Step Start Calculation

[KD6VXI]

In building up VOB,  I've come to ordering step start parts.

My power supply is substantial,  xformer is 4kv at 3A CCS.   Big Dahl.   I haven't measure price or sec R yet,  so if that's necessary,  will be tonight.   

There is approx 27 uF of oil filled across the FWB.   Bridge is K2AW style bricks.  Mercury contactor.

We build a step start to minimize turn on surge and help our components out.   Is there an actual formula anyone has figured out to design around?   Go ogling last night didn't return anything but a bunch of personal designs,  but mostly for way smaller xformers tha  I have.

Thanks,  in advance.

--Shane
KD6VXI

[W2NBC]

http://amfone.net/Amforum/index.php?topic=20055.0

Some good ideas 

[KD6VXI]

That's one Google returned,  but there was nothing that would allow me to calculate R and T values for the xformer and cap used.

I bought a 40 ohm at 100 Watt Dale already,  a timer relay,  etc.   Just wanted to know if their was a formula,  instead of shotgun a value of R,  and then vary T until you get no follow on surge.

This is one of those subjects Stu pops in on,  throws the math out,  and an explanation as well.   Stu,  where art though?   Lol

--Shane
KD6VXI

[w4bfs]

I like and am implementing in my stuff a simple idea that Brett suggested a while back about a simple step start of switch, series r, diode, and small c .... the step start relay has a 120v ac coil ... yes it is operating on dc


       --sw----- r ------ d -------.--------            -----.--- relay contact -------.----
                                           |         |                   |                                 |
                                       relay       |                    |                                |
    120v                              coil        c                   .------- pwr r -----------,
                                           |          |
       -------------------------------------

the sw is 3 position:  off, test, start    test post'n is set up to feed power thru pwr r without powering relay coil for step start function .... this is a good way to be able to test unknown tubes for shorting without destruction from those who are .... this helps define the value of the pwr r and when that is known that also defines the needed step start time regulated by capacitor c ... in addition this is useful for parasitic  testing under controlled conditions

[DMOD]

I always use a 500ms time constant to feed a transistor gate or base that triggers a relay or contactor.

For an RC circuit:

7T = RC.

R then = 7T/C = 3.5/(C).


I don't know how you're going to configure the circuit but if you use an equivalent  series LR circuit after the Power Switch then it becomes

7T = L/R.

R = L/7T; Say your primary winding measures 40H:

R = 11 ohms. Pick a big dude 10 ohm power resistor. The problem wth this arrangement is you get too much AC primary voltage drop.

The better arrangement in my view is shown below as used on my Henry. It uses a relay and a fuse and a resistor. You may have to beef up F1 and R1 for your application. The time constant is about 0.3 seconds.


Phil - AC0OB

[KD6VXI]

Man,  wish I could make a ascii art.   Anywho.

I have a 40 ohm 100 Watt resistor going in series with o e leg of the 240 feed.   This is shorted out by a 3 phase mercury whetted contactor (all 3 legs in parallel to make one big contactor.   Mercury ones have a problem with huge I pulses of FWB C input,  so my solution.   Plus,  I have one).

That is being controlled by one of the octal timer relays with a variable delay from a tenth to ten seconds.

I bought two of the resistors,  I may parallel them to make a 20 ohm 200 Watt, depending on how things go.

Much simpler with sb220 sized stuff.

Thanks for the replies thus far.

--Shane
KD6VXI

[Opcom]

ASCII art is easy with a constant width font like 'courier new'.

Timing: I also prefer a definite time delay. On small stuff the sort of circuit that gradually pulls in the relay as the voltage rises across the transformer primary is OK but that method is mostly a cost-cut.

I have a slightly different opinion about inrush resistors on something this large. Your amp has at least a 12KVA power supply. At 240V that is 50A at full load. 'equivalent' to a 4.8 ohm resistor.
Limiting the inrush to 1/4 to 1/2 that, 12.5 to 25A, seems reasonable and provide a quick charge.

So, a 10 to 25 ohm resistor should be OK. If you got two 40 ohm 100W ones, just parallel them, good to go except the wattage, maybe.

Long term reliability depends on how much of a peak overload the resistor can take time after time without being degraded. That in turn depends on how the resistor is made which varies widely. A 10 Ohm resistor used this way would see peak of 5KW, and about 2800W for the 20 ohm resistor. Even for a fraction of a second that is a big deal. Most resistors have a lot of insulation and if the wire hits 1000-1500 degrees in an instant, the insulation can be harmed.

For these reasons, something like a 120V 600-1000 Watt 'ceramic cone' heating element with open windings is a very reliable choice for inrush limiting on large 240V power supplies.

