I have some real basic questions about using an oscilloscope for testing tube amps.
So far, all my measurements with the scope have been made at the 8R load resistor.
No problem with that. Last night I tried to measure the upper and lower output of a white cathode follower.
I put one probe across the upper valves plate R, and the other probe from the plate of the lower valve to ground.
I turned on the power and…… flames and smoke!!
Well now I know what I did wrong… all my B+ went through a 330R to ground. That 1/2 W resistor just couldn't take 1.15AMPS!
Here are the questions:
Should the scope be grounded via the 120vac plug?
Should the single generator be grounded?
Is there a way to use both probes at the same time to measure output of a WCF?
Or… In general how do I hook up the scope to measure 2 outputs when one of them is not at ground level?
The WCF schematic is attached for reference
So far, all my measurements with the scope have been made at the 8R load resistor.
No problem with that. Last night I tried to measure the upper and lower output of a white cathode follower.
I put one probe across the upper valves plate R, and the other probe from the plate of the lower valve to ground.
I turned on the power and…… flames and smoke!!
Well now I know what I did wrong… all my B+ went through a 330R to ground. That 1/2 W resistor just couldn't take 1.15AMPS!
Here are the questions:
Should the scope be grounded via the 120vac plug?
Should the single generator be grounded?
Is there a way to use both probes at the same time to measure output of a WCF?
Or… In general how do I hook up the scope to measure 2 outputs when one of them is not at ground level?
The WCF schematic is attached for reference
A scope has a common ground.
Putting one probe across plate resistor and the other one across the cathode one means that the "common ground" shorts you power supply.
Most of the time, simply attach the scope ground (the alligator clip at the probes is) to the équipment ground and measure voltages relative to that point.
Yves.
Putting one probe across plate resistor and the other one across the cathode one means that the "common ground" shorts you power supply.
Most of the time, simply attach the scope ground (the alligator clip at the probes is) to the équipment ground and measure voltages relative to that point.
Yves.
sgerus said:
2 channel scopes often have the possiblilty to display the sum of both probes, and to invert the signal of one or both inputs.
Press INVERT for one inoput and ADD, and you should have the signal across the component you want to measure without any smoke.
Both inputs should be set for AC at the same sensitivity.
SveinB
Yves and Svein,
Thanks for the help. Yes, my scope can do add and invert. So to use this method
Probe 1 from ground to the lower plate
Probe 2 from ground to the upper plate
Is that the idea?
Thanks for the help. Yes, my scope can do add and invert. So to use this method
Probe 1 from ground to the lower plate
Probe 2 from ground to the upper plate
Is that the idea?
I would check that each probe is rated for the max voltage. And the scope input itself for 1/10 that, assuming the probe is a 10X attenuation type. Tube voltages with SS scopes often exceed their input specs.
There are some HV differential probes around that work nicely for tube stuff. They have two leads, and you just connect across whatever. (1500 V max on the Tek ones I've seen)
Don
There are some HV differential probes around that work nicely for tube stuff. They have two leads, and you just connect across whatever. (1500 V max on the Tek ones I've seen)
Don
You will find that in general, equipment having a common earth which is internally connected to a reference point (normally ground) results in a whole lot of problems whenever you need to measure audio, and small signals in general. This is because the ground is not only conencted through the usual wiring, but also through the earth wiring, forming loops. This incidentally is the reason why most audio applainces (nearly all solid state) don't have an earth connection, or if they do, it should be only on one component, usually the amplifier. The recent (several years) fad of using full 3-prong DIN earthed conenctors and cables for power has casued a lot of ground loop problems - the savvy designer will use an earth to ground separator, a fancy name for a few diodes and a resistor-cap combo.
Unfortunately, the standard in emasuring devices, including generators, scopes etc, is that the ground is conencted directly to earth potential, which is basically near to mains voltage 0. This means that wherever you hook up your probe ground, you short that poit to earth potential - and if it's not already there or fully floating, this can be a disaster.
What I would recomend doing, is that you get yourself a mains isolator transformer (115VAC to 115VAC or 230VAC to 230VAC depending on where you are on the planet). You can use this either on your scope and emasuring equipment, or whatever you are emasuring, there are subtle differences, depending on what it is you are measuring and what kind of device. In general, try it first with the devices you are emasuring, and of course, ALL other devices attached to it. IF any of them have an earth lug, they should be connected together.
