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


It is currently Sep Sat 18, 2021 9:21 am


All times are UTC [ DST ]





Post New Topic Post Reply  [ 15 posts ] 
Author Message
 Post subject: Oscilloscope clock build
PostPosted: Oct Mon 01, 2018 12:37 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 21349
Location: Warner Robins, GA
I'm looking to build another oscilloscope clock.

I have a

5DEP1
5ADP1
5ADP5
5BP4

I can use those, but are there any other different commonly available 5" CRTs that would make for a good display?

I am thinkling of using this circuit which is a modified version of the CRT driver circuit I used in the two previous clocks which has increased bandwidth versus the original circuit.

Attachment:
Deflection circuit higher bandwidth.png
Deflection circuit higher bandwidth.png [ 15.47 KiB | Viewed 3917 times ]


Now because two transistors are in series I can use a higher B+ voltage, right?

When testing the 5" CRT clock I tried the 5BP4 and I actually liked the very crisp display, but I didn't like the rounded face.

The 5DEP1 I already used once and if y'all have seen my amp projects y'all know my preference is not to do the same thing twice.

The 5ADP5 is of no real use to me so I could use that one, but would require me to use one of my DuMont 304H scopes which is a source of parts for the other working scope.

I have a Tektronix 503 I could just stuff the clock board in, but that would be boring plus it is a Tektronix scope.

So I'm thinking of either the 5ADP5

or

The 5BP4

For both I have a chassis with the power transformer and all mounting hardware for the CRT.

What CRT would y'all suggest I use?

For the 5ADP5 I can use the original CRT power supply circuitry including the focus and intencity circuitry provided the deflection amp circuit has enough drive voltage.

I really don't need another oscilloscope clock, but I do need to test the modified deflection circuit.

If said circuit is that much better than the original I might be tempted to update the deflection circuit in the other two scope clocks.

Whatever CRT I use I plan on ordering the basic parts Thursday when I get paid.


Top
 Profile  
 
 Post subject: Re: Oscilloscope clock build
PostPosted: Oct Tue 02, 2018 4:02 pm 
Member

Joined: Jul Wed 22, 2009 3:07 pm
Posts: 926
Hi Tuberadio,

The circuit as it is now, it has lower bandwidth than your previous version, 1.7MHz vs. 4.8MHz.
Though it has a bit more gain, by about 6dB.
You could use higher supply voltage, but as it is now the emitters of the top transistors are only at +4.5V, so not much gain there. But you could raise the voltage of the basis of the top transistors, that would allow you to raise the supply voltage (assuming similar transistors, which they are not).
You really should simulate those circuits, it would give you the exact answers you are asking.
It would also let you tweak the circuit to your requirements.

Just curious, what is the reason for the asymmetrical emitter resistor at the bottom?

Regards, Peter


Top
 Profile  
 
 Post subject: Re: Oscilloscope clock build
PostPosted: Oct Tue 02, 2018 6:38 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 21349
Location: Warner Robins, GA
Oldwirebender simulated both circuits and the one posted here has a higher bandwidth.

Scroll down to his post in this topic

viewtopic.php?f=19&t=343645

The emitter resistor for the transistor driving the Y2 deflection plate came about when I played around with the original circuit to get more undistorted signal out of it so I could properly drive a 3" CRT running higher B- for better brightness as I found lowering the value of that resistor caused an increase in signal out so that it is closer in level to the signal driving the X2 deflection plate which increased the available maximum output voltage that can drive deflection plates.

Now I do not like the fact that the position controls affect the bias of the deflection amps as that also affects the signal when the output is near maximum.


Top
 Profile  
 
 Post subject: Re: Oscilloscope clock build
PostPosted: Oct Tue 02, 2018 9:41 pm 
Member

Joined: Jul Wed 22, 2009 3:07 pm
Posts: 926
I guess I am behind a few editions!
I was talking about the circuits below.
And yes, those circuits are like apples and oranges, so no wonder I got different bandwidths.

