Forums :: Resources :: Features :: Photo Gallery :: Vintage Radio Shows :: Archives :: Books
Support This Site: Contributors :: Advertise


It is currently Dec Mon 10, 2018 7:38 am


All times are UTC [ DST ]





Post New Topic Post Reply  [ 40 posts ]  Go to page 1, 2  Next
Author Message
 Post subject: Circuit question
PostPosted: Mar Mon 19, 2018 2:55 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 19321
Location: Warner Robins, GA
Wasn't sure exactly how to title this.

I have a Stark LSG-10 RF generator I modded to produce a cleaner sinewave which involved some changes to the cathode follower circuit as seen in this schematic.

Attachment:
Stark LSG-10A currently wired 2.jpg
Stark LSG-10A currently wired 2.jpg [ 114.78 KiB | Viewed 1002 times ]


I'm using this Digital frequency display with the RF generator.

http://www.pongrance.com/fullmanual.pdf

The problem is it loads down the output somewhat and distorts the sinewave a little.

I'm thinking of adding this preamp to the DFD since it has a space for an extra IC chip that isn't connected to anything.

Attachment:
RF preamp .jpg
RF preamp .jpg [ 28.26 KiB | Viewed 1002 times ]


Will that still load down the circuit?

If it does would it be possible to take that chip wire it as a buffer and use that in place of the cathode follower since it is good to 60Mhz which is higher than what the dial can display?

Here's the data for the chip.

http://cds.linear.com/docs/en/datasheet/lt1252.pdf


Top
 Profile  
 
 Post subject: Re: Circuit question
PostPosted: Mar Mon 19, 2018 4:02 pm 
Member

Joined: Feb Wed 07, 2018 6:52 pm
Posts: 359
Location: Stone Mountain, GA
Probably the biggest load is the coax. You need a buffer at the source to drive the coax.

The amp might work with a short wire to the cathode follower.

_________________
Scott.
Powered by infinite improbability drive.


Top
 Profile  
 
 Post subject: Re: Circuit question
PostPosted: Mar Mon 19, 2018 5:13 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 19321
Location: Warner Robins, GA
Currently I only have a couple wires going to the output of the generator and not a coax.

I know that I do need to get the output impedance of the RF generator much lower as well, but with the cathode follower configuration I cannot without messing the sinewave up considerably.

I know that I do not necessarily need a clean sinewave, but I do want it to be as clean as possible.

Now would it be possible to modify that circuit with the chip and just replace the cathode follower with that chip?

If so I can use a voltage doubler circuit consisting of two diodes each one wired reverse from the other which will give me a + unloaded B+ of 8.0598Vdc and a - supply of 8.0598Vdc which should give me an output of 16.1196Vpp not accounting for how far below the rail to rail voltage the chip can actually do.


Top
 Profile  
 
 Post subject: Re: Circuit question
PostPosted: Mar Mon 19, 2018 5:24 pm 
Member

Joined: Apr Thu 14, 2016 8:25 pm
Posts: 411
Location: pensacola fl
I see where you can get a little more drive for the counter. Take it at the top of r8 pin 8 of the 12BH7. The impedance will be lower as well. Then add am MPF102 or similar in a source follower to isolate and drive the counter. You can power the fet from B+ through a dropping resistor and capacitor filter it with both an electrolytic and a ceramic disc or other rf friendly capacitor since electrolytics do not handle rf well.


Top
 Profile  
 
 Post subject: Re: Circuit question
PostPosted: Mar Mon 19, 2018 5:26 pm 
Member

Joined: Jan Thu 01, 1970 1:00 am
Posts: 2221
Location: Saskatoon
It seems very strange that the frequency counter is loading the cathode follower. What does the frequency counter input circuit look like? Typically, they are CMOS with an input impedance of 100's of Megohms.

Try connecting a 22k resistor between the cathode follower and the counter input and see what happens.


Top
 Profile  
 
 Post subject: Re: Circuit question
PostPosted: Mar Mon 19, 2018 5:29 pm 
Member

Joined: Feb Wed 07, 2018 6:52 pm
Posts: 359
Location: Stone Mountain, GA
The freq counter has an impedance of 50k ohms.

I agree a simple FET follower would be a simple solution.

_________________
Scott.
Powered by infinite improbability drive.


Top
 Profile  
 
 Post subject: Re: Circuit question
PostPosted: Mar Mon 19, 2018 5:38 pm 
Member
User avatar

Joined: Jan Fri 02, 2009 11:32 pm
Posts: 566
Location: Equinunk PA 18417
According to the link in the first post, the counter uses a 74HC/AE161 as a prescaler. That chip has 3.5pf input capacitance, directly connected to the counter input. I agree a series resistor and take-off to the counter directly at the tube's cathode, should reduce interaction. 22k might need to be reduced for better high freq. counting.

