Moving on to better multimeter for electronics

[ptricks]

For a bench meter I  still like the old fluke 8840a models.  They were built to last and the displays are better than LCD for the bench. I really don't like handheld meters for the bench because they seem to always be in the way.

Put the handheld where your bench meter is, I'm sure there'll be a net gain in space. 

And then I have to use two hands, one to hold the meter the other to turn the dial to change the setting, or push the buttons or change leads.  Remote readings require another cable and that means the meter that is already going to fall over is even more likely to. The viewing angle on lcd is generally meant for above the meter so on a bench it doesn't face me, it angles up. I can read my bench display from across the room.    There is a reason bench meters are sold .

[joeqsmith]

Reference is a calibrated AWG output which indicates 5.002V RMS on the calibrated 87V at 1KHz. 1dB of reference is what I'm measuring.

I'll run a sweep through it over the weekend if I get a few minutes and post the data. Probably one for another thread - will link it here when done. Will do the 87 as well.

...

This is 5V RMS into 50 ohms terminated at the meter.

For audio I doubt they are working with 50ohms.   To do some sort of A-B comparison, we should run them the same way.   We could drive the generator into the 50ohm, then come off of that with a 600ohm series, then into your meter.   It may be interesting to try it at 50 and 600 just to see if there is a difference.     We could also make up a -20dB 50-600 ohm pad.   I'm not much of an audio buff so your call.   Document how you decide to run it and I will attempt to replicate your results. 

[Specmaster]

For a bench meter I  still like the old fluke 8840a models.  They were built to last and the displays are better than LCD for the bench. I really don't like handheld meters for the bench because they seem to always be in the way.

Put the handheld where your bench meter is, I'm sure there'll be a net gain in space. 

And then I have to use two hands, one to hold the meter the other to turn the dial to change the setting, or push the buttons or change leads.  Remote readings require another cable and that means the meter that is already going to fall over is even more likely to. The viewing angle on lcd is generally meant for above the meter so on a bench it doesn't face me, it angles up. I can read my bench display from across the room.    There is a reason bench meters are sold .

I like you, also like the Fluke 8840A and 8842A and the thing with Flukes is that they rarely need calibration unless you are after accuracy down to the last uV  It has been my experience with the Flukes I own that they hold their calibration well over the years. I'm not 100% sure if these can be calibrated against suitable references using the front buttons alone or not. However I also own some HP and Solartron bench meters where is indeed possible and thus means that given suitable reference signals, you can keep these calibrated yourself very easily. I don't believe this is possible with hand helds however so that's a point to bear in mind. Another is that most handhelds also far more functions then bench meters do.

Fact is that meters are jusy another tool in your kit and it is up to us to decide which one best fits our purposes at the time, I don't consider my kit to be adequate without having both types at my disposal 

[joeqsmith]

" linear past 14khz ", "flat", "straight line"
 
     

Using the signal generator to a 6dB splitter, one leg going to a 50 ohm thru, the other to the scope terminated at 50 ohms for a sanity check.   Set the output to achieve 1.000VAC at 100Hz on the BM869s.  Swept the freq to -0.1, -1.0 and -3dB.   

I only checked the one meter and full disclosure, this is the newer of the two that was dropped, 50,000 life cycles and transient tested to failure.  No spit was used during the repairs, however, to be VERY clear, the meter does not have the same part numbers for the clamp.  I had tried modified the clamp stage to see if I could make it more robust.  If BD139 comes up with a better way to test their meters, I will use the stock meter.  While I damaged that meter as well  during my testing, I used the same parts during the repairs. 

[CDaniel]

The weak link for HF in this portable meters are the probes , obviously unshielded and not designed for this , so the results can be influenced by the testing techniques in many ways .

[joeqsmith]

The weak link for HF in this portable meters are the probes , obviously unshielded and not designed for this , so the results can be influenced by the testing techniques in many ways .

You are absolutely 100% accurate with this statement.  At these frequencies it IS VERY important to use proper cabling.   I was going to buy a new car but instead, picked up some decent cables to make these measurements. 

https://www.thecableco.com/emperor-double-crown-speaker-cable-pair.html

[2N3055]

If I remember correctly, this is similar to what I got on my BM869S.. 
My Picoscopes have 600 Ohm output, maybe a quick run on that?

