ad: wmr-1

Bauer Power On

Discussion in 'Amplitude Modulation' started by KA4KOE, Apr 21, 2017.

ad: L-HROutlet
ad: l-rl
ad: Left-2
ad: Subscribe
ad: Left-3
ad: abrind-2
ad: L-MFJ
  1. AC0OB

    AC0OB Platinum Subscriber Platinum Subscriber QRZ Page

    No, just the power supply voltage buildup at the right side of the two, 4 uF filter caps, no load.

    With a load across the caps, the power supply voltage buildup time would be longer.

  2. K4KYV

    K4KYV Premium Subscriber Volunteer Moderator QRZ Page

    My homebrew HF-300 rig uses 8-30 henry swinging chokes, 25 mfd of filter capacitance, and 30K ohms 200w bleeder resistors in each of the HV power supplies (separate modulator and final). The HV B+ in both supplies builds up to full value in less than a second.

    The 8005 rig (sort of a homebrew version of the G-K 400) uses a common power supply. 10 hy choke and 50 mfd filter cap. Don't remember the value of the bleeder. HV likewise comes to full in a fraction of a second.

    Gates uses 5/15 hy swinging choke. Modified to 25 mfd filter cap and 50k bleeder (IIRC). HV in that rig is also nearly instant on.

    The 120-ohm DC resistance is pretty typical for the choke, but the 8 mfd filter capacitance seems rather anæmic. No way would that amount of capacitance take 2-3 seconds for the voltage to build up to full value.
  3. AC0OB

    AC0OB Platinum Subscriber Platinum Subscriber QRZ Page

    According to the Bauer Schematic, they are using 6 uF for the 1500 volt system and 8 uF for 3000 volts.

    The continuous time transfer function for the LRC circuit (with R = 120 ohms) is given as:

    G = 1/(8e-05 s^2 + s + 1)

    At 37% of the voltage vs time (~ .350 s) curve, the voltage is 1100V,

    At 75% of the voltage vs time (~ 1.25 s) curve, the voltage is 2250V,

    At 95% of the voltage vs time (~ 2.75 s) curve, the voltage is 2750V,

    At 99% of the voltage vs time (~ 5.5 s) curve, the voltage is 2950V.

  4. N2DTS

    N2DTS Ham Member QRZ Page

    Yes, on the air, needs fine tuning, but the keying (ptt) seems to be working great.
  5. KA4KOE

    KA4KOE Ham Member QRZ Page

    I used Don's 3 relay design. It works great, is simple, and even I can understand how it works.
  6. K4KYV

    K4KYV Premium Subscriber Volunteer Moderator QRZ Page

    I'm not sure that's the complete formula calculating the rise time for a power supply filter. The choke inductance, the choke resistance and the capacitor are all in series, with the capacitor shunted by the load resistance. The time constant of the capacitor and load resistance would have to be part of the equation as well. The resonant frequency of the inductance and capacitance standing alone calculates to 17.8~.

    As the capacitor charges, it pulls current through the choke. But the load also pulls current through the choke, and bleeds off charge from the capacitor.

    I need to go back and review the principles of power supply filters, but I'm certain it wouldn't take that long for that power supply to come to full voltage. In fact, there would be some overshoot, with the voltage instantly rising well above the resting output voltage and then settling back down to the steady state voltage.

    One of the often overlooked problems with choke input filters is dynamic regulation. The power supply for a SSB linear, AM modulator or CW transmitter may show very good regulation on the voltmeter, but looking at the output voltage on a scope, you will see it jumping all over the place with a varying load. The way to make the voltage settle down with minimum overshoot is to use a lot of capacitance in the filter, and a low value of load resistance, including the actual load and the bleeder resistor.

    I had a severe problem with dynamic regulation when I first tried the Gates after the CW conversion. I solved the problem to an acceptable degree by increasing the filter capacitance from 8 mfd to 25. The inductance is around 60 hy when the rig is in cw mode; the mod reactor is switched in series with the regular power supply choke filter choke in cw mode to maintain critical inductance in the choke input filter. With the nominal output voltage set to 1300v, the voltage as seen on the scope was jumping from nearly 2000v down to around 800v before it settled down, with the stock 8 mfd of filter capacitor.
  7. W2VW

    W2VW XML Subscriber QRZ Page

    Went looking for thermistor recycle times and found I had mentioned the wrong part number. What I have is a CL101. Yes, they do not limit current as well when they are hot but I do not see a scenario where that would be much of a problem.

    In practice they have allowed me to reduce fuse current rating on my HV supply.
  8. K4KYV

    K4KYV Premium Subscriber Volunteer Moderator QRZ Page

    Maybe we are confusing terminology. By 'soft start' I meant inserting a s mall amount of resistance in series with the primary to allow the capacitor to partially charge at reduced voltage, then a fraction of a second later, short out the resistor and apply full voltage. This reduces the initial surge, by applying the voltage in two steps. Perhaps 'step start' would have been the better term. I have always thought of step-start and soft-start as the same thing, but maybe they are two distinct terms.
  9. K5UJ

    K5UJ Ham Member QRZ Page

    To me soft start and step start are the same thing. I guess a phone rig could be run with one but I'd want the resistor bypassed after about 1/5th second. I think that's long enough.

    There was a mod. to the power supply in the Viking 500 in ER a few years ago that used a thermistor, and the author said after the mod. the fuses could be put back to the EFJ specified current levels. I don't think the thermistor method is no good; I'm just saying that if an operator wants to operate like CB or 2 m. FM with a lot of back and forth quickie transmissions, the thermistor won't be as effective. But for typical roundtable operation it is probably fine.

    GE Ham News some time in the early '50s had a great article, maybe it was a 2 or 3 part series, on power supply ringing. The conclusion was that you could eliminate it (I think they focused on CW operation) by using around 50 mfd capacitance on the supply output. I kind of like a bit of ringing--it imparts a certain sound to the CW that says "vintage rig here." :cool:
  10. AC0OB

    AC0OB Platinum Subscriber Platinum Subscriber QRZ Page

    I think it is the complete formulation for the voltage rise only across the capacitor (see response to Rob, #521).

    Here is the derivation and if you find a better Laplacian expression let me know and we'll put into MatLab:


    Attached Files:

    Last edited: Feb 16, 2018

Share This Page