Questions about required test equipment

[z-adamson]

Yes I did see the sticky and the "show me your test equipment" thread.

That said, I was still wondering a few things.

Some guys have a frequency counter and others don't. What exactly does one do with a frequency counter and how does it apply to receiver/amp/preamp repair and calibration?

Also, what is a spectrum analyzer and how is it used?

What I have so far as I get into the audio repair hobby is...

Fluke 289 digital multimeter
Simpson 260 analog meter
Fluke 123 oscilloscope
Heathkit harmonic distortion analyzer
Tenma 72-490 sine wave / square wave audio signal generator
8 ohm 300 watt dummy loads
Solder and desolder equipment
All the basic hand tools

What else should I get?

 

[Ray Gianelli]

You've already got most of what you need. I've found that probably 80% of repairs can be done with just a multimeter. Scope comes in for signal tracing/waveform analysis. Those 2 pieces are a must, and you're set there.

I have a counter, really the only time I need it is to check/set the 19 kHz pilot on FM.

An oscilloscope displays amplitude versus time. A spectrum analyzer displays amplitude versus frequency... it shows where the frequencies have the most energy. Like the counter, it's rarely used on my bench. But when you need these items, there's really no substitute.

 

[ConradH]

Same- I use my counter for the 19 kHz but also for extra certainty on dial calibration. Otherwise, it stays connected to my audio generator because I like the pretty lights, but it's not really necessary. I'm a big believer in having a decent LCR meter, but most get by without one. $25 will get you a lot of capability with one of the little eBay boards.

 

[z-adamson]

Same- I use my counter for the 19 kHz but also for extra certainty on dial calibration. Otherwise, it stays connected to my audio generator because I like the pretty lights, but it's not really necessary. I'm a big believer in having a decent LCR meter, but most get by without one. $25 will get you a lot of capability with one of the little eBay boards.

http://www.ebay.com/itm/AideTek-DM4...733342?hash=item233a77029e:g:pVgAAOSw44BYi-ZS

An LCR meter like this?

 

[BinaryMike]

The Tenma audio generator doesn't have low enough distortion for use with your THD analyzer. You'll probably want a sinewave source with 0.01% or less THD between 20Hz and 20KHz for that purpose.

Another use for frequency counters: tape deck and turntable speed calibration. Wow and flutter meters don't necessarily include this function.

 

[for_p1]

Yes I did see the sticky and the "show me your test equipment" thread.

That said, I was still wondering a few things.

Some guys have a frequency counter and others don't. What exactly does one do with a frequency counter and how does it apply to receiver/amp/preamp repair and calibration?

Also, what is a spectrum analyzer and how is it used?

What I have so far as I get into the audio repair hobby is...

Fluke 289 digital multimeter
Simpson 260 analog meter
Fluke 123 oscilloscope
Heathkit harmonic distortion analyzer
Tenma 72-490 sine wave / square wave audio signal generator
8 ohm 300 watt dummy loads
Solder and desolder equipment
All the basic hand tools

What else should I get?

For audio work sectrum/distortion analyzer is a main tool. Get most advanced one you can find. For standalone unit I suggest Panasonic vp-7722, or one of Audio Precision units from eBay. You can also use computer based analyzer: Lynx audio interface and SpectraPlus software. Both option will cost you over $1000. Keep in mind that residual distortion of analyzer should be at least 10dB below any device under test.

For RF work in receiver you will need high quality FM generator. There are several models of Panasonic, Hitachi or Leader on eBay.

I would also suggest add RF spectrum analyzer with tracking generator. Nice and inexpensive (under $1500) units are now made now in China.

 

[z-adamson]

The Tenma audio generator doesn't have low enough distortion for use with your THD analyzer. You'll probably want a sinewave source with 0.01% or less THD between 20Hz and 20KHz for that purpose.

EXCELLENT point, I had not considered that. How about the ragtag 101? At 0.05% thd.....is that as good as it gets before getting into big $$? Since this is a hobby and I am only doing these repairs and restorations on my own gear that I will keep and use, big $$ on thid item isn't ideal.

 

[drtool]

In for class.

 

[for_p1]

EXCELLENT point, I had not considered that. How about the ragtag 101? At 0.05% thd.....is that as good as it gets before getting into big $$? Since this is a hobby and I am only doing these repairs and restorations on my own gear that I will keep and use, big $$ on thid item isn't ideal.

0.05% is a very poor level of distortion for generator. You need something between 0.001% (for tube gear) and 0.0003% (for solid state gear) to do reliable audio measurements.

