Test equipment set-up basics

[deci belle]

I'm going to ask one question at a time, and keep it all on this thread. That's where I'm at haha!

My signal generator and o-scope both have chassis-ground binding posts. What do I connect these to?

[jrfrond]

Most signal generators have binding post outputs. You need to make a cable with a dual banana plug on one end, and a male RCA on the other. Why RCA and not 1/4" phone? Simply because most of the readily-available Y-adaptors are RCA, and if you are doing 2-channel gear, not just guitar amps, you will need to use them.

A scope won't work without a ground. You can use the binding post to run a separate cable with an alligator clip to ground the scope to the chassis of the unit you are working on. If the unit has a grounded AC cable, it might already be grounded through the AC though.

[km6xz]

It depends on the output topology. Does the signal generator have a balanced line out put or is it unbalanced with chassis reference? Many of the better generators have the low side isolated from ground allowing work on balances line or unbalanced line equipment.
The two most common connectors for output are dual banana jacks and BNC. Either can be balanced line, for example the Sound Technology generators have insulated BNC jacks for output keeping the low side off chassis ground, A switch is provided for floating or unfloated ground.
Dual Banana is very common so having lots of Banana to BNC, RCA, XLR adaptors or cables are possible to interface your generator to just about any type audio device. The ground terminal is intended for grounding the test instrument. It the cable from the output is chassis connected and the shield of the audio out cable is connected straight through from end to end, do not use the ground terminal on the generator to connect to the chassis of the unit under test or ground loop current will introduce noise and through your readings off. It balanced lines are used for input to the unit under test, the shield can be cut at one end of the cable to remove the problem of ground loops but still provide its shielding from noise.
In that case the chassis ground terminal on the generator can be connected to the unit under test chassis for a safety ground.
What model generator are you using so we can tell you what output circuit is has?

[nosaj]

Mine is a Tenma 72-3060 with a 3 prong cord and a binding plug for ground, but I have been unsucessful at finding a manual for this scope, will try Tenma after the holiday as I just got it yesterday for $70 everything seems to work ok, picks up signal on both channels and the function generator on it works.

Thanks,
Jason

Most signal generators have binding post outputs. You need to make a cable with a dual banana plug on one end, and a male RCA on the other. Why RCA and not 1/4" phone? Simply because most of the readily-available Y-adaptors are RCA, and if you are doing 2-channel gear, not just guitar amps, you will need to use them.

A scope won't work without a ground. You can use the binding post to run a separate cable with an alligator clip to ground the scope to the chassis of the unit you are working on. If the unit has a grounded AC cable, it might already be grounded through the AC though.

[g1]

If your scope has a ground clip on the probe, you don't need to worry about the extra chassis ground. My scope has a BNC connector for the probe so I use the probe ground clip and ignore the extra chassis ground.
Same for the generator, mine has a binding post output (hot & neg) and a third banana jack for the extra chassis ground which I ignore.
If your equipment has 3 prong plugs then the chassis are all grounded anyway, so the extra chassis ground lugs are just convenient extra chassis grounds should you need them.

[jrfrond]

If your scope has a ground clip on the probe, you don't need to worry about the extra chassis ground. My scope has a BNC connector for the probe so I use the probe ground clip and ignore the extra chassis ground.

That small ground clip on the probe can be a REAL pain in the neck when moving around the circuit area, or can even slip off whatever you are clipped to and short something. I never use it and go with either the common ground or the extra ground cable if necessary.

Tenma test equipment is fairly cheap and pretty reliable for shop use. We have a bunch of it scattered around the shop: a couple of scopes, signal generators, all of our scope probes, and a signal tracer. I'd like to add that we rarely purchase any major equipment new. It's often more cost-effective when in need of a GOOD scope (Tektronix, etc.), to purchase one used and send it out for tweaking and calibration. I only really pay close attention to scopes and DMM's, the atter of which are all Fluke in our shop.

[deci belle]

Oh boy! This is good. I've got it all on a bench and I'm figuring out what wires, adapts, etc, I need to cobble together before I actually ask about what to do next.

Thank you jfrond— scope needs to be grounded to the chassis of the unit under test. Check.

km6xz said:

Does the signal generator have a balanced line out put or is it unbalanced with chassis reference?

Hi km6xz! I have an old Heathkit 5218 signal generator that Jazz gave me a heads-up on (ebay). It has two binding posts for the sine wave and two separate posts for the square wave— plus the chassis ground binding post. So my signal generator chassis ground depends…

I've noted those points on using the probe ground clip.

Thanks everyone! I'll be back, soon!

[g1]

Good points from John about the scope ground. My probe has the ground wire coming out the tail end and I forgot I had extended it about an extra 2 feet when it broke long ago, a separate ground to the binding post would be simpler.
I'm using the Heathkit IG-18 which is pretty much the same. I have a dual banana plug connected to 1/4 inch plug and don't use the chassis ground.
In case you don't have the manual: http://www.tauntek.com/Heathkit-IG-5218.pdf
Also, if you want to get all hi-fi about it :Enter Code to Begin Download
or google "greening of the IG-18" (and re-greening for even further updates)
From my understanding the differences between the 18 and 5218 were cosmetic so the mods/upgrades should still apply.

[deci belle]

Yay! thank you for the links to the mod and the manual!

Yes, g-one— I understand that those two models are the same. I did download the re-greening files and that's all i need to do! haha!! Some day I bet I will though. What a modest little unit, all ready to be hot-rodded!! My ambition is to get the hang of selecting frequencies with all those dials because it is a little confusing right now.

That's cool that the ground-wire comes out the back of your probe to keep it out of the way!