[joeqsmith]

If I remember correctly, this is similar to what I got on my BM869S.. 
My Picoscopes have 600 Ohm output, maybe a quick run on that?

I did go ahead and make a 20dB pad attenuator to go from 50 to 600 ohms but have not tried it.   Give it a try with your Picoscope.   Maybe try it with 50 ohms as well.  I doubt my selection of transistors made much of a difference but it wouldn't hurt to have a sanity check. 

The whole meter audio thing is beyond me anyway.  I don't get the whole, I need a perfectly flat handheld meter to look at a non-linear amplifier.   Oh well.  Let's roll with it.       

[maginnovision]

The weak link for HF in this portable meters are the probes , obviously unshielded and not designed for this , so the results can be influenced by the testing techniques in many ways .

You are absolutely 100% accurate with this statement.  At these frequencies it IS VERY important to use proper cabling.   I was going to buy a new car but instead, picked up some decent cables to make these measurements. 

https://www.thecableco.com/emperor-double-crown-speaker-cable-pair.html

Clearly those guys know what they're doing, but they forgot the shields earth connector. Think they used fuel line for the outer insulation?

[Fungus]

Clearly those guys know what they're doing, but they forgot the shields earth connector. Think they used fuel line for the outer insulation?

...once heard they think you’ll agree, that’s a small price to pay given the performance on offer.

[maginnovision]

Clearly those guys know what they're doing, but they forgot the shields earth connector. Think they used fuel line for the outer insulation?

...once heard they think you’ll agree, that’s a small price to pay given the performance on offer.

Indeed. The description got me hooked, now I've gotta find out for myself.

[Fungus]

Indeed. The description got me hooked, now I've gotta find out for myself.

They'll be a good investment once they're fully burned in.

[maginnovision]

Indeed. The description got me hooked, now I've gotta find out for myself.

They'll be a good investment once they're fully burned in.

I'll probably have to pay someone for that, I don't have the equipment for it. I hope the effects of burn in won't be countered by shipping. I'll have to specify ground shipping, right? I'll email the experts.

[joeqsmith]

Using a 1/4 wavelength rule of thumb for a transmission line, at 500KHz that's about 150 meters.   Those are some VERY long DMM leads we are talking about.       I ordered mine custom and should have them today.   The McLaren F1 will need to wait. 

[2N3055]

If I remember correctly, this is similar to what I got on my BM869S.. 
My Picoscopes have 600 Ohm output, maybe a quick run on that?

I did go ahead and make a 20dB pad attenuator to go from 50 to 600 ohms but have not tried it.   Give it a try with your Picoscope.   Maybe try it with 50 ohms as well.  I doubt my selection of transistors made much of a difference but it wouldn't hurt to have a sanity check. 

The whole meter audio thing is beyond me anyway.  I don't get the whole, I need a perfectly flat handheld meter to look at a non-linear amplifier.   Oh well.  Let's roll with it.   

I got a bit of time, made few measurements. 1V P-P from 600 OHm siggen on Picoscope 3406D.
Sweep, from 100Hz to 500KHz BM869S (5V range) and MTX3293 (1V range, and 10V range)

And yes, it's just a curiosity, you don't really need perfectly flat AC bandwidth to repair amplifiers.
But when amplifiers distort, it is nice to measure full energy going into speakers, and that can easily go to 100s of kHz range..
For bandwidth linearity of amplifiers measurement you will need something more specialized than handheld anyways...

[joeqsmith]

It's a bit too hard to tell from the graphs. 

