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


It is currently May Sat 30, 2020 12:35 am


All times are UTC [ DST ]





Post New Topic Post Reply  [ 39 posts ]  Go to page Previous  1, 2
Author Message
 Post subject: Re: Ground loop question
PostPosted: Apr Tue 07, 2020 3:11 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 20994
Location: Warner Robins, GA
Perhaps I can do that, but would it be better to just get one of those Traco +/- 9Vdc isolated power modules as that would provide the proper +/- power supply for the circuit and also for the OP-AMP I use to buffer the vertical and horizontal outputs.

Given the circuit will be in the same box as the converter, is the 75 ohm resistor to ground at the input of the converter really necessary or can I eliminate that and just have the 75 ohm load at the input to the isolator circuit?


Top
 Profile  
 
 Post subject: Re: Ground loop question
PostPosted: Apr Wed 08, 2020 1:46 am 
Member

Joined: May Sun 07, 2017 11:35 am
Posts: 840
Location: Belrose, NSW, Australia
OK if you are going to change the power supply and the output of the diff amp is directly coupled into this circuit you can use the original diff amp circuit as shown, with all 10k resistors and unity gain. Just drop the 75 ohm load on the input to your circuit and put it on the input to the diff amp. This works because you no longer need to drive a back-terminated cable with the diff amp.

The diff amp spec can be relaxed too, even the old LM318 could now be used since it doesn't need to drive a cable. The op amp needs at least 70v/us slew rate and a unity gain bandwidth of at least 5MHz.

_________________
Wax, paper, bitumen, cotton, high voltages - what could possibly go wrong?


Top
 Profile  
 
 Post subject: Re: Ground loop question
PostPosted: Apr Wed 08, 2020 2:55 am 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 20994
Location: Warner Robins, GA
So I would basically keep the OP-AMP used for the brightness and contrast on a single supply, right?

That would greatly simplify things, because instead of a +9 and +12 supply I only need a +/- 9 supply.

Now I could make that with a 7809 and 7909 as that would be cheaper than the Traco module and might be better because the Traco module likes for the +/- outputs to have an equal load in order for both to have the same output voltage.


Last edited by Tube Radio on Apr Wed 08, 2020 3:13 am, edited 1 time in total.

Top
 Profile  
 
 Post subject: Re: Ground loop question
PostPosted: Apr Wed 08, 2020 3:10 am 
Member

Joined: May Sun 07, 2017 11:35 am
Posts: 840
Location: Belrose, NSW, Australia
"If it ain't broke, don't fix it"!

_________________
Wax, paper, bitumen, cotton, high voltages - what could possibly go wrong?


Top
 Profile  
 
 Post subject: Re: Ground loop question
PostPosted: Apr Wed 08, 2020 3:44 am 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 20994
Location: Warner Robins, GA
Agreed.

Just wondered if putting the contrast and brightness OP-AMP on a dual supply would make it better.


Top
 Profile  
 
 Post subject: Re: Ground loop question
PostPosted: Apr Wed 08, 2020 3:59 am 
Member

Joined: May Sun 07, 2017 11:35 am
Posts: 840
Location: Belrose, NSW, Australia
Well if you put the whole circuit on dual supplies you could have the signals all swing around 0 volts and avoid the need for some caps.

_________________
Wax, paper, bitumen, cotton, high voltages - what could possibly go wrong?


Top
 Profile  
 
 Post subject: Re: Ground loop question
PostPosted: Apr Wed 08, 2020 12:47 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 20994
Location: Warner Robins, GA
How would I do that with the circuit?

I'd use a +/- 9Vdc supply.

In the circuit's current state I wouldn't be able to use individual regulators unless I replace the bridge rectifier with a pair of diodes to create the +/- supply.

Now would the OP-AMP used for the contrast and brightness also work for the isolator?