Regards, Peter


Attachments:
CRT_drive.jpg
CRT_drive.jpg [ 47.37 KiB | Viewed 3857 times ]
CRT_drive_bode.jpg
CRT_drive_bode.jpg [ 60.54 KiB | Viewed 3857 times ]
CRT_cascode_drive.jpg
CRT_cascode_drive.jpg [ 54.5 KiB | Viewed 3857 times ]
CRT_cascode_drive_bode.jpg
CRT_cascode_drive_bode.jpg [ 58.75 KiB | Viewed 3857 times ]
Top
 Profile  
 
 Post subject: Re: Oscilloscope clock build
PostPosted: Oct Wed 03, 2018 2:48 am 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 21349
Location: Warner Robins, GA
With your simulation the second circuit seems to have a slight rise in response around 1MHz.

The best way to know is for me to build the circuit and see if it requires less compensation or no compensation.

Now with the first circuit you simulated what would be the result of using the OP-AMP used in the second circuit in place of the TL-082?

A bit more on the resistor. I had put a variable pot in parallel with the original resistor and viewed the clock face on the scope then adjusted the resistor until I got the largest output voltage with a clean display. I then viewed the waveforms from both horizontal deflection plates then adjusted the resistor again until both waveforms were not distorted at their maximum output signal.

I proceeded to measure the resistance and used a fixed resistor of that value.

The goal with this build is to make the circuit have enough bandwidth so that I need little to no compensation.

What I would really like to do is take some real low end scopes that are not rare and use those for scope clocks.

I've got a Supreme and a DuMont scope that uses a 3AP1 CRT. The Supreme I'd like to make into a clock but that would require going to a 3RP1A or other 3" CRT and the DuMont I don't want to do that to because it is a DuMont.

Now if I use the DuMont 304H for the next scope clock I will strip it of all parts not needed and set them aside for my good 304H.

I still don't really want to use a 5ADP series CRT though as they are not cheap to buy and I have a scope that uses that CRT

But I don't want to use a 5DEP1 as I've already done that.

So I could use the 5BP4 as I like the crispness of the display , but I don't know that the deflection plate driver circuit will have enough drive to get full deflection with the B- voltage needed (no less than -2KV) for acceptable brightness.

So I may look for a different 5" CRT that is about the same length as or shorter than the 5P4 which will allow me to use the smaller scope chassis that isn't as deep.

I want to build a scope clock in an old Huntron Tracker, but the display will be too small for any real practical purposes.


Top
 Profile  
 
 Post subject: Re: Oscilloscope clock build
PostPosted: Oct Thu 04, 2018 11:01 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 21349
Location: Warner Robins, GA
I looked at the Supreme scope I have which uses a 3AP1 to see if a 3RP1A would fit and the 3RP1A is too short so that scope cannot be used for a clock.


I put an ad in the classifieds for a display using a 3-5" CRT

EDIT:

I found a Bell & Howell model 34 oscope in what looks to be good condition on Ebay for $59.99 free shipping.

Seller said
Quote:
Vintage Bell & Howell schools DeVry institute oscilloscope model 34 for sale bought at auction. Do not have any past history of this item. Had no way of testing so just selling for parts but might work. All knobs turn smoothly it looks to be in good condition does have cord no bare wires. Does not have original box or manual.


So that is very good as there's a couple caps in the B- supply that are usually shorted and will make the 1V2 glow quite brightly and if left in that state long enough will open the CRT HV winding of the power transformer.

My plan is to leave the B- supply including the focus and intensity controls intact and use the 6.3Vac winding that was for the tubes to power the clock board and CRT driver board. I'll either use a HV diode or perhaps leave the 1V2 in place.

I may measure the voltage with the 1V2 and also with a HV diode as I am wanting to add a string of zeners like I did to my other two scope clocks to keep the B- regulated and need to see if I have enough voltage to do so while keeping the CRT at its normal operating voltages.