 

[z-adamson]

0.05% is a very poor level of distortion for generator. You need something between 0.001% (for tube gear) and 0.0003% (for solid state gear) to do reliable audio measurements.

Good points. Let's say for example, I am testing with 1% distortion in my generator. If I know this up front could I not account for that while interpreting the rest results given by my harmonic distortion analyzer? Seems like I could test a known good amplifier (solid state) with a 1% distortion generator and then get a good baseline. This would be my "zero" so to speak and when doing "real" testing I would then look for distortion above and beyond what I saw when testing known good. Is this a viable approach for DIY?

 

[for_p1]

Good points. Let's say for example, I am testing with 1% distortion in my generator. If I know this up front could I not account for that while interpreting the rest results given by my harmonic distortion analyzer? Seems like I could test a known good amplifier (solid state) with a 1% distortion generator and then get a good baseline. This would be my "zero" so to speak and when doing "real" testing I would then look for distortion above and beyond what I saw when testing known good. Is this a viable approach for DIY?

No this is not a right approach. Usually you first do measurement of closed loop - generator/analyzer. That will show you what is residual level of distortion in your measurement system. It is commonly assumed that this will allow you to measure distortions in device under test if it produces 3 times (10 dB) more distortion than your closed loop measurement above.

In some cases where you use high resolution FFT, you can measure even lower distortion in devices, but it is only possible when distortion spectrum is significantly different from what you see in closed loop. Something like when measurement system only have 2nd and 3rd harmonics above noise floor, but device under test produces significant amount of higher order harmonics. But this only gives you qualitative measure.

If we are talking about audio, then distortion level is 0.01% or less for cheap gear and 0.001% or less for real hi-fi. Your closed loop measurement system should be much better than that.

I also do not like to speak in terms of %. What you need to measure are distortion spectrum components. The all are measured in dB below fundamental. I suggest to have closed loop residual distortion components of your measurement system at least 100 dB below fundamental.

Here is a closed loop distortion measurement of my test setup:

19-20kHz IMD Loop

• for_p1
• Mar 27, 2017

 

[restorer-john]

...Otherwise, it stays connected to my audio generator because I like the pretty lights...

My frequency counter really does much the same-nice long row of lit up red 7 segment LEDs. It's easier to set the exact frequency on the analogue signal generator with the freq counter tied through, via the loop out on the back.

 

[z-adamson]

Thanks for the great post p1. Closed loop....you mean generator and analyzer hooked up directly to eachother, right?

Would a full health heathkit ig-5218 have low enough distortion to be productive in testing for distortion in device? I have a couple leads on them, can't find distortion specs though.

 

[BinaryMike]

EXCELLENT point, I had not considered that. How about the ragtag 101? At 0.05% thd.....is that as good as it gets before getting into big $$?

The good old Tek SG502 might have low enough THD for use with a Heath analyzer. They generally perform better than specified and tend to be available at pretty fair prices. If it's not clean enough, consider an SG505 instead. Or trade the Heath for an analyzer with built-in oscillator.

 

[Alan0354]

You have everything needed already. It would be nice to have a signal generator that has gated input that you can gate the generator on and off. Then have a second cheap generator to generate the gate signal. Now you can do burst signal. This is mighty important to test stability of amplifier. If an amp does not have a lot of phase margin and close to the edge of oscillation, it likely will not show up if you just feed a continuous signal. For one you cannot put too large of the signal before you burn the amp and/or burn the load. With short burst, you can run amp into saturation. Amps tends to get unstable with large signal driving close to saturation.

For repair, you really don't need a spectrum analyzer unless you are going into business. Even at that, it's only good to have, not a requirement. Spectrum analyzer show the output in frequency domain instead of time domain like Oscope. Frequency domain means it show bars that represents the frequency detected. For ideal amp with no distortion, if you input a frequency say 1KHz, the spectrum analyzer will show 1 bar at 1KHz. But all amplifiers distort. Distortion is adding harmonics to the original frequency, usually the dominant ones are 2nd, 3rd, 4th and 5th harmonics. For 1KHz fundamental frequency( the one you input to the amp), the output will show a bar at 2KHz(2nd), 3KHz(3rd), 4KHz(4th)........... The height of the bars show the amplitudes of the harmonics and that can be calculated into total harmonic distortion (THD) as either -##dB or 0.0xxx%. This is mostly useful for designing amp to check the design. For repairing, you get what you get, usually it's just things burned, loose etc. You just fix it and move on.

