Ameritron soft start: How do I stop the arcing?

[357]

Thanks for the reply Nomad,

Here is the schematic. I think its too soon too but it did it from day 1.
Ameritron doesnt update its schematics so its some guess work.
There are 2 2200uf caps on that meter board that the relay runs off of.
So that yellow wire goes to the meter board, they show a bridge rectifier that in real life is 2 diodes and C2 is not 1000uf but 2200uf and they added another 2200uf that isnt shown.

The problem I'm having besides that is the HV is sagging to 3100v which is robbing me of alot of power ATM.
Ameritron isnt returning emails either so I have no clue. I read of people replacing wires with thicker ones but thats just getting ridiculous.
I also read people complain about the arc and the answer was normal operation.
If the relays wer $10 still and drawers full of then that'd be fine I guess but when you have to scour the earth to find one, its disappointing. Not as disappointing as when this tube gets soft mind you.

Maybe I should increase those 2200 to 4700

 

[357]

Time to give it the beans....

 

[357]

Less sag, less grid and proper bias now 150ma. 3850v to 3600 at 1500 cw carrier, I can live with that...65w SSB peaking 1700W
The volt meter must be off because I got 241vac going in at key down. 244 at idle

 

[nomadradio]

That schematic is wrong. Shows a diode in parallel with the coil on the 'run' relay RLY2. Not allowed for a relay with a 120 Volt AC coil.

Weird.

73

 

[357]

That schematic is wrong. Shows a diode in parallel with the coil on the 'run' relay RLY2. Not allowed for a relay with a 120 Volt AC coil.

Weird.

73

I noticed that too.

 

[357]

I'm at a loss on how to time the relay longer

 

[linearone]

I'm at a loss on how to time the relay longer

Why do you need to? Try 25 ohms, try 30.

 

[nomadradio]

Wish I knew how to track the power rating the resistor will need to be against the resistance value. Higher resistance, longer delay. More energy pumped into the resistor. A 30 ohm would need to be bigger, but how much? I would guess around double. The question here is energy, power times duration. Not sure how to calculate that for this circuit.

73

 

[linearone]

Wish I knew how to track the power rating the resistor will need to be against the resistance value. Higher resistance, longer delay. More energy pumped into the resistor. A 30 ohm would need to be bigger, but how much? I would guess around double. The question here is energy, power times duration. Not sure how to calculate that for this circuit.

73

Yes, but you don't have to think about it that hard, the resistor is really only on for a second or two, just get one of those aluminum heat sink finned ones and get yourself a 50 w and you'll be fine

 

[nomadradio]

When overkill is cheap, then why not?

The aluminum-body wirewound chinesium power resistors are a common fleabay item, if you don't mind waiting for it to arrive. We have had good luck with those, only about one or two duds per couple hundred resistors, more or less. The traditional hollow-core ceramic-tube wirewould monster resistors are getting expensive. The aluminum-body types are way cheaper.

73

 

[357]

I decided to just clean the switches. The power and cw/ssb switches were flaky, if I knocked on the panel near the power switch the lights would flicker a bit and the SSB switch was giving different idle currents I had to tap on it to get ssb to stick.
I used a butter knife and a razor to dig the rocker out, polished the contacts with 2000 grit, cleaned them with alcohol and stretched the springs a bit.
Now she's rock solid.

 

[357]

that relay barely arcs now, theres a small flash but really weak. That old relay must have been just worn out

 

[nomadradio]

Congratulations! Gotta love it when a plan comes together. Relays always wear out eventually. Tricky part is predicting when.

73

 

[brandon7861]

Wish I knew how to track the power rating the resistor will need to be against the resistance value. Higher resistance, longer delay. More energy pumped into the resistor. A 30 ohm would need to be bigger, but how much? I would guess around double. The question here is energy, power times duration. Not sure how to calculate that for this circuit.

73

The energy through the resistor needed to charge the cap does not change with resistance. The energy required to charge the capacitor is a function of its capacitance and voltage, not the resistance the current must pass through to charge it. 0.5CV^2=Watt-seconds. Making the resistor larger only increases the time the resiostor has to dissipate the heat.

You probably do not want to know what that instantaneous power is right at first, its mind boggling. The peak power in a 20Ω resistor charging a .001F cap is 800kW! The good news is that its only that high a tiny fraction of a second.

 

[brandon7861]

I'm at a loss on how to time the relay longer

Full charge (well, thats impossible, but close enough) happens at 5 time constants, ie 5*R*C. If you have a .001F cap and want it to take 0.2 seconds to charge, then 0.2/(5*.001)=40Ω