I may take the model 34 scope I currently have (restoring that one) and measure the B- to see what it is at.

I don't feel bad about using one of these scopes as they are still quite common.

Why not put the clock in the one I have?

Because I had one as a teen and from what I remember it was halfway decent so I want to have a fully restored example.


LOL looking on Ebay I see several of the model 34 scopes and I also saw several of the solid state kit scope like what I used for the first scope clock. So both are indeed very common.

I've run into a problem though.

I want to keep the original look with the knobs.

The vertical variable is dual concentric with the V/div switch using a large knob and the vertical variable using a small knob.

Horizontal width is dual concentric with the width using the large knob and the sync using the small knob.

The width is 50K.

The vertical variable is 2K.

It is possible to use the original pots, but I'd rather not.

I'll figure something out.

If the large knobs fit a 1/4" shaft I can use them then epoxy the small knobs to the large knobs.

The vertical position is ok as it is just one pot.

The scope already has an astig pot in the rear so I won't need to add one.

To keep the correct look I will leave the time/div switch with the vernier pot in place.

The banana jacks will be removed and one hole will be used for a small power switch.

One hole might be used for a neon indicator fed off the B- supply.

I decided to fire up the current partially restored scope I have and the brightness of the CRT is ok, but I will need to check the B- voltage to see what it is as the original caps are rated at 1.6KV.


Top
 Profile  
 
 Post subject: Re: Oscilloscope clock build
PostPosted: Oct Sat 06, 2018 10:51 pm 
Member
User avatar

Joined: May Wed 01, 2013 10:05 pm
Posts: 1990
Location: Rayleigh, BC, CANADA
Tube, did you see this: https://www.ebay.com/itm/RARE-CRT-vinta ... SwbVdbhhQO

_________________
Rocco


Top
 Profile  
 
 Post subject: Re: Oscilloscope clock build
PostPosted: Oct Sun 07, 2018 1:17 am 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 21349
Location: Warner Robins, GA
I saw that but figured it would be a little small for my use and it's square which I'm not that fond of even though it works best with the clock as it allows the square menus to display properly.

Now that might work for a scope clock for work but I would have to make a case for it or use a clear case and figure a way to secure it.

Concerning the scope I just got if around 1.6KV isn't enough I can do a half wave rectifier for the B+ and ground one end of the B+ winding instead of the center tap as the other end of the B+ winding connects to the B- winding so that would give higher B- and I can regulate it. Given the B+ will be high I will be able to regulate it like I did in my 3" CRT clock. The benefit there is the display won't change in size or brightness based on the line voltage variation.

That depends on if the transformer insulation can handle that without arcing to the core or another winding.

For the B+ a string of three or four 100 volt zeners will work.

The B- I forget what zeners I used. Thought I had that schematic at home, but I think it is at work on my computer there. It was somewhere around -1.7KV.

I'll basically leave the current focus and intensity circuitry in place as they seem to work ok. If they aren't good enough I will duplicate the circuit used in the 5" CRT clock which I know will work.

This scope due to how it is laid out is quite open both on top of and under the chassis so I have plenty of room to do whatever.

Plus I'll be getting some good 7 pin and 9 pin sockets along with a good compaction socket.


The other reason I want to keep the model 34 I already have is because it has all Bell & Howell marked tubes except the compactron which I believe is GE.


Top
 Profile  
 
 Post subject: Re: Oscilloscope clock build
PostPosted: Oct Mon 08, 2018 7:39 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 21349
Location: Warner Robins, GA
So I took some voltage readings and here's what I got.

Here's the schematic of the power transformer for reference.

Attachment:
Power supply.png
Power supply.png [ 67.13 KiB | Viewed 3741 times ]


From ground to the 1V2 filament pin is 1.095KVrms.

That gives me an unloaded B- of 1.548KVdc if I use a diode instead of the existing 1V2.

Not quite high enough for my liking and desired maximum brightness level.

