The AM Forum
July 23, 2024, 01:55:14 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: high level RF pick-off points to low-capacitance indicators  (Read 11133 times)
0 Members and 1 Guest are viewing this topic.
Opcom
Patrick J. / KD5OEI
Contributing
Member
*
Offline Offline

Posts: 8284



WWW
« on: January 04, 2012, 11:11:37 PM »

With the RF deck finally coming along and actually making power instead of just looking like a blue box, the skullwork comes to the question of where to get RF to drive CRT plates for the mod scope. A trapezoid is desired. Audio will come from modulated B+.

The CRT's capacitance from any deflection plate to 'everything' is about 3-4pF, hard to measure.

The scope the tube came from can make a differential voltage plate to plate of about 600V. However, its a Tektronix, so the peak voltage available is excessive and is only used like that when the horizontal sweep is 'magnified'. I think realistically it needs about 40-50V per inch, and it is a 5" CRT.

I think only about 200V peak is needed, and this corresponds to somewhat less than the peak grid voltage (4-1000 class C, grid driven).

The grid circuit of the 4-1000 is swamped by a 3000 Ohm NI resistor.

The CRT will be mounted below the RF chassis, under the tube socket.

The plate signal needs to be divided down to about 200V peak, and get to the CRT deflection plate. From an accuracy and leveling point of view, it seems to make more sense to take this from the plate because the voltage swing there is constant with a given drive, and the swing at the low-Z end of the pi-network can vary a lot depending on loading.

I'm interested in comments on how to do this, even if it has not been done.
Here are some possibilities:

1:
Plate-compartment capacitive divider from the hot terminal of the PI inductor to GND and tap this off going down through the chassis in a ceramic feed-through to the CRT deflection plate terminal. In order to have a relatively flat response, the bottom cap in the divider must nave a reactance much lower than the capacitance of the CRT plate and the lead going to it. I was thinking top = 3-4pF, bottom = 60-80pF

2:
Plate-compartment capacitive divider from the hot terminal of the PI inductor to GND, and tap this off going down through the chassis in a coaxial cable, to the CRT deflection plate terminal. In order to have a relatively flat response, the bottom cap in the divider must nave a reactance much lower than the combined capacitance of the CRT plate and the coaxial cable section going to it.  I was thinking top = 3-4pF, bottom = 60-70pF

3:
Steal the voltage from the neutralizing probe. It is a 4" tall piece of #12 about 1" from the chimney. It extends below the chassis a bit and is easily gotten to. It has a wire running close to the chassis through a ceramic insulator in the wall of the  grid tuning compartment and from there to the 'cold' end of the tank and also to a 25pF capacitor to GND (same as grid capacitance of the 4-1000). Concern is that the drive level will affect this signal greatly.

4:
There is the method in the radio handbook (attached), where a few turns are used at the plate tank and at the CRT a variable cap tunes the signal. This set is a pi-network though. And I'm looking for a way to avoid having to tune the CRT, as well as concerned about the tuning range of such an arrangement as there will be no bandswitching just to tune the CRT deflection signal. So I do not like this one.

5:
resistive divider. I do have some high value high wattage carborundums but their behavior from 3-30MHz is an unknown quantity.


* trapezoid modulation monitoring.png (32.17 KB, 551x661 - viewed 972 times.)
Logged

Radio Candelstein
W7TFO
WTF-OVER in 7 land Dennis
Contributing
Member
*
Offline Offline

Posts: 2472


IN A TRIODE NO ONE CAN HEAR YOUR SCREEN


WWW
« Reply #1 on: January 04, 2012, 11:27:23 PM »

Try running your output wire from the Pi-net to the output jack through the center of a toroid, then use that as a secondary for your scope.

73DG
Logged

Just pacing the Farady cage...
Opcom
Patrick J. / KD5OEI
Contributing
Member
*
Offline Offline

Posts: 8284



WWW
« Reply #2 on: January 05, 2012, 12:13:06 AM »

you mean a current transformer with one lead as the primary, and then some turns on the secondary?

