HV Valve Bench power supply

Discussion in 'The Projects Forum' started by steve8428, Nov 13, 2015.

  1. steve8428

    Thread Starter New Member

    Nov 13, 2015
    11
    0
    Hello all

    I’m hoping to get some help on a project I’m looking at doing. I want to build a bench power supply for my LAB that will provide a variable DC supply at about 100mA from a 240v ac supply from my bench isolation transformer.

    The reason for this supply is to test valve amp projects before building the final circuit. I wanted the supply to be variable from about 10v to the full 300v dc.

    I have already constructed one design attached in image 1 below which for voltage control from 0v to 300v worked very well. However I have found that the output voltage can swing high when switching on as it is unregulated. I tried the circuit powering a heater element at 45vdc and when switching it on the voltage spiked to 55vdc then back down again. This destroyed the valve filament.

    I am looking at the other circuit in image 2 which is a regulated supply. Has anyone built this and is it stable. What sort of output current could I get from this circuit.

    I have an idea to build the first unregulated circuit for the plate voltages as this can be fairly unregulated, and then the second circuit for the heater elements at a lower input voltage about 80v ac.

    Any help would be much appreciated I am a newbie to this I have experience with HV circuitry but not with designing the circuit.
     
  2. ScottWang

    Moderator

    Aug 23, 2012
    4,855
    767
    Does the circuit you modified by yourself or you copied from somewhere?
    Your original circuit may not have a good current limiting function, I changed the circuit, the R5 decide how much current you want and the mosfet will cutoff when the current caused V_R5 to make Vbe of Q2 turn on and the Vce as short-circuit to stop the Id current.

    R1 is the first current limiting resistor, you need to know how much current you want and to set the values of R1 and R5, the VR1 can be change to 500K.
    image 1_steve8428_ScottWang.gif

    LR8N3 only can be output 15 mA, the input should be in series with a limiting resistor as R103.

    image 2_steve8428_ScottWang.gif
     
  3. steve8428

    Thread Starter New Member

    Nov 13, 2015
    11
    0
    Yes both circuits are copied from a web site. I have constructed the first circuit and it works for over load current but not short circuit. When I put a heater filament across the output that is rated high current the output voltage drops to 0V.

    You have removed the resistor to the base of Q2, is this right. Was this not to protect Q2 when it switches on.

    I change this circuit slightly by replacing the two caps at the start with one 47uF 450V cap. and a NTC current limiter of 2.5 ohms removing R1. So not sure what the current control of this circuit is, I would like over 100mA from it before it cuts out.

    Changing the pot to 500K I can give this a try, at the moment the voltage control if pretty good from 300v to around 50v then the control is very sensitive. I was thinking of using two pots in series one for course control and one for fine control.

    Not too sure about the second circuit. I can't find much on this on the web. I know it has low current control but I thought this configuration would be better.

    I have been looking at a different linear voltage regulator from 0-80vdc. I've found a kit on ebay but cannot find a circuit diagram for it. Will do some more digging later. Thank you for the reply.
     
  4. ScottWang

    Moderator

    Aug 23, 2012
    4,855
    767
    For the first circuit, I assuming that you want to set the current limited at 120 mA, so calculate the R5 as:
    Assuming that the Vbe (sat) = 0.7V ... (Ib = 0.5 mA, Ic =5 mA, Vbe =0.7 V)
    R5 = V_R5/I_R5
    R5 = 0.7 V/120 mA
    R5 = 5.83 Ω
    You may using R5 = 5Ω+0.5Ω+0.33Ω
    You could according the formula to set the output current and R5.
    You should recheck how the R5 and Q2 work to protect Q1. (reread the #2)

    The values of R1 can't be too high, otherwise the V_R1 will be drop too much voltage and the output voltage maybe not enough.
     
  5. steve8428

    Thread Starter New Member

    Nov 13, 2015
    11
    0
    Thanks for the reply and the calculations. i will have a better look over this tomorrow.

    What do you mean by reread #2. (sorry for my ignorance)

    Also can you explain what the zenor does in this circuit. i am assuming it protects the MOSFET gate from over voltage
     
  6. ScottWang

    Moderator

    Aug 23, 2012
    4,855
    767
    It explain how the Q2 and R5 work and you could see it with the formula in #4.

    The zener diode was designed to protect the Vgs of Mosfet, the normally, the rating Vgs=± 20V, some maybe reach up to ±30V, so usually we will using a zener diode as 10~15V to cross on the Vgs to avoid when the Vgs too high then the voltage will destroy the mosfet.
     
  7. steve8428

    Thread Starter New Member

    Nov 13, 2015
    11
    0
    I think I've got it. When the current flow through R5 increase the voltage drop across R5 will also increase. When the voltage drop increases to 0.7v the base of Q2 will also be 0.7 volts and so will switch on fully.

    With Q2 switched on the voltage from collector to emitter will flow and drop the voltage at the MOSFET gate. this will turn off the MOSFET protecting the circuit.

    So I can make a very good variable voltage supply this way but not regulated. If I set the output voltage to say 150V when I apply a load the voltage will drop and need readjusting.

    attached is a copy of the actual circuit built with the one input Cap. Ro is a NTC inrush current limiter of 2.5ohms



    Mod edit:
    The circuit as following was copied from #11.
    [​IMG]
     
  8. ScottWang

    Moderator

    Aug 23, 2012
    4,855
    767
    We are not allow to discuss transformerless issue, so you should adding the isolation transformer, otherwise this thread will be locked.
     
