• You can now help support WorldwideDX when you shop on Amazon at no additional cost to you! Simply follow this Shop on Amazon link first and a portion of any purchase is sent to WorldwideDX to help with site costs.
  • Click here to find out how to win free radios from Retevis!

Using Onsemi FQP13N10 mosfets on RM Italy 203P?

doffo

Well-Known Member
Oct 14, 2012
312
221
53
Hello everyone,
I was able to use some FQP13N10 mosfets on my RM Italy 203P that has been broken for the last 12+ years. I wasn't sure if the Bias needed to be adjusted or should be adjusted for them, but for the time being the amp is working and doing about 150ish watts on SSB. I am only driving it with about 10 watts from the radio.
I was wondering if there are other mosfets equivilant or replacements that would work besides the FQP13N10. A set of 4 Onsemi mosfets that I bought off ebay was $11 + shipping. I have also noticed other FQP13N10L mosfets, but the L at the end I wasnt sure what that denotes.

Thanks for all the help so far. Any input is greatly appreciated.
Have a good day.
 

The "L" usually refers to a "logic" version. The gate is more sensitive, and will drive the FET into full saturation with a lower gate voltage that you might get from the output of a 3.3-Volt logic-chip digital output. These parts are designed to function as efficient on/off switches, so this simplifies driving them from digital chips powered by less than the traditional 5 Volts.

How they compare to the stock version that takes more than 3 Volts to switch fully 'on' as a RF amplifier is anyone's guess. Haven't read any first-hand reports. The IRF 520 also comes in a logic version, the IRL520.

The bias-voltage setting for the "L" part will surely be lower, but that's all I can say without trying it.

73
 
There are several older threads here on the forum. Here is one.

 
  • Like
Reactions: doffo
The "L" usually refers to a "logic" version. The gate is more sensitive, and will drive the FET into full saturation with a lower gate voltage that you might get from the output of a 3.3-Volt logic-chip digital output. These parts are designed to function as efficient on/off switches, so this simplifies driving them from digital chips powered by less than the traditional 5 Volts.

How they compare to the stock version that takes more than 3 Volts to switch fully 'on' as a RF amplifier is anyone's guess. Haven't read any first-hand reports. The IRF 520 also comes in a logic version, the IRL520.

The bias-voltage setting for the "L" part will surely be lower, but that's all I can say without trying it.

73
The data sheet on L versions also has a derated specification for linear operation. That was a bit surprising. I also took my own advice about using the Mouser website part locator to try and find an inexpensive FET that would work in these applications. What I found was, unless you drop to a dissipation at or below 43 watts, you're not going to find anything inexpensive, with a faster switching time. As power handling increases, switching times slow.
 
  • Like
Reactions: Shadetree Mechanic
Is the FQP13N10 now obsolete? I try to avoid using obsolete parts, it's hard to duplicate performance and reliability in the event the devices need replacing in the future.
 
  • Like
Reactions: doffo
Is the FQP13N10 now obsolete? I try to avoid using obsolete parts, it's hard to duplicate performance and reliability in the event the devices need replacing in the future.
Next time around I will try some IRF520 mosfets. Even if some of the power is lost, that is better than a non working amp. Thanks for letting me know the part is obsolete.
 
Also, after checking the gate bias, it was at 1.26v. I will hold off on that amp until I can pick up some trimmer resistors and set the bias correctly. Those mosfets were calling for 2v-4v.
 
As power handling increases, switching times slow.
That's the rule I see comparing the specs between types of switchmode MOSFETs.

At least one fleabay linear using these parts says the upper frequency limit is 14 MHz. Translating the switchmode part specs to predict RF performance is still a by-gum and by-gosh process for me.

73
 
That's the rule I see comparing the specs between types of switchmode MOSFETs.

At least one fleabay linear using these parts says the upper frequency limit is 14 MHz. Translating the switchmode part specs to predict RF performance is still a by-gum and by-gosh process for me.

73
I wonder what the waveform looks like at twice that frequency? The FQP13N10 was the fastest inexpensive transistor I can find on Mouser. It has a turn on time of five nanoseconds and a gate charge of 12 nC. There are many other options approaching half this speed. The interesting thing was, as soon as I switched the package style in my search from TO-220 to TO-247, the search produced no results.