Instead of trying to modify the original circuit I'll just completely rebuild it with all new parts and I'll see if I can find a proper metal box for the circuit which would require using an isolated RCA jack or perhaps I could use an isolated BNC jack and just use an RCA to BNC adapter with the other jacks also being BNC.

A complete rebuild would allow for a much better component layout as well.


I'd use the following power module.

https://www.mouser.com/ProductDetail/TR ... wM2Q%3D%3D

That was the closest they had in a regulated +/- power module. They had a unregulated +/- 9Vdc module, but I didn't want that one.

What would need to change to the circuit given the voltage is now higher than the 9Vdc it was designed for?

I did find an old post where I found one video to oscope adapter that simply fed the video to the Z axis input of a scope which IIRC seemed to work better when I tried it a good while back.

Perhaps I'll try it again and see how that works.

If it works good enough I may not need the brightness and contrast controls and can simply feed the video to the circuit and also to a jack for going to the X/Y display.

Somewhere I found the aspect ratio of my Tektronix 606A X/Y display and what resolution the computer should be set to, but that requires a VGA to X/Y/Z adapter which I would have to build unless there's a commercially available product for a reasonable cost.

EDIT:

I tried connecting the luma signal to both the box and display. I got a much better picture.

With the box the brightness/contrast would change depending what was on the screen. I would have to adjust the bright and contrast controls different for when something had a lot of lighter colors such as white and would have to adjust it different when something had a lot of dark colors in it such as black.

So that tells me the brightness and contrast circuit isn't as good as it should be.

Any ideas on how to improve it?

I do notice that as I turn up the brightness on the display the picture gets brighter and more washed out sort of like a brightness control on a B/W TV does, but the contrast stays the same.

So that tells me I need to eliminate the brightness portion of the circuit and just keep the contrast portion using the display's brightness control for the brightness.

That would also free up an OP-AMP section with enough bandwidth for the isolator so that I only need three chips instead of four.

Now the LM1881 only operates off a single supply so I may just do a regular +9V regulated supply for that and the sawtooth generators unless +12 volts would work better for the sawtooth generators.

Now once I figure out the aspect ratio of the display I may try to figure out what to change in the sawtooth generators to make it work with VGA vertical and horizontal sync signals if that is even possible.

Also for the VGA signal can I just use the green signal for the video or do I need to use resistors to mix the red green and blue?

Also I'd need to incorporate a circuit that would switch an outlet off if video or vertical and horizontal signals are lost. That way the CRT is protected if I lose a signal or something goes wrong with the circuit.

I could do a spot killer, but with the brightness controlled by the display that would only work if the circuit is connected and wouldn't offer any protection if the circuit got disconnected in any way shape or form.

Now the X/Y display does have a normal and inverted X Y and Z input. Would thee be any benefit to making the signals balanced as in would that be another way to eliminate the ground loop or would it provide any benefits far as picture quality is concerned at least for the video?

For the vertical and horizontal I would need to buffer the signals so that the size pots would still work properly without loading down the sawtooth generators.

Also is there a better choice for the TL-082 OP-AMP? Perhaps the same OP-AMP used for the brightness and contrast circuit would work?

So if I do the balanced outputs I will go from three chips to 5 chips. Also with a balanced video signal I may only need the video amp to be unity gain which means greater bandwidth for the OP-AMP video circuit.

Now the contrast control is a 1K pot. Would there be any benefit to using a lower value pot in that location?

What if I went with 100 ohms and added a parallel resistor so that the video input sees 75 ohms?

I figure with the balanced outputs any ground loop would be eliminated. Am I right?

The X/Y display runs a bit warm so I placed a 5" 12 volt fan over the left side where the deflection circuitry is and it didn't affect the CRT at all.

So I'll find a 3" fan and cut a hole for it in the cover then mount it to the top. I won't power it from the display though.


Top
 Profile  
 
 Post subject: Re: Ground loop question
PostPosted: Apr Thu 09, 2020 8:49 am 
Member

Joined: May Sun 07, 2017 11:35 am
Posts: 840
Location: Belrose, NSW, Australia
Whoa, hold up there! Too many questions!!

