Frequency Divider for older Oscilloscopes??

[GlennSprigg]

I can't find enough info about doing this. (Excuse my deteriorating mind!  )
Is there a 'relatively' simple way to create an external Frequency Divider, to say.. measure frequencies on an old 10-Mhz Analog Scope, such as a 'Divide-By-Ten' circuit to measure up to 100-Mhz??  I've seen some 'Counter' circuits that obviously would not retain anything of the original 'WaveForm', and at best could just count the Frequency, beit a simple SquareWave output. Maybe for the 'too-hard' basket? 

Yea, I know. I can hear people screaming... "Just buy a 100-Mhz Scope!"  Sigh.. 

[tggzzz]

You need to define the input you want to observe. Consider especially the amplitude and range of frequencies of interest, e.g. 100MHz+-5MHz.

If you are interested in seeing 0-100MHz on a 10MHz scope, forget it.

If you are interested in 95-105MHz (i.e. 10MHz bandwidth) then you can use RF techniques to mix and downconvert the 100MHz signal to baseband.

Another technique, which only works with repetitive signals, is to sub-sample the input. Many digitising scopes do that, but I suspect building an external sampler will be more than you wish to do.

[David Hess]

https://www.eevblog.com/forum/projects/diy-ghz-sampling-head-for-lt100mhz-scopes/

[Ian.M]

Errr.... *NO* a frequency divider circuit isn't worth doing - just get a 100MHz or better frequency counter.  Its possible to home-brew a 100MHz frequency counter with a PIC18 (using its T0CKI Timer 0 clock input, which is good up to 50MHz if properly handled and can be backed by 24 bits of internal hardware counter - see AN592 for how to get the low byte out of the internal prescaler) + a fast D type flipflop to get it up to 100MHz, guarentee 50% duty cycle at T0CKI and also gate the input and allow AN592 prescaler access.  If you need it to count slow or low level signals you'll need some sort of limiting preamp, and if you need to go above 100MHz, will need a prescaler.   If you aren't already into PICs its better to buy a commercial instrument than to build!

OTOH if you need to actually view fast repetitive signals, you could (in theory) build an analog sampling frontend.  You'll need a very fast sample and hold, a fast trigger circuit, and an adjustable sampling clock gated by a monostable.  Its also a bitch to build a properly compensated input attenuator and/or wide band preamp if you want a wide input range.  See https://en.wikipedia.org/wiki/Oscilloscope_types#Analog_sampling_oscilloscope for a very brief intro + a few helpful links.   However, unless you have most of the parts needed already gathering dust , lots of zero cost free time and nearly infinite patience with setbacks, you'd be better off buying a modern 100MHz or better DSO.

Edit: David and Tggzzz are also suggesting analog sampling front ends.

[2N3055]

Yea, I know. I can hear people screaming... "Just buy a 100-Mhz Scope!"  Sigh.. 

Because what you want is not possible. You have to have 100MHz bandwidth scope, no matter if it is analog, analog sampling, digital sampling or digital real time.

So, yes, you need to get a 100MHz scope to have a 100MHz scope.

You could use current low bandwidth scope as a display for sampling head (that other already pointed to).
Problem is that ANY 100 MHz scope (even the most stupid ones, that you can get for 100 USD new) will be better than any contraption you can make cheaply...

Sorry.

[GlennSprigg]

To all, point taken and understood.   

[tggzzz]

You have to have 100MHz bandwidth scope, no matter if it is analog, analog sampling, digital sampling or digital real time.

Not quite true. Back in the 70s I saw 10MHz waveforms captured and displayed on an XY pen chart recorder

If you have a sampling scope everything up to and including the sampler has to be full speed, but the rest can be as slow as is convenient.

[Kleinstein]

Some older scopes used analog CCD like memories to first sample the signal relatively fast and than play it back at a reduced speed for the slower ADC. In principle this could be use, but those CCD memories, especially the faster ones are hard to get parts.

[2N3055]

You have to have 100MHz bandwidth scope, no matter if it is analog, analog sampling, digital sampling or digital real time.

Not quite true. Back in the 70s I saw 10MHz waveforms captured and displayed on an XY pen chart recorder

If you have a sampling scope everything up to and including the sampler has to be full speed, but the rest can be as slow as is convenient.

That was only display. You had 10MHz scope (sampler type) in front of it. I thought I explained what I meant. It was a 10 MHz sampling scope assembled from two devices, sampling head with S/H and pen chart recorder as display....

[tggzzz]

You have to have 100MHz bandwidth scope, no matter if it is analog, analog sampling, digital sampling or digital real time.

