Scope attenuators 1,2,5, why bother?

[Joerg]

Very wise words.
Yes, but the idea is also to read voltages from the digital display
(as 1.55Vpp for example).
In such a case one could even use continuous variable gain to set sensitivity.

Based on automatic readings? Please don't ...

Got to stoke the wood stove now to make the goose bumps go away.

 

[Joerg]

mm, but the bit shifts will always be 1,2,4,8, etc..


It is just an experiment, I may do it in FPGA again, as that at least gets
me into the > 50 MHz sampling, while PIC will be limited to some kHz (think 25kHz).

Heck, then you might as well go all out and hang one of these puppies up
front, along with some ECL memory:

http://cache.national.com/ds/DC/ADC083000.pdf

 

[amdx]

Very wise words.
Yes, but the idea is also to read voltages from the digital display
(as 1.55Vpp for example).
In such a case one could even use continuous variable gain to set

I did note in your post the digital display of voltage, however is that
pk to pk or rms, ok probably both/either. But how much ringing is
on my squarewave and how much did my snubber knock it down?
Add a couple of moveable cursors to set at the difference between two
points I want measured and display that difference.
Up to 8 cents now,
Mike

 

[Jim Yanik]

My TDS2012 vertical steps 1-2-5, but the timebase steps 1-2.5-5.
Bizarre.

John

21.5 yrs at TEK,I never saw a scope that had anything other than 1-2-5 for
their time-bases.
that included the TDS200,TDS300,500 series.

Perhaps that's some software setting for the 2.5/div TB ranges?
Or the variable setting left in there by somebody?
it really doesn't make any sense otherwise.

 

[John Larkin]

21.5 yrs at TEK,I never saw a scope that had anything other than 1-2-5 for
their time-bases.
that included the TDS200,TDS300,500 series.

Perhaps that's some software setting for the 2.5/div TB ranges?
Or the variable setting left in there by somebody?
it really doesn't make any sense otherwise.

If there's a way to fix it, I can't find it. DEFAULT SETUP doesn't
help.

John

 

[Joerg]

21.5 yrs at TEK,I never saw a scope that had anything other than 1-2-5 for
their time-bases.
that included the TDS200,TDS300,500 series.

But you didn't design that TDS220? Hopefully not ...

 

[Allan Herriman]

Heck, then you might as well go all out and hang one of these puppies up
front, along with some ECL memory:

http://cache.national.com/ds/DC/ADC083000.pdf

The 4:1 demuxed output is slow enough to connect to an FPGA; you don't
need ECL memory for it.

Regards,
Allan

 

[Joerg]

The 4:1 demuxed output is slow enough to connect to an FPGA; you don't
need ECL memory for it.

True. When spending hundreds of Dollars on the ADC another $100 for a
zippy FPGA won't matter ;-)

 

[Jim Yanik]

But you didn't design that TDS220? Hopefully not ...

No,I was a tech in the Indy and Orlando Field Service Centers,repaired and
calibrated analog scopes to 500Mhz,TM500,TV products.
I tried out the TDS220,TDS320,and TDS540 scopes.

 

[Jan Panteltje]

I did note in your post the digital display of voltage, however is that
pk to pk or rms, ok probably both/either. But how much ringing is
on my squarewave and how much did my snubber knock it down?
Add a couple of moveable cursors to set at the difference between two
points I want measured and display that difference.
Up to 8 cents now,
Mike

Yep, 2 cursors is better then one, just hope I have the code space.
It is getting pretty full as it is.
Maybe I will indeed need an ARM....

 

[Jan Panteltje]

Based on automatic readings? Please don't ...

Got to stoke the wood stove now to make the goose bumps go away.

No, Joerg I do not see this as a problem, it makes little difference,
in my view, if you use a pot, or an 11 position switch that sucks anyways,
because often you cannot see what range it is on, _without_ OSD.

If the pot causes the on screen display to flip between 10mV/div, 20mV/Div etc,
it is exactly the same thing _without_ the need for a lot of numbers on
frontpanel.
Or were you referring to overload? I once killed a scope by accidently
touching it to the boost (top) anode of a BW TV (10x probe, about 6kV).

I fixed it myself (Ch1) using Ch2 to see what was wrong in Ch1.
(only had one scope in the repair shop).

 

[Jan Panteltje]

Heck, then you might as well go all out and hang one of these puppies up
front, along with some ECL memory:

http://cache.national.com/ds/DC/ADC083000.pdf

That is a very nice chip, really.
It will also need some very nice memory

 

[Jan Panteltje]

The 4:1 demuxed output is slow enough to connect to an FPGA; you don't
need ECL memory for it.

