3 things I am wondering about:
1. breakdown voltage of safety choke at 10 KW under non-1:1 SWR condition... no risk of flash-over?
2. current rating of the plate blocking caps.... is there an RF current spec on these?
3. Pi-L inductors seem thin, and not plated?
Nice progress!
73, Ed
Safety choke is bonded to chassis at cold end. What I do on my own hb HF amps, (and modified commercial amps) is I install a hb adjustable spark gap, between vac load cap..and chassis. And a 2nd adjustable spark gap between vac tune cap..and chassis. Back in the 70's, I had new cardwell air variable load caps arc and weld the plates. A bitch to repair. That issue was caused by the slow (collins) 3PDT mech TR relay used, and drive applied a split second before the output contacts closed. With no ant on the output for a split second, it's full power into a wide open ( = infinite high swr)..and peak V is through the roof.
This 6M amp will use external high swr fast shut down, with programable SWR trip limits..... but that typ works good for when the swr threshold is just exceeded. ( typ thresholds are 1.5:1 2:1 2.5:1 3:1 ). For a wide open they are not fast enough... hence the adjustable spark gap. Gap set to arc at well below the max peak rating of the vac load cap..... and above the pre-set threshold. IE: threshold set for say 2:1 and adjustable gap set to eq of a 3:1 or 4:1 swr.
The adjustable gap across the vac tune cap will protect the vac tune cap..and again set above the normally calculated peak RF voltage across the tune cap...but well below the max peak V rating of the vac tune cap. The adj gap across the tune cap will also protect the bandswitch on any multiband amplifier. Both caps are tested on a hipotronics hi-pot tester 1st..before install. If they don't meet our specs, they don't get installed, period.
The adjustable spark gap is dead simple to make. I use solid brass screws, like 8-32, 10-32, 1/4-20. ( cdn tire has em up to 3-4" long). Solid brass lock nuts used to adjust the gap, then locked in..sorta like vibroplex paddles...but far more rigid. I use nuts on both sides of the alum plate, + brass flatwashers etc. Sometimes we use thick alum plate, and drill + tap a thread as well as locknuts.
At home depot, they have solid brass acorn nuts. I screw a pair onto the inboard ends of the brass screws..so they face each other. Perfect shape, being domed. (without em, bare threaded end of machine screws are too sharp imo..and inconsistent). Then set the gap with feeler gauges, then hi-pot tested for final tweak (not really required). The idea here is..IF they arc across, and domed ends of the acorn nuts trashed, they are easily replaced.
Vac caps, air caps, bandswitches are too expensive and a bandswitch is a pita to replace. The adj spark gap is dirt cheap, and..'easy' to implement. I also installed one across my 4H dahl HV choke in a C-L-C filter for a HV / B+ experiment. I installed a ceramic spst vac relay + spark gap across the 127 lb hypersil choke. I can shunt the choke on the fly, and turn it into a big C filter. Choke is installed in the B+....and insulated from chassis vis 4 x huge 'glastic' insulators. Under certain software simulated fault conditions, the V can be extremely high across a choke.... hence the cheap insurance by isolating it from the chassis.... + spark gap. Harris, et all, use adj gaps across chokes..so I just re-invented the wheel.
The plate block caps are HEC (high energy corp) 200pf, HT-57 series. The 200 pf cap has the highest CCS RF current rating of the entire HT-57 series.... 15 amps CCS on any freq > 5.11 mhz. 4 are good for 60 amps CCS. Calculated RF current through the block assy is aprx 48 amps. 200 pf HT-57 caps are also used for the bypass caps at cold end of the 30 uh plate choke.
The tube C makes up most of the required tune C... typ on upper HF..and esp on 6M. Plate block current is a function of tube C, loaded B+, and freq. On lower freqs, there is hardly any RF current through the block caps. But on lower freqs, we want the XC of the block assy to be no more than 5% of the plate load Z. On 80-10m amps, same 4 x 200pf HT-57's are used. On 160m-10m, 8 of em are used (2 x rows of 4, or 4 rows of 2)..and sometimes 9 are used ( 3 rows of 3).
On the real high powered stuff, the even bigger HT-59's are used. Then the 250 pf version is used, since it's the highest rated cap in that series...at 35.5 amps CCS on any freq > 16 mhz. Typ 4 of em used...on 80-6m. When more than one block cap used, it has to be strapped carefully, such that the current divides equally between em. Short of it, higher C required to obtain lower XC on the lowest freq..and high current handling required on the highest freq.
Myself, I would have silver plated anything that is copper..or brass. It's an aesthetic thing..as I can't stand the look of plane copper. I use the 'cool-amp' goop from ORE.
https://www.cool-amp.com/I think the strap is .032" thick. Thickness is a non issue on 6m. KM1H still gets 1500 w out of his HB 144 mhz amp..and that uses 1.125" copper tubing in a stripline config. The tubing has turned black, like cold water pipes in a house from 1947. As long as the connections at each end are good, it will last.
It's actually a 'L-PI' network..with an inductor in front of the main PI network. (It can be inserted between anode and block caps..or between block caps + C1 rune cap). The grid to chassis C (43 pf) and the 1st inductor form a step down L network....which steps the plate load Z way down...low enough that a practical PI-net can be designed, with a normal Q. Without the 1st inductor, the loaded Q would be sky high..resulting in sky high current through the main inductor. This trick only works on tubes that don't require a parasitic suppressor. Sri for the diatribe.
Jim VE7RF
Author
Topic: 10 kw CCS on 6M..using the 3CX-6000A7...Part 12 (Read 125 times)