 

[for_p1]

For repair, you really don't need a spectrum analyzer unless you are going into business. Even at that, it's only good to have, not a requirement. Spectrum analyzer show the output in frequency domain instead of time domain like Oscope. Frequency domain means it show bars that represents the frequency detected. For ideal amp with no distortion, if you input a frequency say 1KHz, the spectrum analyzer will show 1 bar at 1KHz. But all amplifiers distort. Distortion is adding harmonics to the original frequency, usually the dominant ones are 2nd, 3rd, 4th and 5th harmonics. For 1KHz fundamental frequency( the one you input to the amp), the output will show a bar at 2KHz(2nd), 3KHz(3rd), 4KHz(4th)........... The height of the bars show the amplitudes of the harmonics and that can be calculated into total harmonic distortion (THD) as either -##dB or 0.0xxx%. This is mostly useful for designing amp to check the design. For repairing, you get what you get, usually it's just things burned, loose etc. You just fix it and move on.

I would not trust repair work to anyone who cannot prove that after repair device matches all original factory specs. And you cannot do it without high precision tools.

 

[Alan0354]

I would not trust repair work to anyone who cannot prove that after repair device matches all original factory specs. And you cannot do it without high precision tools.

OP is doing it as hobby. I don't believe professional repair person needs FFT. You find the problem, fix the problem. All the other work does not require "precision" tool. What he has should be good enough.

I design amps, I need spectrum analyzer ( I have the QA401 which is a 200KHz pc based spectrum analyzer). That is plenty good for designing already. I don't even need a true spectrum analyzer.

You need to have good knowledge, good common sense, not necessary exotic precision equipment.

 

[for_p1]

OP is doing it as hobby. I don't believe professional repair person needs FFT. You find the problem, fix the problem. All the other work does not require "precision" tool. What he has should be good enough.

I design amps, I need spectrum analyzer ( I have the QA401 which is a 200KHz pc based spectrum analyzer). That is plenty good for designing already. I don't even need a true spectrum analyzer.

You need to have good knowledge, good common sense, not necessary exotic precision equipment.

You have a reasonably good analyzer. But I disagree that repair (and more so of really old gear) is limited to fix the obvious problem.

And sometimes it is necessary to fix something that does not present obvious signs of failure - here is an example http://www.diyaudio.com/forums/digi...nger-mod-what-can-50-do-deq2496-part-1-a.html

 

[ConradH]

SA is nice for design but not really necessary for repair and rebuild work. The OEM wouldn't have used it on the assembly line, just a standard THD meter. IMO, a more important tool for catching things is having an automated setup so you can run response curves and THD curves quickly. Otherwise, you just have to be really rigorous with manual equipment. The rule "Explain all anomalies." has served me well, as whenever I've ignored some seemingly minor and unimportant thing, like a rise in THD not present on both channels, it's come back to bite me you-know-where.

Even a super low measurement floor isn't essential. If you can get down below 0.01%, you'll spot most problems, though you may not be able to confidently confirm mid-band numbers. Even an old HP 334A and a modded Heath generator can get you there.

That all said, I still believe he who dies with the most test equipment, wins.

 

[Alan0354]

You have a reasonably good analyzer. But I disagree that repair (and more so of really old gear) is limited to fix the obvious problem.

And sometimes it is necessary to fix something that does not present obvious signs of failure - here is an example http://www.diyaudio.com/forums/digi...nger-mod-what-can-50-do-deq2496-part-1-a.html

Then we agree to disagree. Most problem can be spotted by just simple DC voltage measuring. Good understanding of the circuit is the first and utmost important thing. I just had an experience of brand new diode pack from digikey, two have one diode in it that measure Vbe of 1V or slightly more. I spotted them and replaced them. That was for setting the clipping level and bias. too much drop will create more harmonics. Didn't need to even power up the QA400 to spot that.

There are people believe in fancy equipment, I came from an environment that we make do with minimal and we pride on and produced great results. You should see the PhDs working on optical bench using chewing gum to stick the lens up for experiment and we published papers with that. Nothing is more important than knowing what you are doing. Until 2000, we designed mass spectrometers, semi conductor metrology equip that kicked the butt of competitors with nothing more than a Tek 475 scope, two wavetech generator and a HP8603A pulse generator. We do sub neno sec pulse circuits, ultra low noise circuit way beyond audio equipment. We did it with this little stuff and we were proud of it.

I don't have much experience in audio amp yet, but I believe it's same.

I would not suggest a hobbyist to spend hundreds to get something that is being use ones every other month at best. I don't even use mine ones a month. My suggestion is to spend the time to study electronics, transistors and know them in side out.