Tuesday, October 6, 2009

Central Electronics CE 100V Transmitter

My adventures bringing a CE 100V back to life!

A number of local hams have Central Electronic 100V transmitters, and on occasion I've joined them during their ragchewing roundtables on 80 meters (sans 100V on my part). I enjoyed the sound of the radio as well as its styling, and I thought it might be nice to have one of my own. And thus began my search for a 100V.

I finally found one that a local ham was selling on Ebay. I made a bid...and won it! Fortunately, because the seller was local, I was able to save shipping charges (it is a heavy radio!) and pick it up myself.

When I got it home, I discovered the radio, besides having extensive cosmetic issues, also had operational problems, and I put it to the side while I worked on other projects that were less daunting than tackling the 100V appeared to be.

Finally, I decided to bite the bullet and get the 100V on the air. First thing to do...pull it out of its cabinet and then get to work...

The radio is a marvel of design, and arguably represents the high-water mark of amateur radio transmitter design of the 50's. Which translates into a radio that is large, heavy, and complex (26 tubes!).

(Top View, with VFO Assembly removed)

(Paging Doctor Frankenstein!)

When I looked inside the actual radio, my heart sank...the chassis wasn't dirty, it was oxidized. I believe it must have originally been plated, and this plating had turned an ugly grey color. And in some places, actual rust had appeared!

(Typical oxidation/corrosion on this radio. Labels on the back panel and on the chassis are essentially unreadable.)

(Despite the terrible shape of the chassis, the front panel actually looks pretty good!)

My VFO was very difficult to turn, and felt "lumpy". Per the Tusa notes (well worth a read) on the 100V, I decided to remove the VFO assembly and take a look at what might be going on...

(Note: the VFO assembly is actually fairly easy to remove. You do not need to drop the front panel! Instead, follow the procedure in the Tusa notes (although please note that for step 7, you should unscrew the two bottom mounting posts from the VFO assembly, not from the front panel)).

After I'd removed the VFO assembly, I was curious to know how it looked inside, so I removed the back cover. Whoops! Chunks of foam (and foam bits) tumbled out. Looks like Central Electronics used this foam (3/8 " thick) to act as an insulator to minimize temperature variations within the can. And after 50 years, it was disintegrating.

(Disintegration of the insulation foam in the VFO Assembly!)

I happened to have an old mouse pad lying (1/4" neoprene), so I cut it up and glued it to the inside of the can with some RTV cement. Voila!

(Old mouse pads have many uses, such as...new insulation!)

Now to attack the difficult-to-turn VFO. The Tusa notes recommend repacking the bearings, but from the instructions I'd read (and from the stories I'd heard), it sounded like a real nightmare.

Instead, I decided to see if a shot of WD-40 into the bearings would help to loosen up the old grease...

I held the VFO so that the knob was pointing toward the floor, then applied a quick burst of WD-40 into the "well" (see photo below). The bearings are below this well, and, by holding the knob towards the floor, I hoped the WD-40 would flow down into the bearings.

It seems to have worked. The mechanism turns much more easily now. Sure, I probably ought to repack the bearings (because the viscosity of the grease might give it a bit "smoother" feel). Maybe next year...
(Click on image to enlarge.)

While I had the VFO out of the radio, I was curious to learn how the VFO tracking mechanism worked...

As the VFO frequency is adjusted, the VFO's lead screw moves a core in and out of the main VFO coil, thus changing the oscillator frequency. But there is a secondary adjustable coil, too, whose core is attached (via a rod) to a right-angle bracket that can pivot. This rod moves in and out of the secondary coil according to the height of the "VFO Corrector Adjustment Screws," and allows small corrections to be made to the VFO frequency as the the user tunes over the 1 MHz-wide range of the VFO.

You can get an idea of how the mechanism works from the two photos below:
The lead screw moves the frequency correction assembly (consisting of the screw run through the block) along either direction of the lead screw (depending upon whether the frequency is being adjusted up or down). A rod runs over the bottom of these screws (shown bottom-up in the photo above), which in turn causes the metal right-angle bracket to which it is attached to pivot.

Attached to the other end of this right-angle bracket is a rod which drives the core of the secondary coil in or out, thus correcting the frequency. In the photo below you can see both the larger main coil (on the same axis as the lead screw) and the smaller secondary coil below it (only a couple of turns of this coil are visible).
Pretty clever!