  9. steve8428

    Thread Starter New Member

    Nov 13, 2015
    11
    0
    This is just the sim circuit which did not allow me to add the isolation transformer (not a very good sim). i have got a full lab with isolation transformer and Variac. this circuit is connected to the isolation transformer. i am aware of the dangers of working with HV voltages. My job involves working with over 750Vdc supplies.

    Thanks for the tip hope this is acceptable by the invigilators
     
  10. ScottWang

    Moderator

    Aug 23, 2012
    4,855
    767
    I know what you mean, but you still need to modify the circuit after you simulated, let it has isolation transformer.
     
  11. steve8428

    Thread Starter New Member

    Nov 13, 2015
    11
    0
    Ive edited it in paint to state isolation transformer hope this is ok. If not i will have to look at a different way tomorrow
     
  12. ian field

    Distinguished Member

    Oct 27, 2012
    4,415
    784
    If you want to improve the design with closed loop regulation; The TL431 programmable zener can be used in a casc-o-de stage with a high voltage transistor (or MOSFET). The voltage sensing divider for the Vin pin is done in exactly the same way - just a bigger resistor to produce 2.5V at Vin.
    The high voltage casc-o-de transistor is in common base, the emitter (source) is connected to TL431 cathode. The base (gate) is provided with a stabilized bias voltage somewhere within the voltage rating of the TL431. The collector (drain) acts on the gate of the series pass MOSFET to pull it down and regulate the output voltage.

    Incidentally - with a MOSFET series pass; you loose about 8 -10V off how high you can turn the wick up. This is because the gate has to be higher the source by that much to turn it all the way on.

    On regular voltage regulators; it common to add a charge pump voltage doubler to the secondary side rectifier to generate a "bootstrapped" gate supply above the unregulated rail.

    Its a bit tricky at the voltages you're working with - but the high value resistor won't have to be all that big to supply a MOSFET gate. A 15V zener from gate to source would be advisable protection.
     
  13. ScottWang

    Moderator

    Aug 23, 2012
    4,855
    767
    You forgot to remove the R5(1K) and shorted two pins in the new circuit, it will loss protection function.
     
  14. Veracohr

    Well-Known Member

    Jan 3, 2011
    552
    76
    Don't filaments typically use 6.3V?
     
  15. bertus

    Administrator

    Apr 5, 2008
    15,648
    2,348
    Hello,

    @Veracohr
    The filaments of valves can have different voltages or currents.
    This depends on the valves series.

    Bertus
     
  16. ian field

    Distinguished Member

    Oct 27, 2012
    4,415
    784
    In the UK; most TVs had series heater chains running at 300mA - they still had to make up the power by increasing the volt drop for the bigger valves. I believe the US TVs had 600mA heater chains because of the lower voltage headroom.

    In the UK there was also a popular 100mA series for radios and record players.

    Notice that the European Exx-xxx 6.3V series had various heater current ratings depending on the power rating.
     
  17. steve8428

    Thread Starter New Member

    Nov 13, 2015
    11
    0
    As Ian field stated some of the old radio valves have higher heater voltages. i was originally testing a the 45B5 valve from an old radio which has a heater voltage of 45v at 100 mA. Sadly due to my original circuit not having voltage regulation the voltage spiked over 55 v and destroyed the valve heater.
    This power supply is so I can in future build small radio amps from old radio valves. They might not be good but interesting to make. i have a bigger valve project that i am building but i usually have to wait a long time for parts, so this power supply project fills in the gaps.

    Ian thanks for your suggestion i will need to work this out and look up the components you mentioned as ive never used a programmable zener. i also have another design some one has sent to me that i need to look at and test
     
  18. steve8428

    Thread Starter New Member

    Nov 13, 2015
    11
    0
    I have been given a link to the following site

    http://www.homemade-circuits.com/2011/12/variable-0-to-300-volts-regulated-power.html?m=1

    which has a regulator circuit for 0 - 300 VDC. Circuit is shown below. i am not clear on exactly how this circuit works and if it is any good.

    i believe the 50 v zener is to protect the base of BF337 from over voltage
    the 24v Zener is to protect the base of BF458

    All the supply current and voltage is controlled by BF458 and the output voltage is controlled by the 1M ohm resistor. If the output voltage was to rise the resistor network will see this being across the output supply and so the voltage divider to the MOSFET gate will increase. This will then adjust the voltage through the MOSFET and cause the BF337 to conduct more which will then cause BF458 to adjust the output voltage to the original set voltage.

    Have i got this correct and will this work
    upload_2015-11-15_9-45-40.png
     
  19. steve8428

    Thread Starter New Member

    Nov 13, 2015
    11
    0
  20. ian field

    Distinguished Member

    Oct 27, 2012
    4,415
    784
    There were still plenty of old hybrid TVs about when I started out in TV repair, the availability of horizontal power output valves gave me grand ideas about building an amplifier. A transformer for the PL509 40V heater proved to be a bit of a hurdle and I didn't want the inconvenience of dropper resistors for a heater chain. Apparently there was such a thing as an EL509 - heater; 6.3V 2.5A.
     
Loading...