• The move to the new server is done. There are some software and database maintenance updates in process. This has us passing the hat around to help out. We appreciate any donations. Seriously, even a dollar helps. The payment page may be found here - https://www.audiokarma.org/support.html

Equipment for bench testing

Jon_Logan said:
Well yeah, if you can make the plunge.

I bought one used from my ex employer. It was clean and was a dual domain unit (analog and digital) with wow and flutter and a few filters. Now if you ask me when I'm going to use the W&F function again, I can't answer that. It relegated my Sound Technology 1701 to a closet. The oscillator on them is very clean. The residual THD is very low. It's procedure writing capabilities are awesome. It's a very great piece of test gear. You'll need a scope, computer and APIB interface to run it, but I have absolutely no regrets.

There is a gotcha though. AP is trying to push their newer products and last I heard they refuse to work on the System One's. They might let you trade it in towards a more current unit.

For Analog measurement, I don't think there's better out there.
Click to expand...


Jon,

What do you think of the new AP stuff? Their rep took me for a test drive, and I felt like an old fart preferring my trusty old System One DD/DSP. Obviously, the newer boxes can do certain things exceptionally well. Still, I prefer the whole measurement paradigm of the One/Two. I figure that if all the box is doing is post process an acquired sample, I can already do that in many different ways.

But, the AP guys are smart, and I am sure much care went into the design. I figure that if someone can help me to get it to "click" in my head, I might get to like the new approach.

-k
 
Raccoon1400 said:
My bench has been swamped with power amps and I need a few more things to bench test them.

I have:
signal generator
tektronix 2215A scope 60MHz
2 8ohm dummy loads from digikey: FVE300-8.0-ND
fluke 115 RMS meter

I still need a proper set of scope probes. They need to have the 10x switch.
I'm looking at ones on digikey, but am not quite sure what I need, any suggestions?

I am going to get another pair of dummy loads.

Is there anything else I need?
Click to expand...


Sure, a nice Audio Precision System Two or an HP Dynamic Signal Analyzer would be a really fun way to drop ten or twenty grand. But, those are luxuries. If you are serious about testing power amps, there is somewhat more modest equipment that I would consider critical. One can do very basic tests and troubleshooting with pretty much what you have now. However, if you want to properly understand circuit function, or do circuit development and optimization, or research new devices and topologies, etc etc, I believe a step up in capability is needed. In no particular order:

- An audio band voltmeter. Your Fluke only measures out to about 1 KHz. MUST HAVE.

- A burst generator to create non-continuous waveforms or, better, an arbitrary waveform generator. My first choice would be a stand-alone unit, but it can be done via a computer. It is also fairly easy to build a box that will attach to the output of your signal generator and create bursts, but this method is limited in flexibilty. MUST HAVE.

- A tunable audio filter, so that noise and distortion can be tested according to various standard standards. For example, it is pointless to measure amplifier noise without a defined signal bandwidth. This filter can be purchased or, if you are good at it, you can design and build one yourself. MUST HAVE.

- A thermocouple probe. MUST HAVE.

- A digital storage scope is very helpful for understanding clipping behavior and transient instabilities, and it lets you document and share your work. IMPORTANT.

- A reactive dummy load. Purely resistive loads are, of course, a requirement, but it is also very valuable to have at least one load that simulates a speaker. This is easy to build yourself. IMPORTANT.

- No bench is complete without an LCR meter with a choice of test frequencies. (At least 120Hz and 1KHz.) IMPORTANT.

I'd be happy to answer any questions on the above, discuss affordable strategies, alternatives, etc.

-k
 
Last edited:
no bench is complete without one of these ...:D:D
A1384a_ex.jpg
 
Ken, what constitutes a reactive dummy load??

The "Blue ESR meter" only generates one frequency I believe, correct?
I also have a Fluke 179. how does that measure up??
 
ken kantor said:
Jon,

What do you think of the new AP stuff? Their rep took me for a test drive, and I felt like an old fart preferring my trusty old System One DD/DSP. Obviously, the newer boxes can do certain things exceptionally well. Still, I prefer the whole measurement paradigm of the One/Two. I figure that if all the box is doing is post process an acquired sample, I can already do that in many different ways.

