The AM Forum
March 29, 2024, 01:16:07 AM *
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: Gates BC1G crystal oscillator  (Read 4056 times)
0 Members and 1 Guest are viewing this topic.
k7mdo
Contributing
Member
*
Offline Offline

Posts: 420


« on: April 28, 2019, 08:16:16 PM »

I'll try to keep this short.

I am in the process of converting my Gates BC1G to 75 meter phone for local AM rag chews. 

I have all of Timtrons information for a BC1T but there is a little piece that may be missing and I am novice enough to only know to ask!

The oscillator module has two 12BY7A tubes, one is the crystal oscillator and the other is a buffer/amplifier.

The original frequency was 1580 kc and I have easily been able to bench test it to 1950 kc.  In fact it looks to be fine there but when I install a 3870 kc crystal, no oscillation. 

I see how to alter the oscillator section by changing a couple of capacitors but the little "tank circuit" of the buffer/amplifier is a little more of a mystery.  It is comprised of a slug tuned little phenolic based coil with a 100 pf capacitor in parallel.  I tried grid dipping the unit with little success but I know from other reading that some folks have removed "some" of the turns off of the little coil.

I think I can handle this but there is another thing I see in there that has not been mentioned in the literature, namely there is what appears to be a "capacitive cap" at the top of the junction of the coil and capacitor.  It is a 1 1/8" pigtail of stiff wire leading off into space.

I think if I remove the coil/capacitor combination I will be able to characterize its values with my test equipment but "what about that dang "pig-tail?"

Also I have to eventually dig into the final output tank circuit but I am saving that for later.

Who has struggled with this before?

Thanks, 73, Tom


Logged
DMOD
AC0OB - A Place where Thermionic Emitters Rule!
Contributing
Member
*
Offline Offline

Posts: 1768


« Reply #1 on: April 28, 2019, 11:44:22 PM »

The oscillator is a broadband oscillator with a capacitive voltage divider on the grid.

You may have to reduce C4 to 150 pF in order to get the oscillator to function.

The buffer-amp is a tuned plate circuit as shown below.

The tuned circuit is a parallel resonant circuit. If you want to use the present C9 then you need an L3 coil measuring about 13 uH, or attempt to adjust it to minimum. You need to measure the L3 coil out of circuit and determine if you can get down to 13 uH. Otherwise, you may have to roll your own.

If you use a 220 pF capacitor for C9, then you can use a lower inductance such as ~ 8.4 uH.

I would try changing C9 to 68 pF first before doing any coil winding. A 68 pF capacitor in parallel with a 27.2 uH coil resonates at 3.7 MHz.


Phil - AC0OB

 

* BC-1G XV2 Stage.pdf (102.91 KB - downloaded 144 times.)
Logged

Charlie Eppes: Dad would be so happy if we married a doctor.
Don Eppes: Yeah, well, Dad would be happy if I married someone with a pulse.NUMB3RS   Smiley
k7mdo
Contributing
Member
*
Offline Offline

Posts: 420


« Reply #2 on: April 29, 2019, 01:50:24 PM »

Great help Phil!  I now have more confidence to move ahead....  I had already searched the old Radio Handbooks and knew C4 would need to change and probably C3 but I will start with your suggestion at about 150 pf.

I think I will first try the 68 pf on L3 to see what I get.... then if it isn't satisfactory I will remove L3 to see what it measures and see what I can to get the output needed.

I have set up a "test" bench with an old Heath power supply so that I can see what the output looks like on the scope and have recorded output levels of the original 1580 kc set up.

 
I may have some useable replacements for the L3 coil in my "warehouse" from other projects.


There is still the mystery of teh 1" wire leading "no where" off of the top of L3 at its junction with C9.  The diagram shows a dashed line kind of leading away from C11 which has a note: "connection determined by frequency" and so I have to sort this out yet.

I will take small steps (and the easier route) first and report.

