• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Cascoded DN2540 CCS down to 0.3 mA

Do you think that that cascoded version using a pair of DN2540 would be a good choice for a current sink that will be used for ~75 volt drop at 6-10mA?

The DN2540 cascode should work well (i have used many as "tails" for LTPs). However, if you are after the best sonics you should substitute IXYS 8N50D2s for the DN2540s. The proof is in the listening. Walt's circuit yields a small but significant improvement over and above the standard cascode. Caution is advised if using Walt's circuit for a plate load. I experienced over voltage and eventual failure of the small capacitor on start-up. In that case I would advise using a B+ delay or revert to standard cascode CCS.
 
Hi,

It was in the context or your post directly above mine.

"Eventual failure of the capacitor" 1uF/63V

So what would happen if the capacitor was uprated to prevent this occurring? Would that prevent the failure? Or simply move the failure to another component?
 
You could use a high voltage capacitor, but the gate-source oxide of the upstairs FET is also vulnerable, if 100V caps are known to be at risk.

zener diodes can be used to protect both of these positions, but watch out for the capacitance of these diodes, at the working voltage. This can undo some of the good work of the cascoding.

The IXYS FETs are better, but with these upstairs, the downstairs FET gets only a modest increase in VDS. The value of COSS of the DN2540 is still falling rapidly until VDS gets to >12-15V.

I suspect the overvoltage damage in anode loads is caused by the slew-rate (V/S) of the rising HV supply. The rising edge couples through Cgd, which is at max when there is 0V on the supply. Using vacuum diodes, especially damper diodes of the kind that take 30s to get up to 400V, might improve things. But a zeners are a more reliable fix.
 
The failure only occurred in the one instance where I used this CCS as a plate load and then it took a few months (occasional loud pops on start-up). The MOSFETs where not damaged. Makes more sense to me to manage the B+ on start-up than to use a big *** 600V cap in that position. I haven't had any issue using them with LTPs or as current sources for CCS fed shunt regulators (disclaimer: all those units have delayed B+).
 
  • Like
Reactions: LuvValve
In this case a single TO-220 Depletion FET (example DN2540) will be enough.

Use a good sized heatsink, and keep the Vds at 20-30V, to keep the DC drift of current under control.

Use a trimmer potentiometer in series with a fixed value resistor, to prevent accidental excessive current adjustment.
 
Did you try with a non-cascode (single FET) ?

The cascode gives good overall performance, but if both FETs are DN2540, the lower FET is in Triode Mode, and low current can never be achieved...... but with only one it should work better.

If you tried single, and it was still too high, another choice for 10mA & low power dissipation: BSP135 (infineon). SMT SOT-223 package - but it is easy to solder, and you can get adapter PCBs.

Use BSP135 for low current testing and lower Vin-Vout, and DN2540 when higher current is needed.