the 1539 are graphed to show fT>=250MHz @ Ic=10mA.
and specified at fT~= 400MHz typ @ Ic=50mA.
Indicates that driver bias should be around 10mA or higher to get sufficient speed to control the FETs. R22 & 23 will be around 22r to 56r depending on Vgs of the FETs.
£2/pair seems not too bad a price if one pair can drive two dual die FETs. But keep them cool. BIG sinks!!!
and specified at fT~= 400MHz typ @ Ic=50mA.
Indicates that driver bias should be around 10mA or higher to get sufficient speed to control the FETs. R22 & 23 will be around 22r to 56r depending on Vgs of the FETs.
£2/pair seems not too bad a price if one pair can drive two dual die FETs. But keep them cool. BIG sinks!!!
AndrewT said:
I thought they were quite a bargain! I know the Vgs of the FETs in that circuit now, so I'll work out exactly what I need for R22/23 to get the best out of the drivers. And think I've got perfect heat sinks that I can bolt each side of the driver/bias assembly.
Lumba Ogir said:
Thanks. Intriguing! I like the look of that...
Dalbani are out of stock on the 2SA1539's, but have arranged for 2SA1405 instead, which are almost indentical. Looking at the data sheet for the 1405 it looks to me like 40mA is the best comprimise between fT and dissipation. This should correspond to about 400MHz. To achieve the full 500MHz we're looking at 70-80mA.
For the 2SC3954 it's 30mA to get 400MHz, but 70mA again to get the full 500MHz
Only experimentation will tell whether the extra dissipation will be worth the extra speed and if it's needed....
I've got no problem with driving them hard, there's room for good heatsinking on the drivers. They're both thermally coupled to the bias transistor too.
For the 2SC3954 it's 30mA to get 400MHz, but 70mA again to get the full 500MHz
Only experimentation will tell whether the extra dissipation will be worth the extra speed and if it's needed....
I've got no problem with driving them hard, there's room for good heatsinking on the drivers. They're both thermally coupled to the bias transistor too.
AndrewT said:
I'd stick to the 2SC3503/2SA1381 myself, and run them at 10mA. Apologies to speed junkies.
I wouldn't cascode anything.
I would replace the collector resistors with CCSs, although this would require additional psu voltages due to the inconveniently low Vgs of the FETs. Then raise the emitter resistor values by a factor of 10 to bring the bw back below 10MHz. Then reduce the gate resistors and hope to ditch the DG capacitors. Actually, I'd probably ditch the lateral FETs and use VFETs and solve the problem that way.
That's what I would do if I were intent on using CFPs and FETs.
But I don't expect anyone to concur.
I see what you're saying, but I'm going to go ahead and try the different transistors for a couple of reasons:
It's would be informative to see if they do have the desired effect. I'll try running them at different currents to find out if speed really is an advantage in the drivers, and if so, how much.
Secondly, I've got quite a few of theses FETs hanging around, so going for something different isn't an option for the moment. The circuit has to fit the devices, as it were. I'm also a little reluctant to go to all the extra work of adding another set of rails for the CCS's if an easier method can be found to achieve a good result. I'm not ruling it out, however...
Sorry for the lack of work over the weekend, it was rather a busy one, and my new transistors only turned up this morning. I should find some time this evening though.
It's would be informative to see if they do have the desired effect. I'll try running them at different currents to find out if speed really is an advantage in the drivers, and if so, how much.
Secondly, I've got quite a few of theses FETs hanging around, so going for something different isn't an option for the moment. The circuit has to fit the devices, as it were. I'm also a little reluctant to go to all the extra work of adding another set of rails for the CCS's if an easier method can be found to achieve a good result. I'm not ruling it out, however...
Sorry for the lack of work over the weekend, it was rather a busy one, and my new transistors only turned up this morning. I should find some time this evening though.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.