One thing you should do with your Z / Luma signal is clamp it, there is an output from the LM1881 you can use to drive an active clamp. That will fix your brightness / black level shift issues.

I'll see if I have time to review your circuit and questions and come up with something.

It really deserves to have a PCB laid out for it. I NEVER build perf board prototypes any more, it's so cheap to get prototype PCBs made.
I use Protel / Altium, it picks up a lot of your mistakes before a board is made and the prototype usually works first go.
Do you have access to any similar tools?

A very good free tool I highly recommend is LTSpice. You can model your circuit and measure voltages, waveforms etc. before you build it. Well worth spending the time getting to learn it.

_________________
Wax, paper, bitumen, cotton, high voltages - what could possibly go wrong?


Top
 Profile  
 
 Post subject: Re: Ground loop question
PostPosted: Apr Thu 09, 2020 11:31 am 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 20994
Location: Warner Robins, GA
I honestly think the brightness black level shift issues were caused by the brightness/contrast circuit as when I connected the luma signal with an RCA Y splitter to the X/Y display and the circuit, I got a much better picture with no brightness change whatsoever between things with a lot of light colors and things with a lot of dark colors.

The way the brightness contrast circuit works I think is the video signal is applied to the - OP-AMP input and the second OP-AMP is wired up as a buffer and if fed a DC voltage which is controlled by the brightness pot and its output feeds the + input of the video amp OP-AMP. Probably not the best way to do it.

But with the modified circuit the only brightness control is the intensity control of the display itself.

The contrast control (adjusts the applied video signal to the video amp OP-AMP) will then be the only control used as that will vary the video signal.

I'll post an updated schematic in a bit.

I do have a decent sized piece of perfboard which has the solder pads on one side. I'm thinking it's at least 4" X 4". I'll use that to build the new circuit on and I'll lay it out better which also may include bypass caps on the power rails of each chip if that is necessary.

I found way to simplify the circuit by using fully differential OP-AMPs

http://www.ti.com/lit/ds/symlink/ths4131.pdf

I'll need a SMD to DIP adapter which Mouser has.

https://www.mouser.com/ProductDetail/Ch ... KwkQ%3D%3D

I can use that for the X and Y outputs, but I still need a buffer for the size pots which are 10K multi turn panel mount.

So I'm not saving any ICs, but I am going to get better performance.

EDIT:

I had sent an updated schematic to my home email from work but it never got there so ill post it later today at work.

Now using that differential OP-AMP chip, can I still connect the input ground to the - input with the 100 ohm contrast control paralleled with a 300 ohm resistor to make the 75 ohm input load and have the sync separator circuit still work properly?

I suppose one of the sync separator chip pins can be used to determine if there's a video signal, right?

So I'd then need a circut that would turn a relay off in the absence of a video signal.

I can get an IEC jack and I have a chassis mount standard single 120 volt wall plug so everything would be self contained.

For the power transformer Im thinking a small 12Vrms torroidal would be best as that type keeps most of the field in the core and would have less potential effect on the circuit.

For what it is the original circuit is ok, but I don't think it was ever desgned to be the type of high quality I am expecting as it was originally for displaying a video signal on an oscilloscope and I just happened to find that an X/Y display worked ok for that.


Top
 Profile  
 
 Post subject: Re: Ground loop question
PostPosted: Apr Fri 10, 2020 3:40 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 20994
Location: Warner Robins, GA
Here's the updated schematic.

Attachment:
Video to oscope adapter 1.jpg
Video to oscope adapter 1.jpg [ 407.01 KiB | Viewed 332 times ]


Attachment:
Video to oscope adapter 2.jpg
Video to oscope adapter 2.jpg [ 479.41 KiB | Viewed 332 times ]


The nice thing about this circuit now is I can use it with most any X/Y display or scope regardless of if it has balanced or single ended inputs and it doesn't matter if the device inverts the signals either.