Thanks again, 73, Tom
Logged
W1NB
Member

Offline Offline

Posts: 72


« Reply #3 on: April 29, 2019, 08:10:05 PM »

I suspect the mystery wire is a “gimmick”, a short stub of wire used to add a small amount of capacitance. Probably used to fine tune the parallel resonant circuit.
Logged
k7mdo
Contributing
Member
*
Offline Offline

Posts: 420


« Reply #4 on: April 29, 2019, 11:27:05 PM »

Some progress!

I replaced C4 with a 200 pf and now the oscillator actually "oscillates" on frequency, 3880 kc.  However, as you might expect the core of L3 has to be backed off as far as it will go and the total output is about 1/5 of the original output achieved with the 1580 kc crystal...  ie  It has lower output.

My oscillator has a variable (140 pf) cap for C-9 but in the current configuration of components it does virtually nothing on rotation.  There is an additional cap C-11 that is in parallel with C9 and it appears (un measured yet) to be 100 pf per the diagram I have.  thus my capacitor range on the tuned circuit is 100-240 pf in parallel with the L3 coil.


I think the next step is to cut out C-11 to drop the capacitance and see what happens. 

Unfortunately the L3 coil will be a bear to remove if it becomes necessary and yes, I think the wire hanging in free space is a "gimmick" wire.


Tom





Logged
DMOD
AC0OB - A Place where Thermionic Emitters Rule!
Contributing
Member
*
Offline Offline

Posts: 1768


« Reply #5 on: April 30, 2019, 12:57:09 AM »

I have the complete exciter-driver assembly for a BC-1T and the exciter circuitry is identical but the BC-1T schematics and parts list have more detail, so here goes.

The L3 coil is variable from 105 to 205 uH. The C9 capacitor is variable from 6.7 to 140 pF.

Remove any other capacitor such as C11 from across C9 because you have too much capacitance across the L3 coil. Suggest we not concern ourselves with pigtails and such as we need to clean up the XV2 Buffer-Amp plate load circuit.

So as per the schematic and my MatLab calculator you need to adjust L3 to 168 uH (about mid-range), and adjust C9 to 10 pF and that will give you 3.880 MHz.

Of course you can go slightly lower in inductance and slightly higher in capacitance as well, so I would experiment with jockeying both L3 and C9 for the highest RF output.

For example, if C9 = 15 pF and L3 = 112 uH, you still get 3.880 MHz.


Quote
I replaced C4 with a 200 pf and now the oscillator actually "oscillates" on frequency, 3880 kc.

Also, it wouldn't hurt to lower C4 to 150 pF.


Phil - AC0OB  

* BC-1G XV2 Stage.pdf (104.31 KB - downloaded 129 times.)
Logged

Charlie Eppes: Dad would be so happy if we married a doctor.
Don Eppes: Yeah, well, Dad would be happy if I married someone with a pulse.NUMB3RS   Smiley
k7mdo
Contributing
Member
*
Offline Offline

Posts: 420


« Reply #6 on: April 30, 2019, 05:48:36 PM »

Good information as always Phil.  I removed C-11.

I think I am getting stuck with doing something to lower the L3 value...  with it adjust "as low as possible" and C9 at its lowest capacitance I get a good solid signal out of the oscillator/buffer amp but no where near the signal strength that I got originally with the 1580kc crystal and all original parts...  I think on the scope today I measured a clean signal but only about 5 volts peak to peak...  as original I was getting maybe 10 times that on the scope.  

I don't really know what the "minimum" value of drive that will be required by the 807's.

I am going to experiment with some lower frequency crystals to see if the coil/cap combination comes into range of adjustment tonight....  

On another note, it seems like the cathode resistor R7 should have an RF bypass cap?

EUREKA

Ignore the above winging.....  I removed L3 from the circuit board, unwound 1/2 of the windings (with trepidation), found that the soldering iron with flux would bare and solder all 5 tiny strands at one time by experiment... replaced the modified L3 and was astounded at the success!!!!  I now have the same amount of drive I got with 1580 kc crystal in the original configuration!  Lots of drive...  maybe 50 volts peak to peak coming out of the unit now at 3880kc.... !!!!