Honestly, although this is OT, I don't see the value in that generator circuit mod. It increases the generator's output impedance by a factor of 5 which will make it even more sensitive to a load.

edit: fet follower is good idea

_________________
-Richard


Top
 Profile  
 
 Post subject: Re: Circuit question
PostPosted: Mar Mon 19, 2018 6:03 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 19321
Location: Warner Robins, GA
The FET would work I think.

I get the following voltage range from the cathode follower.

13.15Vpp at 327.5KHz on band A
255mVpp at 30.7973MHz on band E

Not concerned with band F as the counter won't go that high.

Unfortunately doe to the oscillator circuit I will never get the output levels to be equal across all bands and all frequencies.

Yes Rich IU agree about the cathode follower mod increasing output impedance.

Was trying to avoid that, but at the time the only way I knew to solve it was properly biasing the cathode follower which meant increasing the cathode resistance and also applying a bias voltage to the grid.

Here's the stock schematic. Mine has two 3pF caps one conencted between band E and ground and the other betwen band C and ground.

Attachment:
stark_lsg-10a_rf_signal_generator_sch.png
stark_lsg-10a_rf_signal_generator_sch.png [ 111.06 KiB | Viewed 963 times ]



Now I can solve all my problems by getting a N3ZI DDS VFO kit for $99, but I don't quite have that much to spend yet.

I do still plan on installing the preamp on the board as the counter can be used for other things besides just this RF generator.


Top
 Profile  
 
 Post subject: Re: Circuit question
PostPosted: Mar Mon 19, 2018 6:13 pm 
Member

Joined: Feb Wed 07, 2018 6:52 pm
Posts: 359
Location: Stone Mountain, GA
Try putting it back to original. It will have a lower output impedance and may drive the counter without an additional buffer.

With the high resistance o the cathode, that is probably why you are getting much lower signals at high frequencies.

_________________
Scott.
Powered by infinite improbability drive.


Top
 Profile  
 
 Post subject: Re: Circuit question
PostPosted: Mar Mon 19, 2018 7:01 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 33558
Location: Maryland 20709, USA
Tube Radio wrote:
I have a Stark LSG-10 RF generator I modded to produce a cleaner sinewave...
There's no advantage in an RF generator to producing a "cleaner" sinewave.

That might be advantageous in an audio generator, but not in an RF generator.

- Leigh

_________________
73 de W3NLB
http://www.AtwaterKent.info
Click "Grebe Stuff" for Synchrophase info


Top
 Profile  
 
 Post subject: Re: Circuit question
PostPosted: Mar Mon 19, 2018 7:18 pm 
Member

Joined: Dec Fri 08, 2017 5:40 pm
Posts: 162
Location: 44141
Thank you Leigh.

_________________
Bruce Hagen


Top
 Profile  
 
 Post subject: Re: Circuit question
PostPosted: Mar Mon 19, 2018 7:33 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 19321
Location: Warner Robins, GA
Leigh wrote:
Tube Radio wrote:
I have a Stark LSG-10 RF generator I modded to produce a cleaner sinewave...
There's no advantage in an RF generator to producing a "cleaner" sinewave.

That might be advantageous in an audio generator, but not in an RF generator.

- Leigh


I know that, but need a relatively clean sinewave for the counter to work properly as the stock output of the Stark was more along the lines of how a 1/2 wave rectified DC voltage looked on band A and maybe B with the rest looking a bit better and at some frequencies had other oddities with the waveform such that the counter would often not display the correct frequency and the output wouldn't always supply enough drive for the counter to display properly.

The cleaner sinewave made the counter work much better and it displays all the correct frequencies.

SHenion, the lower output is present at the oscillator plate and grid as well and is just how the simple oscillator is.

I wonder if the tube doesn't have as much gain as the frequency increases or maybe as the coils get smaller the coupling gets lower and lower which would produce less output.

I do know that the sinewave on the oscillator plate and/or grid is a clean sinewave so it is possible to produce a clean sinewave. I'm thinking perhaps one of those chips could serve as a replacement for the cathode follower as it would be able to drive a low impedance load and maybe I could figure out some sort of filter that will reduce the output of the lower bands, but that would introduce a phase shift.