EDIT:

Here's how the raster looks. Not sure if it looks right or not.

Attachment:
20200410_133757.jpg
20200410_133757.jpg [ 312.73 KiB | Viewed 327 times ]


Are there any sawtooth generator circuits that would be able to take the sync pulses and make them into perfect sawtooth waveforms in some way shape of form or is the best I can do going to be an oscillator circuit that uses the sync pulses to hold an oscillator to a specific frequency.

The current method of sawtooth generation does work, but I do think it can be improved.


Top
 Profile  
 
 Post subject: Re: Ground loop question
PostPosted: Apr Fri 10, 2020 11:34 pm 
Member

Joined: May Sun 07, 2017 11:35 am
Posts: 840
Location: Belrose, NSW, Australia
Now you are getting into the whole realm of locked oscillators, AFC etc.

Look at some early transistor TV circuits. Here is one from the late '60s that works well:
Attachment:
T26_vert.png
T26_vert.png [ 356.13 KiB | Viewed 325 times ]

Leave out Q51 and associated bits. Inject your sync at C505.
Q54 onwards also gets left out. Sawtooth appears across C508.

However, the linearity looks fine to me, I guess it depends on what you are going to do with this.

_________________
Wax, paper, bitumen, cotton, high voltages - what could possibly go wrong?


Top
 Profile  
 
 Post subject: Re: Ground loop question
PostPosted: Apr Fri 10, 2020 11:54 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 20994
Location: Warner Robins, GA
Oh ok so the top which is normally not on the screen is fine?

Also as I look at the display from the bottom to the top the scan lines start to gradually get farther apart to the point where it looks like at the top there's space for at least 1/2 the thickness of a scanline.

So that means the vertical sawtooth generator may not be exactly proper, right?

I can view the vertical sawtooth and horizontal sawtooth on a scope and take a picture then post it here later tonight if that helps.

So I can incorporate your circuit with the video input ground and contrast pot ground (serves as 75 ohm load resistor as well) connected to the 390 ohm resistor instead of that resistor going to ground, right?


Top
 Profile  
 
 Post subject: Re: Ground loop question
PostPosted: Apr Sat 11, 2020 12:53 am 
Member

Joined: May Sun 07, 2017 11:35 am
Posts: 840
Location: Belrose, NSW, Australia
I suggest you breadboard that circuit, the oscillator will free-run, and confirm you can get the linear sawtooth you are looking for.

It's effectively what is called a relaxation oscillator. To get a linear sawtooth at the greater amplitude required (this complete circuit amplifies it quite a bit), you need to charge a cap slowly through a current source and have the oscillator discharge it for you. So maybe the Height control should be replaced by a current source.

If you don't use a current source to charge the cap, the sawtooth will be exponential instead of linear.

You can simulate a current source by charging with a resistor from a high voltage.

If you haven't already done so, download LTSpice. Spend a little time learning it. Look at the videos. You'll thank me!

https://www.analog.com/en/design-center ... lator.html

_________________
Wax, paper, bitumen, cotton, high voltages - what could possibly go wrong?


Top
 Profile  
 
 Post subject: Re: Ground loop question
PostPosted: Apr Sat 11, 2020 4:23 am 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 20994
Location: Warner Robins, GA
I seen one circuit that used a 555 timer for the sawtooth generators and used the sync signals to trigger them.

I wonder if I can find some sort of sawtooth generator like one of those cheap function generator boards on ebay and make it trigger from the sync pulses.

EDIT:

Here's the vertical and horizontal waveforms.

The interesting thing is that the horizontal waveform changes some on the bottom as I vary the H size pot.

Now the scan lines being slightly farther apart the closer I get to the top of the image, is that the vertical or horizontal causing it?