I can see how that could develop some serious voltage. Won't the ferrite ring act like an RF choke?
Logged

Radio Candelstein
W7TFO
WTF-OVER in 7 land Dennis
Contributing
Member
*
Offline Offline

Posts: 2472


IN A TRIODE NO ONE CAN HEAR YOUR SCREEN


WWW
« Reply #3 on: January 05, 2012, 12:28:29 AM »

Usually it's self-inductance doesn't matter, provided you use the correct core type, as it is now a broadband transformer with the secondary winding.

73DG
Logged

Just pacing the Farady cage...
k4kyv
Contributing Member
Don
Member

Offline Offline

Posts: 10037



« Reply #4 on: January 05, 2012, 11:31:37 AM »

Try running your output wire from the Pi-net to the output jack through the center of a toroid, then use that as a secondary for your scope.

That's how I do it (it works equally well with link coupled transmitters). The output from the winding on the toroid feeds through about 10' of RG-59 to a small (25-50w size) fixed link coupled coil (JEL series IIRC), with a tuning capacitor across the main coil.  One side of the coil is grounded and the other feeds the vertical input to the scope directly, by-passing the internal vertical amplifier altogether.  I have coils for 160, 80 and 40, selected with a small  rotary switch.  The tuning cap is used as the "vertical gain" to adjust the height of the pattern, which may be set for envelope or trapezoid display.  Don't remember how many turns I wound on the toroid, but the toroid coil matches the input link to the tuned circuit perfectly. I couldn't get enough rf voltage from the toroid to feed the vertical plates directly through the RG-59, but the tuned circuit allows me to deflect the spot completely off the screen at resonance.

The only disadvantage to my system is one more thing to tune to resonance when I QSY or change bands.

It is always problematic to have the monitor scope located some distance away from the transmitter.  The same goes for the external VFO.
Logged

Don, K4KYV                                       AMI#5
Licensed since 1959 and not happy to be back on AM...    Never got off AM in the first place.

- - -
This message was typed using the DVORAK keyboard layout.
http://www.mwbrooks.com/dvorak
W0BTU
Member

Offline Offline

Posts: 228



WWW
« Reply #5 on: January 05, 2012, 02:24:09 PM »

4:
There is the method in the radio handbook (attached), where a few turns are used at the plate tank and at the CRT a variable cap tunes the signal.

That is almost exactly how I've always done it (like the "NOTE:... " diagram at the lower right), with the parallel tuned circuit connected to the vertical deflection plates, with the following exception: the pickup link is in my antenna coax switch box; 1:1 2-turn links air-coupled, very close coupling at each end. All coil diameters about 1".

I never had to switch taps or anything. The variable capacitor adjusted the amount of vertical deflection from 80-10 meters just fine.
Logged

73 Mike 
www.w0btu.com
Opcom
Patrick J. / KD5OEI
Contributing
Member
*
Offline Offline

Posts: 8284



WWW
« Reply #6 on: January 05, 2012, 10:13:28 PM »

Thanks for this advice. If I use either the toroid or the 1:1 2-turn 1" coil, I take it the effect of this extra circuit on the RF/tuning/loading is not very large. I guess #12 wire, and /or, what formulation of toroid material?

In one way I am lucky because the CRT will be only a few inches from the coupling circuit. Maybe I can use 300 Ohm TV line to run to the def. plates.

The core or transformer methods both would allow balanced feed to the deflection plates and that is very agreeable for those picky about deflection, focus, and geometry. Not that it is that critical but it minimizes those issues. This also will allow me to jig up the CRT and coupler on the workbench and work things out the easy way by just feeding RF there from any source of sufficient power.
Logged

Radio Candelstein
W0BTU
Member

Offline Offline

Posts: 228



WWW
« Reply #7 on: January 05, 2012, 10:38:12 PM »

If I use either the toroid or the 1:1 2-turn 1" coil, I take it the effect of this extra circuit on the RF/tuning/loading is not very large. I guess #12 wire, and /or, what formulation of toroid material?

I used #20 or #24 plastic insulated wire, and I didn't use a toroid. The scope tuner parallel tuned circuit--that was connected to the vertical deflection plates--used Air-Dux and if I recall, a 100 pF air variable. The insulated wire link was wrapped around the center of the Air-Dux.

Quote
... the CRT will be only a few inches from the coupling circuit. Maybe I can use 300 Ohm TV line to run to the def. plates.