The cold resistance of a 1KW 120V heater is about 12 Ohms, and a 600 ohm one about 20 ohms.
A 240V heater with the same resistance would be >3KW size, thus the 120v ones are a better choice.
http://www.processheating.com/index.php?page=Coneheater2

There is data here on Nichrome heaters and wire from which those are made:
https://en.wikipedia.org/wiki/Nichrome

None of this is perfect. There are different alloys of Nichrome, but they are similar and it has always worked for me.

[N1BCG]

Is the power supply getting turned on and off with each transmission or is it on for the duration of your operating time?

[W1RKW]

I once had an amp that would kill it's power switch over time. It didn't have a step start so I made one. Took a very simple and inexpensive approach. Didn't want to modify the amplifier so I built something that was external and the amp could plug into. It basically consisted of an electrical quad box, with a duplex outlet and a light bulb. The box held the electronics, relay and the load was a 200w light bulb. Both the box and light bulb were mounted on a small piece of plywood. I picked my timing resistor based on the time the light bulb stopped dimming.  Eventually settled on about a 10sec delay before the relay closed to short out the light bulb. It worked for years and stopped eating switches.  I think when all was said and done, the cost was about $25.

[KD6VXI]

Patrick,  thanks for the explanation....   That's what I was trying to wrap my head around.

Yeah,  it's 12 kva,  ccs.

This amp is cathode keyed.   10K resistor in the CT on stby and a diode drop string for setting ZSAC at 6kv switched in around the resistor on TX.

--Shane
KD6VXI

[Opcom]

It's the fast keying that I have some concern about here, but you said you are using a capacitor input, so you ought be exempt from ringing.

With my unit's LC filter I could get some ringing that may have enough voltage to injure components. I don't do CW but would prefer the amp be able to handle an accomplished fist. The good of the many, and all that.

What are you going to do for TR switching at high speeds? I don't have the TR switch type/speed quite figured out yet.

[KD6VXI]

Patrick,

I don't do CW or VOX.   I have a foot switch and toggle or push button switches.   Boom mounted Mike on the am setup.

The D104  was moved to the ssb desk,  which is where the sb220 will live,  as well. 

I keep making a promise to learn CW.   I've tried,  for years.  Best I've done was Gordon West's Radio Shack tapes,  but never hit anywhere near 5 wpm.  I've contemplated pc based CW,  but seriously...   

So,  no high speed switching here.

--Shane
KD6VXI

[DMOD]

Patrick,  thanks for the explanation....   That's what I was trying to wrap my head around.

Yeah,  it's 12 kva,  ccs.

This amp is cathode keyed.   10K resistor in the CT on stby and a diode drop string for setting ZSAC at 6kv switched in around the resistor on TX.

--Shane
KD6VXI

If the amp is cathode keyed, then why not just let the HV build up at the beginning of each power-turn on with an Inrush Circuit?

Otherwise you would have a surge current at every hot switching, severe contactor arcing, and a delay in RF power buildup on transmit.

The issue seems to be selecting the right component values to handle the turn-on surge current.

Phil - AC0OB

[KD6VXI]

That's what I'm trying to do,  Phil.   HV will be on all the time.

Just trying to figure the maths involved to actually design the step start inrush limiter,  rather than just do what everyone does.

--Shane
KD6VXI

[w3jn]

If you're leaving the HV on all the time, a NTC thermistor in series with the primary legs is the way to go.  Doesn't get any simpler.

[DMOD]

That's what I'm trying to do,  Phil.   HV will be on all the time.

Just trying to figure the maths involved to actually design the step start inrush limiter,  rather than just do what everyone does.

--Shane
KD6VXI

Let's do an example using the following values for the relay inrush circuit below:

Total Filter Capacitance = 27 uF at the appropriate voltage,
Total Bleeder resistance = 100k at turn on (no other load at turn-on),
Inductance of Primary of Power Transformer = 40H (measured).
Transformer secondary current capability = 3 Amps Max.


Time Constant for RC circuit is TC = RXC = 2.7s, which is the point at which 63% of the voltage has risen to its maximum value (4kVX1.4 = 5.6kV. 0.63X5.6kV = ~ 3500 V).

We'll use that same TC value for the Relay Inrush Circuit.

TC = L/R1, (neglecting transformer resistance).

Solving for R1, R1 = 40H/2.7s = ~ 15 ohms.

You may want to use two 30ohm@100W resistors in parallel since heating will occur for > 2.0s.

F1 = 2.5XMax Transformer Sec Current (Rule of Thumb) > = 7.5 Amps SB