another mosfet fail...

[sonoma]

have you tried hooking it up to an external power supply to see if any thing changes?
does the radio still have the original electrolytic caps in it. as old as that radio is you could have some leaking off which will also cause a problem.

 

[Robb]

OK . . . so this is a base station radio w/built-in power supply . . .

First thing I would do is check the power supply voltage output. Chances are, it has been turned up too high and then the output of the supply collapses under load ('keying'). Seen this a couple of times before. Make sure that the supply isn't turned up beyond 12.5v.

Had worked on a local operators radio similar to this one. Seems that it just didn't work when keyed. Guy had turned up the power supply voltage; so I just dialed it back and it worked fine after. They come with a cheap switching-type supply.

Not saying that is the problem; but there is a good chance that the power supply just won't be able to put up with the extra load of the MOSFETS, since they will draw far more power than the bipolar outputs.

May well have to ditch the stock supply and put in a stronger one . . .

 

[dozerman]

have you tried hooking it up to an external power supply to see if any thing changes?
does the radio still have the original electrolytic caps in it. as old as that radio is you could have some leaking off which will also cause a problem.[/Q

OK . . . so this is a base station radio w/built-in power supply . . .

First thing I would do is check the power supply voltage output. Chances are, it has been turned up too high and then the output of the supply collapses under load ('keying'). Seen this a couple of times before. Make sure that the supply isn't turned up beyond 12.5v.

Had worked on a local operators radio similar to this one. Seems that it just didn't work when keyed. Guy had turned up the power supply voltage'; so I just dialed it back and it worked fine after. The come with a cheap switching-type supply.

Not saying that is the problem; but there is a good chance that the power supply just won't be able to put up with the extra load of the MOSFETS, since they will draw far more power than the bipolar outputs.

May well have to ditch the stock supply and put in a stronger one . . .

yeah it had a built in supply.

 

[dozerman]

Robb, the supply is set at 13.8 volts. when i got it i turned it down from where it was. its always run well at that setting. i will turn it back out of curiosity. but since putting the 13n10's back in, ive had no problems from the power supply, dimming or oscillation. sonoma, ive replaced a bunch of the electrolytics since ive had it. replaced a few more tonight. its inline right now, talking on it to see if any problem presents itself. if nothing else i will just undo the mod and put it back stock

 

[BBB]

Cheap $4 MOSFETS just plain SUCK. Too many failures over time in the base station. I'm gonna fix the blown Chinese MOSFET radios I have and sell 'em all

Would love to have a 2sc1969 conversion kit for the Stryker 955 and I would reconsider

 

[dozerman]

Cheap $4 MOSFETS just plain SUCK. Too many failures over time in the base station. I'm gonna fix the blown Chinese MOSFET radios I have and sell 'em all

Would love to have a 2sc1969 conversion kit for the Stryker 955 and I would reconsider

Yeah, for the most part I've not had luck with mosfets. The 13n10 is the exception. They work out pretty well.

 

[Robb]

Cheap $4 MOSFETS just plain SUCK. Too many failures over time in the base station. I'm gonna fix the blown Chinese MOSFET radios I have and sell 'em all

Would love to have a 2sc1969 conversion kit for the Stryker 955 and I would reconsider

Wow; I don't think I've ever paid more than $1/ea for the IRF-520 MOSFETs that I've bought and used. I also bought them from Mouser and they were the Vishay brand. Never, ever had one single failure yet. Had also bought some 13n10s as well, and they have also worked out w/o failure. Also used the ERF-2030 MOSFET, and they are all still working too.

Make my own biasing devices as well to compliment the installs, since those ERF biasing devices are getting scarce.

They are all sensitive to static; so you must take those precautions. All of the devices are stored in anti-static bags; a must. I also use a 16" x 24" anti-static mat for installing IC's and MOSFETs. It is placed front/center on my electronic repair bench. They are fairly inexpensive and available on ebay . . .

 

[BBB]

Wow; I don't think I've ever paid more than $1/ea for the IRF-520 MOSFETs that I've bought and used.

Yeah, the 955's have (4) MOSFETs between both the driver & final sections

FYI, If I were to decide to keep running these cheap devices, I would not run a power supply over 12.5 VDC Max. I'm starting to believe they just can't handle anything over that for any length of time and start to breakdown internally.

 

[Robb]

Yeah, the 955's have (4) MOSFETs between both the driver & final sections

FYI, If I were to decide to keep running these cheap devices, I would not run a power supply over 12.5 VDC Max. I'm starting to believe they just can't handle anything over that for any length of time and start to breakdown internally.

Short of redesigning that radio's PA stage, you can also make sure that the radio hasn't been so peaked as to get every single watt out of it. That alone will push the MOSFET into failure, as they will give every last watt before they explode. I've experimented and seen them do it myself.

