Print

[VR2AX]

The maximum allowable specified current through a single capacitor will be frequency dependent: usually the max specified current will be lower on 160 than on 80, and considerably less than that allowed on 20.

The actual RF current will depend on the load presented at the tuner output to the feed line. Generally, on 160 and possibly also on 80, several lower (equal) value capacitors in parallel will handle the required current better than a single value unit of equal total capacitance.

[VE7RF]

I buy the HEC brand (high energy corp) brand  170 pf ceramic doorknob caps from MFJ, they have the cheapest prices.
They are in their catalog.  They are N750. (7.5 kv rated..and good for  4 amps CCS on 160m...( each cap).  They are used in the Ameritron  AL-82/1200/1500 amplifiers to PAD.... BOTH the tune AND load caps.

3 x 170 pf = 510 pf.    Machine screw em to the chassis.  Use 1/2" to 1" wide copper strap and connect the tops of all 3 of em..then run the strap off to where ever it needs to go.   Ok, now you have heat sinked both ends of each cap.
It won't drift.  End of problem.

The reason 170 pf caps are used as padders on tuners and amplifiers, is that 170 pf is the highest value you can get..in a N750 cap.    Next size up is the 200 pf cap, and that 200 pf cap has a TC of N3300.    The 200 pf cap will handle 5 amps CCS on 160m..which is 1 amp more than the 4 amp rating of the 170 pf cap...BUT the 200 pf cap has a much lousier TC.

[VE7RF]

I buy the HEC brand (high energy corp) brand  170 pf ceramic doorknob caps from MFJ, they have the cheapest prices.
They are in their catalog.  They are N750. (7.5 kv rated..and good for  4 amps CCS on 160m...(each cap).  They are used in the Ameritron  AL-82/1200/1500 amplifiers to PAD.... BOTH the tune AND load caps.

3 x 170 pf = 510 pf.    Machine screw em to the chassis.  Use 1/2" to 1" wide copper strap and connect the tops of all 3 of em..then run the strap off to where ever it needs to go.   Ok, now you have heat sinked both ends of each cap.
It won't drift.  End of problem.

The reason 170 pf caps are used as padders on tuners and amplifiers, is that 170 pf is the highest value you can get..in a N750 cap.    Next size up is the 200 pf cap, and that 200 pf cap has a TC of N3300.    The 200 pf cap will handle 5 amps CCS on 160m..which is 1 amp more than the 4 amp rating of the 170 pf cap...BUT the 200 pf cap has a much lousier TC.

[NO9E]

I have three data points here.

Mfj998 is an L tuner designed for legal limit. I used it for a few years, including 80m ladder line fed dipole for 160m. No damage of any high power components. The caps are small and apparently pretty good. Mfj may have extras.

I needed a cap for shunt feeding a 100ft tower on 160m. A few button caps overheated. I bought one inch cap and it works well.

For matching a 160m inv L I tried a few caps. A barrel type - hollow tube one inch long and 1/8 inch diameter never failed.

Ignacy NO9E

[W9IQ]

Yes.  The conventional "wisdom" seems to be to use a 1:1 with a multiband antenna, but I was never able to achieve a match across all bands with the 1:1 and my ATR-30 tuner.  I used to use this antenna on 160 through 10, but these days just on 160, 80, 60, 40, 30.

Here are my calculations:

Frequency = 1.825 MHz
Load = 321+j255
Inductor Q = 100
Capacitor Q = 1000
Inductance = 13.21 µH in series
Capacitance = 654.6 pF shunting the load
Tuner Efficiency ≈ 96.6% (-0.15 dB)
Input Power = 1500 watts
Total Power Dissipated ≈ 51 watts
Inductor Dissipation ≈ 45.6 watts
Total Capacitor Dissipation ≈ 5.4 watts

The above dissipation figures are based on a 1500 watt carrier. Most transmission modes have a duty cycle of <50% so the average dissipation will be proportionally less.

Since the discussion revolves around the heating of the paralleled 470 pF doorknob cap, with a Q of 1000 it has a current of ~4.7 amps RMS and a dissipation of ~4 watts.

Given the high heat dissipation of the inductor, it may be possible that it contributes to the drift of the doorknob capacitor value if there is not adequate separation and cooling.

The overall dissipation could be reduced by presenting a more favorable load impedance. This possibly be done by using a hybrid 1:9 balun although the effect on other bands is not known at this time. The other possibility would be to use a fixed pre-tune circuit that is switched in for 160 meter operation. This could potentially reduce the overall power loss in the capacitor.