In my setup, I have a dedicated power strip with a ground fault interrupter hanging off one side of the isolation transformer. In essence, it simulates exactly the connection of earth and 0 in the nomal mains system, except that the transformer isolates it from the mains system. The ground fault interrupter is there to protect against live to earth shorts, just like a regular mains one would be.
Still, this is not the ideal solution, a differential probe or running the scope differentially should be better if more awkward. The reason an isolation transformer may not always be ideal is elakage to the mains through the interwinding capacitance of the transformer. A properly constructed one will be quite good in this regard (a split bobbin od dual winding C-core construction), but still this may interfere when you conenct the scope probe ground to a high impedance point in your circuit. Even so, I find the use of an isolation transforemr practically mandatory - especuially if you use a PC as part of your measuring system, it also has an internal ground to earth conenction which tends to signifficanlty worsen the accuracy, noise floor and dynamics (depending what you are emasuring).
Unfortunately, the standard in emasuring devices, including generators, scopes etc, is that the ground is conencted directly to earth potential, which is basically near to mains voltage 0. This means that wherever you hook up your probe ground, you short that poit to earth potential - and if it's not already there or fully floating, this can be a disaster.
What I would recomend doing, is that you get yourself a mains isolator transformer (115VAC to 115VAC or 230VAC to 230VAC depending on where you are on the planet). You can use this either on your scope and emasuring equipment, or whatever you are emasuring, there are subtle differences, depending on what it is you are measuring and what kind of device. In general, try it first with the devices you are emasuring, and of course, ALL other devices attached to it. IF any of them have an earth lug, they should be connected together.
In my setup, I have a dedicated power strip with a ground fault interrupter hanging off one side of the isolation transformer. In essence, it simulates exactly the connection of earth and 0 in the nomal mains system, except that the transformer isolates it from the mains system. The ground fault interrupter is there to protect against live to earth shorts, just like a regular mains one would be.
Still, this is not the ideal solution, a differential probe or running the scope differentially should be better if more awkward. The reason an isolation transformer may not always be ideal is elakage to the mains through the interwinding capacitance of the transformer. A properly constructed one will be quite good in this regard (a split bobbin od dual winding C-core construction), but still this may interfere when you conenct the scope probe ground to a high impedance point in your circuit. Even so, I find the use of an isolation transforemr practically mandatory - especuially if you use a PC as part of your measuring system, it also has an internal ground to earth conenction which tends to signifficanlty worsen the accuracy, noise floor and dynamics (depending what you are emasuring).
True !smoking-amp said:
1/10 (X10 attenuation) probes having higher impedance introduce less load error in the measured circuit.
Oh ! How do they do that ? A differential amp somewhere ?
Yves.
"Oh ! How do they do that ? A differential amp somewhere ?"
Yes, they have a 100X or so (has a button for two ranges) attenuator followed by a differential amplifier. I can't find it on Ebay at the moment or I would give a link. It is a rectangular black plastic box (about 6 inches by 2 inches by 3/4 inch) with red and black HV clip leads coming out one end and a cable to the scope BNC conn. and a power supply dongle on the other end. Mainly made for working on switching power supplies I think. The more common one has 5 MHz response I think, and there is a fancier one with 100 MHz response I think (maybe a gray box for that one).
Don
Yes, they have a 100X or so (has a button for two ranges) attenuator followed by a differential amplifier. I can't find it on Ebay at the moment or I would give a link. It is a rectangular black plastic box (about 6 inches by 2 inches by 3/4 inch) with red and black HV clip leads coming out one end and a cable to the scope BNC conn. and a power supply dongle on the other end. Mainly made for working on switching power supplies I think. The more common one has 5 MHz response I think, and there is a fancier one with 100 MHz response I think (maybe a gray box for that one).
Don
ilimzn said:
HMMMMMM. It is very likely I am misunderstanding you, but I don't think I fully agree. Most lab scopes' earthing lug is internally connected to the probe's shield, so regardless if you use a isolation transformer or not, you still have the issue that the probe common is earthed. You have the option, with said iso-xfmr, of "floating" the scope (by not connecting the earthing lug or third prong to ground), but this is most certainly a safety hazard, as the case of the scope can become energized.