A bipolar transistor won't do that. But then again, they also don't have anywhere near the same power output either. And if this MOSFET radio is being used to drive an amp to saturation to attain full output; then you can forget about using it if it is pushed up so hard to meet that requirement constantly. In particular, if a long key down is being done.

A MOSFET will run long at the same output of a bipolar trans w/o incident, or will have more output for a shorter period of time if turned up. But I wouldn't expect them to get BOTH out of them. No heat sink is going make that much of a difference either.

In fact, if four MOSFETS are running at a cool 85w/peak; then they should be doing well with a moderate key down time. Running them @ approximately 21w/ea should be fine with a sufficient heat sink and less than a minute key down time. Sure, you can get 30w/ea out of a MOSFET; but why should that be done when most high output amps really require no more than 100w total input power? 85w would be the best way - IMO. Best to leave some headroom.

The voltage supply isn't really the issue, as the radio is spec'd to be run @ 13.8v.

 

[dozerman]

Short of redesigning that radio's PA stage, you can also make sure that the radio hasn't been so peaked as to get every single watt out of it. That alone will push the MOSFET into failure, as they will give every last watt before they explode. I've seen them do it myself.

A bipolar transistor won't do that. But then again, they also don't have anywhere near the same power output either. And if this MOSFET radio is being used to drive an amp to saturation to attain full output; then you can forget about using it if it is pushed up so hard to meet that requirement constantly. Especially if a long key down is being done.

A MOSFET will run long at the same output of a bipolar trans w/o incident, or will have more output for a shorter period of time. But I wouldn't expect them to get BOTH out of them. No heat sink is going make that much of a difference either.

All of the Galaxy cb's I've made dual final, I only had a few to blow up and that was when I was just starting doing it. But they were mosfet to begin with and I swapped the 520's for 13n10's. after this 11b fiasco, I'm done with "upgrading" older boards to be mosfet. Yesterday I replaced to 10000 uf cap in the power supply and had good power at idle, I tried to key and now everything is dead. The transformer is probably shot. Or the transistors that feed the supply. Lol got a sweet 11b for parts if anyone needs one lol

 

[Robb]

All of the Galaxy cb's I've made dual final, I only had a few to blow up and that was when I was just starting doing it. But they were mosfet to begin with and I swapped the 520's for 13n10's. after this 11b fiasco, I'm done with "upgrading" older boards to be mosfet. Yesterday I replaced to 10000 uf cap in the power supply and had good power at idle, I tried to key and now everything is dead. The transformer is probably shot. Or the transistors that feed the supply. Lol got a sweet 11b for parts if anyone needs one lol

Running the voltage up to the limit of a MOSFET limit is yet another sure way to get max power out as well as push the envelope of what a MOSFET can deal with. A 13n10 requires LESS voltage to the gate than a IRF-520 will require by one whole volt.

Typical IRF-520 gate voltage to keep it running cool would be 3.5v, where a 13n10 would need no more than 2.5v to achieve the same output. Leaving the 13n10 at the same setting of 3.5v as the IRF-520 - would be one cause for its failure.

Running a 13n10 @ 3.5v is really pushing it for that device - IIRC. Look at the spec sheets for each device and you can see it for yourself . . .

 

[dozerman]

Running the voltage up to the limit of a MOSFET limit is yet another sure way to get max power out as well as push the envelope of what a MOSFET can deal with. A 13n10 requires LESS voltage to the gate than a IRF-520 will require by one whole volt.

Typical IRF-520 gate voltage to keep it running cool would be 3.5v, where a 13n10 would need no more than 2.5v to achieve the same output. Leaving the 13n10 at the same setting of 3.5v as the IRF-520 - would be one cause for its failure.

Running a 13n10 @ 3.5v is really pushing it for that device - IIRC. Look at the spec sheets for each device and you can see it for yourself . . .

lol I never said I ran them at the voltages of a 520. The ones I put in this base were all set at 1.5 v. They are the L suffix. The ones I put in galaxy cb's 919-979 are all the gets without the L suffix. They do better at 2.5 as you said

 

[nomadradio]

The letter "L" added to the part number generally means "Logic", as in 3-Volt and 5-Volt digital logic signals. Instead of the typical 3 to 3.5 Volt needed for the common version, these are meant to switch on at around 2 Volts so you can drive them directly from the output pin of a microcontroller, or a logic-gate chip.

Pretty sure I got some "IRL520" parts a while back. Never did use them for anything.

Bought a batch the '13N10L' version a while back. Thought it would be interesting to compare to the 'normal' version. They're still in the plastic tube, all 25 of them. At least I know where they are.

Another project slides off the back burner and down behind the stove.

73

 

[Onelasttime]

Novel idea since you can still get the bjt finals for it put them in and run an external amp. It is not as if we can not get any bjt finals and driver's. Soon you will have no choice but since that day is not here put her back to stock and be done witht he popcorn like Mosfets! If you like fet's so much get a BLF188XR or two and build a nice amp for it.