Hello,
It is a high voltage switcher NPN in TO-220 package, 400V 12A. I have the "L" suffix ST13009L, can't find any model online. It is an old-ish part but still listed as active on Digikey, however no spice model for it on ST's site... weird. Still looking but if you guys have one in your archives that would make my day! I'm using LTspice.
Regards.
It is a high voltage switcher NPN in TO-220 package, 400V 12A. I have the "L" suffix ST13009L, can't find any model online. It is an old-ish part but still listed as active on Digikey, however no spice model for it on ST's site... weird. Still looking but if you guys have one in your archives that would make my day! I'm using LTspice.
Regards.
Found similar here http://bordodynov.ltwiki.org/
`
Sorry it took so long😳
`
.MODEL MJE13009 NPN IS=418.55f BF=21.619 VAF=100 IKF=11.864 ISE=5.9443p NE=1.4275 BR=.2103 VAR=100 IKR=4.4990 ISC=701.45p NC=1.270 NK=.6367 RB=82.769m RC=25.254E-3 CJE=3.1796n VJE=1.5 MJE=.11593 CJC=516.27p VJC=1.5 MJC=.39769 TF=12.894n XTF=10 VTF=10 ITF=1 TR=64.234u Vceo=400 Icrating=12
.MODEL MJE13009_ NPN (IS=5.36792p BF=22.7911 NF=0.85 VAF=10 IKF=5.53474 ISE=8.27274e-13 NE=3.68432 BR=2.27911 NR=0.75 VAR=4.68261 IKR=8.44426 ISC=8.27274e-13 NC=3.65698 RB=0.938365 IRB=0.1 RBM=0.1 RE=0.0001 RC=0.0350521 XTB=0.114306 XTI=1 EG=1.12482 CJE=4.50839n VJE=0.4 MJE=0.23 TF=1e-08 XTF=2.77017 VTF=6.06314 ITF=0.001 CJC=5e-10 VJC=0.95 MJC=0.457977 XCJC=0.799439 FC=0.8 TR=1.76449u PTF=0 Vceo=400 Icrating=12)
Sorry it took so long😳
Hey no big deal, I'm patient and not in a hurry at all ! Thanks a lot, the first model seems to be the ducks guts like you say down there 😎
Wondering about the one with the underscore suffix, is it another format for other simulators?
I have bookmarked the library page as well. Thumbs up!
The underscore one is an alternative fit model both run in LTspice. The main difference between them is VAF (Early voltage) - the first one has a high VAF and the second a low VAF. The ST10039 datasheet does not have enough information to decide which is better. The second one seems to have more parameters fitted so I'd use it. You'd have to measure a transistor to decide if a VAF=10 is better than a VAF=100.
The alternate model gives a DC current gain which is almost an order of magnitude higher than the first one which is closer to the datasheet specification (26-39). As there is about the same ratio between the Early voltage parameters and the simulated current gain for the two models it may be related ? My knowledge of spice parameters is basically non-existent so I can't tell what's wrong with the alternate model; I'll use the first one for the moment.
Thanks again!
- Joris
Thanks again!
- Joris
Hi Joris,
Yes, second thoughts, use the first model with VAF=100 for now. The first model, with VAF=10, will modulate the current gain by a factor of 10 with a collector voltage of 100V.
I assume you want to simulate an amplifier and where collector voltage can get to 100V. You said you have an 'L' device. A simple test with a voltage supply can settle which VAF applies to your device. You need
1) a LV bias voltage (1.5-5V) for the base-emitter (with a series base resistor ~1-10k), and
2) say a 5V collector supply that can be increased later to 12V or higher. Bias for 5-10mA of collector current with 5V on the collector.
Then increase the collector voltage and note the change in collector current. A VAF=10 the Ic doubles when going from 5V to about 15V. If your Ic increases by say 10% then use the VAF=100 model.