Top
 Profile  
 
 Post subject: Re: Circuit question
PostPosted: Mar Mon 19, 2018 7:57 pm 
Member
User avatar

Joined: Jan Fri 02, 2009 11:32 pm
Posts: 566
Location: Equinunk PA 18417
I like the idea of a simple fet buffer assuming it has enough drive for your counter and cables. Please consider that a buffer's input should be connected by the shortest wire possible, preferably mounted inside the gen close to the tube, otherwise you're making even more trouble for yourself.

Pardon me being OT again, I also notice you changed the rf feedpoint from the oscillator's plate to its grid. I don't understand the circuit well enough, why was that change made? Also, why those two 3pf caps? Did those bands need padding?

Truth is, I've contemplated some sort of amp/buffer scheme for my poor little Heathkit rf gen. It's 120mv or less output is plenty for radio servicing but there are times I want more.

_________________
-Richard


Top
 Profile  
 
 Post subject: Re: Circuit question
PostPosted: Mar Mon 19, 2018 8:24 pm 
Member

Joined: Jan Thu 01, 1970 1:00 am
Posts: 2221
Location: Saskatoon
The reason why I suggested the isolation resistor is that it's possible that you are overdriving the CMOS input to the point where the protection diodes are starting to conduct, and that may be the reason why it's causing distortion of the waveform.


Top
 Profile  
 
 Post subject: Re: Circuit question
PostPosted: Mar Mon 19, 2018 9:01 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 19321
Location: Warner Robins, GA
I do know the input impedance of the counter is at least 50K.

I could try a series resistor, but I'm thinking it might be best if I were to replace the cathode follower.

Would it be possible to take the preamp circuit, make it a buffer and increase the two 10K resistors to 1Meg while still being able to properly handle a 40MHz signal?

The original circuit shows C10 as a 3pF cap and that may have been their attempt at reducing the output as the frequency went lower or perhaps to provide less loading, but it also affected the waveform at the lower frequencies so I found a 30pF cap produced the best sinewave across the whole range of frequencies aside from the highest band which was already lower in output and is even lower after the mod, but that band covers frequencies I'll never use and I got an FM generator for 88-108MHz.

I might try disconnecting the 30pF cap tomorrow and see what the waveform looks like on the plate and grid of the oscillator.

The modulation scheme is not very good so I may not use it to reduce load on the small power transformer the generator has.

Wonder if it is possible to for instance install a 6BA6, 6BE6 or other 7 pin pentode in place of the audio oscillator pentode then use a diode to produce a - voltage to control the gain like what AGC would do. I could then feed a FET buffer. Wonder if I could just feed the - voltage to the triode grid and wire it up for gain then feed a FET buffer? The idea there would be to keep the output voltage fairly constant.

There is a wide voltage range though so it may not work like I think it would.


Top
 Profile  
 
 Post subject: Re: Circuit question
PostPosted: Mar Mon 19, 2018 10:30 pm 
Member

Joined: Jan Thu 01, 1970 1:00 am
Posts: 2221
Location: Saskatoon
Tube Radio wrote:
Would it be possible to take the preamp circuit, make it a buffer and ...

Tube Radio wrote:
Wonder if it is possible to for instance install a 6BA6, 6BE6...

Would it be possible, for instance, to just try the resistor, or are you determined to make this thing as complicated as you possibly can?


Top
 Profile  
 
 Post subject: Re: Circuit question
PostPosted: Mar Mon 19, 2018 10:59 pm 
Member

Joined: Feb Wed 07, 2018 6:52 pm
Posts: 359
Location: Stone Mountain, GA
The problem is not the waveform, it is the counter input. It does not have proper input conditioning or hysteresis.

Put the generator back to stock, then figure a better input to the counter. I'd use a buffer with feedback to add a bit of hysteresis.

As is, you put a 50ohm cable on it and it will load the signal way down and probably ring.

_________________
Scott.
Powered by infinite improbability drive.


Top
 Profile  
 
 Post subject: Re: Circuit question
PostPosted: Mar Tue 20, 2018 2:00 am 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 19321
Location: Warner Robins, GA
BobWeaver wrote:
Tube Radio wrote:
Would it be possible to take the preamp circuit, make it a buffer and ...

Tube Radio wrote:
Wonder if it is possible to for instance install a 6BA6, 6BE6...

Would it be possible, for instance, to just try the resistor, or are you determined to make this thing as complicated as you possibly can?


No it's just that any resistance unless it is a high value will load the output down somewhat and for a 22K resistor I'd most likely have to increase the output on the higher bands for the counter to display right.

