transimpedance amplifier

[Gluca]

Ciao,

I was playing with spice to simulate a IV stage for my dac and it seems spice goes nuts. Whats wrong with the schematic? (forget the DC offset at the input of the dac) How do I calculate the gain of the stage? OK, too many questions.

Gianluca

[Arnulf]

Why are you using a current source in cathode circuit with fixed bias ?

[Gluca]

I think the problem is not in the CCS but in the choke. I am under the impression I really need a small resistor there. That's weird because a see a couple of schematics like this and they have been reported to be very OK. Mah???

[Michael Koster]

Originally Posted by Gluca

I think the problem is not in the CCS but in the choke. I am under the impression I really need a small resistor there. That's weird because a see a couple of schematics like this and they have been reported to be very OK. Mah???

Your triode is set up to act like the top device in a cascode. The anode current is simply a repeat of the cathode current, which is 10mA plus/minus the current modulation from the DAC.

You need a resistive load, either in place of the choke or in parallel otherwise the gain will be dependent only on the choke reactance and DCR.

[Gluca]

Noted. I swapped the choke with a R (2k) and it is working as expected. But impedance at the DAC port is in the 4-500ohm ballpark ... way too high.

[tomchr]

It wouldn't hurt to include some series resistance with the choke. A few 10's of ohm is probably appropriate. A lot of times spice gets confused when you have lossless inductors and current sources coupled too closely.

~Tom

[cetoole]

Originally Posted by Gluca

Noted. I swapped the choke with a R (2k) and it is working as expected. But impedance at the DAC port is in the 4-500ohm ballpark ... way too high.

Well yeah, what do you expect when using a tube with such low transconductance? Try something else, triode strapped high slope pentode (E55L, 7788) or just super triode (we437a, 6c45p).

[Michael Koster]

Originally Posted by cetoole

Well yeah, what do you expect when using a tube with such low transconductance? Try something else, triode strapped high slope pentode (E55L, 7788) or just super triode (we437a, 6c45p).

Yes, the impedance seen by the DAC would be approximately 1/gm. 6c45p, 7788 or D3A in triode, etc. should give you 20-25 ohms at the right OP.

[Bandersnatch]

hey-Hey!!!,
Run the CF as a pentode. Just to note, type EF184 will deliver less than 100R ( if you can set up an idle point with greater than 10 mA/V ). Type E55L is just going too far imo; D3a is much more efficient and easier to come by( sez as he adds to E55L collection...). You'll get better PSRR with a pentode CF, and the g2 supply is quite simple: dropping R and bypass cap.
cheers,
Douglas

[cetoole]

Other issues you will see with that setup is no way to set the cathode voltage to ground, or whatever reference the DAC requires, and the distortion of one of those triodes with a 2k plate load will be rather high. I would think the way to go would be to servo the grid or possibly plate with a fixed grid reference. D3a is a very nice tube and probably a good choice. It would be interesting to compare the results with triode and pentode connection.