Shopping for used scope to use as curve tracer

[Electron123]

I'm thinking of buying a used scope to connect to a go/no-go breadboard curve tracer. I have a new digital one but, from what I've discovered and read, it can't be used as a curve tracer. So before I go shopping, can anyone give me tips on what I should look for to make sure a scope can work as a curve tracer? I plan to use it for trouble-shooting and don't need anything fancy, so I'll be looking for budget equipment.

[bdunham7]

Assuming that you just need a workable XY display mode, most any old analog CRO would work.  However, many or most DSOs would also work if you configure them properly.  What scope do you have that won't work with your tracer?

[Electron123]

I currently have a Rigol DS1102.

Most of the used scopes online don't come with probes.  Will any set of basic, passive probes due? And are the barrel connectors all the same or do I need to get something model-specific?

[james_s]

By "barrel" connections I'm assuming you mean BNC, those are standardized across almost all scopes save for some really exotic stuff.

Lots of decent dirt cheap probes, the common P6100 can be had for about $20 a pair and work just fine. I do typically glue the X1/X10 switch to the X10 position though, I have no idea why almost all cheap probes are switchable.

Not all scopes have XY mode so do confirm that is there before you buy something. Does your DS1102 not have XY mode? It's pretty crap on every DSO I've tried, but should be adequate for a curve tracer.

[Fungus]

Will any set of basic, passive probes do?

Yes.

[mag_therm]

A crt non-storage scope will need to sync repetitive XY draw , with flyback, at a rate depending on persistence.
Rate of > 10 ms/div  would be OK  on the old Eico TR_410 here. That TR-410 which has XY plot ability , but no Ext X gain adjustemnt, needing an external amplifier.
There is presently another thread about XY on old Leader scope that you could read.

Edit : The new Instek GDS-1202B here has XY but I did not try it yet.
It does display the Pen Drive output from the vintage HP141T very well, just using the X timebase locked to the pen drive return.
You might be able to use you DSO like that?

[wasedadoc]

I currently have a Rigol DS1102.

That does have X-Y mode.  Have you tried it?

[Electron123]

That does have X-Y mode.  Have you tried it?

Yes, I was up until 1am trying to get it to work - no luck.  I also found a thread on another site complaining about a similar problem (https://forum.allaboutcircuits.com/threads/curve-tracer-displays-gabage-on-my-rigol-dso.186950/#post-1757858)

[wasedadoc]

That does have X-Y mode.  Have you tried it?

.. trying to get it to work - no luck.

That doesn't tell us much!

FWIW a few months ago I tried an Octopus type circuit with my Rigol DS1054Z.  Admittedly not the same model scope as yours but I did get acceptable results.  ISTR that switching it out of auto memory depth mode and going down to a low number was a key step. 

[bdunham7]

That does have X-Y mode.  Have you tried it?

Yes, I was up until 1am trying to get it to work - no luck.  I also found a thread on another site complaining about a similar problem (https://forum.allaboutcircuits.com/threads/curve-tracer-displays-gabage-on-my-rigol-dso.186950/#post-1757858)

It isn't all that easy and on a scope without persistence, it may be even harder.  The big difference between an analog CRO and most DSOs is that the DSO does not use a continuous display of the signal--there is still triggering, an acquisition period, memory depth and a lot of blind time.  If you are using an 'octopus' type device driven by 60Hz line power, you can by start trying the line trigger on your scope and a total horizontal time (acquisitiion period) of at least 20ms, so probably 2ms/div. 

[Wallace Gasiewicz]

I was eventually able to get traces on my Agilent digital scope, using an analog BK transistor trace unit.
The traces were terrible. I never did get to the point that I was able to find something like a  Z axis for blanking the return trace.
I do not recall all the adjustments I made to make it work, it was a while ago.
It is not as easy as using an analog scope, which the tracer I have was actually made for.
It would be great if someone could develop something to do a trace easily on a digital scope. It would let you record and save the various traces.

Any old analog scope with X, Y, and Z inputs would work just fine. A DC input must be there also. I think all scopes have a choice for DC vs AC input.  Low frequency response scopes would be preferable, so cheaper scope would be just fine,  and of course one with a big clear tube screen. There are even old Heathkit scopes without a timebase that would work fine.
Storage not necessary.

I do not know what output impedance your tracer has, as I recall the old tracers were meant to be directly connected to scope input.
So, I do not know if you need a 50 ohm input on your scope for these purposes...
The input to the scope is basically a fairly slowly changing DC value, not really any sort of HF signal, so freq response is not important at all.

[Wallace Gasiewicz]

Just another thought....does anyone have any experience with using a cheap computer based scope for this purpose??
I do not know very much about these computer based scopes at all myself.

[mag_therm]

1) If the abscissa ( ...I had to spell check it !) on  X axis  is independent var and does vary linearly with time, the 'scope can be used in normal sweep mode.
If scope is storage, then one trace only is needed. For non storage, repeated sweeps of DUT are needed usually with a linearly rising sawtooth, with falling trigger on the flyback.
This case covers a lot of engineering problems. Example is sweep of frequency in spectrum analyser.

2) Second case would be with abscissa not varying linearly with time in DUT , when a true XY plot is required. Examples, switching devices and magnetics.

[wasedadoc]

So, I do not know if you need a 50 ohm input on your scope for these purposes...

Not just a case of Octopus type tracers not needing a 50 ohm input. If the scope does have that and it can be switched off, you must do so.  If it cannot, find another scope.

