Waveform signal inversion...

[Externet]

Hi all.
I would appreciate suggestions to invert an existing sawtooth signal

/|/|/|/|/|/|/|/

onto a

\|\|\|\|\|\|\|\

Preferably with the use of a common emitter transistor circuit
Note no amplification (gain) is desired
Note the sawtooth signals are 18 Vpp riding on DC
The bottom (valleys) of the waveforms are at a +28 V level
The tops (peaks) of the waveforms are at a +46 V level
The frequency span it has to work is 1 Hz to 30 MHZ
The power supply available is +87 V DC

Sorry, unable to indicate input nor output impedances. They are simply unknown, but both guessly suspected to be in the neighborhood of 5K Ohms.

Sorry, unable to attach or link the graphic yet. My old posts do not count as minimum merit.

Miguel

 

[55pilot]

OK, take a step back, and start with the big picture. What exactly are you trying to achieve?

1Hz to 30MHz is a HUGE span. Those are some strange voltages. What is the application?

Knowing the load impedance is critical. It will make a difference between this being a hard problem to it being a virtually impossible one.

In either case, this is not going to be a simple or easy problem. What is your experience level? What resources do you have at your disposal to solve this problem?

---55p

 

[Externet]

Thanks for responding, 55pilot

OK Stepping back, but not much to go...

http://i588.photobucket.com/albums/ss323/Innernet/62and63.png

To achieve : I have waveform "62" and need to create "63" from it.
The huge span, yes, it has to be that way. It is the horizontal sweep for electrostatic deflection on an oscilloscope. Guessing the I/O impedances... ? Will say ~5Kohm.
Voltages... That is what they have to be, no workarounds.
My experience level is high enough; I have the tools, means and test equipment to do it, and my own plan; but wanted to explore other opinions before starting 'my way'

Ignore the 3ms period; that is a setting for troubleshooting. Real operation goes up to 50ns

Miguel

 

[55pilot]

Now I understand what you are doing. The next question is why?

It is obviously do-able, but not straight-forward, especially at the higher frequencies. By the time you design, implement, test and refine the design, you would be into it for a lot of time and money. Unless you have some really unique issues, I can not think of a reason for doing that. If you are repairing a broken scope or adding functionality to it, you are better off buying a replacement.

Another thing to consider. If you do this to the scope, how do you know that it works exactly right? If you do not get the slopes and voltages exactly right, you will be seeing an incorrect signal on the scope and not know it. Depending upon what you are doing with the scope, that could be really bad.

---55p

 

[Externet]

Thanks.

Why... because the integrated circuit that outputs the 2 horizontal deflection plates sweep signals is kaput only for that output I want to re-create from tapping the good signal and inverting it.

There is no replacement IC U800. Pure Tektronix unobtanium part 155-0241-02

I have/see no other choice. Well, the other choice is dump the Tek2465 oscilloscope, and am willing to try the inversion. Yes, unique 'forced' issue.

So, going back; any circuit suggestion?

 

[55pilot]

Can you buy a dead scope and salvage the part from it? That would be my plan A, plan B and plan C.

What you are trying to achieve is not easy to do and your bigger goals are better achieved by finding a salvage part or replacing the scope. This circuit will not be easy to design or get working. If the end result is not perfect, your screen image will be distorted and the scope will be useless.

---55p

 

[Externet]

Hi.
It is very useless now
Plans A,B,C do not click Those flawed chips have a built-in epitaph to die soon, very proven. To transplant the output section of an older model oscilloscope that does not use such IC is another not too far choice.

Miguel

 

[Externet]

Tektronix 2465

Done, works.
Workaround the IC U800, Tektronix unobtanium part 155-0241-02

Now to fit my canteen style contraption circuit within the guts, with smaller power resistors and heat sink. It does not run hot at all, and there is room for it.

Much better than a dead 'unfixable' oscilloscope.
Miguel
-sorry , poor quality pictures-

 

[Icom]

U800 fix

Hello.
I've the same problem with the U800 hybrid...
Can you explain how you solved it? I'm very interessed ...

Thank you.
Davide.

 

[vidiot]

Can you post the circuit you used?