Thanks again to Phil (DMOD)....  !!!!
Logged
DMOD
AC0OB - A Place where Thermionic Emitters Rule!
Contributing
Member
*
Offline Offline

Posts: 1768


« Reply #7 on: May 01, 2019, 12:21:01 AM »

Well, we don't know if we're getting enough voltage from the oscillator stage XV1 to the grid of XV2.

The original C4 capacitor was 300 pF at 1.5XX Mhz. Now we're going up twice that Freq. so 150 pF I felt was a good starting point and maybe 100 pF would be even better. With C4 = 100 pF we still have 150 pF across the crystal.

If you are using an FT 243 crystal, in my experience, they prefer 50 pF across 100k. So you may still have too much capacitance across the crystal. Remove the current C4 200 pF cap and see what happens.

For the XV2 stage putting a 0.01 uF across the cathode resistor should reduce feedback voltage and increase gain slightly.

Putting a capacitor in series with C9 reduces the total capacitance across L3. Placing a 220 pF in series with C9 will get you 0 to 86 pF. But again, for C9 = 15 pF and L3 = 112 uH, you should get resonance at 3.880 MHz.

For the 807 grids, you need at least 100 volts p-p in class C with a grid bias on them of -47 volts. What is the 807 grid bias voltage?

Also, make sure the XV2 screen voltage is at least 105 volts.


Phil - AC0OB


* BC-1G XV2 Stage.pdf (107.36 KB - downloaded 106 times.)
Logged

Charlie Eppes: Dad would be so happy if we married a doctor.
Don Eppes: Yeah, well, Dad would be happy if I married someone with a pulse.NUMB3RS   Smiley
k7mdo
Contributing
Member
*
Offline Offline

Posts: 420


« Reply #8 on: May 01, 2019, 11:23:53 PM »

Phil, I think we are there!  The output of the oscillator now measures 200 volts peak to peak.  I thought I would summarize what was done for someone else in the future.

1. I ended up removing L3 from the circuit board and removing 1/2 of the windings.  
2. I changed (out of my junk box) the C9 Variable capacitor measured at 140 pf and installed a 100 pf one that I had with the same "foot print".  I think it would have been easy enough to simply cut off some of the original capacitor's plates to achieve the same thing but I had the smaller one. I think even a 50 pf would probably have worked fine.
3. I changed capacitor C3 to 12 pf and C4 to 150 pf.
4. I removed C11 entirely and the "gimmick" wire. C11 paralleled L3

At this point I can easily peak the FT-243 crystals to an output of 200+ volts peak to peak and have tested my two main frequencies 3870 and 3885 by comparing the output when switching the crystal selector switch on the unit with no apparent change in tuning.


 I made a "test' harness for the unit and using an old Heathkit power supply was able to bench test the changes as I progressed.  I will keep that harness and if anyone else wants to use it, let me know.

 



* gates oscillator.JPG (144 KB, 640x424 - viewed 219 times.)
Logged
k7mdo
Contributing
Member
*
Offline Offline

Posts: 420


« Reply #9 on: May 02, 2019, 10:14:46 PM »

Even though others have not referenced the next stage of this conversion, it is to adjust the L8 (L4 on some schematics) coil between the RF driver and the 833a's.  I am able to dredge up a lot of information on the internet but no one seems to explain the process other than to have you remove a bunch of turns from the original and move the taps a bit.  

I am more inclined to make my own L8(L4) from coil stock from the warehouse.  I have an "equivalent" diameter coil stock length to the L8 coil in the Gates.

The L8(L4) in the Gates has 45 turns and measures 81 uH.  If I follow the internet directions I remove turns until there ae only 30 left and that would measure 46 uH.  I am able to make an equivalent measurement with my coil stock with just about the same number of turns.