Once I get the preamp built for the counter that won't be an issue and I can try the resistor.

But I had figured if there was a way to make the output more even in level across all the frequencies that would be more beneficial than just trying to make the counter work properly with the RF generator as is.

I also want the lower impedance output like it should be. Am thinking 50 ohms or if I used one of the example circuits shown on the datasheet for the chip I could have a 75 ohm output impedance.

I can put it back stock, but then the output on the two lower bands goes back to looking like a 1/2 wave rectified waveform and the other bands look weird again. Personally I don't like that as I prefer a nice clean as possible sinewave like better RF generators produce.

What I may do tomorrow with the Stark is insert the scope probe tip in each coil and see how strong the signal is for each coil.

I guess in a way this is a see what I can do to improve this circuit kind of thing.


Think I'm trying to make a silk purse out of a sow's ear, but it's all I have to work with at the moment.



I can say this. The Stark is way better than the EICO 315 I have as I tried to fix the cathode follower in the EICO and got such poor results that I gave up because the oscillator itself was driven so hard that it was distorting.


Top
 Profile  
 
 Post subject: Re: Circuit question
PostPosted: Mar Tue 20, 2018 12:41 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 19321
Location: Warner Robins, GA
So looking at both schematics this morning I noticed what I thought was an error on my part until I realized that's how mine was wired from the factory. It has to do with the cathode follower grid.

The one schematic shows C10 connected to the grid and the junction of R6 and R14 connected to the other side of R10/C8.

I tried that and it didn't make any difference other than to increase the output to where I was seeing distortion again on band A.

Apparently what was done with mine was a design change in an attempt to correct the issue of severe overdriving of the cathode follower or maybe they found it caused the output levals to be more equal for each band.

What I may do is disconnect C10 then use my scope to measure the output level and see what it looks like and see how much of an amplitude change I still have.

I will also try the resistor coupling the counter just to see if it will work.



EDIT:

Here's some measurements.

Attachment:
Measurement.jpg
Measurement.jpg [ 18.88 KiB | Viewed 878 times ]


Now looking at the last set of measurements wheere the scope probe is in the band E coil the output is low, but the output also does not vary much. Now if I amplified that the voltage variation gets larger, although I can more easily work with that signal level versus the signal from the oscillator grid.

As I go lower in frequency though the sinewave starts to get not as clean looking as it once was, but it isn't bad and I can live with that.

Now if I remove the audio tube the outpout increases some due to the lighter loading on B+ and the power transformer which increases the output slightly.

Now if I wanted to add AGC I would need to step up the voltage I get from a wire placed in band E coil then use that signal after going through a 1 meg resistor and diode to form some AGC voltage then use that voltage to control an active device, although I am thinking that perhaps I could use the preamp circuit I posted a schematic of in the first post and figure a way to control the gain based on a rectified voltage or even figure a way to control the gain in the feedback circuit based on frequency.

What is the average maximum output voltage of good quality RF generators?

I want to get mine to where it is the same output level so that I am sure I have enough signal level for any alignments I need to do. Also the output impedance needs to be 50 ohms if possible.


Top
 Profile  
 
 Post subject: Re: Circuit question
PostPosted: Mar Tue 20, 2018 2:20 pm 
Member
User avatar

Joined: Jan Fri 02, 2009 11:32 pm
Posts: 566
Location: Equinunk PA 18417
Edit: you added to your post as I was writing.

There is no R14 in the original schematic, which you also posted. Just so we're all on the same wavelength, this is your schematic with all deviations from the original marked. The change in output attenuator raises the generator's output impedance from 500/5 ohms (low output) to 10k/10k ohms, changes not desireable for an rf source. The follower's bias was apparently raised quite a lot. You mentioned doing this. Then, there is the change to C10's value as well as connection point to take oscillator output off of a part of the tuned grid circuit instead of its plate. Finally, what is the purpose of the two added 3pf caps?
Attachment:
Stark TubeRadio mod questions.jpg
Stark TubeRadio mod questions.jpg [ 129.12 KiB | Viewed 875 times ]


edit: changed schematic markup to include C10

_________________
-Richard


Last edited by richfair on Mar Tue 20, 2018 3:20 pm, edited 2 times in total.

Top
 Profile  
 
Post New Topic Post Reply  [ 40 posts ]  Go to page 1, 2  Next

All times are UTC [ DST ]


Who is online

Users browsing this forum: No registered users and 8 guests



Search for:
Jump to:  
























Privacy Policy :: Powered by phpBB