[km6xz]

The probe ground return is not intended to be a common for chassis ground currents, if there are any, it is for signal shielding and ground reference for the signal. Very frequently the shield is a floating commons connection isolated from chassis ground on a good scope. The ground clip lead is short because with 10-100megohms input Z, the extra length of a lead will cause ringing in the measurements and forms a low pass filter that drops accuracy at the higher frequencies. For probing around low Z circuits(speaker out, power supplies etc) it won't matter but looking at grids or fast rise time signals like digital, the usable bandwidth and calibration of the scope-probe combination will be compromised. Good low Cx probes are easily damaged and are expensive, over $100 each, way over $100 if you have a wide bandwidth scope.
So, the answer is....use a longer lead if you are not concerned with frequency response performance of the scope or looking at really high-Z circuits. The original short lead will be the only one that will give allow the specs of scope to be met.
The fastest way to limit probe live is drawing an arc on high voltage test points, usually happens approaching the probe to the voltage, high Z test point, about .5-1mm away. The arc is due to parasitic oscillations that the probe and circuit set up. Same with DMMs, one of the common ways to kill one.
Take care of your test gear and it will take care of you.
Find out the power supply, input circuits and shielding circuits of every piece of gear you have so you know whether the ground is floating for signal return, or chassis return, and AC mains grounding. The mixing grounding and shielding schemes can introduce errors, noise, and cause premature failure of probes or test instruments. All the better gear have selectable ground reference, by switches or jumpers, which helps being consistent with your bench.
If you have ever tried to trace down noise in audio systems you know that confusing shield grounds and mains grounds will introduce more problems than solutions. Small signal ground references should not be the same return as chassis, any voltage drop between chassis, due to IR losses, becomes high level noise current as far as an unbalanced input circuit is concerned. You do not want that to appear at the input of your amp, mixer or effect. Balanced line systems often do not even connect shields straight through for this reason, and a shield is effective connected to only one side of the cable length so not having continuity eliminates one of the ground current paths from signal input but allowing Mains grounds to be connected for safety.
You will see gear from a wide range of grounding scheme eras so it is important to understand all the standards that have been used over the years and how to interface those with your test gear.

[deci belle]

Find out the power supply, input circuits and shielding circuits of every piece of gear you have so you know whether the ground is floating for signal return, or chassis return, and AC mains grounding. The mixing grounding and shielding schemes can introduce errors, noise, and cause premature failure of probes or test instruments. All the better gear have selectable ground reference, by switches or jumpers, which helps being consistent with your bench.

You will see gear from a wide range of grounding scheme eras so it is important to understand all the standards that have been used over the years and how to interface those with your test gear.

Check!

[deci belle]

It's been a month since posting last.

The amp's volume is @ 4, where I usually have it set when I play it. 440Hz (1st string, 5th fret) is one of the notes that sets off ringing oscillation when I play notes on my guitars with a pick.

I hooked the sigGen @ 440Hz into the scope's "X" terminal and amp's normal channel input.
The amp's speaker out is hooked to the dummy-load. The sine-wave amplitude is set to 1v rms with the coarse dial and -20db with the fine dial.

The 10x probe is in the dummy-load's BNC output and hooked to the scope's "Y" terminal.

The scope is set to .5 mS and the channels are set to .5v/div.

The dummy-load is singing away and the DMM is reading 26vac off the dummy-load @ 8Ω.

The amp's wave looks like a wicked, sinuous square wave with little hooks poking up, half-way up the steep slopes.

I smell something hot and it seems like the dummy-load… I've been fiddling with the dials on the scope for about 10 minutes and then the amp's mains fuse (2.5A) blew.

I don't know what to make of the wave I was looking at and I don't know why the fuse blew. I just turned off the amp and checked the fuse right away.

Any observations on what I've done or what I should have done with my test setup?

[g1]

Not sure what scope you have but assume it is dual channel and when you say x and y inputs you mean ch.1 and ch.2. Was one of your scope traces showing a clean sine wave from the generator?
Sounds like it was clipping pretty good at the output. 26VAC into 8 ohms would be about 85 watts with a clean sinewave. However, you had some square wave (clipping) at the output so your meter may not have given you a proper reading. What is power rating of the amp under test?
How many watts can your load handle? (may want to consider a fan)
Fuse blowing during high power testing is probably easier to do than blowing fuse while playing amp cranked through speakers .
Try again with same amp settings but start with signal generator at zero output, slowly bring up the generator signal level till the output starts to clip, then see what the signal generator level is for future reference.

[deci belle]

Hi g-one! The scope is a Tenma 72-320 (20MHz). Ya, that's 8Ω 200w dummy load into ch1 and sigGen into ch2. I left the amplitude at 1V and lowered the db to -30 to start.

This old Showman could be around 70w or so— some day I'll do that test. It's clipped at 4/vol at around 27V… at 10 it's 36V.

The sweep is set to .2 mS/frame. I have two peaks with a trough in between to fill the screen top to bottom w/ ch1's developed flat-top even with the overlapping clean sine wave by adjusting the variable volts/div dials set differently for each channel to get both waves virtually the same size (ch1 is @ 1V/div and ch2 is @ 50mV) just for fun. Is that what I should be doing?

…just getting familiar with everything at this stage— not in any position to actually analyze anything in particular yet.

Also, I don't know how to read the sigGen's gauge, g-one— the 5218 has three scales: two for volts and one for db.
It doesn't seem to be reflecting what's going on with the amplitude dials: the dials are at 1V/-30db and the gauge says .75 or 1.5 rms volts/-10db. I had previously calibrated the 5218 with the scope (the inner trim-pots), and everything seemed good to go.

[deci belle]

Thank you for answering the question about the amplitude dials and how they correlate to the gauge, g-one. Now I see how the coarse dial sets the range for which scale to read and the inner (fine) adjustment is used to dial in the voltage within that range.

I'll start a new thread about the scope's variable volts/div question.