If you don't have a variable voltage supply for the collector voltage then you could use two 9V batteries with 10mA or so bias. You are just after ball park readings. BTW 10mA keeps heating low so the current doesn't drift too much with time.
BTW2 you shouldn't change the VAF value in the models to suit your readings - because most of the parameters of these SPICE models are built around the VAF value which is chosen at the start of the parameter fitting process.
Yes, second thoughts, use the first model with VAF=100 for now. The first model, with VAF=10, will modulate the current gain by a factor of 10 with a collector voltage of 100V.
I assume you want to simulate an amplifier and where collector voltage can get to 100V. You said you have an 'L' device. A simple test with a voltage supply can settle which VAF applies to your device. You need
1) a LV bias voltage (1.5-5V) for the base-emitter (with a series base resistor ~1-10k), and
2) say a 5V collector supply that can be increased later to 12V or higher. Bias for 5-10mA of collector current with 5V on the collector.
Then increase the collector voltage and note the change in collector current. A VAF=10 the Ic doubles when going from 5V to about 15V. If your Ic increases by say 10% then use the VAF=100 model.
If you don't have a variable voltage supply for the collector voltage then you could use two 9V batteries with 10mA or so bias. You are just after ball park readings. BTW 10mA keeps heating low so the current doesn't drift too much with time.
BTW2 you shouldn't change the VAF value in the models to suit your readings - because most of the parameters of these SPICE models are built around the VAF value which is chosen at the start of the parameter fitting process.
Thanks Ian for the follow-up. Like I said I have no knowledge of model parameters so I won't mess with the models you supplied.
I do have a jig I made for extracting small signal data from power mosfets that can serve for the measurements you suggest. This is to be the pass device in a high voltage regulator and will often be subjected to collector-emitter voltages ranging well over 100V, from 125V - 150V; so determining the correct parameter related to the Early effect can be relevant. My jig has a large heatsink, I could even measure at the expected current.
But even then, a ten-fold difference in DC gain for the second model hardly seems correct.
I do have a jig I made for extracting small signal data from power mosfets that can serve for the measurements you suggest. This is to be the pass device in a high voltage regulator and will often be subjected to collector-emitter voltages ranging well over 100V, from 125V - 150V; so determining the correct parameter related to the Early effect can be relevant. My jig has a large heatsink, I could even measure at the expected current.
But even then, a ten-fold difference in DC gain for the second model hardly seems correct.
Hi. Each model is defective. The second model can be improved by making VAF=500. Another disadvantage of the second model is the low nf value. This value should be close to 1. It is impossible to simply replace this value with this value because the voltage BE and the current gain behavior of the transistor depend on it. Due to the small value, the steepness of the transistor is overestimated. It is more correct to make a combined model of the two models and try to add consideration for the quasi-saturation of the transistor. If there are volt-ampere characteristics, this can be done.
.MODEL MJE13009 NPN
.MODEL MJE13009 NPN
- AF = 1
- BR = 1.16
- CJE = 5.29084E-09
- CJS = 0
- EG = 1.11
- FC = 0.5
- IKF = 10
- IKR = 1.38
- IRB = 0.0141
- IS = 5.43E-12
- ISC = 1.71E-09
- ISE = 6.1E-15
- ITF = 0.001
- KF = 0
- MJC = 0.405552
- MJE = 0.351655
- MJS = 0.33
- NC = 1.48
- NE = 1.16
- NF = 1.1
- NR = 1.11
- PTF = 0
- RB = 0.0989
- RBM = 0.0999
- RE = 0.0381
- TR = 0
- VAF = 381
- VAR = 95.5
- VJC = 0.367838
- VJE = 0.56191
- VJS = 0.75
- VTF = 1
- XCJC = 1
- XTB = 0
- XTI = 3
- BF = 1.520000E+01
- CJC = 4.645170E-10
- RC = 3.780000E-02
- TF = 1.000000E-08
- XTF = 1.000000E+00
Last edited: