Correct me if I am wrong but the increased drift with the 2SK170BL would only afftect those requiring constant (absolute) B+ V, for most tube designs a slow +_ 1V drift wouldn't be a big deal especially after everything has warmed up and SS. Or is there more to it than that?
What about a 2SK170BL with a slipon heatsink?
The GR version is diificult to come by.
What about a 2SK170BL with a slipon heatsink?
The GR version is diificult to come by.
lsk170a are available.
I am not sure of the temp compensation mechanism improvement that results from using a 117 instead of a 170.
I certainly found when looking at low voltage Salas Regs that the different devices all had a quite different effect on drift with temperature.
Could someone explain why the 117 is better than the 170 in this temp comp respect?
I am not sure of the temp compensation mechanism improvement that results from using a 117 instead of a 170.
I certainly found when looking at low voltage Salas Regs that the different devices all had a quite different effect on drift with temperature.
Could someone explain why the 117 is better than the 170 in this temp comp respect?
Thanks all of you. Please look at the above
Characteristics of
BF244
VDG Drain-Gate Voltage 30 V
VGS Gate-Source Voltage - 30 V
ID Drain Current 50 mA
IGF Forward Gate Current 10 mA
Tstg Storage Temperature Range -55 to +150 °C
BF245
VDG Drain-Gate Voltage 30 V
VGS Gate-Source Voltage -30 V
IGF Forward Gate Current 10 mA
PD Total Device Dissipation @TA=25°C
Derate above 25°C
350
2.8
mW
mW/°C
TJ, TSTG Operating and Storage Junction Temperature Range - 55 ~ 150 °C
2SK170
Gate-drain voltage VGDS 40 V
Gate current IG 10 mA
Drain power dissipation PD 400 mW
Junction temperature Tj 125 °C
Storage temperature range Tstg 55~125 °C
2SK117
Gate-drain voltage VGDS 50 V
Gate current IG 10 mA
Drain power dissipation PD 300 mW
Junction temperature Tj 125 °C
Storage temperature range Tstg 55~125 °C
Regards Dade
Characteristics of
BF244
VDG Drain-Gate Voltage 30 V
VGS Gate-Source Voltage - 30 V
ID Drain Current 50 mA
IGF Forward Gate Current 10 mA
Tstg Storage Temperature Range -55 to +150 °C
BF245
VDG Drain-Gate Voltage 30 V
VGS Gate-Source Voltage -30 V
IGF Forward Gate Current 10 mA
PD Total Device Dissipation @TA=25°C
Derate above 25°C
350
2.8
mW
mW/°C
TJ, TSTG Operating and Storage Junction Temperature Range - 55 ~ 150 °C
2SK170
Gate-drain voltage VGDS 40 V
Gate current IG 10 mA
Drain power dissipation PD 400 mW
Junction temperature Tj 125 °C
Storage temperature range Tstg 55~125 °C
2SK117
Gate-drain voltage VGDS 50 V
Gate current IG 10 mA
Drain power dissipation PD 300 mW
Junction temperature Tj 125 °C
Storage temperature range Tstg 55~125 °C
Regards Dade
Last edited:
Regarding the temperature drift. Think about how much current goes through the voltage reference made of a current source (jfet) and a resistor. If the current source is set for low current, a temperature variation will be translated into a higher voltage variation because the resistor has a high value. Higher current will force you to use a smaller resistor in the voltage reference, and Vout will not vary as much with temperature. But then you need to deal with the higher dissipation. I don't see why 2sk117 and 2sk170 would be inherently different when it comes to tempco. Both will have a sweet spot in terms of tempco, and it will not be at Idss. Have a look in the attached image at how the current changes in a jfet as a function of temperature. Pretty obvious which operating point is the best.
Edit: I forgot to mention. In this position, one of the most important specs for the jfet is Vgs(OFF). Both the 2sk117 and the 2sk170 have a Vgs(OFF) of between -0.2V and -1.5V. There are not many jfets that have such small Vgs(OFF). J201 is another, but its Idss is much smaller at between 0.2mA and 1mA.
Edit: I forgot to mention. In this position, one of the most important specs for the jfet is Vgs(OFF). Both the 2sk117 and the 2sk170 have a Vgs(OFF) of between -0.2V and -1.5V. There are not many jfets that have such small Vgs(OFF). J201 is another, but its Idss is much smaller at between 0.2mA and 1mA.
Attachments
Last edited:
For Q2 Supertex N3 TO-92 package, could be useful for space saving, haven't tried it though. BF245A,B,C too much pinch off to work under just Vbe correctly. Maybe some low side IDSS samples of A grade but you got to measure they are under 0.6V. Don't know why K117GR was more stable than K170BL, maybe has different Gos or its zero tempco curves crossing point is different (no such charts from Toshiba), but it is surely steadier in test with alike currents and trimmer. It also goes negative drifting like BF245. K170 gains. In GR too.
This is a regulator, your CCS is an active current load. Different things. Any raw DC source as high as adequate is acceptable input to the reg. In case of any weird interactions you can use a 0.1uF/400V decoupling capacitor across reg's DCin connector. Especially if with long cables coming in. Keep in mind that this is a new circuit and we don't have enough feedback yet to what it may not like so to tell many things in advance.
Salas, I am reading through your post 2061 about the SSHV2.....how to determine what current to set with R4 across R5 the T.P.??...for example, if I set my C4S to be 20 mA per channel....do I need one SSHV2 per channel??....what dummy load should I use....should I set R4 first, then R5??
I will start to shop for parts next week....looking forward to try SSHV2.
Thanks again.
I will start to shop for parts next week....looking forward to try SSHV2.
Thanks again.
Use what your load gets +20mA headroom. By your example, you can make double mono with 0.4V reading on TP or single with 0.6V reading on TP. Use dummy that gets as many mA as your load gets at nominal B+. If you got 20mA real load at 200V for instance, use 10K 10W test resistor load. Set R4 first.
Can you define the low ripple roughly, as in the prototyping what were the incoming incoming ripple roughly (per sim like psud) ? I think I remember we discovered that SSHV1 did better with more pre-filtering than initially thought.
As far as the earlier post regarding using this reg with a CCS, I remember reading one of Olsen's presentations where there was a benefit to using a shunt reg before a CCS as far as isolating the music signal from the power supply. But remember you will lose 20V's headroom just inserting the reg, which may in many preamp designs be too much to allow the CCS to do its job. But most likely the 20V's can be gained back via loosening up the pre-filter in an existing design. But we probably need to nail down the max incoming ripple a bit (I imagine is depends on the load and shunt current.)
You mentioned some prefering CLC to CRC. One last question, does the output impedance of the prefiter matter at all?