Some YC156 amp FAQs:

What is a YC156 and why did you use it?
The YC156 is a custom tube made for a medical equipment OEM for use in MRI systems. It is similar to a 3CPX5000A7, except it has a grid flange for direct chassis mounting. Another way to look at it: it's a 3CX15000B7 with a 5kW anode cooler (same internals and curves). I used the YC156 because it is very rugged, easy to use mechanically, has reasonable heater requirements (15V, 15A), and most of all, I have a batch of them. You can see a copy of the tube data sheet here. pdf 738k)

It looks like you put a lot of money in that amp! How much did it cost and how long did it take to build?
I build things mainly to have fun and use the parts I already have, not because I really want the result. It cost me about $100 in parts and took about 2 years, working off and on, to build.

If you don't want it, can I buy it?
You would probably be happier with an amp you made yourself, with the features you want.

I am interested in the details of your water cooled dummy load. What values of resistors did you use and how did you connect them?
The resistors are 100 ohms/800 watts and are similar to the Florida RF P/N 31-1005.

Other companies such as KDI and RF power products make similar parts. The eight resistors are connected in series two at a time to form four 200 ohm resistors. These are then connected in parallel to form one 50 ohm/6400 watt load. The copper chill plates have manifolds at each end as you can see in the photos, and are cross-drilled with 2 or 3 passages under the resistors. The ends of the passages are then plugged with brass screws to seal the assembly. I have a thermocouple/temp controller set at 50C which opens a solenoid cooling water valve. This is all expensive stuff, but water cooling is the best if you can find it or afford to buy it.

What chimney does the YC156 take?
If you want to use an Eimac chimney, the SK-306 works well.

What is the cathode input impedance of the YC156?
I have found it to be in the range of 25 to 45 ohms for class AB2 operation on 20m and 40m. The tube needs a tuned circuit on the input to provide a clean and symmetrical drive waveform to the cathode at all points of the RF input cycle. I'm using pi networks with a Q of about 4.

So, let's get to the point. How much juice will that monster put out?
7000W PEP SSB or 5000W CW with 120W to 170W drive is no problem. See the test results page for more info.

How do you measure RF power?
The basic power measuring tool I use is an HP power meter with a thermal type sensor. These are very accurate and respond to true average power. I sample the high level RF with a directional coupler which has been calibrated with a network analyzer to within .1dB at the frequency of interest, and then measure the sample with the power meter. For peak readings I substitute a calibrated 50 ohm scope for the power meter.

Why only two bands, 40m and 20m?
Band switching high Q tank components in an amp this size is not an easy task. Reliably switching tens of RF amps, at thousands of RF volts, with minimal loss for multiple bands was a challenge I didn't feel up to. Two bands would make the switching simpler, and 40, 20 are the bands I most often use. As it is, I had to put three Kilovac HC-2 vacuum relays in parallel to handle the current, and pot the assembly in a special two part silicone for flashover protection. I've been lucky with this arrangement so far, but I'm still keeping my fingers crossed.

Why not just lower the tank Q?
This tube has a lot of output C. Mounted in my chassis, it measures 58pF cold. This sets a lower limit on 14MHz Q that under some conditions is still quite high (~20). In situations where the tune cap has a fairly high minimum C, the technique of putting an inductor between it and the anode can help quite a bit. Unfortunately, with my minimum tune C of 10pF, and 58pF of tube capacitance, this trick was of minimal benefit.

So building a multiband YC156 amp that covers 10m might be tough?
In a word, yeah. This tube was designed to produce high peak powers in pulsed service. High power = lower anode load Z; lower anode load Z = lower Q. At maximum ham power levels you're looking at a tube/tank circuit that's going to get pretty hot.

You're using 7kV, why not lower the anode voltage to get a lower Z and a lower Q?
Doing that might help some, but it will also increase grid current, reduce gain, and hurt linearity. You need to find a happy balance between these, and I tend to favor the higher anode voltages. This issue may be even worse when you consider that we're using surplus tubes, which may have been rejected for having high grid current, as I believe some of them were.

Where are the knobs? Real radios need to have knobs.
I can see your point, but this amp uses stepper motors to turn the caps and was designed to be remotely controlled (size, blower noise). I guess I could have used optical encoder knobs instead of switches, but they would have had a wimpy feel in addition to making the firmware more of a pain. I don't consider programming fun, it's just one of those necessary evils you have to deal with on a project like this.

Where did you get the silver plated coils?
I plated copper tubing with Cool-Amp rub-on powder. The plating probably isn't as thick, and doesn't look as nice as a real plating job, but I had some of the stuff sitting around and decided to give it a try.

I see labels all over this project, how did you make them?
I used a Brother PT-9200DX label printer. It laser prints any kind of text or graphics you want on a laminated adhesive label tape. The tapes come in a variety of print and background colors. In this amp they are all black print on either a white or clear tape.

Why are the anode blocking caps mounted on a fiberglass wall?
The wall was just a convenient place to mount them. It is there is to keep hot air from the tube from heating the components in the tank circuit area. Tube exhaust flows in a short path from the anode out through the filter in the front of the amp. Pressurized air from the bottom compartment also blows through the hole in the tube deck under the blocking caps to keep them cool. The wall is made out of a large piece of insulating material so that it will be transparent to RF, and not interfere with the currents circulating between the tube and the tank circuit.

Where did you get a dummy load big enough?
I made a water cooled load from eight 800W flange mount film resistors mounted on copper chill plates. It is capable of dissipating 6kW on CW, and somewhat more than this PEP on SSB. I have a temperature controller which opens a solenoid valve on demand, running tap water through the load and into a drain. You can see a picture of it here .

What kind of T/R switch does this amp have, and can it do QSK?
I'm using a common vacuum relay, a Kilovac HC-2 (similar to Jennings RJ-1A). Although I haven't done anything in the design that would prevent QSK operation, no steps were taken to enhance the switching speed either. I don't intend to use QSK, mainly because I feel there's just something wrong with cycling a vacuum relay that many times.

How do you limit current in the HV circuit in case the tube arcs or the B+ flashes to ground?
Like most designs, I have a resistor in the B- line and clamp diodes to keep the B- from going negative. I was careful to use components designed to handle the huge current involved: 700 A peak rated diodes, and a non-inductive resistor with an energy rating of 7000 joules (my cap bank holds over 1200 J). I recommend using film resistors specifically designed for energy absorption, never wirewound or ceramic encapsulated types, unless of course you enjoy replacing the resistors and repairing collateral damage from the flak. Kanthal-Globar is a good source for these resistors.

What other features does the amp have?

  • Remote control and metering (amp is too loud to have in the same room while operating)
  • Protection from grid overcurrent, anode overcurrent, reflected power, high voltage low, and loss of air pressure faults
  • Preset tune and load settings for each band
  • Anode-cathode phase detector to aid in tuning; I can extend this to do fully automatic tuning in the future if desired
  • Automatic power up/power down heat/cool sequence
  • LCD display of settings and fault status
  • High voltage supply step-start
  • Heater soft-start
  • High isolation (optical and magnetic) of metering and control circuits from high voltage circuit. This is overkill, and the parts used are unusual and expensive, but I had them in the junk box, so into the amp they went
  • Stepper motor limit sensing, with auto-home on power up
  • Fast amp shutdown during fault via FET switch in cathode circuit