Assuming the "common" of sgerus amp is connected to earth, his basic issue is one of two things: either he is attempting to measure the voltage between two elevated points, or he is measuring the voltage of a single-ended point with common (earth) as the reference. If the former, the solution is either a differential probe (proper way to do it, big $$$$$) or using the invert-add trick. One must still be sure not to violate the voltage capabilities of the scope/probe system with this method. If the latter, just connecting the tip of the probe, leaving the shield off, does work well for basic measurements where noise is not a concern. Even in those instances where noise was a concern, I have had good results with making the shield connection and completing the ground loop. Ground loops are mostly 60Hz in nature, and can be minimized by using the same outlet for the amp and scope.
Keep in mind, sgerus is attempting to measure the speaker output, so a little noise isn't going to destroy his basic measurements.
If it were me, I would connect the signal generator up as normal, and the scope probes in invert-add. Then experiment with the generator's shield, either floating or connected, for lowest noise. This assumes the amp common is grounded.
Perhaps you could clarify the statements about the GF interrupter; these devices are quite misunderstood, and i'm not sure you are using it as intended, or effectively.
No problem with that...
I'm attempting to measure the output from the upper and lower parts of a cathode follower. I just converted my pentode input stage to a 6sn7srpp / 6sn7 white cathode follower.
That all working fine... now I just learning or attempting to learn what the WCF stage is actually doing.
So what I think I've learned here is:
1. the probe ground clips have to ground because the scope is grounded.
2. use the invert and add method
3. use a differential probe
4. use the probe without the ground clip, because the amp and scope are grounded via the 120VAC plug
Mission Accomplished:
Probe 1 on the cathode of the upper Valve = 30V pk-pk
Probe 2 in the grid of the lower Value = .46V pk-pk
Probe ground clips not used, amp and scope plugged into same outlet.
Both probes set to 5V per unit, 1K sine wave input.
Then I switched display to ADD (no need to use invert in this case
because the singles are naturally opposite of each other)
The new combined sine wave was 29.5V pk-pk
So my White Cathode Follower output is about 98% of the input voltage.
Probe 1 on the cathode of the upper Valve = 30V pk-pk
Probe 2 in the grid of the lower Value = .46V pk-pk
Probe ground clips not used, amp and scope plugged into same outlet.
Both probes set to 5V per unit, 1K sine wave input.
Then I switched display to ADD (no need to use invert in this case
because the singles are naturally opposite of each other)
The new combined sine wave was 29.5V pk-pk
So my White Cathode Follower output is about 98% of the input voltage.
zigzagflux said:
You are correct, this is why I have mentioned that the best first step is to isolate (including earth) the actual device you are measuring, and leave the scope and rest of the measuring setup earthed. However, in some cases you can't do that (example, earthed signal generator and earthed scope means scope probe and SG ground are at earth potential so floating the device you are emasuring does not solve the problem of using scope earth and hot as diferential inputs, in fact if you attempt this it will likely smoke what you are emasuring) where you need t he ADD-INVERT or diff probe.
In cases of sensitive measurements, a good isolation xformer is the only way to prevent ground loops (parallel ground paths through earthing in the mains stirp and actual ground clips and wires through the device you are measuring). This is not a solution for sgerus' problem but just general advice.
For instance, I use a signal generator, distortion analyser, scope and PC for audio measurement. The SG supplies the signal, the output of the measured amp (or whatever) goes to one ch of the scope and the input of the distortion analyzer (it is used to null the fundamental), the output of the DA goes to the other scope channel and PC sound card input, where the distortion residual is looked at using FFT software. The scope, the SG, the DA, and the PC all have grounds internally earthed, but the cable runns and probes make for many additional paths for grounding. So, the signal generator has an earth separator (alowes about 1.4V difference between SG earth and DF/scope earth), and the thing I am measuring has an isolation Xformer (the one with the ground fault interrupter). SG, scope, DF and PC are all connected to a common mains strip. The PC has no other galvanic connections to anything else, and it's monitor uses the PC's power supply for power.
I was just looking and it seems that a 25MHz 1400V differential probe is only a little over $300. I might have to get one of those. It would be nice to be able to probe across anything that I want without worrying so much about voltage ratings.
- Status
- Not open for further replies.