The AM Forum
March 28, 2024, 04:37:00 PM *
Welcome, Guest. Please login or register.

Login with username, password and session length
 
   Home   Help Calendar Links Staff List Gallery Login Register  
Pages: [1]   Go Down
  Print  
Author Topic: CE-100V Low Power Out  (Read 4904 times)
0 Members and 1 Guest are viewing this topic.
k6jca
Guest
« on: October 02, 2009, 09:25:02 PM »

I'm trying to get a rather beat-up CE-100V transmitter back on the air, and I'm stuck...

I'm testing it on 80 meters, and I'm only getting about 5 watts out.  I've checked the drive level to the grids of the PA tubes, and it looks fine (it was almost exactly the same as I measured on a different 100V that had full-power out).

I've checked the tubes' DC voltages -- they all look fine (although the cathode voltages are a bit light -- they measure 0.04v in lieu of 0.1 VDC).

I've also swapped out the PA tubes (with 2 different pairs) -- no change in performance.

So I'm scratching my head.  Any thoughts as to where else to look?

By the way -- there is one mystery: the schematic shows a "coaxial capacitor" (C402A).  I can't find anything that looks like this.  Anyone know what it is?

Thanks!

- Jeff, k6jca



Logged
WQ9E
Member

Offline Offline

Posts: 3282



« Reply #1 on: October 02, 2009, 10:40:24 PM »

Jeff,

What is the panel meter reading?  Does it show normal idle current with no drive?  The low cathode voltage compared to the other rig indicates a lack of cathode current which appears as a voltage drop across the .5 ohm cathode resistor.

If it shows normal idle current, try it on 40 and see if it is an output network or switching issue.

Is the cathode resistor itself OK?  I don't have my manual out and the scanned schematic from BAMA is in multiple pieces so I have not traced out the metering circuit.  But the cathode (pin Cool should show about 1/2 ohm to ground.

If the DC voltages are OK, the cathode is near ground, and the excitation level is proper then it sounds like an issue with the output network.

The coaxial capacitor looks like it was used to standardize "stray" capacitance between units so that the broadband couplers would work properly.  Trace the circuit from the parasitic suppressors to the common on the bandswitch for the input to the couplers and if it is there you will find it.  I don't recall this component from when I worked on mine but that was several years ago.  According to the schematic it is adjustable and probably looks like a little piston trimmer.  CE made some nice equipment but from their CE-10 through their CE-200V the manuals were consistently bad (in my opinion).

Rodger WQ9E
Logged

Rodger WQ9E
k6jca
Guest
« Reply #2 on: October 03, 2009, 06:41:48 AM »

Jeff,

What is the panel meter reading?  Does it show normal idle current with no drive?  The low cathode voltage compared to the other rig indicates a lack of cathode current which appears as a voltage drop across the .5 ohm cathode resistor.

If it shows normal idle current, try it on 40 and see if it is an output network or switching issue.

Is the cathode resistor itself OK?  I don't have my manual out and the scanned schematic from BAMA is in multiple pieces so I have not traced out the metering circuit.  But the cathode (pin Cool should show about 1/2 ohm to ground.

If the DC voltages are OK, the cathode is near ground, and the excitation level is proper then it sounds like an issue with the output network.

The coaxial capacitor looks like it was used to standardize "stray" capacitance between units so that the broadband couplers would work properly.  Trace the circuit from the parasitic suppressors to the common on the bandswitch for the input to the couplers and if it is there you will find it.  I don't recall this component from when I worked on mine but that was several years ago.  According to the schematic it is adjustable and probably looks like a little piston trimmer.  CE made some nice equipment but from their CE-10 through their CE-200V the manuals were consistently bad (in my opinion).

Rodger WQ9E

Thanks for the reply, Rodger.

The panel meter doesn't show much movement in either the watts or the current position, but, given the state of misadjustment of the various pots and coils when I got this unit, I've been ignoring its readings and instead depending upon measurements I can make with other equipment.

Power output on all bands is either non-existent or very low (measuring with an external wattmeter), and so I've decided to tackle 80 first.  I've essentially got its signal path peaked up all the way to the PA grids, and the PA grid signal (at the test point within the PA cage) is the same level, and frequency, as I measure on a friend's working radio (both measured with the same scope).

The resistance-to-ground at the PA cathodes (they're in parallel) actually measures about 1 ohm to ground instead of the expected 0.5 ohms, which might be due to either one of the two parallel 1-ohm resistors being bad, or resistance drift with age (or...?).  I'm not too worried about that at the moment, because, although that'll cause the current reading on the meter to be high, I don't expect it to have much effect on power.

But, given that it is reading 1 ohm, this means that the combined plate current is only about 40 mA, in lieu of the 200 mA I'd expect (assuming that the manual is correct in stating that I should measure 0.1 v at the cathodes (that is, across 0.5 ohms)).