A note about the frequency correction adjustment. I would recommend that you start at the end of the VFO that has the largest "positive" (rather than negative) delta from the dial frequency. In other words: if the offsets at either end of the dial are both positive, start at the end that has the largest positive delta. If both of the offsets are negative, start at the end that is closest to the dial frequency, and if one end is positive and the other negative, start at the positive end.

"Zero" your dial at this frequency by moving the black line on the clear plastic to overlay the "0" on the dial (there's a screw a few inches below the VFO knob that let's you do this). Then, moving the VFO in 500 KHz increments, adjust the frequency using the "VFO Corrector Adjustment Screws" per the Tusa Consulting note on VFO Recalibration.

Problems that I've run into:
  1. Meter not working. No movement, at all. I opened up the meter and discovered that one of the "spiral springs" that attach to the armature had opened up. It was a real pain to repair, but repair it I did. (By the way, my meter is about 1 mA Full Scale, and has a resistance of 47 ohms). I also added a pair of diodes (1n5818) hooked antiparallel across the terminals of the meter (to protect the movement from burning out), as well as a 0.1 uF cap -- if you add the two diodes you must include this cap, otherwise your meter may read low in the "Watts" position.
  2. VFO Sticking/Hard-to-Turn (See discussion above)
  3. VFO Not Tracking (See discussion above)
  4. Lack of the -120V Blocking Bias in STBY mode. A 1uF/200V electrolytic cap that was attached to this line (via a 10 ohm resistor in the power-supply section) was shorted to ground. (Note: neither this cap, nor the 10 ohm resistor, appear in the schematics). These parts were probably added to slow-down the transition between STBY and Transmit. I didn't have a 1uF with a high enough voltage rating in the junkbox, so I instead used a 2 uF cap.
  5. Inability to Null Carrier in Sideband Modes. The meter would remain pegged to the right in Null mode irrespective of any adjustments I made to the two Carrier Balance knobs. I measured the forward-voltage of the four original germanium (CK715) diodes in the modulator plug-in module, and the voltages varied wildly (from 0.234 volts to 0.632 volts). I replaced these with HP 5082-2063 (Schottky?) diodes that I had in one of my parts' bins (their Vf was 0.34 volts, and matched within millivolts for all 4 diodes). Works fine now.
  6. Wattmeter: Reads too low, and cannot adjust far enough. Resistance values had drifted over time, and one of the 100 ohms resistors had drifted to 109 ohms. Replaced with 100 ohms, and now can adjust with the pot (although it's almost at its limit).
  7. RF Ammeter: Reads too high, and cannot adjust far enough. The resistors have apparently drifted. The voltage-divider resistors are in an extremely awkward location, so their replacement is very difficult. Instead, I added 120 ohms in parallel with R147/R148, and that brought the voltage into range to allow correction using the pot.
  8. No X-Axis movement on Monitor Scope. Replaced V21 (6U8A)
  9. Low-frequency Noise in Audio, Eventual Loss of Sideband Suppression. I traced this to a leaky cap in the audio phase shifter module -- one of the symptoms was a high DC voltage at an output (pin 8) of the phase shifter. Cap C130 was leaky (but the leakage couldn't be measured with a DVM) -- I replaced this 4711 pf mica cap (actually measured 4739 pf) with 4731 pf consisting of a 4300 pf mica and a 470 pf mica in parallel. (Update: I've discovered that this problem is discussed in the 100V article by Charlie Talbott, K3ICH, in the August, 1996 issue of Electric Radio, and I've implemented one of his mods, which is to insert a 1uF, 400V cap betweenV7 pin 1 and the phase-shift network PS-2 socket's pins 2 and 6 (to isolate C128 and C130 of the phase-shift network from the B+ voltage on V7's plate)).

With these issues resolved, I've deemed the 100V ready for the air:

The 100V in its operating position!

Although the 100V is now up and running, there are still...

Problems I've yet to resolve:
  1. 8 MHz Oscillator cannot be adjusted to be exactly 8.000 000 MHz (it remains too low). Even with the adjustment cap at minimum value.
  2. FSK Adjustment does not span 100-900 Hz. Instead, it only seems to have a range of about 150 Hz.
  3. PA "Idle" Wattage (in SSB Xmit, no voice) should be in the range of 60 watts (per the recommendation of others) -- mine is more around 30 watts. (Central Electronics added adjustment pots for both the driver bias and the PA bias adjustment to their 200V transmitter, but these parts are not in the 100V, and to add them involve more surgery than I'm willing to undertake at the moment).