I measured from the far left red wire on the schematic to the 1V2 filament pin and got 1.492KVrms

That gives me an unloaded B- of 2.109KVdc.

Here's the CRT circuit.

Attachment:
CRT.png
CRT.png [ 101.55 KiB | Viewed 3741 times ]


So I'm thinking of putting a string of zeners across C31 to regulate the voltage.

I'll most likely go with the voltage I chose for the other scope clock I did with a 5DEP1 as it provides good brightness in a brightly lighted room and the deflection circuitry fills up the CRT properly and I can get a bit of overscan as well. I could go higher with B-, but I cannot think of any real reason to.

Now given B- will be higher I might have to play around with the focus and intensity circuitry, although I don't think that will be necessary.

If I do need to play around with it I'll try a neon indicator in series before or after the focus control like I did with the other two scope clocks as it provides about a 120Vdc drop (if I use a panel mount indicator with neon bulb or 90Vdc if I use a plain neon bulb), serves as a pilot light and also lets me know B- is present.

I see two issues.

1. The B+ supply will be 1/2 wave rectified which means larger filter caps.
2. Unloaded B+ will be 565.6Vdc.

Now given I'll be using the improved circuit with the four transistors I suppose that I can run at least 400Vdc B+, right?

If so I could use four 100 volt zeners in series or two 200 volt zeners then a power resistor to drop the voltage.

I'll need 600 volt electrolytic caps though.

If I could just get B+ to an unloaded peak under 525Vdc my choices for capacitors is much wider and much less expensive.

I can get a Sprague Atom 20uF 600 volt for $22.95 which is a bit expensive, but it is cheaper than the $30 Mouser wants. I know it will be an excellent capacitor though.

https://www.tubesandmore.com/products/c ... ectrolytic

I've found one that might work 25uF 800V $12.25 which would also be a good capacitor, but I have to find a way to secure it.

https://www.tubesandmore.com/products/c ... ectrolytic

Now would a 20uF or 25uF cap followed by a resistor and a 20uF or 25uF cap with a zener across it provide enough filtering of the B+ supply?

I was looking at film caps as they would last a lot longer than electrolytics, but those are quite expensive and much more than I would want to spend.

EDIT:

Here's the initial schematic.

The two unmarked resistors will be figured when I build the circuit.

Attachment:
CRT driver 5DEP1 improved bandwidth.png
CRT driver 5DEP1 improved bandwidth.png [ 67 KiB | Viewed 3646 times ]


Now I havw to figure out where I'm going to mount the clock board so that I have access to the serial port without having to remove the scope from the cabinet each time I need to access it.


Last edited by Tube Radio on Oct Mon 15, 2018 11:34 am, edited 1 time in total.

Top
 Profile  
 
 Post subject: Re: Oscilloscope clock build
PostPosted: Oct Mon 15, 2018 2:15 am 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 21349
Location: Warner Robins, GA
So while asking a question about the OP-AMP and a suitable sub I saw where the feedback for the OP-AMP was drawn wrong. The 10K resistor from pin 1 is shown going to ground instead of going to pin 2 where it should be. I'll update the schematic in the previous post with the correct schematic.

I also discovered a problem.

The scope uses dual controls for three things and all three are a switch and a pot.

The scope on the vertical position control uses two knobs on one shaft in order to keep the same look of the other dual controls.

The switch shafts are slightly larger in diameter than the vertical position control shaft.

So that means the knobs will fit on a standard 1/4" shaft, but will not be fully centered on the shaft.

I'm thinking maybe I can use a small piece of heatshrink on the shaft so the knobs fit properly, but the heatshrink might be too thick.

I tried to find some .1uF 3KV caps and could not find any at a reasonable price so what I may do is this. In the other model 34 scope I used .1uF 3KV caps to replace a couple .1uF caps in circuits powered by B+ as that's what I had on hand so I may replace two of those with .1uF 630 volt caps so that I have the necessary HV caps for the B- supply.