Block RAM?

 

[Allan Herriman]

Block RAM?

Demux in logic to a suitable width (perhaps 64 bits @ 375MHz), then
write into block ram.

Many of the FPGA families fast enough for this work also contain fast
multipliers, which would make the digital gain adjustment trivial. The
ADC083000 only has 6ish ENOB, so digital gain adjust will only get you
so far (i.e. you would probably need finer steps than 1,10,100 on the
analog PGA).

The FPGA would be StratixII or III, Virtex4 or 5, etc. The Altera
ones might be better, as they have the really large (~100k) dual port
rams. Truely useful sized rams (~Mbytes) would need to be off-chip,
of course.

Some of the FPGA families support PCI-e, if you wanted to get serious
update rates on a plug in card for a desktop computer.

Regards,
Allan

 

[Joerg]

No, Joerg I do not see this as a problem, it makes little difference,
in my view, if you use a pot, or an 11 position switch that sucks anyways,
because often you cannot see what range it is on, _without_ OSD.

If the pot causes the on screen display to flip between 10mV/div, 20mV/Div etc,
it is exactly the same thing _without_ the need for a lot of numbers on
frontpanel.
Or were you referring to overload? I once killed a scope by accidently
touching it to the boost (top) anode of a BW TV (10x probe, about 6kV).

I fixed it myself (Ch1) using Ch2 to see what was wrong in Ch1.
(only had one scope in the repair shop).

OSD? Sure, but not 1.55V ...

 

[Joerg]

That is a very nice chip, really.
It will also need some very nice memory

It will also require a really nice bank account ;-)

 

[Joerg]

No,I was a tech in the Indy and Orlando Field Service Centers,repaired and
calibrated analog scopes to 500Mhz,TM500,TV products.

Ah, good. Didn't want to step on anyone's toes here. I guess as a tech
it must not have been too much fun repairing one of these plastic
bread-box scopes.

I tried out the TDS220,TDS320,and TDS540 scopes.

IMHO the TDS220 (and some others) were a real disgrace for Tektronix.
The first time I used one my faith in that company was shaken quite a
bit. I guess they should have kept some of the older engineers.

[...]

 

[Jim Yanik]

No,I was a tech in the Indy and Orlando Field Service
Centers,repaired and calibrated analog scopes to 500Mhz,TM500,TV
products.

Ah, good. Didn't want to step on anyone's toes here. I guess as a tech
it must not have been too much fun repairing one of these plastic
bread-box scopes.

I tried out the TDS220,TDS320,and TDS540 scopes.

IMHO the TDS220 (and some others) were a real disgrace for Tektronix.
The first time I used one my faith in that company was shaken quite a
bit. I guess they should have kept some of the older engineers.

[...]

I didn't work on the digital scopes.

the entire TDS series is module-exchange only.(I believe the modules are
now repaired at Beaverton and DC service centers,the last remaining US
service centers)

IIRC,TDS200s were only two boards and a display in a case.
The TDS200 PS was a throwaway(purchased part),but I acquired one to play
with,it was just a bad cap and diode in the backlight section.TEK didn't
even have schematics for that PS. IIRC,in some spare time,I also fixed a
2430 PS for the tech that serviced that line.He had already installed a new
PS,so he saved the old PS as a spare for test.

I liked working on switchers...

 

[Jan Panteltje]

OSD? Sure, but not 1.55V ...

Why not, it would be an incredible coincidence if it was 2.000000V ;-)

 

[whit3rd]

Yes, but the idea is also to read voltages from the digital display
(as 1.55Vpp for example).
In such a case one could even use continuous variable gain to set sensitivity.

This sounds like a good idea, from the cost viewpoint; a cascade
of three AGC type amplifiers (Gilbert cells) will give
logarithmic gain programmed by a voltage source (8 bit DAC?).
You can use a phantom input channel to feed a calibration
source, to get past the thermal drift, and a sample/hold
circuit to hold the gain constant during an accumulate/calibrate
cycle.

But for human interface reasons, you want a human-set screen
range that DOESN'T change until the human changes it. It's
OK to have the internals do some gain-riding with the signal,
but I'd get dizzy watching the screen readjust all the scale values
in realtime.

That little grid overlay on the oscilloscope screen is a GREAT
simplification for lots of uses. I'm not happy thinking I have
to digitize with a cursor instead of just reading it off (no hands!).