Also I connected the video out of the box to the X/Y display using a capacitor so that the intensity control had no effect and I got a better picture. So I moved a window over to it that has a lot of white and I saw the brightness change some. So both of those things tells me thee's a problem with the video amp circuit. Perhaps it has to do with how the brightness control works.

Attachment:
1 Vertical.jpg
1 Vertical.jpg [ 332.57 KiB | Viewed 315 times ]


Attachment:
1 Horizontal.jpg
1 Horizontal.jpg [ 352.96 KiB | Viewed 315 times ]


EDIT:

Here's the 555 timer idea.

http://www.electronixandmore.com/projec ... index.html

Here's something a little more advanced for going from VGA to an oscope.

Only showing that because the sawtooth generators are better.

http://www.electronixandmore.com/projec ... index.html


Top
 Profile  
 
 Post subject: Re: Ground loop question
PostPosted: Apr Sat 11, 2020 7:47 am 
Member

Joined: May Sun 07, 2017 11:35 am
Posts: 840
Location: Belrose, NSW, Australia
In the second example you'll see the current source (the PNP transistor) charging the capacitor. As I suggested. That's done to improve the linearity.

Re the 555, two transistors do the same thing in this case!

_________________
Wax, paper, bitumen, cotton, high voltages - what could possibly go wrong?


Top
 Profile  
 
 Post subject: Re: Ground loop question
PostPosted: Apr Sat 11, 2020 4:40 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 20994
Location: Warner Robins, GA
Yes they do. I had figured the 555 might produce a more linear sawtooth.

That said If you look on the original schematic I have the TL-082 outputs going directly to 68uF capacitors which feed 10K pots. That I am thinking would be a capacitive load and if so that could be the problem with the waveforms couldn't it?

Also I could have used a much lower value for the horizontal given it runs at a higher frequency.

Now before I go building the circuit, since the outputs are balanced and are feeding balanced inputs that would mean I only need half the input voltage for the same deflection amount or video, right?

The .1uF bypass caps at the + and - supply pins of each chip, is that a proper value?

Also for future reference the X and Y designations are reversed and was something I had screwed up on the original schematic. The vertical and horizontal designations, however are correct.

I did some calculations and the last 2N3904 in the horizontal causes a phase shift of maybe 3 to 4 degrees starting at 20KHz due to that stage's coupling cap on the base.

Now could that be causing any issues with the horizontal sawtooth?

Now if I replace the .1uF cap with a .22uF cap the phase shift starts at 10KHz.

So I may put a .22uF cap there when I rebuild the circuit.

For the 9 volt regulator I may use one of these.

https://www.mouser.com/ProductDetail/TR ... cAcw%3D%3D

EDIT:

Now if i wanted to make this work for VGA signals all I would need to do is find the specific frequencies for the resolution I want to use and change the value of the two caps that get charged, right? The sync separator chip wouldn't be used as VGA has the separate sync signals already.


Top
 Profile  
 
 Post subject: Re: Ground loop question
PostPosted: Apr Tue 14, 2020 1:27 am 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 20994
Location: Warner Robins, GA
I did notice a focus issue with the display.

There are barely visible horizontal lines about 1" apart scrolling slowly up the screen.

The space between one set of lines is perfectly focused.

The space between the next set of lines is slightly out of focus.

Would the grounding issues with the original circuit cause that issue?

Also for the new circuit I plan on using some ground wire from a piece of romex to serve as the ground buss so that the grounds are as low of a resistance as possible so that the least amount of voltage is dropped in the grounds.

That said going to a dual supply for the OP-AMPs will lower the current through the grounds.

Now for the vertical and horizontal size pots I plan on trying to find some panel mount pots with a screwdriver adjust as once I get the size adjusted I shouldn't have to change it again.

Honestly I wish I had a larger X/Y display as it would be easier to see from a distance as this display pretty much has to be right in the way to be viewable.

