Modulation Trans Protection

[w9jsw]

If I use a separate PS for the modulator in an 813 rig (pair modded by a pair), is it sufficient to fuse the primary CT to protect the mod trans in the event the plate PS fails? Or should I use a more active protection scheme like a current sense circuit that trips a relay if the mod cathode current gets too high?

If a fuse is adequate, I currently use the K1JJ fuse approach using 1-2 strands of RG213 braid between posts on the main PS. Same on the mod trans CT?

I am moving my 813's rig BC iron to a 4-400/3-500 rig. Looking at the stash, I have a pair of plate transformers with chokes that are separately capable of 300ma. One will do 1800V and the other will do 1600V. So higher for plate and lower for modulator would be the plan.

Should make a nice 350W backup rig.

John

[K8DI]

If I use a separate PS for the modulator in an 813 rig (pair modded by a pair), is it sufficient to fuse the primary CT to protect the mod trans in the event the plate PS fails? Or should I use a more active protection scheme like a current sense circuit that trips a relay if the mod cathode current gets too high?

If a fuse is adequate, I currently use the K1JJ fuse approach using 1-2 strands of RG213 braid between posts on the main PS. Same on the mod trans CT?

I am moving my 813's rig BC iron to a 4-400/3-500 rig. Looking at the stash, I have a pair of plate transformers with chokes that are separately capable of 300ma. One will do 1800V and the other will do 1600V. So higher for plate and lower for modulator would be the plan.

Should make a nice 350W backup rig.

John

I'm not under the impression that excessive modulator current will necessarily be the first issue if you lose the RF plate supply.  Without that supply, the effective load on the mod transformer greatly increases, and you'll see high voltage spikes/arcing/insulation damage, and possibly oscillation, which may or may not cause an overcurrent.  Spark gaps (which you already have, right?) should protect the mod transformer. For an extra layer, what about an ordinary relay, whose NO contacts are in the mod PS's ground connection, with a series resistor, just like the PTT (in series with it, in fact). Drive the coil through a big resistor from the RF B+ supply...if the RF supply dies, the mod supply gets disconnected. You could use the second set of contacts as a lockout that would prevent re-closing without some deliberate reset action, too.

Ed

[K1JJ]

Hi John,

"a nice 350W backup rig"   -    a nice backup rig indeed!  


A combination of the techniques you described has worked well for me.

I especially like using the 1-2 fine strands on ceramic pillars of  RG-213 coax shield in series with the HV lead of the supply. I have saved my diode rectifiers many times when the 240V breaker was not fast or sensitive enough to trip first.  

The time tested 5-25 ohm resistor in series with the HV lead is a good idea too.  Also, I use PTT step starts on ALL supplies here to ease the strain on PS parts.

In addition, I have built up a 4-1000A Hall effect trip system. The DC plate circuit, grid and screens are all monitored for excessive DC current. The PTT is disabled during a current fault. This is a slower response, but is plenty fast for tubes which can usually handle a slow response... unlike SS.

I use spark gaps  on my mod xmfr, mod reactor and screen mod choke. I have never used the fine wire idea for the mod iron, but why not?

These spark gaps have load snubbers with vac relays during unkey across them to absorb any collapsing magnetic fields. No sparking.

And of course, a six pole sequencer using  both attack and release variable timing is used to find the smoothest key up and unkey performance.

T

[DMOD]

If I use a separate PS for the modulator in an 813 rig (pair modded by a pair), is it sufficient to fuse the primary CT to protect the mod trans in the event the plate PS fails? Or should I use a more active protection scheme like a current sense circuit that trips a relay if the mod cathode current gets too high?

If a fuse is adequate, I currently use the K1JJ fuse approach using 1-2 strands of RG213 braid between posts on the main PS. Same on the mod trans CT?

I am moving my 813's rig BC iron to a 4-400/3-500 rig. Looking at the stash, I have a pair of plate transformers with chokes that are separately capable of 300ma. One will do 1800V and the other will do 1600V. So higher for plate and lower for modulator would be the plan.

Should make a nice 350W backup rig.

John

Are you looking for Overvoltage protection as well as Over Current Protection or both?

For OCP I prefer the "fine wire" fuse to the Primary's CT for speed of protection.

For reverse voltage protection on the secondary, I still prefer the Zener string and resistive loading.

Phil AC0OB

[w9jsw]

Thanks for the suggestions, guys!

Spark gaps - yes, on primary and secondary plus mod reactor. On this new PS transplant, I will have on the mod secondary. I read that spark gaps on the mod primary are not necessary - is that the consensus here?

20-ohm glitch in the B+ for both. Fine wire fuse in the B+ for both. Fuse before and after the caps in the mod PS.