And the grid voltages are measuring right around -42 volts, as spec'd by the manual.  But could the manual be wrong?

Screen and plate voltages seem fine, too, per the manual's table.

But I'll re-verify these measurements later today.

The output network (which, essentially, is just the plug-in coils) visually seems fine, and it ohmed out OK.  But I'll check it with a grid-dip meter later on, too.  However, given that it is supposed to be broad-band (and thus low-Q), I'm not sure I'll get much of a reading.  But it can't hurt to try.  (By the way, I did trace from the parasitic suppressors in an attempt to find the coaxial capacitor -- nothing found (yet), but, like you, I've noticed that the schematic doesn't necessarily match the actual wiring, and it's possible that this component was never installed.)

Well, if worse comes to worse, I suppose I can go to my friend's house and make some more measurements (within the PA chassis) on his unit, but it'll require taking his transmitter out of its case (and possibly lugging mine over, too), and I am NOT looking forward to that.

Thanks again for your help, Rodger!

- Jeff, k6jca
Logged
k6jca
Guest
« Reply #3 on: October 03, 2009, 07:01:43 AM »

One additional note...

The grid drive level that I'm measuring to the PA's is about 33 vpp (on both transmitters).  The manual spec's 33 vrms, which would mean that I'm only at about a third of the level that I should be.  (All measurements made in CW mode, by the way).

However, as I stated earlier, this drive level gives near full-output on my friend's 100V.  And I'm wondering if the spec'd screen-grid voltage of -42 vdc is incorrect, and if it should be higher.  After all, a 33 vpp signal has a peak value of only 16 volts, which is much lower than the 42 volts that would be required to drive the control-grid into conduction.

Just wondering...

- Jeff, k6jca
Logged
WQ9E
Member

Offline Offline

Posts: 3282



« Reply #4 on: October 03, 2009, 08:51:29 AM »

Jeff,

What are you reading for screen (pin 4) voltage.  This should be a high positive voltage.

Even on 80 meters, the capacitance of your scope probe (I assume you are using a 10X low cap probe since a 1X would be very problematic) is going to detune the grid a little bit.  What does the actual waveform look like?  There are a lot of signals mixed around in the 100V and make sure that what you are getting is drive actually on 80 meters instead of an off frequency signal passed through from another stage.

Depending upon the meter used, accurately measuring the very low resistance of cathode resistors is difficult.  I have a Simpson low resistance add on for my 260 that does a good job but with regular meters make sure you are balancing out any resistance in the leads themselves and avoiding error from poor contact at the measurement points.  In any case, the extra .5 ohm is not enough to create sufficient additional degenerative feedback to drastically reduce your power.

But bottom line is if the tubes are good (which you tested by substitution), static DC voltages are correct (or close), cathode is close to ground, and you have drive then you should be developing significant input.  The only case I can think of where you could have near normal tube and input conditions with low output and low cathode current would be extremely heavy loading with the output tank still in resonance.  The odds of that on all bands seems very low so I would be more suspect of a static voltage being off or insufficient drive at the operating frequency.

As an experiment, temporarily reduce the grid bias (using the bias adjust control) to see if your indicated plate current comes up to normal indicated idle current.  If so measure the screen and negative grid bias with it drawing normal idle current.
Logged

Rodger WQ9E
k6jca
Guest
« Reply #5 on: October 03, 2009, 10:56:20 AM »

Jeff,

What are you reading for screen (pin 4) voltage.  This should be a high positive voltage.

Even on 80 meters, the capacitance of your scope probe (I assume you are using a 10X low cap probe since a 1X would be very problematic) is going to detune the grid a little bit.  What does the actual waveform look like?  There are a lot of signals mixed around in the 100V and make sure that what you are getting is drive actually on 80 meters instead of an off frequency signal passed through from another stage.

Depending upon the meter used, accurately measuring the very low resistance of cathode resistors is difficult.  I have a Simpson low resistance add on for my 260 that does a good job but with regular meters make sure you are balancing out any resistance in the leads themselves and avoiding error from poor contact at the measurement points.  In any case, the extra .5 ohm is not enough to create sufficient additional degenerative feedback to drastically reduce your power.

But bottom line is if the tubes are good (which you tested by substitution), static DC voltages are correct (or close), cathode is close to ground, and you have drive then you should be developing significant input.  The only case I can think of where you could have near normal tube and input conditions with low output and low cathode current would be extremely heavy loading with the output tank still in resonance.  The odds of that on all bands seems very low so I would be more suspect of a static voltage being off or insufficient drive at the operating frequency.

As an experiment, temporarily reduce the grid bias (using the bias adjust control) to see if your indicated plate current comes up to normal indicated idle current.  If so measure the screen and negative grid bias with it drawing normal idle current.

Hi Rodger,

If I recall, the screen voltage was around 340 volts.  I plan to verify again shortly.

