Salas:
Given the range of current that this project supports, would it be ok to use the SOT-89 version of the DN2540 for the cascode device? It can handle drain current of 170mA at considerably less power, but is cheaper and takes up less board space. I figured you might have a sim and could validate if it would be acceptable.
Also, no problem using 2SK880 as JFET? I assume GR grade?
Given the range of current that this project supports, would it be ok to use the SOT-89 version of the DN2540 for the cascode device? It can handle drain current of 170mA at considerably less power, but is cheaper and takes up less board space. I figured you might have a sim and could validate if it would be acceptable.
Also, no problem using 2SK880 as JFET? I assume GR grade?
Last edited:
The top one in the cascode will have to dissipate (vin-vout)*Iccs. That means a small sink for the TO-220 version already. Electrically its the same. Could replace the bottom one which only sees the top one's VGS for VDS. 2SK880GR is SMT K117GR essentially so no problem there.
If Vout is say around 260v, can we find a SMT substitute for the 1381 transistors? I could probably get this thing pretty small...
Those have a share of dissipation, (260V/2)*2.2mA each. With that in mind you may look around for analogous SMT.
I have been looking at thermal parameters and its interesting to note how the relatively bad junction-case thermal resistance of the DN2540 comes into play. It's 8.3 versus a max of 1 for the IRF! Even the 10M90S fairs better at 3.1.
I am assuming Tj = 80C, Ta = 30C, Rcs = 0.5, and the two Rjc above. I get 4W capability for the DN and 9.6W for the IRF! This is pretty dramatic!
I wanted to put Q1 and Q3 back to back on the same 3.7C/W heat sink (the 38.1mm version of the heatsink used on the official boards here). My application has 6.5W on Q3 and 1.32W on Q1. Given the above calculations, I think I am ok?
I am assuming Tj = 80C, Ta = 30C, Rcs = 0.5, and the two Rjc above. I get 4W capability for the DN and 9.6W for the IRF! This is pretty dramatic!
I wanted to put Q1 and Q3 back to back on the same 3.7C/W heat sink (the 38.1mm version of the heatsink used on the official boards here). My application has 6.5W on Q3 and 1.32W on Q1. Given the above calculations, I think I am ok?
Using it for labor/testing PSU ?
Hello Salas,
Great work, but I do not read the fully thread.. 🙂
It´s possible to use the SSHV2 as a labor/test PSU?
For that I have to put the Pots to a front plate of an housing and the PSU must be stable without any load. Any drawback about longer wireing ( 10cm ) in this case ?
I need max 300V and 0,12A. But mostly I will only supply one or two small signal tubes like ECC83 or similar. For that a fine tuning of the CSS will be necessary ( 0-20mA ). I have a nice toroidal transformer 250V/120mA and 6,3V 3A for this project.
Regards
Karsten
Hello Salas,
Great work, but I do not read the fully thread.. 🙂
It´s possible to use the SSHV2 as a labor/test PSU?
For that I have to put the Pots to a front plate of an housing and the PSU must be stable without any load. Any drawback about longer wireing ( 10cm ) in this case ?
I need max 300V and 0,12A. But mostly I will only supply one or two small signal tubes like ECC83 or similar. For that a fine tuning of the CSS will be necessary ( 0-20mA ). I have a nice toroidal transformer 250V/120mA and 6,3V 3A for this project.
Regards
Karsten
Hmm, I would be reticent. This shunt supply is hot especially when without load since at least 20mA should be running anyway, although stable, and solid state when designed for simplicity without active protection mechanisms can go out much easier than tubes. You may try it in such a general role, but there is no reliability record for its suitability as yet.
Hi Salas,
Thanks for your assessment about the usage of your SSHV2 as a labor/test PSU. I feared that the SSHV2 is perfect for supply a more or less fixed Load.
Are you know a schematic which can be used for my usage? In general all these good folks here which experiment with valves need a stable and secure HV PSU. For that my hope is to get tips to build my own.
Thanks a lot, an have a nice start in 2015
Regards
Karsten
Thanks for your assessment about the usage of your SSHV2 as a labor/test PSU. I feared that the SSHV2 is perfect for supply a more or less fixed Load.
Are you know a schematic which can be used for my usage? In general all these good folks here which experiment with valves need a stable and secure HV PSU. For that my hope is to get tips to build my own.
Thanks a lot, an have a nice start in 2015
Regards
Karsten
A CCS fed Shunt Regulator consumes a lot of power.
That develops a lot of heat, either in the load and/or in the CCS/Reg.
For a variable supply into a variable load I'd suggest a series type regulator, if you need a regulated supply.
