The FTDI based USB-GPIB adapter...

[Shannon Parks]

Well, I received mine Monday from Spark Fun Electronics. I fired it up Monday night and played with it for about half an hour, but will need more time to get it going. Right now I'm using Tera Term Pro and the adapter is directly connected to the Boonton. The Boonton is throwing back Error 22, meaning it doesn't like the syntax. Makes sense since I don't know what I'm doing yet. :) I wouldn't rush out and grab one yet.

Shannon

[WA4SWJ]

Shannon,

OK - I'm waiting with baited breath.

What does that mean anyway - that one has eaten fishing worms??

Regards,

[mesherm]

It is actually "bated" breath. The "bated" meaning the same as "abated" or held back. A fancy Shakespeare way of saying your holding your breath in anticipation. Yellow_Light_Colorz_PDT_09

[WA4SWJ]

Just having a little fun with our wonderful English language.

I am looking forward to the results of Shannon's tests. I've got a Boonton 1121 waiting to be tried out. It powers up fine and seems to be working, but I just haven't hooked it up to anything yet. It looks ratty on the outside but the interior is in very good condition. It will be fun to watch the computer put it through its paces once I get what I need to make that happen. Maybe after the July 4th Holiday I'll get some time to crank it up. Manually first then on to automation!

Happy 4th everybody!

[Shannon Parks]

OK, had a smart idea today and realized I should use 'NI Spy' on the lab machine. It shows the DLL calls to the machine that I basically already knew was happening, but it jarred my mental block. Yellow_Light_Colorz_PDT_05

Seems to work fine. Nifty device! Not sure if it is worth talking my boss into using in his lab, but perfect for this application as I would never drop $500 for a National Instruments USB-GPIB adapter. Hopefully I'll post a new version in July that supports this device - don't buy one until I do it, though!

Ed, the Boonton is real easy to get the feel for. The 300VDC limit on the inputs makes it easy for us to use and not blow up. On the middle controls, select a frequency, then an output level, then hook it up to whatever (including itself, using a BNC-BNC cable). Select the analyzer functions from the left hand side, such as level or distortion. You might look at 1V at 10Hz, 1kHz, 20Khz and 140kHz - to see if your source matches the analyzer. If they agree (within a tolerance), you are in good shape! Distortion at 1V should be below .05%.

Shannon

[Shannon Parks]

Got the Beta code working over the Fourth. Works well! Just need to optimize my code at this point. So the software currently supports the National Instruments line of GPIB controllers (and 100% compatible) and this Prologix USB-GPIB controller.

The nice thing about the Prologix unitis that it NI PCI and USB based GPIB controllers are $500 new, and still retain a lot of value on Ebay. While I use a $40 16-bit ISA NI GPIB card in the lab computer, I'm one blown motherboard away from having to upgrade. I realize that using an ISA card isn't possible for many people, too. Now with a notebook, the Prologix controller and the luggage Boonton, you have a portable test center! I'm itching to sweep my Klipsch speakers in-circuit.



Notice the new controls for using the Prologix controller (plus a debug window that will be omitted). It is basically controlled exactly like a serial port, as the FTDI chip is indeed a serial to USB interface. Then the Atmel microcontroller does all the GPIB level control. Serial --> USB --> GPIB. Loading the FTDI drivers is a snap. You just plug it in, point at the drivers, install, point at the drivers again, and install. No reboot needed.

I won't post the distributable until I have the code cleaned up. If anyone needs it, email me.

Shannon

[WA4SWJ]

Shannon,

I finally had a few moments to fire up the 1121 and here is what I got at 1 volt output:

10 Hz .0052%
1kHz .0040%
5kHz .0043%
10kHz .0053%
20kHz .0064%
40kHz .0087%
60kHz .0120%
100kHz .0203%
140kHz .0371%

Not too bad. So, my 1121 seems to be working fine. It's rough on the outside but it's clean inside and works otherwise. The sine wave output looks great on my 'scope.

Hope you're having a great weekend.

