2

I have a design (see page 3) that uses a JFET to linearize the charge current for a capacitor used as an integrator. I've been using the BSR56, but that's been discontinued.

I am used to looking at MOSFET datasheets to determine substitution parameters, but I am unfamiliar with JFETs. What parts of the datasheet can I compare to obtain the best match?

Marcus Müller
  • 94,373
  • 5
  • 139
  • 252
nsayer
  • 1,573
  • 1
  • 18
  • 35

1 Answers1

1

I note that the circuit diagram (page 3) indicates Q1 (BSR56) can be replace with a 0 ohm link. This suggests it's function is not that critical.

Having said that Q1 is setup in a constant-current configuration so you could take a look at the data sheet for the BSR56 https://pdf1.alldatasheet.com/datasheet-pdf/view/50812/FAIRCHILD/BSR56.html and review the FET characteristic curves (see example figure from http://www.learningaboutelectronics.com/Articles/What-is-IDSS-of-a-FET-transistor) to determine ID at the operating point and look for another device with a similar ID and "flatness" of the curve in the Saturation Region.

With regard to other parameters probably just match the common sense stuff such as N channel JFET, don't exceed max ratings e.g. Pd>0.25W, ... There are some "matching" tools on websites that might help e.g. https://alltransistors.com/mosfet/crsearch.php?&struct=JFET&polarity=N&pd=0.25&uds=40&id=0.05&rds=25&caps=SOT23

mhaselup
  • 1,486
  • 9
  • 12
  • It's not critical - without the linearization, the circuit would be functional, but not as accurate. I'm hoping to find a substitute that would operate with the as close to the same current limit as possible so as to not have to redo the R and C values. The circuit is fed 5 volt pulses that are anywhere from 0 to 1µs wide once per second and the hope is an approximate 3.3 volt output range. – nsayer Apr 13 '20 at 04:06
  • I've had a trawl around and I can't find a datasheet for the BSR56 with the FET characteristic curves, this may be because it appears to be categorized as a switching transistor. Are you trying to repair a board, do you have access to an original where it would be possible to empirically determine Id and calculate a matching resistor capacitor pairing for an alternative device. I'm sure there are lots of alternative JFETs with similar characteristics e.g. https://www.nxp.com/docs/en/data-sheet/PMBFJ111_112_113.pdf that could be workable with chosen RC but the time constant is going to be off. – mhaselup Apr 13 '20 at 07:32
  • I'm not trying to repair the board. I'm the original designer. I copied this chunk of the circuit from someone and just sort of swapped out the RC values ad hoc until I was close enough. This time, I want to be a little less empirical, so I am trying to understand better how to design this little linear integrator circuit. In principle I don't mind going with a completely different part and making appropriate changes to suit (so perhaps I worded the question badly). but if I do, I want to better understand what I'm doing. – nsayer Apr 13 '20 at 23:47
  • I played around with circuitlab and it looks like I can get the range I want with a J310 JFET (which DigiKey says can be directly substituted with an mmbfj309) by changing the resistor out for a 560Ω one, FWIW. But that doesn't really get me closer to understanding how to design the current limiter. I gather dV/dT = I/C, so if I want 3.3v in a µs with a 1 nF cap I need a constant current of 3.3mA. How do I use the datasheet to figure out the correct gate bias resistor? – nsayer Apr 14 '20 at 00:32
  • Can you send me a link to the datasheet for the device you are planning to use and I'll take a look. – mhaselup Apr 14 '20 at 00:33
  • https://www.digikey.com/product-detail/en/MMBFJ309LT1G/MMBFJ309LT1GOSCT-ND/1139804/?itemSeq=323375788 – nsayer Apr 14 '20 at 00:37
  • From that supplier I would recommend https://4donline.ihs.com/images/VipMasterIC/IC/ONSM/ONSM-S-A0000713519/ONSM-S-A0000713519-1.pdf?hkey=EC6BD57738AE6E33B588C5F9AD3CEFA7. For 3+mA this could be biased at about VGS=-0.3V which equates to 100 ohms. Please take a look at the datasheet and refer to diagram HD00869 which shows ID vs VGS, the "flatness" of the -0.3V curve vs VDS is your current source effect. – mhaselup Apr 14 '20 at 06:13
  • Let us continue this discussion in chat. – mhaselup Apr 14 '20 at 06:19