Far easier to use a Variac supplying a suitable step up isolation transformer.
Dial up the voltage you require to test your assembly.
Then build the fixed voltage version that your circuit needs.
I have two outputs from my Variac. to a wall socket fixed to the case/box and to a LV isolation transformer with two sets of 4mm sockets for the dual secondaries. I can power mains fed equipment and/or supply an adjustable Vac+Vac to any circuit, including a plug in module for AC to DC PSU.
That develops a lot of heat, either in the load and/or in the CCS/Reg.
For a variable supply into a variable load I'd suggest a series type regulator, if you need a regulated supply.
Far easier to use a Variac supplying a suitable step up isolation transformer.
Dial up the voltage you require to test your assembly.
Then build the fixed voltage version that your circuit needs.
I have two outputs from my Variac. to a wall socket fixed to the case/box and to a LV isolation transformer with two sets of 4mm sockets for the dual secondaries. I can power mains fed equipment and/or supply an adjustable Vac+Vac to any circuit, including a plug in module for AC to DC PSU.
Last edited:
Hi all, I could use some help with a SSHV2.
I somehow can’t get it to work properly.
I had twice a smoking R3, I have four boards, now using the third (will look at the two damaged boards later).
CCS section seems to be ok, followed the test procedure from post #4479.
Input voltage is 250V and load resistor is 6k5 ohm, aiming for a output voltage of about 220VDC@35mA.
One way to get the board to work is to keep touching the Q1 heatsink (Vout = 220V @ 38mA). The moment I stop doing that the voltage on the TP goes up from 0.55V to 2 to 3V and the output voltage drops to 0V.
Same happens when I keep the multimeter pen on the left TP (connected to R3&R5), on disconnecting the pen from the TP output voltage drops to zero.
Any suggestions are more than welcome.
Regards, Peter
I somehow can’t get it to work properly.
I had twice a smoking R3, I have four boards, now using the third (will look at the two damaged boards later).
CCS section seems to be ok, followed the test procedure from post #4479.
Input voltage is 250V and load resistor is 6k5 ohm, aiming for a output voltage of about 220VDC@35mA.
One way to get the board to work is to keep touching the Q1 heatsink (Vout = 220V @ 38mA). The moment I stop doing that the voltage on the TP goes up from 0.55V to 2 to 3V and the output voltage drops to 0V.
Same happens when I keep the multimeter pen on the left TP (connected to R3&R5), on disconnecting the pen from the TP output voltage drops to zero.
Any suggestions are more than welcome.
Regards, Peter
Q1 drain is at Vin potential. Touching its sink can eventually hurt you. Use 0.1uF HV cap right across the input connector, maybe you got an oscillation in the test. Touching means changing impedance.
Maybe voltage doubler(s) without active parts at all. Like in this.
https://www.youtube.com/watch?v=WLYeUssYMFM
Problem solved
I have Q1 isolated from the heatsink or was it not supposed to be isolated?
Anyhow you were right on the spot, it was an oscillation remedied by a cap on the input, thanks for the help.🙂
I attached scope pictures of input (blue) and output (yellow) of the SSHV2 does it look ok to you?
Regards, Peter
I have Q1 isolated from the heatsink or was it not supposed to be isolated?
Anyhow you were right on the spot, it was an oscillation remedied by a cap on the input, thanks for the help.🙂
I attached scope pictures of input (blue) and output (yellow) of the SSHV2 does it look ok to you?
Regards, Peter
Should be isolated from the small sink for danger reasons, not electrical.
There is no oscillation on those pictures but interference. Use the 20MHZ BW Limit option in CH menu. B in bold will appear next to the channels vertical sensitivity indication. Turn the reg off at a point while the probe (X1) is still attached. So to see what's there on the screen additionally or not. Sweep through slower timebase too.
There is no oscillation on those pictures but interference. Use the 20MHZ BW Limit option in CH menu. B in bold will appear next to the channels vertical sensitivity indication. Turn the reg off at a point while the probe (X1) is still attached. So to see what's there on the screen additionally or not. Sweep through slower timebase too.
Salas it was indeed mostly interference due to the lack of a proper ground connection on the 1st floor of our house. I had been thinking to make a proper ground on the workbench for some time already so I finally did that and the measurements are a lot cleaner (and the workbench safer).
Again blue = SSHV2 raw input and yellow = SSHV2 output (vert. 5mV/div iso 20mV/div of previous post).
Again blue = SSHV2 raw input and yellow = SSHV2 output (vert. 5mV/div iso 20mV/div of previous post).
Last edited: