Favorite step-start power supply circuit?

[W1AEX]

My Viking III rig switches on two separate plate supplies, up to 800 vdc for the RF deck, and up to 1500 vdc for the modulator deck, each time I key it. Both supplies are controlled by the same variac. About 1 out of 10 times it catches the primary voltage at just the right time and comes up with a resounding "Thunk" that really can be exciting if you're not expecting it.

I'm looking for a nice simple step-start circuit that I could use to insert a 1/4 to 1/2 second delay when the rig is keyed. On the big rig, I cheated and used a nice Potter Brumfield module that had a calibrated pot to handle this. That rig comes up nice and soft with a distinctive click - click that never "Thunks". I've searched the forum here and found some discussion about this topic, but haven't come across anything definitive, other than some people like them and others prefer to live with the "Thunk". I found a few circuits floating around in various electronics forums, but wonder if anyone has a "favorite" circuit for this function?

Thanks!

Rob

[WBear2GCR]

For repeatability a 555 timer chip to control a larger transistor or two that runs a relay is a good way to go... I'd shield the 555 and probably use an optoisolator to keep the control transistor away from the 555 voltage wise, in case of RF going round-n-round, that and a PCB built copper box for the little circuit...

For longer delays the RC time constant on the base of a transistor works nicely for a not so consistent delay time in the seconds...

               _-_-bear

[WD5JKO]

  I am always amazed at how versatile the 555 timer can be. There is a CMOS  version too, I think it is ICM555.

For more of a kit soft start relay, try this link:

http://www.harbachelectronics.com/main/page_products_other_parts.html

Look for "SS-100 Universal Soft-Start" kit for $27.00

Also don't rule out the solid state relay which I've already described. When done right you will "hit the switch" dead one every time; no more occasional groan.

Jim
WD5JKO

[Opcom]

I like the thunk. It's one of the best sounds a big radio makes. Just build for it! As long as the power transformer is not in danger.. The NCL-2000 leenyar, now that is a grunting box.

but if I step start the HV supply, I use a bridge rectified and lightly filtered DC relay that gets its voltage supply from across the primary of the transformer. The usual same series resistor applies to limit primary current to about 3X the normal TX amount. I wish to always step start filaments, but not there yet. Simple and no pesky IC chips to cook up.

[K4TLJ]

What is wrong with an NTC resistor such as a CL40? I put two (one in each 120V primary) of an SB-200 and stopped the lights dimming when the amp was switched on.

[WD5JKO]

What is wrong with an NTC resistor such as a CL40? I put two (one in each 120V primary) of an SB-200 and stopped the lights dimming when the amp was switched on.

 If sized right for the application, these devices are great for infrequent use every ten minutes or so such as turning on a linear amp. If the application is a big AM setup where we PTT the B+ by switching the AC primary voltage, then the NTC varistor is a bad choice since the current surge reduction will be limited only for the first time we key the transmitter. These devices might take several minutes too cool off when the transmitter is unkeyed.

  These NTC varistors will drop some voltage even when hot. Depending on the application, this drop might matter since the B+ will be lower.

  I looked again at the Harbachelelectronics, and they sell several different step-start relay kits. They look simple in design without any IC chips to fail. The consequence of using these however with fast break-in conversation is that the first 1/2 second or so of the transmission will be at reduced power output.

  So what does the group think of the options:

1.) relay or switch, and just hit it clunk-groan and all

2.) Step-Start with 1/2 second delay at reduced B+

3.) Use a NTC thermistor

4.) Solid State relay (zero crossing type)* so we hit the power every time without the occasional groan

5.) Use full time B+ like a linear and switch the HV with PTT using a vacuum relay

6.) Use full time B+ like a linear and switch RF amplifier with grid blocked keying


* will likely need resistive load across transformer primary to maintain triac of scr conduction over full line cycle. This could be a incandescent lamp (HV ON).


Maybe we can look at the pros and cons of each approach.

73,
Jim
WD5JKO

[AB2EZ]

Here (attached) is the step-start circuit I added to my KW-1 (which incorporates T-R switching of the 240 volt primary of the plate transformer). Note that the KW-1 employs a choke-fed power supply.

I happened to have the following on hand: a suitable 12volt DPST relay; and a 15,000 uF 75 volt capacitor (I use these in my Class E transmitters)

Note the following:

I was concerned about the possibility that the step start relay might fail to close; and that the step-start current-limiting resistor would overheat... and possibly start a fire. Therefore, I attached a thermal fuse (inexpensive) to the 10 ohm resistor with a nylon cable tie. If the resistor gets warmer than 100C, the thermal fuse blows, and opens up the path through the resistor. In my KW-1, I do not have the center tap of the primary of the plate transformer connected to neutral. Therefore, if the thermal fuse opens up, the transformer is disconnected. Naturally, (not shown in the attachments below) there are regular fuses on each of the two hot leads that go to the transformer's primary.

I tested this feature by placing a piece of paper (actually the business card of my former boss) between the relay contacts. After about 2 seconds or so, the thermal fuse opened up.

Also shown below is the primary transformer current vs time, when the 240 volts is applied to the primary of the plate transformer...  as measured with a current probe. Note that after the first 2 cycles of 60Hz AC, the current into the primary of the plate transformer reaches a steady state value. [This is true for my KW-1, it may take longer to reach steady state with other power supply designs]. For this reason, 160 milliseconds was plenty of delay for the step-start circuit.

Stu

[W1VD]

From a different thread - see post by Joe W3GMS:

http://amfone.net/Amforum/index.php?topic=19796.msg140729#msg140729

[WBear2GCR]

The problem with the capacitor charging enough to pull the relay is that it is not reliable for fast switching, fine for turn-it-on/turn-it-off situations. But where you want a 0.5sec delay reliably for keying and un-keying, re-keying, the cap won't do that trick. The improvement on that is to use a smaller cap with a higher impedance charge to saturate the base of a transistor, causing the relay to pull in. That still has the same issue of depleting the cap's charge when fast cycling.

That leaves the solid state relay, with or without a "cycle counter" for a trigger... depends on if you just key the primary or if you are shunting a step start resistor or not...

Or else the brute force methods mentioned...

                  _-_-bear

[W1AEX]

Thanks for all the data in the thread. I'll mess around with the ideas here and get it to come up in a more civilized click - click mode in short order!

:O)

[K5UJ]

I like the thunk.

Me too.  As long as the gear isn't hurting.  One of the coolest things about working a strapping rig is hearing the thunk followed by the slight sound of breezes blowing over the mic if the guy has a fan or something in the shack.

[W1AEX]

One of the coolest things about working a strapping rig is hearing the thunk followed by the slight sound of breezes blowing over the mic if the guy has a fan or something in the shack.

Yup, especially if they are going into a selective fade and you get that sweet sounding "platform shift" effect.

Below are some small mp3 files captured off the air that show the hard start-up sound of the Viking III running 150 watts output, compared to the soft start of my 4-400 running close to a kilowatt back in 1989. The 4-400 would come up with about 200 watts output momentarily, then click up to about 650 watts output. I much prefer the "click - click" to the abrupt "bang!"

[KM1H]

I use a groan and thunk on all the HB KW amps from HF to UHF. I prefer long term reliability to being cute with SS.

The Harbach barely handles the SB-200 size amp and blows the undersized cement resistors on a serious amp.  The design is OK, just needs more robust components and Id never use cement resistors. There is nothing like a good wirewound.

Carl
KM1H

[WA1GFZ]

Put a resistor in series with the HV transformer just low enough to limit inrush to a safe level. Take a SPST relay with a 115 (or 240) VAC Coil. Connect the coil in parallel with the transformer primary. Connect the contacts across the resistor. K.I.S.S. The relay shorts out the resistor when the voltage across the resistor drops low enough to pull in the relay.
The voltage across the resistor drops when the filter caps have charged.