cause a particular design to fail with the new part.

 

apologies for missing the point no offense, took it upon myself from what you said that it was a suitable substitute and in brief I stated in the response to #12 that based on that, and summarily based on what his input was that I would likely be going for the actual part. No slander intended of any kind. I need help from this forum Apologies to you and Ronv!!!

 

It depends on what parameters are important for a particular application. Many applications are not too picky about small changes in parameters -- that's one of the hallmarks of a good design -- so if you replace a component with something that is close (preferably "better" in the specs that matter the most), then it will probably work.

But "better" has to be viewed in the context of what is important for that application. I killed an IC design once because we moved to a smaller process and, due to schedule and budget constraints, I didn't do detailed FET-level simulations of the logic since we were keeping the device dimensions exactly the same (and the costumer was made aware of that and the risk involved, but they needed prototype chips in hand by a certain date to meet their contract obligations). But the process was enough faster that the logic failed to meet set-up and hold time requirements and so didn't function at all. Strangely, that wasn't a huge problem because the prototypes were delivered on time and that was what mattered contractually and the contract (their contract with their customer) allowed for a possible respin of the design, which we did (for free since our normal guarantee is that the design will be functional or we will do the redesign (but not the refab) for free although performance against spec is not guaranteed) and the second run worked beautifully. But it was still a bitter pill and a valuable lesson for me.

If you CAN use the original part, then you generally should. If you can't, for whatever reason, then you try to find one that is as close as possible, particularly if you don't know which parameters the design is most sensitive to, and then you test the result as well as you can.

I am not knowledgeable when it boils right down to it I am just tinkering, but I'm thinking, before coming here that because this FET is on the power supply that all things in that section have to be exact or there will be problems. That is the jist of your response yes?

 

Apologies? I did not detect anybody feeling attacked or offended.

The gist of our conversation is that without the actual schematic, we're all making educated guesses. You can probably get away with using the 6N part but nobody is making guarantees here.

 

I am not knowledgeable when it boils right down to it I am just tinkering, but I'm thinking, before coming here that because this FET is on the power supply that all things in that section have to be exact or there will be problems. That is the jist of your response yes?

The answer, as is very often the case in engineering, is that "it depends". If the power supply (or any circuit) was designed well then there should be very few components whose exact parameters are critical and you should be able to replace them with components that are close (either a bit "better" or a bit "worse") and have the circuit still function (though possibly not quite as well). How far you can push that depends on the circuit and on the component in question. In some cases you can replace a small signal transistor with a power transistor and the thing will work fine while in other cases if you replace a transistor with its usual recommended replacement the thing won't work (often indicating either a poor design or a design that is particularly high performance).

 

I don't think there is a case where the old one will perform better than the new one.
The new one has 25% lower on resistance. So assuming it spends more time on than switching it will be better than the slower one.

Really and this would be plainly, a good thing in a switch mode power supply of any type calling for the 5n80 FET I'm confused hope you can tell that from my responses. Someone says yes, somebody else says no, has anyone made this type of swap out and faired? I just don't want to screw anything up And apologies Ronv actual query I made was basically if it would be, lack of better words, good enough I got that answer from you and MrChips #12 as well as WBahn have just pointed out precautions Thanks all I understand that this is a judgment call

 

So the question is why do you want to make the switch in the first place?

If the original design calls for FQPF5n80 and it worked well and it is readily available and it is reasonably priced then why switch?

If there is some reason to believe that FQPF6n80 will give better performance without adverse consequences then why not give it a try?

 

Not a problem.

When you said switch mode power supply and I saw an 800 volt FET I assumed it was used as a switch in the primary side. That could have been a bad assumption on my part. If it is used in some very high frequency application - maybe not. I just couldn't envision that application.

 

So the question is why do you want to make the switch in the first place?

If the original design calls for FQPF5n80 and it worked well and it is readily available and it is reasonably priced then why switch?

If there is some reason to believe that FQPF6n80 will give better performance without adverse consequences then why not give it a try?

Sorry for just now responding but it really was a question placed to see if it could work I had it laying around but have recently ordered the part something dumb I should know better sometimes the old man has a brain #art guest I should do as others have said and "KISS" thanks for the input though for real I will be calling on you again I'm sure!

 

Not a problem.

When you said switch mode power supply and I saw an 800 volt FET I assumed it was used as a switch in the primary side. That could have been a bad assumption on my part. If it is used in some very high frequency application - maybe not. I just couldn't envision that application.

No sir you were right in that assumption it is a switch I think... on the primary side here is a pic of the board something I should have done in the beginning component rests at Q10 and I imagine its responsible for switching output to +12 and +24 not really sure but learning Oh this image was taken from an online source not my own but useful the arrow points to it

 

Yeh, looks like.
Be aware the original problem is probably not the FET. It may be just the weak link. Usually the problem is a bad capacitor(s). Check them for any leakage or puffiness on the tops before you just change the FET.