But, the AP guys are smart, and I am sure much care went into the design. I figure that if someone can help me to get it to "click" in my head, I might get to like the new approach.

-k
Click to expand...

Hi Ken,

I started on their System One's, then a System Two Cascade, then a 2700. They actually got better in their analog specs with succeeding models. I'm sure you know the Cabot guy that is their chief/partner technical guy. Their digital side is very strong. The MLS signal for loudspeaker testing was pretty cool. The System Two Cascade and better models had a deeper memory for better distance range and resolution for their MLS test.

I don't think they (AP) ever got slack in their analog performance.

Is the Reps name "John" something?

Hi again, I kind of misread your question. When we switched from the AP SYStem One's to the System Two's, to the Cascade's, then the 2700's, the main reason was for the DSP horsepower that the more expensive models had. The 2700's had a Dolby Digital encoder built in that was central to the testing I was doing. We had kind of the left analog testing as a core test and most of the testing had changed to Digital Domain tests and the System One, while good, didn't have the memory depth or the 192kHz serial digital I/O.

As far as analog measurements go, I have never found the System One lacking. I've never had it cal'd and my looped back numbers are still frighteningly good. I'm used to their "old" software. I still use APWIN 2.24 and don't feel a reason to change. I use MatLab to generate "special" files such as Mega multi tones (60+ tones) and shaped tone bursts and load them into the Digital domain buffer. The System One is only a little bit challenged I find and have found a lot of work-arounds to get similar digital performance to the System Two Cascades on up. There's a sharp English guy that works for them.. Tom something. He was really good at digital signal generation, such as chirps, etc.
 
Last edited:
laatsch55 said:
Ken, what constitutes a reactive dummy load??

The "Blue ESR meter" only generates one frequency I believe, correct?
I also have a Fluke 179. how does that measure up??
Click to expand...


- The "Blue ESR meter" is not an LCR meter, and cannot be used as one. Fortunately, decent LCR meters are inexpensive these days.

- I dunno, I'll just go to the Fluke spec sheet right now... 50KHz. Not bad.

- I was thinking about something like:

http://www.ebay.com/itm/VINTAGE-Hew...839?pt=LH_DefaultDomain_0&hash=item1c227845df

- My approach to dummy loads:

http://www.stereophile.com/reference/60

Somewhere, I have some photos showing different power rating loads I have constructed. I'll look.

- An alternative approach to dummy loads:

http://www.aikenamps.com/spkrload.html

- A good example to justify their use:

http://www.stereophile.com/content/mesa-engineering-baron-power-amplifier-measurements

- Oh, another thing that is great on a power amp test bench is a "Kill-A-Watt." Hardware stores have them for under $20, and they are incredibly useful.

Hope this all helps!

-k
 
Many thanks! I had a Two Cascade on my bench for years, and a 2700 for only a couple of weeks. Both wonderful machines, but too expensive for me to afford when I went off to start my new venture. But, the One DSP I bought used is fine for what I need most of the time.

Any experience with the 525? That's the new one I tried and didn't care for. (Although the 1 MHz FFT option might make it worth while all by itself!)

I don't remember the recent rep's name, off hand. I used to deal with the great Mike Houge. These days I don't have much contact with the company, though I do chat with Rich Cabot every long while.

-k

BTW- are you free to say who you were doing the Dolby work for? You clearly are an experienced pro.

QUOTE=Jon_Logan;5319004]Hi Ken,

I started on their System One's, then a System Two Cascade, then a 2700. They actually got better in their analog specs with succeeding models. I'm sure you know the Cabot guy that is their chief/partner technical guy. Their digital side is very strong. The MLS signal for loudspeaker testing was pretty cool. The System Two Cascade and better models had a deeper memory for better distance range and resolution for their MLS test.

I don't think they (AP) ever got slack in their analog performance.

Is the Reps name "John" something?

Hi again, I kind of misread your question. When we switched from the AP SYStem One's to the System Two's, to the Cascade's, then the 2700's, the main reason was for the DSP horsepower that the more expensive models had. The 2700's had a Dolby Digital encoder built in that was central to the testing I was doing. We had kind of the left analog testing as a core test and most of the testing had changed to Digital Domain tests and the System One, while good, didn't have the memory depth or the 192kHz serial digital I/O.