I don't plan to start out with a mod reactor unless one comes along. Mod trans will be a Thordarson 11m78 that is rated for current on the secondary. 

I have a snubber that clamps across the mod secondary when PTT is dropped. It is a NC vac relay. Another K1JJ classic design (tm)

I also have a negative peak limiter on the B+.

Belt and suspenders?

John

[W3SLK]

I didn't design or build it but the HN-500 uses a one-shot latching relay in the HV section of the modulator. It and in this case the rheostat are on ceramic stand-offs in series with the supply. The rheostat is used to set the trip current. I know this worked well before and I thought it activated when I lost the supply to the RF deck, (that is a story in and of itself). Protect your investments well. Some of this iron I am learning can't be found any longer!

[K1JJ]

Thanks for the suggestions, guys!

Spark gaps - yes, on primary and secondary plus mod reactor. On this new PS transplant, I will have on the mod secondary. I read that spark gaps on the mod primary are not necessary - is that the consensus here?

20-ohm glitch in the B+ for both. Fine wire fuse in the B+ for both. Fuse before and after the caps in the mod PS.

I don't plan to start out with a mod reactor unless one comes along. Mod trans will be a Thordarson 11m78 that is rated for current on the secondary.  

I have a snubber that clamps across the mod secondary when PTT is dropped. It is a NC vac relay. Another K1JJ classic design (tm)

I also have a negative peak limiter on the B+.

Belt and suspenders?

John

Hi John,

Yes, the diode/power resistor NPL on the B+ line is good protection for the mod iron in case of -100% negative modulation. As you know, it will cause splatter if leaned on to increase your audio. With an indication diode, I use it as a safety net and guide when I am nearing -100% modulation... and try to let it flash only occasionally.

The mod iron snubber you described was probably the main reason I was able to get rid of unkey sparking on the 4-1000A mod iron spark gaps. Sequencing helped, but when I was finally able to dissipate the energy across the spark gap during unkey, the energy was quenched and the severe gap sparking problem went away entirely. I'll bet the snubber alone without sequencing may have solved the problem... but I still use both.

As for spark gaps on BOTH the pri and sec of a mod xfmr...  Years ago I acquired a 10KW 800 pound RCA mod transformer gorilla and matching reactor. (I since sold it)  It was a beast and professionally built for BC use.  It had large 1" ball spark gaps on the pri, sec and on the reactor. Three sets.   The Tron told me it is a good idea to cover all bases.  I don't have a technical reason why spark gaps are not required on the primary.  But maybe gaps on the primary will keep the energy from going thru the pri to sec by quenching it at the pri...  up front.  This assumes the energy goes from pri to sec and not the other way... :-)

And yes, as Pat said, the fine wire in series with the HV lead can be FAST at lower voltages.     BUT sometimes with big supplies the fuse connection can slow down and be sustained when the arc floats slowly upwards above the pillars. I've seen it last a full 2-3 seconds before quenching.  Make the pillars at least 2" apart for BIG supplies.  And l keep the surrounding area clear of other components for at least +- 4" to be sure there are no collateral damage fires.  A commercial HV fuse (in a vacuum maybe?) would be a better idea, but these HV fuses are expensive.

Check it out:
https://www.google.com/search?q=high+voltage+fuse&rlz=1C1GCEA_enUS908US908&oq=high+voltage+fuse&aqs=chrome..69i57j0i512l9.4098j0j15&sourceid=chrome&ie=UTF-8


T

[KB2WIG]

"  And l keep the surrounding area clear of other components for at least +- 4" to be sure "

How about something like a coffee can shield?  I'm assuming that the metal would contain the plasma/shrapnel...... i was thinking of a test tube, with the JJ fuse inside, but then I started to think this would actually help retain the plasma path. May be some metal'fabric cloth' for a shield.

klc

[w9jsw]

Fortunately I will be running 1800V and 1500V supplies. That and transformers that can only provide ~300ma should temper the enthusiasm.

I bet your 4KV supply draws quite an arc!

Sounds like a plan.

Anyone have a schematic for a HN-500? Can't seem to find one. Who made it? Interested in examining the latching relay approach.

John

Kevin, you on the air yet?

[W3SLK]

John, the HN-500 is my HB 813sX 810's. The power supply has the rheostat and latching relay directly inline with the B+ to the modulators (810's) at 1500VDC if memory serves me. Anyway the relay drops the B+ out at overmodulation peaks (~450mA), and a button on the front panel resets it. The relay and rheostat are isolated from ground so in essence when you overmodulate, it creates a peak voltage that activates the relay and kills the B+. I drew it out a long time ago. I would have to pull it apart. Maybe I can look at it this weekend.

[w9jsw]

Don't bother. I get the idea.