I tried my original BM869s, with probemaster leads and 50 600 ohm 20dB pad.  Again, setting meter to 1.0VACRMS as a reference.

limit, freq, measured voltage

0.1 dB,  149KHz, 1.0108V   Fell out high
1.0 dB,  373.1KHz,  0.8925V
3.0 dB, 456KHz, 0.709V

I then went back to the 50ohm termination using the original BM869s
0.1dB, 127KHz, 1.0116V
1.0dB, 385KHz, 0.8933V
3.0dB, 486KHz, 0.7078V


Staying with the 50ohm, I repeated the test using the new 87V that again, has 50,000 cycles on the switch.
0.1dB, 26KHz, 0.988V
1.0dB, 72KHz, 0.892V
3.0dB, 135KHz, 0.707V

**** Added Fluke 189 #1 from an old friend of mine  *****
0.1dB, 98KHz, 0.9883V
1.0dB, 160KHz, 0.8908V
3.0dB, 223KHz, 0.7089V

Fluke 189 #2
0.1dB, 70KHz, 0.9886V      Not as good as the worse of the two BM869s
1.0dB, 147KHz, 0.8928V
3.0dB, 213KHz, 0.7098

Fluke 189 #4
0.1dB, 98KHz, 0.9883V
1.0dB, 148KHz, 0.8910V
3.0dB, 202KHz, 0.7083V

A flat 14khz is good enough for a lot of audio related tasks, but if it has to be straight line 20khz to 30khz then it's has to be a 100khz meter >189, 89IV, 289, 8060A 

Based on the above, we know the Fluke 189 is the definition a straight line, and we can see the BM869s has a flatter response.  I guess we have to conclude that the BM869s is in fact straight line as well.   

...
The whole meter audio thing is beyond me anyway.  I don't get the whole, I need a perfectly flat handheld meter to look at a non-linear amplifier.   Oh well.  Let's roll with it.   

...
And yes, it's just a curiosity, you don't really need perfectly flat AC bandwidth to repair amplifiers.
But when amplifiers distort, it is nice to measure full energy going into speakers, and that can easily go to 100s of kHz range..
For bandwidth linearity of amplifiers measurement you will need something more specialized than handheld anyways...

I assumed the repair techs would use a scope or network analyzer for this, not a meter.   Then again, I've just been schooled on the definition of a straight line.  Not to be pedantic, well, maybe a little, I always thought a line by definition would be straight.       How does it get any straighter than a line??  Once again, I'm lost in the world of audio...

[joeqsmith]

While I'm thinking about it, someone should check their UNI-T UT61E.  That meter should be very fast as well.  In fact so fast that any high frequency blows right through it, literally!     I wouldn't be too surprised to find out its the fastest meter out there today.   I would run mine, but the BW was traded for added robustness. 

« Last Edit: Today at 09:04:04 am by 2N3055 »

I knew I should have grabbed those graphs.   

[Electro Detective]

Not sure what leaked cleaning solutions are being inhaled    but the AC bandwidth specs in the Fluke user and calibration manuals for 87-1, 87V and 28-11 -specifically- state 20khz max,
depending on the range/s.

If it's typo error and should be 200khz, then we all have dud meters randomly performing from 14 to 101 khz
and better jump on that Fluke Warranty asap before they get pushed out of business by Brymen and Uni-T 
or flog the used ones back to the unwary pawn dealers and Negative Feedback fearing Ebay sellers. 

---------------------

My verifications were simple as simple can be:

A linear/straight line performance (or whatever correct term floats boats) audio signal generator 10hz to 30khz hooked up directly to the 87 meter,
with similar results using either general purpose hook up leads/spaghetti or shielded audio cable.

Then the sig gen hooked up to to a brand new out of the box customers power amp (Crown 2400? heavy mofo) and amp outputs to 87 meter

And yes an oscilloscope was in on the action too to ensure the sig gen and amp were doing their thing
and a Tabor 8020 function generator with sine and all the other wave/pulse goodies, HP distortion meter,
as well as variable 0 to 32 ohm dummy load and halogen lamps for thrashing real audio gear to their limits

Come on gents, these are kiddie tests, the Fluke 87X will perform as specified by Fluke from 0 to 40 volts AC +-2.0%
with high frequency response rolling off a lot sooner once you hit on 400 to 1000 volts AC
 

For the fastidious: All contact points cleaned beforehand with isopropyl alcohol,
then after the tests lubed with WD40 (as recommended by Fluke) and some mule spit    to protect from tarnishing and crustiness during storage

i.e. I have a frozen pack of humble pie in the fridge, and state there is no way anyone will pull anything past 20khz on their stock 87x
at any level above 400mV AC, in actual real world average user conditions that Fluke caters for. 