Attachment:
B&H model 34.jpg
B&H model 34.jpg [ 237.79 KiB | Viewed 3667 times ]


Also the CRT is un-shielded which makes it more susceptible to magnetic fields. I have the shield from the other Heath solid state kit scope I got for parts and I'll use that for this clock to ensure the CRT is shielded.


Top
 Profile  
 
 Post subject: Re: Oscilloscope clock build
PostPosted: Oct Tue 16, 2018 11:10 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 21349
Location: Warner Robins, GA
Here's some pictures of the scope.

It arrived today in very good condition.

I almost hate to make it a scope clock.

I took the pictures after removing the front screws.

Now either the vertical input jacks or horizontal input jacks will be where the two pushbuttons will go and the jacks not used for the switches are where the power switch and neon indicator will go. Or I may do a power switch and fuse then use the sync jack for the neon indicator. That way there will be no unused holes and it will be properly fused.

The green graticule will be removed and used for my good model 34 since it doesn't have one. The bezel around the CRT will stay in place.

Attachment:
1.jpg
1.jpg [ 98.78 KiB | Viewed 3629 times ]


Attachment:
2.jpg
2.jpg [ 82.54 KiB | Viewed 3629 times ]


Attachment:
3.jpg
3.jpg [ 85.45 KiB | Viewed 3629 times ]


Attachment:
4.jpg
4.jpg [ 75.91 KiB | Viewed 3629 times ]


Attachment:
5.jpg
5.jpg [ 64.53 KiB | Viewed 3629 times ]


As can be seen in the following two pictures there's lots of room on top of and under the chassis for any circuitry I will need.

Also as is typical with every 5DEP1 CRT I have seen the label had fallen off and was in the cabinet. I set it back on the CRT for the picture.

Attachment:
6.jpg
6.jpg [ 89.96 KiB | Viewed 3629 times ]


Attachment:
7.jpg
7.jpg [ 145.11 KiB | Viewed 3629 times ]



As is typical with this model of scope the first B- cap is shorted. I brought it up slowly on a variac and around 50vac saw some sparks in the 1V2 so I disconnected the first B- filter cap and proceeded to power it up again. This time after adjusting the controls I got the trace you see here. Looks better in person though.

Attachment:
8.jpg
8.jpg [ 101.34 KiB | Viewed 3629 times ]


The parts not needed (tube sockets, switches, pots ETC...) will be removed and I will keep in a box as spares for the nearly fully restored model 34 scope.

This one also has a full complement of Bell & Howell tubes except for the GE compactron.

Judging by its looks I suppose the scope was used and cared for quite well until no longer needed and then put away some place where it wouldn't get damaged, scratched ETC...

The idea is to mount the deflection board on top of the chassis as close to the rear of the CRT as possible so the wiring to the deflection plates is as short as possible and I will most likely use twin lead for the wiring.

I've thought about the circuit then remembered how as I reduced the value of the collector resistors in the original deflection circuit then reduced the emitter resistors to keep proper bias the bandwidth improved which was noticed by a much improved image on the CRT. So what I'm thinking of is this.

Ditch the circuit I was going to use with the four transistors and do it like this.

Reduce the collector resistors to where I can have very good bandwidth then reduce the emitter resistors to keep proper bias and run that stage on a much lower B+ perhaps 100Vdc or less. Then have the collectors drive some HV power transistors. That will work provided the power transformer HV winding can handle the B+ current.

Then again I'll have nearly 600Vdc B+ and would have to drop that way down.

Before I go ripping out circuitry and building circuitry I am going to try the original circuit. I can connect the input to the horizontal width control and get DC coupling, but the vertical is a bit more difficult.

If I can find a way to make the following circuit DC coupled I might just be able to do it.

Attachment:
Scope circuit.png
Scope circuit.png [ 104.13 KiB | Viewed 3629 times ]


The reason for DC coupling is because with the typical AC coupling these low end older scopes have the image tends to move around some and is worse with some clock modes.