As far as analog measurements go, I have never found the System One lacking. I've never had it cal'd and my looped back numbers are still frighteningly good. I'm used to their "old" software. I still use APWIN 2.24 and don't feel a reason to change. I use MatLab to generate "special" files such as Mega multi tones (60+ tones) and shaped tone bursts and load them into the Digital domain buffer. The System One is only a little bit challenged I find and have found a lot of work-arounds to get similar digital performance to the System Two Cascades on up. There's a sharp English guy that works for them.. Tom something. He was really good at digital signal generation, such as chirps, etc.[/QUOTE]
 
Hi Ken,

I used to work for Dolby.
I started there when noise reduction was the big money maker. I worked in the Consumer division Licensing area. Mostly testing NR and Pro-logic stuff on analog THD measuring gear, such as hp, and Kenwood stuff. Then DVD's were introduced and Dolby was part of the standard, so the System One's started showing up and we could now graph stuff with great performance. Woo-hoo!. We could also look at the S/PDIF bitstream and look at the state of the aux data flags. Pro/consumer, pre-emphasis or not, etc.

Then I moved to an implementation group and there, the evaluation boards were I2S or 192kHz serial data, or whatever, so everybody needed System Two Cascades. Then to get test time down, the 2700's were used in the Implementation group. It had the on board Dolby Digital encoder, so you could script a whole check out procedure, practically and get a high throughput. The only thing was, every submitted sample was kind of a one off, so canned procedures didn't really nail it.

I've used their little all-in-one tester with the LCD screen. I didn't like that. The ATX (?) one for industrial environments, that was okay, but I didn't get a lot of time on that one. For me, the System One is all I need. I haven't tried their newer (after 2700) stuff. They did demo a mainframe with plug in solution (a la National Instruments), but we were stocked with 2700's. I heard they introduced an HDMI I/O based thing, but that's after my time.

Mike Hogue, yes I remember him. I always used to talk to Jim Williams up there. He used to work for Hard Truckers or Ultrasound. I think he's not there anymore.

There's also that Prism system from the UK. Fair price. The only down side I remember was for open loop sweeps. It did not have a counter tuned graph function, so tape decks could not be swept. That was a few years ago. They might have that capability now?
 
Last edited:
laatsch55 said:
Thanks for the tips Ken, appreciate it. At least I know I don't know enough to make those decisions yet.
Click to expand...

Have fun, and take it a step at a time. Hey, you may get totally bored after 2 or 3 days of watching clipped sinewaves on your scope, or you may sell all your possessions to buy an Audio Precision System Two, an HP3562A, a Tektronix DPO70000 and a Bruel & Kjaer PULSE....

-k
 
I've watched sine waves for 5 years now. Never gets boring. I have a Leader LBO-308S, good little scope, and an Eico 377 Sig Gen, but nothing to measure distortion. I'm getting to where folks want to know the before and after on the White Oak/ Phase Linear conversions. HP 8903B??
 
Jon_Logan said:
Hi Ken,

I used to work for Dolby.
I started there when noise reduction was the big money maker. I worked in the Consumer division Licensing area. Mostly testing NR and Pro-logic stuff on analog THD measuring gear, such as hp, and Kenwood stuff. Then DVD's were introduced and Dolby was part of the standard, so the System One's started showing up and we could now graph stuff with great performance. Woo-hoo!. We could also look at the S/PDIF bitstream and look at the state of the aux data flags. Pro/consumer, pre-emphasis or not, etc.

Then I moved to an implementation group and there, the evaluation boards were I2S or 192kHz serial data, or whatever, so everybody needed System Two Cascades. Then to get test time down, the 2700's were used in the Implementation group. It had the on board Dolby Digital encoder, so you could script a whole check out procedure, practically and get a high throughput. The only thing was, every submitted sample was kind of a one off, so canned procedures didn't really nail it.