- Glenn W9IQ

[VR2AX]

Also remember that 470 + 180 in parallel means that 470/650 of the current passes through the fixed and 180/650 passes through the variable

[K8AC]

Mfj998 is an L tuner designed for legal limit. I used it for a few years, including 80m ladder line fed dipole for 160m. No damage of any high power components. The caps are small and apparently pretty good. Mfj may have extras.

Ignacy NO9E

I had an MFJ998 for a short time years ago.  Melted a relay when using this antenna on 160M.  VE7RF mentioned MFJ as a source for 170 pF doorknob caps and I was able to locate those in the MFJ catalog.  You can't find them on the website, though.   

[W1VT]

MFJ used to have parts on the website but that went  away with one of their web site major revisions.

[W9IQ]

Also remember that 470 + 180 in parallel means that 470/650 of the current passes through the fixed and 180/650 passes through the variable

Very true, but that was already taken care of in my calculations. The total current through the ~655 pF shunt capacitance is ~6.5 amps.

- Glenn W9IQ

[K8AC]

Here are my calculations:

Frequency = 1.825 MHz
Load = 321+j255
Inductor Q = 100
Capacitor Q = 1000
Inductance = 13.21 µH in series
Capacitance = 654.6 pF shunting the load
Tuner Efficiency ≈ 96.6% (-0.15 dB)
Input Power = 1500 watts
Total Power Dissipated ≈ 51 watts
Inductor Dissipation ≈ 45.6 watts
Total Capacitor Dissipation ≈ 5.4 watts

The above dissipation figures are based on a 1500 watt carrier. Most transmission modes have a duty cycle of <50% so the average dissipation will be proportionally less.

Since the discussion revolves around the heating of the paralleled 470 pF doorknob cap, with a Q of 1000 it has a current of ~4.7 amps RMS and a dissipation of ~4 watts.

Given the high heat dissipation of the inductor, it may be possible that it contributes to the drift of the doorknob capacitor value if there is not adequate separation and cooling.

The overall dissipation could be reduced by presenting a more favorable load impedance. This possibly be done by using a hybrid 1:9 balun although the effect on other bands is not known at this time. The other possibility would be to use a fixed pre-tune circuit that is switched in for 160 meter operation. This could potentially reduce the overall power loss in the capacitor.

- Glenn W9IQ

Interesting info - thanks, Glenn.  Not something I could figure out for myself.  I have loads of room in the tuner cabinet and while it is vented, I have plenty of room to install a low volume muffin fan if necessary.  Regarding a more favorable load impedance: I guess I could try changing the ladder line length to improve that.  Too cold here in the mountains this time of year to try that now.  I've located some new 100 pF doorknobs from an estate sale and will try that solution first.  I'll report back here in a few days with the results.  Thanks to all for the comments and ideas

[VR2AX]

Also remember that 470 + 180 in parallel means that 470/650 of the current passes through the fixed and 180/650 passes through the variable

Very true, but that was already taken care of in my calculations. The total current through the ~655 pF shunt capacitance is ~6.5 amps.

- Glenn W9IQ

👍

[W4JFA]

If my calculations are correct, using the smith chart, if you either add or subtract 1/8 wavelength of the ladder line (figure in velocity factor), you will end up with pretty close to no reactance. You will be left with mostly the resistance.

Your matching network may match the load much better/easier and you possibly won't need to do anything with a padder.

This was using the figures from a post above of R-315 +J-255.

There are variables such as losses and the "actual" impedance of your ladder line, but this should be very close.

Bob

[K8AC]

Follow-up:  My 100 pF doorknob caps from eBay came on Saturday and I was able to get them installed in the tuner before the 160 contest got underway again.  These caps were NOS, of RCA manufacture, from 1957 and packed in waterproof military pouches.  They were marked N750 or something to that effect.  I used 5 of them in parallel.  Over 11 hours of CW at near 1,500 watts on 160, there was no noticeable drift in value, watching the SWR indicator.

73, Floyd - K8AC

[W1VT]

If you are buying on ebay, it may be just as cost effective to buy five small capacitors as it is to buy one large one if you factor in shipping.
The smaller capacitors give you a degree of flexibility with respect to the final value.
A lot of small capacitors gives you the option of testing parts and choosing the best ones for your project. 
Better to weed out the bad parts at the beginning instead of blowing them up in the amp!

[VR2AX]

The small value capacitors have greater 'resistance' (impedance) to RF current. So more of the current is diverted through the parallel high value ones (which present a lower 'impediment' to RF current flow). This is ok but emphasises the need for all components to be able to withstand whatever current/voltage they are individually required to handle.