Not quite true. Back in the 70s I saw 10MHz waveforms captured and displayed on an XY pen chart recorder

If you have a sampling scope everything up to and including the sampler has to be full speed, but the rest can be as slow as is convenient.

That was only display. You had 10MHz scope (sampler type) in front of it. I thought I explained what I meant. It was a 10 MHz sampling scope assembled from two devices, sampling head with S/H and pen chart recorder as display....

The OP wants to see a 100MHz signal on his 10MHz scope. That is possible with a sampling front end.

[2N3055]

You have to have 100MHz bandwidth scope, no matter if it is analog, analog sampling, digital sampling or digital real time.

Not quite true. Back in the 70s I saw 10MHz waveforms captured and displayed on an XY pen chart recorder

If you have a sampling scope everything up to and including the sampler has to be full speed, but the rest can be as slow as is convenient.

That was only display. You had 10MHz scope (sampler type) in front of it. I thought I explained what I meant. It was a 10 MHz sampling scope assembled from two devices, sampling head with S/H and pen chart recorder as display....

The OP wants to see a 100MHz signal on his 10MHz scope. That is possible with a sampling front end.

That is what I'm saying too. And that comprises a sampling scope with required bandwidth.
His original question whether he can down-convert the signal like it is done in direct heterodyne receiver,  and look at it on 10MHz scope and preserve pulse shape of original signal. He can't.

It wasn't discussion of how sampling scopes work.

[tggzzz]

You have to have 100MHz bandwidth scope, no matter if it is analog, analog sampling, digital sampling or digital real time.

Not quite true. Back in the 70s I saw 10MHz waveforms captured and displayed on an XY pen chart recorder

If you have a sampling scope everything up to and including the sampler has to be full speed, but the rest can be as slow as is convenient.

That was only display. You had 10MHz scope (sampler type) in front of it. I thought I explained what I meant. It was a 10 MHz sampling scope assembled from two devices, sampling head with S/H and pen chart recorder as display....

The OP wants to see a 100MHz signal on his 10MHz scope. That is possible with a sampling front end.

That is what I'm saying too. And that comprises a sampling scope with required bandwidth.
His original question whether he can down-convert the signal like it is done in direct heterodyne receiver,  and look at it on 10MHz scope and preserve pulse shape of original signal. He can't.

It wasn't discussion of how sampling scopes work.

In the first post he is after frequency, not wave shape.

In the second post I noted you can simply downconvert to below 10MHz.

[2N3055]

In the first post he is after frequency, not wave shape.

In the second post I noted you can simply downconvert to below 10MHz.

I went back an read it again. And now I must say you are 100 % correct about that. I was confused..
Sorry about that and thank you for being patient with me and setting me straight..
All the best,
Sinisa

[tggzzz]

In the first post he is after frequency, not wave shape.

In the second post I noted you can simply downconvert to below 10MHz.

I went back an read it again. And now I must say you are 100 % correct about that. I was confused..
Sorry about that and thank you for being patient with me and setting me straight..
All the best,
Sinisa

No problem. We all make mistakes, but too few of us admit it

Besides, some mistakes are more interesting than others

[Jorge Ginsberg]

Hello Mr. GlennSprigg

I am new to this forum and just read your question.
I am an Electronics Engineer but I am already retired. I live in Argentina and my English is very bad.  I am writing to you with the help of "Google Translator".
As your post is more than 1 year old, before answering I would like you to tell me if you have already solved your question so I don't write in vain.  Best regards from Buenos Aires. 

(My sister lives in Australia, in Melbourne)

[Ian.M]

@Jorge,
First, welcome to these forums.

To all, point taken and understood.   

I would understand that as: Glenn accepts that his original question had no cost-effective practical solutions.  However if you have built an analog sampling front end or other device to down-convert high frequency signals for display on a lower bandwitdh scope, we'd all be interested in seeing your project.   

[joeqsmith]

Welcome Jorge,   

I don't mind seeing some of these more technical threads revived and am interested in hearing about what you have to offer.   

Some time ago I had downloaded the patents for LeCroy's Digital Bandwidth Interleaving and attempted to write a simulator to demonstrate how it works.  Roger Delbue from LeCroy chimed in the comments and offered a few bread crumbs.   The video starts out showing a home made Arb which can be used for this purpose.   In the end of the video I show a down converter used to monitor the output of a simple oscillator I was playing with.   



At 12:40, I show another simple down converter using a DRO.   