I've used their little all-in-one tester with the LCD screen. I didn't like that. The ATX (?) one for industrial environments, that was okay, but I didn't get a lot of time on that one. For me, the System One is all I need. I haven't tried their newer (after 2700) stuff. They did demo a mainframe with plug in solution (a la National Instruments), but we were stocked with 2700's. I heard they introduced an HDMI I/O based thing, but that's after my time.

Mike Hogue, yes I remember him. I always used to talk to Jim Williams up there. He used to work for Hard Truckers or Ultrasound. I think he's not there anymore.

There's also that Prism system from the UK. Fair price. The only down side I remember was for open loop sweeps. It did not have a counter tuned graph function, so tape decks could not be swept. That was a few years ago. They might have that capability now?
Click to expand...

Wow. I probably saw you one of the times I got to visit/tour the various facilities. What an amazing braintrust. Are you still in the Bay Area?

I got the Prism in for an extended eval. That was Julian Dunn's baby, if I remember correctly. I didn't ever settle into it the way I did for AP, although it was superb gear. It had more of an orientation towards digital signal analysis, with less stuff useful for loudspeakers and acoustics than would be ideal for my apps. But, hey, probably 95% of what I need can probably be done with a good soundcard. Every once in a while I need to measure in the digital domain, and that's what nerdy friends are for.

-k
 
Hi Ken,

Yes, Julian Dunn. A sad story. I didn't buy into all of his theories about the impact of jitters, but I respected him and am sad that he's not around. I agree that the Prism's didn't have that great interactive feel of the AP's.

Yup, still in the Bay Area. Hayward. It gives my wife and I about an equal commute. She works in the city, I work in the south bay.

I was going to drop by your place in Berkeley when I dropped by Meyer Sound a few months ago, but I had a rented truck and blew it off. You still have a lab there?

At the D, we looked at a lot of soundcards. Lynx, Creative, M-Audio, RME, etc. Many were not bit perfect in the digital domain. Lynx and RME were, but were sometimes finicky with their software and/or clock settings. There was a card that was cheaper than either, bit perfect and had great analog sections, CardDeluxe. I don't know what's available now, but a CardDeluxe with a good (software controlled?) attenuator in front of it would be a great core to a card based test instrument.

Do you remember a guy named Brian Elliott? I used to do some part time work for him. He's a buddy of Seigfried Linkwitz, both doing the open baffle designs. Brian was buying some NHT 12" for a sub. He really liked'm.
 
Raccoon1400 said:
My other multimeter measures frequency.
for the 4 ohm loads, I will use two pairs of 8ohm.
Click to expand...


Ahhh, but if you get the 4 ohm dummy loads you can put them in series for 8 ohms or parallel for 2 ohms, but maybe testing at 2 ohms isn't appropriate for the typical amp you're working on?
 
BinaryMike said:
I'm looking pretty hard at a Stanford Research Systems model SR1 audio analyzer to replace my old Sound Tech analyzer. Here the link: http://www.thinksrs.com/products/SR1.htm

Would you happen to have an opinion on the relative merits of that product?
Click to expand...

After I left the D, I showed an ex coworker the Standford Research box. My ex colleague liked it too and tried to show it around "The D", but they now seem hell bent on reducing/eliminating test hardware and seem to now be opting for soundcard or file based testing.

I've some other SRS stuff and thought it was good. I've used a synth and a lock in amplifier and they seem to be well designed with good specs. Audio Precision started getting "out there" with $27k+ pricetags. For that price range, the Prizm is in there too.
 
laatsch55 said:
I've watched sine waves for 5 years now. Never gets boring. I have a Leader LBO-308S, good little scope, and an Eico 377 Sig Gen, but nothing to measure distortion. I'm getting to where folks want to know the before and after on the. HP 8903B??
Click to expand...

I don't have any idea what the "White Oak/ Phase Linear conversions" are, but if you basically need THD, then I'd recommend just getting a good soundcard I/O and PC. Something like the TASCAM US144. Then you can use any one of about a million freeware or payware apps that do all kinds of audio measurements, including THD. Sure, the AP is perhaps a little better, but not significantly, for this kind of thing. Certainly my current portable laptop rig can do far better distortion measurements than the HP gear I relied on 15 years ago, at a tiny fraction of the price.

-k
 
You must log in or register to reply here.
Back
Top Bottom