I rest my case Your Honor   

[The Soulman]

#stoppolutingtheforumwithsecondrankhalvearsedcomedy
#AD536AKHrulez (yes rules with a z, because # uknow.)

[joeqsmith]

Not sure what leaked cleaning solutions are being inhaled   
...

My verifications were simple as simple can be:

A linear/straight line performance (or whatever correct term floats boats) audio signal generator 10hz to 30khz hooked up directly to the 87 meter,
with similar results using either general purpose hook up leads/spaghetti or shielded audio cable.

Then the sig gen hooked up to to a brand new out of the box customers power amp (Crown 2400? heavy mofo) and amp outputs to 87 meter

And yes an oscilloscope was in on the action too to ensure the sig gen and amp were doing their thing
and a Tabor 8020 function generator with sine and all the other wave/pulse goodies, HP distortion meter,
as well as variable 0 to 32 ohm dummy load and halogen lamps for thrashing real audio gear to their limits

Come on gents, these are kiddie tests, the Fluke 87X will perform as specified by Fluke from 0 to 40 volts AC +-2.0%
with high frequency response rolling off a lot sooner once you hit on 400 to 1000 volts AC
 
...

I rest my case Your Honor   

Let me just hook of this DMM to verify what we are seeing with the scope...   Did you spit on the meter while testing for good measure too?     

[Electro Detective]

No, and I won't loan the meter out to people testing suspect TVs, barbeque starters, bug zappers...   

I've got a 101 loaner for that 

Anyways, have fun with it, and looking forward to a 'JQS audio adventures testing frequency response of worthy multimeters'  Youtube video perhaps?   

[maginnovision]

Keysight 1252B:
312kHz for 1dB
605kHz for 3dB

1282A:
223.1kHz for 1dB
269.4kHz for 3dB

Bluepoint DMSC683A(I don't think this is even RMS):
2.701kHz for 1dB
6.393kHz for 3dB 

Fluke 98II(Technically it's all math but it does call the function a multimeter):
2.86MHz for 1dB
5.963MHz for 3dB

Keysight U1620A(dedicated meter inputs, I think this is similar to 1242C?):
150kHz for 1dB
193kHz for 3dB

For comparison these were with no attenuation, just fed by a sig gen High-Z output. 1V RMS signal output. Signal generator is close enough for this test. Also just used some probemaster leads. Good enough for me.

[joeqsmith]

The scope should be able to read below a 1/10 of a dB, should have more than one channel, should be flatter than a meter.  Why put a meter across the amp? Were you trying to impress someone with jotting down some numbers for the brand new amps response and felt the meter was giving you better answers than the scope?    I would have been laughing my ass off. 

Sorry, I have no plans to enter the world of the audio buffs.   

[joeqsmith]

I tested my Keysight 1252B and got 312kHz for 1dB down and 605kHz for 3dB down. My 1282A 1dB was 223.1kHz and 269.4kHz for 3dB, quite a bit sharper fall off. For comparison these were with no attenuation, just fed by a sig gen High-Z output. 1V RMS signal output. Signal generator is close enough for this test. Also just used some probemaster leads. Good enough for me.

That's impressive.   Looks like I had made a graph showing the UT61E's frequency response before and after mods.  That thing really peaks up before rolling off.   After mods, it just rolls off. 

[maginnovision]

I tested my Keysight 1252B and got 312kHz for 1dB down and 605kHz for 3dB down. My 1282A 1dB was 223.1kHz and 269.4kHz for 3dB, quite a bit sharper fall off. For comparison these were with no attenuation, just fed by a sig gen High-Z output. 1V RMS signal output. Signal generator is close enough for this test. Also just used some probemaster leads. Good enough for me.

That's impressive.   Looks like I had made a graph showing the UT61E's frequency response before and after mods.  That thing really peaks up before rolling off.   After mods, it just rolls off.

I think the 1252B was closest to flat and the 1620A had the worst response. I didn't actually log it all though, just watched as I went through.