I don't feel so bad about making this one a scope clock as it is so common. I saw several on Ebay with wildly varying prices.


Top
 Profile  
 
 Post subject: Re: Oscilloscope clock build
PostPosted: Oct Thu 18, 2018 10:18 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 21349
Location: Warner Robins, GA
I took the sparkfun version of the scope clock with the bad RTC and made it to where I could connect it to any scope.

I connected it to the original circuit and the display was ok.

I got a better display when connecting to the variable gain controls, but the horizontal is reversed.

Also I cannot connect where the clock will be dc coupled to the horizontal because the position pot feeds a dc voltage through a 2.2 meg resistor to the grid.

EDIT:

Thought about trying to connect the board there anyways so I decided to measure the voltage on the grid. It is 82Vdc and while it is fed through a 2.2 meg resistor and would not hurt the clock I knew it would throw the horizontal position off more than what the control could compensate for plus it would affect the bias of that stage and make it not operate correctly.

Anyways before I could take a photo of the displayed clock I switched the time/div switch not realizing it didn't leave the horizontal input floating so the DAC of the clock is damaged.

I had a bit of swimming in the image so I installed the shield from the other parts solid state B&H scope and that fixed the swimming issue.

Attachment:
20181018_203433-600x800.jpg
20181018_203433-600x800.jpg [ 81.95 KiB | Viewed 3597 times ]


With only replacing the two B- filter caps which came out of the other good model 34 and will be replaced by .1uF 630V caps given I only used the 3KV caps because I had them on hand and the voltage at that point is less than +400Vdc I get a quite good displayed sinewave.

Attachment:
20181018_203512-600x800.jpg
20181018_203512-600x800.jpg [ 77.15 KiB | Viewed 3597 times ]


So I can either build the deflection circuit in the first post of this topic or I can build the original deflection circuit lowering the value collector and emitter resistors to minimize the Miller capacitance and to keep bias proper and run it on low B+. I could then drive four power transistors for the deflection plates which will have a lower value of collector resistor provided the power transformer HV winding can handle the B+ current.

The second idea might allow me to get rid of the OP AMP altogether which would definitely improve the bandwidth a bit.

The lower impedance of the power transistors will make the capacitance of the deflection plates not matter nearly as much if at all. I would then mount the transistors using transistor insulators to the rear CRT support bracket which would make the leads to the deflection plates as short as possible and provide some sort of heat sink plus it would keep the transistors close to the same temperature.

The problem with that idea is this.

I would need to drop nearly 600Vdc down to a bit less than 400Vdc which would mean a large resistor giving off lots of heat.


Top
 Profile  
 
 Post subject: Re: Oscilloscope clock build
PostPosted: Jan Mon 14, 2019 3:58 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 21349
Location: Warner Robins, GA
Well it looks like this project is on hold indefinitely.

The maker of the Dutchtronix scope clock I use has stopped making them.

He posted on the website that he is going to make the info available to make the boards soon, but that would mean I either have to make the board myself or find someone who can. Would also possibly mean I have to program the atmega chip myself unless he is going to provide pre-programmed chips.


Top
 Profile  
 
 Post subject: Re: Oscilloscope clock build
PostPosted: Jan Tue 15, 2019 1:34 pm 
Member

Joined: Jul Wed 22, 2009 3:07 pm
Posts: 926
Have a look at this site:
https://github.com/kosma/bakingbread

DISCLAIMER: I have not tried that!

Peter


Top
 Profile  
 
 Post subject: Re: Oscilloscope clock build
PostPosted: Jan Tue 15, 2019 2:51 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 21349
Location: Warner Robins, GA
Quite interesting.


Top
 Profile  
 
Post New Topic Post Reply  [ 15 posts ] 

All times are UTC [ DST ]


Who is online

Users browsing this forum: No registered users and 6 guests



Search for:
Jump to:  
cron


































Privacy Policy :: Powered by phpBB