[Electron123]

OK, I'm probably doing something wrong here.  The scope doesn't seem to do anything no matter what I do with the octopus (even when I unplug it).  I've attached a schematic of the octopus I'm trying to use, a photo of it on my breadboard, a screenshot of what I'm seeing on my scope, and a data dump of the parameters.  The power source I'm using is a wall wart with 7.5 VAC, 350mA.  The pictures were all taken with the octopus attached to a resistor (DUT). The red and black allegator clips in the breadboard photo are the octopus probes. The black one is connected to the scope's ground terminal.  The scope probe next to the red clip is channel 2 and the other is channel 1.  I've tried all the suggestions generously provided in this thread.

[bdunham7]

Try taking the scope out of XY mode and just show CH1 and CH2.  You should be able to get the 60Hz applied voltage signal on CH1 and then the DUT current on CH2.  You may have to use very different voltage settings to get good signals.  Once you have stable individual signals, go for the XY.

[CatalinaWOW]

For best results as a curve tracer on an analog scope, don't just look for XY mode, but make sure it has Z axis, external beam intensity control or other method to turn the beam off during retrace intervals.  Not all do.  My old Heathkit was one example.  The same concept would apply on a DSO. 

How desperately you need this depends on the details of how your curve tracer works.  Some may just step back down to zero and the only issue will be a possible bit of hysteresis.  Others jump directly from the end of one track to the next and this can be a visual pain.   Possible to ignore but not desirable.

[Electron123]

>Try taking the scope out of XY mode and just show CH1 and CH2.

Hmmm.  Is this is what I'm supposed to be getting?

[bdunham7]

For best results as a curve tracer on an analog scope, don't just look for XY mode, but make sure it has Z axis, external beam intensity control or other method to turn the beam off during retrace intervals.  Not all do.  My old Heathkit was one example.  The same concept would apply on a DSO.

I've seen this mentioned a few times, but typically a basic curve tracer doesn't use any blanking and for many components the retrace will be pretty much overlaid on the forward trace.  For switches like triacs and so forth the retrace will be separate, but they don't blank it--the octopus and other basic curve tracers don't have any output which could be used for the Z-axis.  Even full-featured all-in-one-box curve tracers only use CRT intensity modulation in certain modes.

[bdunham7]

>Try taking the scope out of XY mode and just show CH1 and CH2.

Hmmm.  Is this is what I'm supposed to be getting?

That needs some adjustment!  Change the settings until you get 1-2 cycles on the screen horizontally and both channels are about 3-4 divisions high.  Turn the BW limit on and reduce your memory depth.  Perhaps change your probes and scope to 1X, that signal looks really weak and noisy.  What is your tracer circuit like?

[Wallace Gasiewicz]

Your scope needs something to trigger on. A dig scope does not have a true x/y setting it is very difficult to set it up for an appropriate trigger on either input when in xy mode.
I do not know if you have a line trigger.
I am using an old BK 501A curve tracer, It feeds directly to the BNC scope inputs without probes.


I just looked at my notes pertaining to my Agilent 54810 scope;
For curve tracing it works on "LINE" Trigger just fine.

Settings on BK Curve Tracer:
Sweep Volts 20             Vert Sensitivity    2mA Div             Step 10uA step

Scope settings :

1 vs 2 with vertical being Ch 1 This is in my "Math" menu

Horizontal setting 200 uSec/ div, even though you might think you are using only x and y, you need to set this.
Trigger:   LINE       or pick your poison, view the signals and adjust the trigger level. I think your tracer has a 60 HZ output, perhaps use this.

CH 1 DC 20 mV / div
CH 2 DC 500 mV / Div

Acquisition: 
Depth 2048 pts
Sample Rate   250 KS/ Sec

Digital filter 9 bit BW  Limit ON, this helps a lot.

Do NOT USE AVERAGING

After all this the curve part is almost unusable, the trace goes up too fast.
If I use my old analog scopes, the curve is much easier to use.

Just an observation from the BK manual...HF scope not required but scope should be 10  KHz   HA!!!

[Electron123]

Some progress, though somewhat sideways.  Part of the problem was with the hook tips on my probes, so I replaced them with alligator clips and got a better signal.  But it's still not right (see attached).  Here are the settings I'm using.  They're not quite what you recommended but seem to get the best result:

1. both probes on 1x (reset scope to 1x too)
2. both probes with BW on 20MHz (only option)
3. Ch1 horizontal scale: 1V
4. Ch2 horizontal scale: 5mV
5. Trigger type: video, sync: line.  There's an option for number of lines but that doesn't have a noticeable effect
6. Acquisition mode is "normal" (it's disabled)
7. Memory depth is 6k for each channel (lowest)
8. Sample rate: 500Sa/s. It looks like there's no way to directly set the sample rate, the manual just says to adjust the horizontal scale.

[wasedadoc]

This is a sample of what I can get with a Rigol DS1054Z and Octopus type circuit.  12 Volt transformer. 50 Hz. Series circuit of 1.5k current limit resistor, 3.9 Volt Zener as DUT and 100 ohm current measuring resistor.  Channel 1 probe switched to x1, ground to Zener anode and tip to Zener cathode.  On scope, CH1 settings are 20 MHz BW limit on and probe gain x1.

Channel 2 probe also switched to x1. Ground to Zener anode and one end of 100 ohm. Tip to other end of 100 ohm.  On scope, CH2 settings are 20 MHz limit on, Invert on, probe gain 0.01 and Unit set to A.

Trigger set to AC.  Memory depth 60k. (6k much the same). Timebase slow.