Sure can. :-)

I wish I could remember more details. I don't have the original scope tuner, only the link in the antenna coax switch. I have some parts gathered to replace the original scope tuner.
Logged

73 Mike 
www.w0btu.com
Opcom
Patrick J. / KD5OEI
Contributing
Member
*
Offline Offline

Posts: 8284



WWW
« Reply #8 on: January 06, 2012, 12:21:52 AM »

I really do not have a link in the common sense, it is all 50 Ohms or coax-based stuff.

Do I have it right for the proposals for the link and the toroid?

If the shortest 300 Ohm lead is used, is it possible to dispense with the tuning cap?

It seem to have value for amplitude adjustment but there is a concern that a resonance will couple enough juice to arc it or burn the coil.


* RF coupling to CRT.png (29.04 KB, 800x800 - viewed 768 times.)
Logged

Radio Candelstein
W7TFO
WTF-OVER in 7 land Dennis
Contributing
Member
*
Offline Offline

Posts: 2472


IN A TRIODE NO ONE CAN HEAR YOUR SCREEN


WWW
« Reply #9 on: January 06, 2012, 02:44:10 AM »

I'd add a swamping resistor in series to linearize it, and a load resistor to keep it tame.  If needed.

Did the punch arrive OK?

73DG
Logged

Just pacing the Farady cage...
k4kyv
Contributing Member
Don
Member

Offline Offline

Posts: 10037



« Reply #10 on: January 06, 2012, 09:24:21 AM »

Wouldn't you need a midtap on the coil, or a dual section tuning capacitor with rotors grounded to chassis, to establish an rf ground reference for balanced feed to the deflection plates as depicted in the diagrams?
Logged

Don, K4KYV                                       AMI#5
Licensed since 1959 and not happy to be back on AM...    Never got off AM in the first place.

- - -
This message was typed using the DVORAK keyboard layout.
http://www.mwbrooks.com/dvorak
Opcom
Patrick J. / KD5OEI
Contributing
Member
*
Offline Offline

Posts: 8284



WWW
« Reply #11 on: January 08, 2012, 11:50:44 AM »

I'd add a swamping resistor in series to linearize it, and a load resistor to keep it tame.  If needed.

Did the punch arrive OK?

73DG

The punch arrived safe and sound and I thank you. It's an amazingly large punch alright. Hopefully the circuit will be worked out and tested in a few weeks, but if you need it back right away I can go ahead and punch it, but I would prefer to make sure the coupling and the CRT power supplies etc. are going to work well first.

I want to understand about the swamping in series, etc. because of the situation is driving a 3-4pF capacitive load (defl. plates) + the 300 Ohm TV set twin lead, so the reactance of that (10pF?) is some 5300 Ohms @3MHz, and for series resistance to linearize a frequency response, the swamp ought to be several times the load and this might bring voltage issues seeing the CRT needs about 200V to deflect. Plus, as the frequency increases, the Xc will become less, and the voltage on the plates will also decrease, so that at 30MHz, it may be 530 Ohms, could be 10dB down. So I do not understand what is the object of series resistance.

I have loads, such as a 25K, 2500, and other values of Carborundum resistors.

If series resistance is used, the load ought to be at the transformer, right? R1 position, not R2.
R3 and R4 would be the series swamping resistors, correct?



Wouldn't you need a midtap on the coil, or a dual section tuning capacitor with rotors grounded to chassis, to establish an rf ground reference for balanced feed to the deflection plates as depicted in the diagrams?

That is a very good point. It would seem to depend on the overall electrical balance and mechanical symmetry of the circuit. It can be done without much trouble. If the CRT is close to the transformer, maybe it will not need the capacitor for tuning and therefore no balancing. If it has to be balanced, then where to put the ground? I'm wary of making a ground loop right under the final amp chassis.

Don't CRTs generally tolerate small imbalances of deflection voltages as long as the net DC balance is correct? Many cheap mod scopes just hit one plate with the RF. I think it's a cheezy way to do it but no one ever complains about the patterns. "test equipment" scopes always go balanced.

In that junk haul yesterday there were some link couplers from SCR-274's, that is the ceramic coil with a rotating link inside it. Can that link can be made 1-turn of heavier wire, and develop voltage across the plate coil for the CRT? A load might be needed on the plate side, not sure about it. Sorry no good pic of this part.