Regards,

[Shannon Parks]

I finally had a few moments to fire up the 1121 and here is what I got at 1 volt output:

Hi Ed,

Those look real good. Here's my 1120 at 1V output next to yours:

Code: Select all

        Ed's   Shannon's
10 Hz  .0052%   .020%
1kHz   .0040%   .015%
5kHz   .0043%   .014%
10kHz  .0053%   .016%
20kHz  .0064%   .012%
40kHz  .0087%   .017%
60kHz  .0120%   .017%
100kHz .0203%   .025%
140kHz .0371%   .040%

I printed off page 5-15 from the Boonton 1121 manual. It has table 5-13, Residual Distortion Test Record. This seems to be a very good Boonton checkout of both the source distortion and analyzer error. Sure, we're not doing a NIST calibration here, but if somethings out of whack, this test will show it 98% of the time. I'll see which of the other tests is best for level accuracy, too.

I just used a single jumper from source to analyzer for these tests. Basically, I just:
-set the source level
-then frequency
-engaged the proper low pass filter (500kHz means none, I guess - just press the 'on' filter again to turn them all off)
-select DIST for distortion measurment then press the 'dB' button to measure in dBs

Just did this test right now and I failed 8 tests!! Just a few dBs off. Anyhow, this can be tweaked, plus I have a slew of new 5534s which should have higher performance that the stock ones (just the Japanese cheapo's from Mouser, but the datasheet shows awesome specs). But I'll just swap out the current lab one with my $69 Boonton 1120. Yes, it is ugly and missing a few 'teeth', but it just now passed all the Distortion Tests.

Shannon

[Shannon Parks]

Hi Ed,

I just used Table 5-4, Analyzer AC Level Accuracy Test Record (page 5-11) for level testing. Instead of a NIST Calibrator, we just jumper it again. Do a 16V test instead of 30V tests, and omit the two 300V tests. My ugly Boonton looks very good on these tests, though reads a hair more than 1% low at 50kHz, 16V with 15.81V. No big whoop as I never push the output that hard unless I'm doing some tests with an inverse RIAA filter, for example.

Shannon

[WA4SWJ]

Shannon,

A question about your software - I haven't had time to load and try it and I don't have a GPIB interface yet.

The oscillator in the 1120/1121 obviously has some distortion as evidenced by the numbers we're both getting. It seems that the software should have a built in subtraction calculation to subtract the oscillator inherent distortion from the amplifier measured result thus yielding the actual distortion of the amplifier under test and not the sum of the oscillator and amplifier distortion. Perhaps you have done this already. It would be interesting to run a table of each frequency used for testing with the oscillator's distortion component at that frequency as a subtraction in the Excel spreadsheets. Maybe even in 1 Hz steps (but that's likely a bit of overkill). In my case my 1121's oscillator distortion seems pretty low so it might not make very much difference. But, I am interested in making the amp perform at its best so, from a pure measurement standpoint, it would be interesting - not that I could hear any difference between a few hundred's of a percent. Just an interesting engineering exercise.

Hopefully that little GPIB board is working OK for you. I might buy one in a couple of weeks when I have more time. Work gets in the way of hobbies!

BTW - what are 5534's? I'm assuming from your comment that they are op amps. Made by whom?

Regards,

[Shannon Parks]

It seems that the software should have a built in subtraction calculation to subtract the oscillator inherent distortion from the amplifier measured result thus yielding the actual distortion of the amplifier under test and not the sum of the oscillator and amplifier distortion. Perhaps you have done this already.

Hi Ed,

I did have a calibration feature in my program back in the previous version that I used with the HP8903B and VB6. It did track all the source THD+N and level corrections. I will incorporate this feature for THD, but haven't yet. One annoying issue with this feature is that you would be required to recalculate the offsets every time you wanted to change the source level (eg your source puts out less distortion at 1V than 50mV). For quick testing, calibration is probably unnecessary, but for final data, it will be worth it. It just slowed me down with the Poseidon's, so I hadn't added it yet.

That will be the nice thing about the 'Find Power' or 'Find THD' features. You can find the 1W, then calibrate at that source level and test. Or the 10W, or the 1%, etc.

Concerning the 5534, this guy's site is an amazing resource for audio opamp info:
http://www.dself.dsl.pipex.com/ampins/webbop/5532.htm

They have a big 'S' on them - made by Signetics. Great specs even when compared with the fancy opamps we have these days from AD & Burr-Brown.

Shannon