[Jorge Ginsberg]

I am sorry to say but I too can only offer a few "crumbs"....
Many ...MANY years ago (between 1980 and 1982), I built a prescaler that obtained an almost exact copy of an analog signal but with a frequency 100 or 1000 times lower.  That mean that it was possible to observe a 100 Mhz signal as if it were a 100 Khz signal.  It was entirely made with Shotky-TTL technology, something already obsolete or almost obsolete today.
I can give the idea here but unfortunately I have sold and given away everything. Schematics included.
It happens that I have been retired for 15 years now and as soon as I started to see this equipment coming from China, at such low costs, I realized that if I didn't sell my lab quickly I was going to be left with old or obsolete junk in my hands.  Therefore, I hurried to sell everything.  Now I just read about electronics. Nothing more than that.
I regret not keeping the schematics.  I was convinced that they were worthless obsolete things.  I never imagined that someone would want to build an analog prescaler, being that nowadays you can get good sampling oscilloscopes for less than 400 dollars....
If you still want to know how I did it and you want to develop it, just let me know and I will tell you -at least what I remember- how to do it.
Greetings from Argentina

[Jorge Ginsberg]

what you want is not possible. You have to have 100MHz bandwidth scope, no matter if it is analog, analog sampling, digital sampling or digital real time.

So, yes, you need to get a 100MHz scope to have a 100MHz scope.

You could use current low bandwidth scope as a display for sampling head (that other already pointed to).
Problem is that ANY 100 MHz scope (even the most stupid ones, that you can get for 100 USD new) will be better than any contraption you can make cheaply...

Sorry.

2N3055 : You are wrong. (nice nickname)
It is perfectly possible to make a circuit that takes samples of the 100 MHz signal and with those samples make an exact copy but at 10 Khz, so that they can be observed by a cheap oscilloscope. And you can do it in analog form. No memory, no counter, no microprocessor. The only thing you could use is some high speed TTL chips and even not even that. And it's a relatively simple project, but very critical in terms of the semiconductors, capacitors and resistors you employ.
The problem here is not the technology but the cost.
If China sells you an oscilloscope capable of sampling a signal at a rate of 2 Gs/s, for a ridiculous price of 400 dollars, it is silly to work so hard to make a homemade prescaler that will surely cost more than that.
Just look here and be amazed at the price:
https://www.rigolna.com/products/digital-oscilloscopes/1000z/
Look also here:  https://www.renhotecic.com/HANMAKET-DOS1102-110MHz-Digital-Oscilloscope-2channel-Oscillograph-1Gsa-s-7-Tft-LCD-Osciloscope-Kit-Better-Than-Ads1102cal-0004577
But from the technological point of view there is no problem to do it.  And I know very well what I'm talking about...
The only justification for such a project is the enthusiasm for electronics and the pleasure of doing it.  Nothing more than that.

[2N3055]

what you want is not possible. You have to have 100MHz bandwidth scope, no matter if it is analog, analog sampling, digital sampling or digital real time.

So, yes, you need to get a 100MHz scope to have a 100MHz scope.

You could use current low bandwidth scope as a display for sampling head (that other already pointed to).
Problem is that ANY 100 MHz scope (even the most stupid ones, that you can get for 100 USD new) will be better than any contraption you can make cheaply...

Sorry.

2N3055 : You are wrong. (nice nickname)
It is perfectly possible to make a circuit that takes samples of the 100 MHz signal and with those samples make an exact copy but at 10 Khz, so that they can be observed by a cheap oscilloscope.
The problem here is not the technology but the cost.
If China sells you an oscilloscope capable of sampling a signal at a rate of 2 Gs/s, for a ridiculous price of 400 dollars, it is silly to work so hard to make a homemade prescaler that will surely cost more than that.
Just look here and be amazed at the price:
https://www.rigolna.com/products/digital-oscilloscopes/1000z/
But from the technological point of view there is no problem to do it.  And I know very well what I'm talking about...
The only justification for such a project is the enthusiasm for electronics and the pleasure of doing it.  Nothing more than that.

You might know what you're talking about but you didn't seem to understand what I did talk about. 
You simply misunderstood. English is not native language for either of us.

Please read again: you cannot look at high frequency signal on a low bandwidth scope with a PRESCALER ( a divider like the one used in frequency counters).
Which was OP original question.
And you can't.

And few sentences later I say " You could use current low bandwidth scope as a display for sampling head (that other already pointed to)."
Which is obviously true because thousands of scopes like that exist...
At which time you're looking at 100 MHz signal with your homemade 100MHz+ sampling scope consisting of a homemade sampling head and a low bandwidth scope used as a display unit for sampler.

At home it is not too hard to make a sampling head that will sample at 1 GS/s (1GHz). But today you can buy cheapest scope that already samples at that rate, and for the price of component to make sampling head (unless you already have the parts). And then you still have only sampler and no input stages, attenuators etc...