And yes, it's a x10 probe.  I think the capacitance spec is 8 pf.  By the way, the waveform looks very nice -- 3.8 MHz sig.

I'm not too worried at the moment about the PA cathode resistance for exactly the reason you mentioned.  But I'll make a 4-wire resistance measurement, just to see what it actually is.

By the way -- do you know what the idle current (or idle wattage, if using the panel meter) should be for the PA?

Anyway... I'm off to the "lab" to take more measurements!

73,

- Jeff
Logged
k6jca
Guest
« Reply #6 on: October 03, 2009, 11:00:19 AM »

(By the way, I notice that you have (or had) a PRC-47.  You might be interested in some mods I made (take them with a block of salt).  They're posted at:  http://k6jca.blogspot.com/    - Jeff)
Logged
WQ9E
Member

Offline Offline

Posts: 3282



« Reply #7 on: October 03, 2009, 12:44:03 PM »

Jeff,

I still do have my PRC-47 and I bookmarked your site for future reference, thanks!

The idle current should result in an indication of 50 to 60 watts input on the meter with the bias set for normal operation. 

Rodger WQ9E
Logged

Rodger WQ9E
k6jca
Guest
« Reply #8 on: October 04, 2009, 02:07:48 PM »

Hi Rodger,

I made more measurements yesterday on both my 100V and my friend's 100V and discovered that he actually had about twice the drive to the PA's that I had.  So this morning I went through and repeaked all of the appropriate transformers -- now my max output is about 70 watts in CW mode (with both the "CW Carrier" and the "Power" controls at max).  But this works out to about 195 watts input power if I measure the PA current and convert to power by multiplying by 600 (by the way, the cathode resistance is actually 0.48 ohms).  I need to determine input power via this method, for the moment, because the 100V meter is WAY off, and I can't bring it into calibration with the pot.

I also noticed that my friend's 100V power input, in SSB mode (idle), was about 70 watts.  By comparison, my SSB Idle power is about 27 watts.  I believe this is because he has a higher screen-grid voltage than I do (363 volts (his) vs. 330 V (mine) -- his rig has solid-state plug-ins for the rectifier tubes), and a slightly lower PA grid-bias voltage (-39v (his) vs. -41 (me)).  Regarding the grid-bias -- using an external power supply I can force my idle-power to be 70 watts if I lower the bias voltage to -34 volts (and 60 watts if the voltage is -36 volts). 

I'm not sure if I should start fooling around with screen and control-grid voltages or just leave well enough alone -- my rig's measurements are close to the values mentioned in the manual, but...I have to wonder how accurate the manual is.  And because there aren't any pots for controlling either of these voltages (at least that I know of), it'd be a matter of playing around with resistance values, which I'm hesitant to do, given the inaccessibility of some of the components.  (But I do wonder if I should be seeing more power out for an input power in the range of 175 - 195 watts, and if changing these voltage might improve the efficiency (I'll try another pair to tubes, too, just in case...)

Still many items on the fix-it list, though:  sticky VFO mechanism, VFO off frequency by about 100 KHz, Internal Wattmeter reads incorrectly, RF Ammeter reads incorrectly, and...who knows what else.  On the plus side, I've repaired the meter (one of its internal "spiral springs" had opened up), replaced an "open" resistor in the 2nd mixer, fixed the balanced modulator so that carrier (in ssb mode) can now be nulled (replaced the original CK715 diodes with matched Schottkys), and opposite-sideband suppression isn't too bad.

Things are coming along (finally).  Thanks again for your help!

- Jeff, k6jca

Logged
WQ9E
Member

Offline Offline

Posts: 3282



« Reply #9 on: October 04, 2009, 04:20:24 PM »

Jeff,

It sounds like you are almost there.  Congratulations on fixing the meter; I have repaired a few with success but I have also had many failed attempts when one of the spiral springs was the issue.

Once you are sure of the metering I would suggest running your 100V at around 45-50 watts input at idle via adjustment of the control grid bias.  On CW it isn't as critical but for SSB or AM you need to set the operating point close to the original specifications to achieve the excellent audio that the 100V can produce.

There is a good website covering the PTO on the 100V and it really isn't that difficult to rebuild.  Locate your local bearing supplier first just in case you need some in a hurry to finish your work.

You will probably have this one "airworthy" soon so time to start thinking about a receiver.  I have its later (and I believe somewhat cost reduced) 200V twin paired with my NC-400.  It makes a nice all mode station-I remember someone was using their 100V on the digital modes just to prove they could.

Rodger WQ9E
Logged

Rodger WQ9E
Pages: [1]   Go Up
  Print  
 
Jump to:  

AMfone - Dedicated to Amplitude Modulation on the Amateur Radio Bands
 AMfone © 2001-2015
Powered by SMF 1.1.21 | SMF © 2015, Simple Machines
Page created in 0.056 seconds with 19 queries.