If I mock up the coil (45 turns) paralleling the variable capacitor (C7 250pf or C13 on some schematics), I get (with an REA analyzer) a resonance at about 1000 hz.  Which makes some sense as the original frequencies for these Gates centered on the broadcast band.


Here is where I get a little behind the knowledge area!  No one references changing the C7 variable cap along with the coil L8 to achieve resonance somewhere near 3880 hz.  I am also completely novice as to the selection of the taps for the grids of the 833's.  seems like they just use a segment of the L8 coil and mysteriously move it "down a tap" or so after removing 15 turns from the coil.  


I dug through a bunch of old Amateur Radio Handbooks from the 1950's but they seems to ignore the topic as though it should be obvious!?


Insights?


* coil L8.JPG (4137 KB, 3776x2520 - viewed 258 times.)
Logged
DMOD
AC0OB - A Place where Thermionic Emitters Rule!
Contributing
Member
*
Offline Offline

Posts: 1768


« Reply #10 on: May 05, 2019, 01:33:23 PM »

The Mighty BC-1G/ (L4)/BC-1T (L8) High Q Driver Coil is a sight to behold and one to duplicate, but never to modify, IMHO.

If the OP decides to resonate the coil with a variable capacitor, he has 9 inductance value options because of the numerous taps.

The self capacitance of the driver coil is ~ 2.7 pF so if you calculate the resonating capacitance needed you should subtract about 2.7 pF from it.

If one were to select, as an example, the 27 uH tap for resonating with a variable capacitor of ~ 60 pF, make sure the turns are shorted from the 27 uH tap through to the top tap.


Phil - AC0OB



* Migty L8 (L4) driver coil.pdf (105.06 KB - downloaded 107 times.)
Logged

Charlie Eppes: Dad would be so happy if we married a doctor.
Don Eppes: Yeah, well, Dad would be happy if I married someone with a pulse.NUMB3RS   Smiley
k7mdo
Contributing
Member
*
Offline Offline

Posts: 420


« Reply #11 on: May 06, 2019, 01:31:04 PM »

If I read you right, you are suggesting not removing any turns from the coil (just tapping it down some and effectively shorting the "unused" portion)...  and in that case, the paralleling cap (which is 240 pf) could be persuaded into resonating at 3880...  possibly by just moving the taps on the coil.  Probably in the case of my original setup which has a .002 in parallel with variable, I need to remove it (the .002) entirely and possible even series it with the 240 pf to get to a resonance range with the new coil tap positions..

It seems most people who posted on the internet did remove coil windings but they never mention the capacitor....  little help.


Sorry Phil if I am misreading your response.


Logged
DMOD
AC0OB - A Place where Thermionic Emitters Rule!
Contributing
Member
*
Offline Offline

Posts: 1768


« Reply #12 on: May 08, 2019, 12:57:55 AM »

If I read you right, you are suggesting not removing any turns from the coil (just tapping it down some and effectively shorting the "unused" portion)...  and in that case, the paralleling cap (which is 240 pf) could be persuaded into resonating at 3880...  possibly by just moving the taps on the coil.  Probably in the case of my original setup which has a .002 in parallel with variable, I need to remove it (the .002) entirely and possible even series it with the 240 pf to get to a resonance range with the new coil tap positions..

It seems most people who posted on the internet did remove coil windings but they never mention the capacitor....  little help.

Sorry Phil if I am misreading your response.


You got it. Why butcher it up or remove windings with so many taps available?

Remove any extraneous capacitance across it and apply the correct capacitance to tune it.

With that many taps, you have 9 choices, but I would use the 27 uH tap with the capacitance shown and short out all other windings from 27 uH to the top as per the schematic.


Phil - AC0OB

* BC-1T BC-1G Migty L8 (L4) driver coil.pdf (105.06 KB - downloaded 97 times.)
Logged

Charlie Eppes: Dad would be so happy if we married a doctor.
Don Eppes: Yeah, well, Dad would be happy if I married someone with a pulse.NUMB3RS   Smiley
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 0.043 seconds with 19 queries.