* RF coupling to CRT_2.png (11.71 KB, 777x374 - viewed 747 times.)

* RF coupling to CRT balanced.png (14.89 KB, 771x379 - viewed 764 times.)

* RF coupling to CRT balanced goniometer.png (148.36 KB, 771x1000 - viewed 757 times.)
Logged

Radio Candelstein
W7TFO
WTF-OVER in 7 land Dennis
Contributing
Member
*
Offline Offline

Posts: 2472


IN A TRIODE NO ONE CAN HEAR YOUR SCREEN


WWW
« Reply #12 on: January 08, 2012, 01:29:23 PM »

Good news on the shipment.

I think you are on the right track, using inductive coupling.  Either a toroid or regular solenoid will be more stable over QSY than a cap coupling.

Balanced may not be necessary, especially considering the short distance from source to load.

As with all the variables involved, it will be designed in situ, 'cause "The map is not the territory".

73DG
Logged

Just pacing the Farady cage...
Opcom
Patrick J. / KD5OEI
Contributing
Member
*
Offline Offline

Posts: 8284



WWW
« Reply #13 on: January 13, 2012, 01:57:36 AM »

I hope this weekend to start working on the CRT circuit - the HV power supply and voltage divider, CRT controls, etc.

If I am lucky though, the HVAC here will be upgraded Saturday. An additional register will be added in the dining room and a return air with two 12x24 filter grilles will also be added. The current return is 12x24 and will remain, so that will be a 3X capacity (6 sq FT filter area) compared to the way it is now (2 sq FT) and the negative static pressure in the intake plenum should drop to half, resulting in a greater air volume. Furnace blowers do not like resistance in the intake. It is 60 in here now, I have to turn the heat on I guess.

back to the project, I got to do the calculations for the divider. also have to find some dual pots for beam positioning and a way to mount them all insulated since they may be some 500-1500V from GND.
Logged

Radio Candelstein
Opcom
Patrick J. / KD5OEI
Contributing
Member
*
Offline Offline

Posts: 8284



WWW
« Reply #14 on: April 01, 2012, 07:50:01 PM »

update - The rotating link couplers from SCR-274's can be refitted with a heavier wire that will take a a KW worth of current, but - the bearings/contacts will not tolerate it for AM cariers. There was a smell but ceramic and metal does not make smoke at that power level. A different means has to be found to rotate the coil. more later.

Three power transformers from scopes were found on the shelves and a nice one stood out having been tested making 2800V@2mA and a highly insulated 6.3V winding. A bridge should make 3900V and a doubler should make 3500/7000V (duncan simulator).

A 5400V 3mA power supply came from another shelf. It looks like a 6x4x5" oil cap and is potted but has 4 terminals, HV+/- output, and 110V input. It apparently has a tube rectifier inside because the HV does not come up until the primary is about 1/2 way up. The hope was to supply less volts to it and trim the 5400 down a bit. the total requirement is 10KV at 1.5mA most of which goes through the voltage dividers. If the 5400V supply is used, some capacitor clamps will have to be found that will hold a 6" tall capacitor case for it, and at least 1' wide ones as it is a bit heavy.

So the values are not precisely close to required voltages yet putting these in series. 5400+3900=9300, 5400+7000=12400. 9300V might work but IMHO the conditions should be made as close to those in the Tektronix 545 as possible. There are other monolithic capacitor-looking HV supplies here but they are physically too large, oil filled, and make too much voltage at 12-30KV, and it may not be a good idea to mount an oil filled part with the terminals on the side.

Before, taking some voltage from the rig was being considered, but not any longer because it has 'variacs'. Those voltages could be set anywhere from 2-3.5KV, and a CRT of this 'precision' type needs a fairly constant proportion of voltages to work well. The 545 regulates all of those voltages.

need
-1450
-1350
+225
+350
+500
+8650

The Zeners are in the schematic "because'.. I don't know if that is a reliable idea at all. I'd prefer VR tubes for stabilizing the lower voltages. I have only 1-2mA of current. What to do between -1450 and -1350 seems like a place for a Zener diode, but will it pop around those kinds of circuits and potentials?


* mod scope circuit cropped for Tucker.gif (10.37 KB, 784x670 - viewed 730 times.)
Logged

Radio Candelstein
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 0.12 seconds with 19 queries.