StoppTidigare
Full Member level 2
- Joined
- Jul 16, 2002
- Messages
- 148
- Helped
- 1
- Reputation
- 2
- Reaction score
- 0
- Trophy points
- 1,296
- Location
- Northern earth
- Activity points
- 1,703
Hi all, I've been analyzing the philips BFG425W-transistor, just for fun of course... Being on my vacation ....Finally having time for some serious work, not that stupid work that I have to do for living. Anyway...
Been fiddeling with the newest Aplac, and Genesys 2003. I've been astounded about the differences in the simulators output.
With Aplac I've done a Ic as a function-of-Vce-sweep, varying Ib, and found out that when fixing Vce=2V and doing a Ic as a function of Ib sweep. The difference in Ic is approx. 9-10mA for that particular Ib !
Futher more sweeping S21, It goes down to 0dB at 10GHz at vce=2V ans Ic = 20mA.
According to datasheet ft is equal to 25GHz for that particular bias. (ft is when gain has gone down to 0dB, when load is replaced with a short, according to flatulent in a previous discussion. I hope I remember what he wrote correct...)
Something is either very wrong with the model provided by philips, or Aplac is really out of line...Its not physical to have such a big differnce between S21 being 0dB with a matched load and current gain being 0dB att 14GHz above!(ft=24GHz).
Comparing the plots of the S-params swept from 40 to 3000MHz as is done in the datasheet, leaves a lot to which for. the difference was BIG, except for S11 and S12. The S21 info was far away from the datasheet that it could have been a complete differnet transistor.
Aplac also starts to sweep from a couple of kHz suddenly, though is specified to run from 40-3000MHz. Very strange!
I wonder if the S-params from the datashett is from an actual measurement or from an implementation of the model ? If so it could be the case that Aplac is right and the model given is shitty....
I ran the same simulations with Genesys 2003, except the Ic=f(Ib) with Vce=2V.
It seems that genesys 2003 does not use Spice3f5 for solving the non-linear equations in the Ic-Vce sweep, because it cannot converge to zero at the beginning of the sweep.
For doing this sweep, u need access to HARBEC, their non-linear solver, which I belive is homemade....
S21 is 0dB, at ic=20mA and Vce=2V, at around 8.9GHz, that is 1.1 GHz away from Aplac's prediction !!!
Also here I think it was that S12, and S11 was closest to the datasheet.
I wish that I could test with ADS, and MWO, as well as Ansoft Designer to compare their outputs...
maybe its good for genesys to become mixed up with ADS....
If somebody is interested I can upload a word-doc. with my simulations and the output I got, but not tomorrow, I mean later today (its half past three in the morning), because I'll have the opportunity to work with an excavator. we're draing our house. Hydraulics is really fluid electronics ! Fun !
regards,
StoppTidigare
Been fiddeling with the newest Aplac, and Genesys 2003. I've been astounded about the differences in the simulators output.
With Aplac I've done a Ic as a function-of-Vce-sweep, varying Ib, and found out that when fixing Vce=2V and doing a Ic as a function of Ib sweep. The difference in Ic is approx. 9-10mA for that particular Ib !
Futher more sweeping S21, It goes down to 0dB at 10GHz at vce=2V ans Ic = 20mA.
According to datasheet ft is equal to 25GHz for that particular bias. (ft is when gain has gone down to 0dB, when load is replaced with a short, according to flatulent in a previous discussion. I hope I remember what he wrote correct...)
Something is either very wrong with the model provided by philips, or Aplac is really out of line...Its not physical to have such a big differnce between S21 being 0dB with a matched load and current gain being 0dB att 14GHz above!(ft=24GHz).
Comparing the plots of the S-params swept from 40 to 3000MHz as is done in the datasheet, leaves a lot to which for. the difference was BIG, except for S11 and S12. The S21 info was far away from the datasheet that it could have been a complete differnet transistor.
Aplac also starts to sweep from a couple of kHz suddenly, though is specified to run from 40-3000MHz. Very strange!
I wonder if the S-params from the datashett is from an actual measurement or from an implementation of the model ? If so it could be the case that Aplac is right and the model given is shitty....
I ran the same simulations with Genesys 2003, except the Ic=f(Ib) with Vce=2V.
It seems that genesys 2003 does not use Spice3f5 for solving the non-linear equations in the Ic-Vce sweep, because it cannot converge to zero at the beginning of the sweep.
For doing this sweep, u need access to HARBEC, their non-linear solver, which I belive is homemade....
S21 is 0dB, at ic=20mA and Vce=2V, at around 8.9GHz, that is 1.1 GHz away from Aplac's prediction !!!
Also here I think it was that S12, and S11 was closest to the datasheet.
I wish that I could test with ADS, and MWO, as well as Ansoft Designer to compare their outputs...
maybe its good for genesys to become mixed up with ADS....
If somebody is interested I can upload a word-doc. with my simulations and the output I got, but not tomorrow, I mean later today (its half past three in the morning), because I'll have the opportunity to work with an excavator. we're draing our house. Hydraulics is really fluid electronics ! Fun !
regards,
StoppTidigare