Comments on KW-1 fuse-blowing issues

[AB2EZ]

I realize that not many people who read this board have a KW-1... but I can pass along some suggestions that may also be applicable to other transmitters

My KW-1 does not have the original Collins plate transformer. It has a plate transformer of the type that is also used in a Johnson Desk Kilowatt transmitter. Although I would have preferred to have the original Collins transformer... a nice thing about the transformer I do have is that it has both: 2900 VAC outputs (either side of CT) and 2350 VAC outputs. I moved two wires to allow me to use the 2350 VAC outputs. This resulted in a reduction of the HV DC from well over 2500 Volts to around 2100 Volts (with a 240VAC line voltage). I adjusted the modulator (a pair of 810's) bias accodingly. I load the transmitter to 300mA of plate current (instead of the normal 400mA). This produces 630W of input power and 375W of output at carrier. All in all... I think the reduced stress on the transmitter is a good idea.

With respect to the fuses:

I had a fuse-blowing problem with the fuses in the primary of the plate tranformer. I measured the primary current that flows when the transmitter is switched from standby to transmit. With the reduced secondary voltage (2350VAC), the primary current peaks at 10A for one half cycle of 60Hz, and then settles down to around 4A (even with a choke input plate supply). I added a step start circuit, with a time constant of around 0.25 seconds. Since then, I have been using 10A fuses, with no fuse-blowing problems.

Stu

[w3jn]

A NTC thermistor in the primary helps also.  I put one in line on my GPT-750 and there's no more intermittent "kachunk" from the transformer on keyup.

[AB2EZ]

Using NTC thermistors seems like a simple, inexpensive, modern solution. Obviously, they are widely used in such things as TV monitors, high power audio amplifiers, etc.

Is there a "dark side" to the use of NTC thermistors instead of a step-start relay?

Stu

[W2PFY]

.

[w3jn]

The dark side is that with quick break-in the thermistor may not have time to cool down sufficiently.  Still, MUCH simpler and easier than a step-start.

[W3GMS]

Stu,

I posted a thread awhile back on inrush circuits on our switching supplies we designed for our Enterprise Servers at Unisys.  Like John said, thermistors are a very inexpensive solution but you must give them enough time to cool down otherwise the inrush limit is severely compromised.  A good solution places a shorting device across the thermistor 50 or so milliseconds after AC is applied.  That way the thermistor does not get hot and you always will have inrush protection even when doing rapid style break in like Ralph, W3GL  .  Thermistors are rated for various levels of AC current so you can pick the one that fits the bill.  Under high inrush conditions, normal wirewound in many cases do not have a high enough fusing current rating and poof they open.  Thermistors properly selected don't do that and they are cheap.  In some cases I used Triac's to short out the thermistor after the inrush occurred.  In other cases I used a relay depending on the application.  I used a zero cross opto isolator to drive the gate of the triac and it worked very well.  The disadvantage of the triac solution is they can produce some broadband noise.  In the switching power supply business we used a balanced twin section line filter and that took care of the conducted noise issue with the triac.  Its wasn't a big deal to add the filter since the filter was required anyway to get pass conductive FCC criteria.  If you use a contactor to break the primary voltage on receive then you really don't have to worry about any noise from the triac. 
Some folks mistakenly think you don't need inrush protection with a choke input type filter and this is not necessarily so.  Upon application of DC to the filter capacitor it looks very close to a short which will cause the inductor to saturate.  Initially the peak current is limited by all the DC resistances in the path along with the AC line impedance and the characteristics of the transformer.   

Regards,
Joe, W3GMS   

[AB2EZ]

Joe

An interesting suggestion. However, when adding the bypass relay, you end up with all of the components of a step-start circuit. So the choice between a thermistor and a wirewound resistor boils down to the issue of whether the thermistor is more robust than a wirewound resistor.

As an aside...

I was worried about what would happen if the bypass relay didn't close properly. What I did was to place a 100C thermal fuse in mechanical contact with the step-start wirewound resistor (using a nylon cable tie)... and I placed the thermal fuse in series with the main contactor's control circuit. If the bypass relay doesn't close, the step start resistor heats up... then the thermal fuse blows, and the main contactor opens.

To test this, I placed a business card between the contacts of the bypass relay, and pushed the transmit switch. In about 1 second, the thermal fuse blew... and the transmitter shut down. The thermal fuses are less than $1.00 each.

Stu

[W3GMS]

Stu,
I am sure if you size the wire correctly on the wirewound you will be ok but the Thermistor has every advantage I can think of over a ww in that application.  If the relay does fail, the thermistor would go to it's minimum resistance and you would have nothing melt down so to speak and still be on the air!   If you look at the plate voltage meter during key up you should see the rise in voltage when the relay kicks in.  If you see a slow change in plate voltage indicative of the relay failing, you would observe a more gradual change in final plate voltage as the thermistor heats up.  An open wire wound resistor would leave you with no inrush protection only a slight delay in TX voltage being applied which may be more difficult to detect.  I think Digi Key sells the thermistors design for inrush limit applications. 
Have fun and I am sure whatever solution you come up with will be a good one.
Regards,
Joe, W3GMS
 

[Todd, KA1KAQ]

WA2PJP has designed a little step-start for his KW-1, Stu. Not sure how he did it, but perhaps he'll check in with a run down. I know it works well and I'll b adding something similar here soon. These old rigs don't much care for quick-draw break-in, especially with any amount of voltage involved. Found this out the hard way, unfortunately.

[KB2WIG]

" The thermal fuses are less than $1.00 each. "

They can be found at the roadside; look for a coffee maker. They are used in line with the heater elements.  Most elements to a full size coffee maker are rated for at least 10A.


When I smoked, I would pick up the odd bic lighter off the ground and rip the flint out of it fer my Zippo.  I saved about $0.45 if I found 10 or so. Just another money $aving tip We have here.

klc

[Steve - WB3HUZ]

With a few hundred thousand lighters and you'll be rich!