D
Deleted member 548388
Sorry for bothering you again, Rod, what voltage should this cap be rated for? This components grow rapidly with increasing voltage rating and my PCB is rather small^^
I tested a bunch of DN2535/40 with the DCA75: some of them are recognized correctly as depletion mosfets, some others as voltage regulators, or as defective components.
I read somewhere that the DN2534/40 doesn't work properly under 15mA. It seems to be true, I had the same issue trying to feed a zener diode with 1mA using the DN2540 or DN2535, and both overheated. I cannot find that post again, perhaps some of you have better knowledge?
I read somewhere that the DN2534/40 doesn't work properly under 15mA. It seems to be true, I had the same issue trying to feed a zener diode with 1mA using the DN2540 or DN2535, and both overheated. I cannot find that post again, perhaps some of you have better knowledge?
The problem with low currents and DN2540 applies to cascoded pairs of them. In cascode: increasing Vgs (to get lower current) with larger source resistor is negated by the upper FET's increasing Vgs, which increases the Vds of the lower FET (increases current).
You should be able to get 1mA with a single DN2540, but it may not be very stable with temperature.
The Datasheet SOA should be carefully compared to your circuit voltages, and heatsinking applied if the power burn is more than 500mW. Running more than 325V across DN2540s can cause voltage failures - transients at power OFF are often responsible.
You don't say what the voltage is, but if the SOA is respected, 1.3 to 1.7mA or so can be directed to a zener with a LND150 at IDSS - no source resistor at all. Use a resistor to drop some voltage, if it is too high.
You should be able to get 1mA with a single DN2540, but it may not be very stable with temperature.
The Datasheet SOA should be carefully compared to your circuit voltages, and heatsinking applied if the power burn is more than 500mW. Running more than 325V across DN2540s can cause voltage failures - transients at power OFF are often responsible.
You don't say what the voltage is, but if the SOA is respected, 1.3 to 1.7mA or so can be directed to a zener with a LND150 at IDSS - no source resistor at all. Use a resistor to drop some voltage, if it is too high.
Hello Rod, you're right, I would have better explained more about my case.
First of all, I was using the DN2535/40 TO220 version.
The maximum voltage on the DN2535/40 is 50V or 150V depending on where it is used.
I enclose here the schematic of the application: I tried to use the DN2535/40 as a CCS in the power supply, but it didn't work. The DN2535/40 was acting as a short, overheating itself, and the zener diodes. So I replaced the DN2535/40 with a cascode CCS ad the issue was solved. Then I made the mod to the audio output source follower, loading it with a CCS instead of a resistor. The CCS was made out of a DN2535/40. This mod was working very badly, my friend Flavio had to select a couple of DN2535/40 out of a bunch of ten and the sound came out correctly, better than using the resistor.
Some DN2535/40 are not recognized correctly by the DCA75, they all were purchased from official sellers like Mouser, Digikey, RS Components.
Any hint?
First of all, I was using the DN2535/40 TO220 version.
The maximum voltage on the DN2535/40 is 50V or 150V depending on where it is used.
I enclose here the schematic of the application: I tried to use the DN2535/40 as a CCS in the power supply, but it didn't work. The DN2535/40 was acting as a short, overheating itself, and the zener diodes. So I replaced the DN2535/40 with a cascode CCS ad the issue was solved. Then I made the mod to the audio output source follower, loading it with a CCS instead of a resistor. The CCS was made out of a DN2535/40. This mod was working very badly, my friend Flavio had to select a couple of DN2535/40 out of a bunch of ten and the sound came out correctly, better than using the resistor.
Some DN2535/40 are not recognized correctly by the DCA75, they all were purchased from official sellers like Mouser, Digikey, RS Components.
Any hint?
Attachments
@marigno, in the picture is CCS with DN2540(sot89) and BSS159(sot23) in cascade, current 2mA, voltage drop 90V, used as CCS in my HV regulator for tubes.
DN2540 sometimes gives strange results on testers, which does not mean that it is not good and DN2540 is easily destroyed with careless handling and static electricity even before it is soldered to the PCB, so you need to be careful with it.
I have been working with depletion mosfets for over ten years and they have all been ordered from Mouser and have always been ok so I doubt you are getting bad ones from Mouser.
DN2540 sometimes gives strange results on testers, which does not mean that it is not good and DN2540 is easily destroyed with careless handling and static electricity even before it is soldered to the PCB, so you need to be careful with it.
I have been working with depletion mosfets for over ten years and they have all been ordered from Mouser and have always been ok so I doubt you are getting bad ones from Mouser.
Attachments
D
Deleted member 548388
@heatsink12 ,in my tube IV stage I used IXTP08N100D2(on heatsinks) and two 2SK3557-7 in parallel at 15mA, 150V voltage drop, much better result than with two DN2540. I only protected 2SK3557 with a 10V zener diode. I recommend the use of two 2SK3557 because their Ids vary a lot and should be matched in that case.
The solution I presented in the post 15 was much better, a huge difference in sound, a more open sound image, more details ... and since then I no longer use jfets in CCSs even for a voltage drop of 2.5V(two BSS139) .
Instead of BSS159, you can also use BSS126, but pay attention to max Ids.
The solution I presented in the post 15 was much better, a huge difference in sound, a more open sound image, more details ... and since then I no longer use jfets in CCSs even for a voltage drop of 2.5V(two BSS139) .
Instead of BSS159, you can also use BSS126, but pay attention to max Ids.
Attachments
@grunf
Hello Grunf, I must apologize, I didn't receive any mail reporting your answer, I see it only now. In the meanwhile I discovered what happened: I handled the mosfets without being grounded, and I destroyed them. I bought them on Mouser, so they were good items. After handling them, they acted like shorts.
In my house, I get a lot of static electricity, so I must be careful. After destroying an LSK489, a couple of weeks ago, I finally understood.
I took note of your CCS at #15, thank you for sharing it!
Hello Grunf, I must apologize, I didn't receive any mail reporting your answer, I see it only now. In the meanwhile I discovered what happened: I handled the mosfets without being grounded, and I destroyed them. I bought them on Mouser, so they were good items. After handling them, they acted like shorts.
In my house, I get a lot of static electricity, so I must be careful. After destroying an LSK489, a couple of weeks ago, I finally understood.
I took note of your CCS at #15, thank you for sharing it!
That's true! I would solve the problem only in my lab, while the whole house needs a humidifier, but installing it in the utility room means all appliances are to be dismounted and moved. I wish to have in front of me the man who installed the water heater and the A/C machine, never seen a job more sloppy than that. I'll go for a "local" humidifier to put in the lab. By now I made myself a wrist strap and never destroyed any other component yet!
You can bias on a mosfet with a multimeter and it will read shorted until you reverse bias it. I found this out 25 years ago, I biased on a fet used on a TV, while I was testing it as a new part. The first test on the multimeter showed it open. I tried it again and it was shorted, switched the leads and it was open again.
I just tried it with a IRFZ44N in my hand, with a Fluke 87 in diode mode, and positive lead to gate, negative lead to source, and then move the positive lead to the drain and it will read 0 Ohms, and will for some time. Reverse the leads on the drain and source and the transistor will read open again.
I use the DCA75, and it fails to recognize some DN2540. Some other 2540 are recognized as OK. The 2540 not recognized by the DCA75 acted as a short instead of a CCS. My fault, no doubt about it, as I mishandled them. You cannot test FETs and mosfets using a normal multimeter, not even the best of them, not even trying to polarize using your finger between gate and drain, sometimes it works, but not always.