And making a 10-20 GHz sampler at home is not that trivial..not to mention 4-8 GHz front end... That one would be useful though..

Regards,

[Jorge Ginsberg]

what you want is not possible. You have to have 100MHz bandwidth scope, no matter if it is analog, analog sampling, digital sampling or digital real time.

So, yes, you need to get a 100MHz scope to have a 100MHz scope.

You could use current low bandwidth scope as a display for sampling head (that other already pointed to).
Problem is that ANY 100 MHz scope (even the most stupid ones, that you can get for 100 USD new) will be better than any contraption you can make cheaply...

Sorry.

2N3055 : You are wrong. (nice nickname)
It is perfectly possible to make a circuit that takes samples of the 100 MHz signal and with those samples make an exact copy but at 10 Khz, so that they can be observed by a cheap oscilloscope.
The problem here is not the technology but the cost.
If China sells you an oscilloscope capable of sampling a signal at a rate of 2 Gs/s, for a ridiculous price of 400 dollars, it is silly to work so hard to make a homemade prescaler that will surely cost more than that.
Just look here and be amazed at the price:
https://www.rigolna.com/products/digital-oscilloscopes/1000z/
But from the technological point of view there is no problem to do it.  And I know very well what I'm talking about...
The only justification for such a project is the enthusiasm for electronics and the pleasure of doing it.  Nothing more than that.

You might know what you're talking about but you didn't seem to understand what I did talk about. 
You simply misunderstood. English is not native language for either of us.

Please read again: you cannot look at high frequency signal on a low bandwidth scope with a PRESCALER ( a divider like the one used in frequency counters).
Which was OP original question.
And you can't.

And few sentences later I say " You could use current low bandwidth scope as a display for sampling head (that other already pointed to)."
Which is obviously true because thousands of scopes like that exist...
At which time you're looking at 100 MHz signal with your homemade 100MHz+ sampling scope consisting of a homemade sampling head and a low bandwidth scope used as a display unit for sampler.

At home it is not too hard to make a sampling head that will sample at 1 GS/s (1GHz). But today you can buy cheapest scope that already samples at that rate, and for the price of component to make sampling head (unless you already have the parts). And then you still have only sampler and no input stages, attenuators etc...

And making a 10-20 GHz sampler at home is not that trivial..not to mention 4-8 GHz front end... That one would be useful though..

Regards,

OK.  It may have been a problem with the language.  I don't rule it out.  It is true that English is not our native language. I am using "Google translator" to write this.
Let's start over.
There is no need to build a device at home that samples at a rate of 1 or 2 Gs/s.  Transforming a 100 Mhz signal into a 10 Khz signal is possible by sampling at the rate of one sample per cycle.  That is: if your signal is 100 Mhz, your sampling will also be 100 Mhz. Even your sampling can be slower, as long as you keep the synchronism. And this, is the more critical thing in the design.
It's a pity I didn't save the circuit but I did this more than 30 years ago and I remember perfectly how I did it, the only thing I don't remember is the circuit but any electronics hobbyist could redesign it. 
From an economic point of view it is not worth it but if you have fun doing it it is fully justified.

Regards from Argentina

[2N3055]

what you want is not possible. You have to have 100MHz bandwidth scope, no matter if it is analog, analog sampling, digital sampling or digital real time.

So, yes, you need to get a 100MHz scope to have a 100MHz scope.

You could use current low bandwidth scope as a display for sampling head (that other already pointed to).
Problem is that ANY 100 MHz scope (even the most stupid ones, that you can get for 100 USD new) will be better than any contraption you can make cheaply...

Sorry.

2N3055 : You are wrong. (nice nickname)
It is perfectly possible to make a circuit that takes samples of the 100 MHz signal and with those samples make an exact copy but at 10 Khz, so that they can be observed by a cheap oscilloscope.
The problem here is not the technology but the cost.
If China sells you an oscilloscope capable of sampling a signal at a rate of 2 Gs/s, for a ridiculous price of 400 dollars, it is silly to work so hard to make a homemade prescaler that will surely cost more than that.
Just look here and be amazed at the price:
https://www.rigolna.com/products/digital-oscilloscopes/1000z/
But from the technological point of view there is no problem to do it.  And I know very well what I'm talking about...
The only justification for such a project is the enthusiasm for electronics and the pleasure of doing it.  Nothing more than that.

You might know what you're talking about but you didn't seem to understand what I did talk about. 
You simply misunderstood. English is not native language for either of us.