Other notes and Comments:
  • The schematics can be inaccurate! I've found additional parts, and I've found parts missing, when comparing the actual circuitry to the schematic.
  • Tusa Consulting has a number of notes on the CE 100V and 200V transmitters. You can find these notes here.
  • A previous owner had replaced the two batteries internal to the Speech Limiter module with two AA-size alkaline batteries, and had mounted their holder on the outside of the Speech Limiter's case.
  • To get at the tubes and adjustments beneath the fan in the audio section, just loosen the transformer screw and tilt the fan bracket up...

  • Some other useful data...
100V I.F. Mixing Scheme
(Click on image to enlarge)

Some manual copies have impossible-to-read voltage charts. Here are clearer copies (thanks to Jon, K6JEK).

Tube Voltage Chart
(Click on image to enlarge)

RF Voltage Chart
(Click on image to enlarge)

Articles on the 100V in Electric Radio magazine:
  • "Restoration of the Central Electronics 100V," Dennis Petrich (K0EOO), Electric Radio, Number 20, October 1991.
  • "Observations on the Central Electronics 100V & 200V," Charlie Talbott (K3ICH), Electric Radio, Number 88, August 1996.
(There may be additional articles in Electric Radio. These are the two that I've found.)

Phase Network Simulation [11 March 10]:

I ran a SPICE simulation on the CE 100V's Phase Network and compared the two outputs. Here's the plot (and the schematic):

(Click on image to enlarge)

The solid line is amplitude, while the dashed line is phase.

Sideband suppression versus phase error can be calculated with this equation:

- 20$\displaystyle\log_{10}^{}$|tan($\displaystyle{\frac{\delta}{2}}$)| (Where "delta" is the phase error.)

So a phase-error of 1 degree (from the ideal phase-shift of 90 degrees) will result in about 41 dB of sideband-suppression; a phase error of 2 degrees: 35 dB; while a phase-error of 10 degrees will result in only about 20 dB of sideband suppression.

(Note: I haven't included in this simulation the small-signal resistances presented by the grids of the tubes that the phase network output drives (I don't know what they are). For ease of calculation, I've assumed that they're infinite. Also, because I don't have any SPICE models for tubes, I just used a transistor as the driver.)

[LTspiceIV, the program that I used for my simulations, is free, and it can be found here.]

Standard Caveat!

I may have made a mistake in any of the above, so use at your own risk!


Alan, K2WS said...

Hello Jeff,
It's been a few years since we xch'ed emails. It was back in the days when you did some wonderful mods on PSDR! I'm Alan, K2WS.
I have a 100V that was re-furbished by Bill, W0YVA. Of course it does breakdown occasionally and I've repaired the driver stage coils on 40, repaired the Audio PSN, etc. I've had it on 75 ESSB & AM, 40, 20 and 10 meters! What fun. The old girl sounds exceptional and most hams don't realize they're talking to a half century old rig, hi.
I run it with a 75A-4 and the old Dow-Key relay I used when I was in high school.
Let me know if I can help out Jeff. When not on the 100-V I'm running a 5000A these days. Let me hear from you.
73, Alan K2WS

Jeff said...

Hi Alan,

Good to hear from you again! I'm glad to hear you're enjoying both your 100V and 5000A. I hope to have my 100V on the air soon. Just a couple more things to do...



Andre said...

Hi Jeff!

I ran across your blog this weekend as I just acquired a Central Electronics 100V transmitter today. I'm not sure if it works yet, and I'm trying to take all the proper and necessary safety precautions before firing the transmitter up. I'm basically in the "gather information" mode. I have similar problems that you've run into as well, particularly with the VFO knob. My chassis is in fair shape, and if you'd like pictures of the back panel, let me know, or visit my blog at: http://signalofkc0mmy.blogspot.com

It's encouraging to know that other people are still using these transmitters! Hopefully, mine isn't "junk."

Anonymous said...

are the phillips head screws original in the VFO?
I think someone has had it apart before you?

Anonymous said...

Hi I just acquired a 100v....but I don’t hear anything when I turn it on...just fans! What am I missing? Thanks

Jeff said...

You should hear the single fan. Other than that (because it's a transmitter-only), I don't think you should hear anything else. But please let me know if otherwise.

- Jeff