That makes me think I'm doing something wrong. The gate charge was the only specification related to switching times that Mouser allowed me to search. I believe I used 12 nC and less. I've noticed that the gate charge is not always directly proportional to the switching speeds. One part had a gate charge of 10 nC however, its turn on switch time was seven nanoseconds. Perhaps I should be increasing that number slightly and see what turns up.
 
  • Like
Reactions: Shadetree Mechanic
I have been doing a little research in this area myself. I am trying to understand what parameters to "match" when putting together Mosfet pairs and quads.

Also, regarding the upper "operating" freq. of a Mosfet, I ran across a post from a guy who said he used to design Mosfets for Motorola and he said to take the reciprocal of the sum of the Rise and Fall times and convert it to MHz and that would give some indication. I don't know. I am just passing that along - take it with a grain of salt.

I put together a spreadsheet of "possible" replacements - TO-220 case style, relatively close Input & Output capacitance, Upper Operating Freq., etc. I am getting this data off of the datasheets. Most of the specs track but there are a few exceptions - FQP13N10 being one of them, which the Lincoln II+ uses.

Also you will see that for the same part #, the specs can be wildly different (IRFZ24N) between brands.

And lastly, for the elusive RM3/RM4, looking at the majority of the pics, it is my opinion that it is a Vishay or IR (Infineon) Mosfet. They are the only two brands that have an oblong "cutout" in the tab.

Please shoot holes in this, I want to learn. I can do that best by having conversations with people smarter than I am.

I am in the process of getting a test bed to try out the different Mosfets and will put the results here.

I am attaching my spreadsheet.

P.S. - Stryker put out in one of their bulletins that if you change the 2 finals in the 955HPC v2 (same as AT-5555nII) from Vishay IRF520 to IRFZ24N, you will gain about 20 watts.
 

Attachments

  • Mosfet for Radio.xlsx
    9.7 KB · Views: 45
  • 9500257920001.pdf
    435.1 KB · Views: 49
Last edited:
  • Like
Reactions: doffo
Wish I could remember who told me to use 1 over risetime+falltime. Pretty sure the gain won't be impressive at that frequency, but it's a guideline and an upper limit.

Still trying to wrap my mind around how gate charge and Miller capacitance figure into it. Less is better, but how much less?

Would be cool if somebody figured out what those two parameters are for a RF-designed MOSFET. Would give you something to compare, anyway.

73
 
I have been doing a little research in this area myself. I am trying to understand what parameters to "match" when putting together Mosfet pairs and quads.

Also, regarding the upper "operating" freq. of a Mosfet, I ran across a post from a guy who said he used to design Mosfets for Motorola and he said to take the reciprocal of the sum of the Rise and Fall times and convert it to MHz and that would give some indication. I don't know. I am just passing that along - take it with a grain of salt.

I put together a spreadsheet of "possible" replacements - TO-220 case style, relatively close Input & Output capacitance, Upper Operating Freq., etc. I am getting this data off of the datasheets. Most of the specs track but there are a few exceptions - FQP13N10 being one of them, which the Lincoln II+ uses.

Also you will see that for the same part #, the specs can be wildly different (IRFZ24N) between brands.

And lastly, for the elusive RM3/RM4, looking at the majority of the pics, it is my opinion that it is a Vishay or IR (Infineon) Mosfet. They are the only two brands that have an oblong "cutout" in the tab.

Please shoot holes in this, I want to learn. I can do that best by having conversations with people smarter than I am.

I am in the process of getting a test bed to try out the different Mosfets and will put the results here.

I am attaching my spreadsheet.

P.S. - Stryker put out in one of their bulletins that if you change the 2 finals in the 955HPC v2 (same as AT-5555nII) from Vishay IRF520 to IRFZ24N, you will gain about 20 watts.
I have done this mod and it works. I am having good results with IRFZ24's replacing other FET's.
 
  • Like
Reactions: Enterprise312ok
I have done this mod and it works. I am having good results with IRFZ24's replacing other FET's.
Do you have a brand preference for the IRFZ24 or at least a reliable source to buy them them? Tempted to use them instead of the IRF520.
 

dxChat
Help Users
  • No one is chatting at the moment.