Please read again: you cannot look at high frequency signal on a low bandwidth scope with a PRESCALER ( a divider like the one used in frequency counters).
Which was OP original question.
And you can't.

And few sentences later I say " You could use current low bandwidth scope as a display for sampling head (that other already pointed to)."
Which is obviously true because thousands of scopes like that exist...
At which time you're looking at 100 MHz signal with your homemade 100MHz+ sampling scope consisting of a homemade sampling head and a low bandwidth scope used as a display unit for sampler.

At home it is not too hard to make a sampling head that will sample at 1 GS/s (1GHz). But today you can buy cheapest scope that already samples at that rate, and for the price of component to make sampling head (unless you already have the parts). And then you still have only sampler and no input stages, attenuators etc...

And making a 10-20 GHz sampler at home is not that trivial..not to mention 4-8 GHz front end... That one would be useful though..

Regards,

OK.  It may have been a problem with the language.  I don't rule it out.  It is true that English is not our native language. I am using "Google translator" to write this.
Let's start over.
There is no need to build a device at home that samples at a rate of 1 or 2 Gs/s.  Transforming a 100 Mhz signal into a 10 Khz signal is possible by sampling at the rate of one sample per cycle.  That is: if your signal is 100 Mhz, your sampling will also be 100 Mhz. Even your sampling can be slower, as long as you keep the synchronism. And this, is the more critical thing in the design.
It's a pity I didn't save the circuit but I did this more than 30 years ago and I remember perfectly how I did it, the only thing I don't remember is the circuit but any electronics hobbyist could redesign it. 
From an economic point of view it is not worth it but if you have fun doing it it is fully justified.

Regards from Argentina

Yes, Jorge, you are correct. That technique is how sampler scope works. It has sample and hold circuit that takes quick ( 1 nsec) sample of signal at input at precise moment. And plots it on output device. Than it takes sample at different time. And plots that. And if you  do this 1000 times, you will have reconstructed input signal. You can do that sequentialy (equivalent time sampling) or randomly and measure time (random interleaved sampling, RIS). That is how sampling scopes work. It doesn't have to be one sample per period, it can be much less. You can literally plot on a paper plotter, as long as input signal is repetitive and newer changes.  And that is problem with analog sampling scopes, they work only for repetitive signals or for eye diagrams....

Best regards,

[bdunham7]

Yes, Jorge, you are correct. That technique is how sampler scope works. It has sample and hold circuit that takes quick ( 1 nsec) sample of signal at input at precise moment.

I think the method he is talking about is slightly different.  Instead of storing and plotting the samples to reconstruct the original waveform, I believe what he describes simply samples the waveform at a slightly different frequency than the input signal (or possibly a harmonic or subharmonic frequency) and then renders that to the screen in real time.  The effect is that the original signal is aliased down to the difference, or something like that--so the scope is actually displaying a much lower frequency signal that is hopefully correlated with the input signal in a known manner.  Maybe you could call that a sampling downmixer?  Although it can be demonstrated for known signals, I don't think it works too well in the general case.

[Jorge Ginsberg]

No bdunham7. That is not what I am trying to say.   "2N3055" has understood me better what I am trying to say.
What I am saying is that it is possible to take copies of a 100 Mhz signal (to give an example) and display them at a much lower frequency, so that they can be viewed on an inexpensive low bandwidth oscilloscope. 
If you take a single sample per period and repeat the same thing for 10,000 periods, when you reconstitute the signal, it will be an exact copy of the original signal, but its frequency will be 10,000 times lower.  By doing that it is possible to observe a 100 Mhz signal as if it were a 10 Khz signal, something that any cheap oscilloscope can do.
And it is not difficult to achieve.  The only thing necessary is to have elements that work properly in the UHF band.  I did it with BFY90, MPF102, BF960 transistors and  and some TTL-Shotky flip-flops. And the resistors should be non-helical so they have almost no inductances.
And I could see UHF signals on a simple Telequipment D61 oscilloscope.
And I think that's what GlennSprigg is looking for , only that when he made his first query he expressed himself badly because he mentioned a "prescaler" which is used only to measure frequencies and not to observe waveforms, although if we go by the concept of the term "prescaler", what we are doing on the oscilloscope is conceptually something similar: moving to the low frequency domain something we can't observe or measure in the high frequency domain.

[Jorge Ginsberg]

Yes "2N3055". Finally we agree.
Note that even if a 100 MHz signal does not keep its shape all the time and has variations from time to time, even so, thousands of periods are still repetitive.
For example, suppose a 100 MHz signal has a change of shape every thousandth of a second, yet, between one change and the next, there are 100,000 identical periods between one change and the next.  And those 100,000 different periods can be sampled.