Attachment:
20200413_200619.jpg
20200413_200619.jpg [ 389.51 KiB | Viewed 247 times ]


EDIT:

Decided to buy a VGA to s-video/composite converter as I did not want to be limited to only computers with a S-video output.

https://www.amazon.com/gp/product/B001N ... UTF8&psc=1

I had selected a much cheaper one, but it wouldn't be here till some time in May.

I'm hoping it is worth the price I paid.

With the S-video output of the device I can still just use the luma signal to get a better B/W picture.

I do not know if it has any size adjustments other than overscan/underscan, but the ability to fully adjust the size would be beneficial for the 5" Sony B/W TV I have. For this composite to X/Y converter I would not need size adjustments.

I do at some point want to make a circuit that will work directly on a VGA input and generate the necessary sawtooth signals, but switching between resolutions would need to be manual unless I build an elaborate circuit that can auto detect the resolution and put out the desired sawtooth signals.

Now I'm sure the VGA CRT monitors have some sort of circuit like what I need, but they most always only have one main board and I'm reluctant to probe around a monitor board due to the potential to come in contact with the HV to the CRT.

Unless it's possible to get an older monitor and disable the CRT circuitry.

Even then I don't know what signals are output and if it isn't sawtooth waveforms I cannot use the circuit.


Top
 Profile  
 
 Post subject: Re: Ground loop question
PostPosted: Apr Mon 20, 2020 12:08 am 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 20994
Location: Warner Robins, GA
Here's what I'm thinking.

When I build the circuit I'm going to install three of the type shorting links like what is used on computer motherboards so I can disconnect the video from the sync separator chip then I'll install a socket for the two caps that are part of the ramp generators.

I'll then try different cap values and see what if any works for the desired VGA resolutions I want.

I'll also use the proper resistors to take the VGA color signals and make them monochrome which also serves to let the computer know something is connected to the VGA port.

If I can get it working with the VGA, I'll add a 3PDT switch to select VGA or composite then I'll add a two pole rotary switch with enough positions for the desired VGA resolutions and composite video.

Thing is for the video I might be better off using a DPDT relay on the board so the video signals don't have to go to the switch.


Top
 Profile  
 
 Post subject: Re: Ground loop question
PostPosted: Apr Tue 21, 2020 9:13 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 20994
Location: Warner Robins, GA
The adapter came in yesterday.

I connected it up and set the VGA resolution to 640 X 480 using the luma signal of the S-video out of the adapter and the picture is a good bit better than the S-video output of my laptop.

I also have the composite output of the adapter going to the 5" Sony B/W TV and its picture looks great as well.

What I'm thinking of doing is this.

I have two solid state scopes I can use to feed the VGA vertical and horizontal sync to and can pick off the sweep waveform from somewhere in the horizontal circuitry of both scopes.

I can then use the resistors to mix the three colors properly and feed that to the Z axis input of the display with the vertical and horizontal sawtooth waveforms being fed from the scopes to the X and Y inputs.

That way I can see how the display will look being fed a straight VGA signal.

If it isn't any better than the S-video output or I cannot use higher than 640 X 480 resolution due to text being too small to read I'll just not worry about the converter being able to handle VGA and will just mount the adapter in the box the converter will be in.

Nice thing about the adapter is that as long as it is powered on it's producing a raster so that will protect the display's CRT.

I'll also then include switches to where the luma signal can be sent to a jack and I will also have a jack for the composite output and also a jack and switch for a composite input. That way the box isn't just limited to a VGA input and X/Y/Z output.

The adapter is powered from 5Vdc. I don't know the amperage though. So I can include a +5Vdc regulator on the main board.


Top
 Profile  
 
Post New Topic Post Reply  [ 39 posts ]  Moderators: 7jp4-guy, Mr. Detrola Go to page Previous  1, 2

All times are UTC [ DST ]


Who is online

Users browsing this forum: No registered users and 5 guests



Search for:
Jump to:  


































-->


Privacy Policy :: Powered by phpBB