Most of the long scale meters have a very high input impedance because of the accuracy issues. Typically 100 M Ohms or more. My Fluke 8506 is listed as 10,000 MOhms up to 20V range. The HP 3456A is listed as 10 to the 10th Ohms. Input loading should not cause an error on these instruments. The leakage of the test probe wires is probably higher. 3 1/2 digit meters are typically 10 M Ohms input Z.
A well designed autocal routine can work very well. its dependent on a few very stable standards and then well worked out ratioing. The internal references need to be checked but those are mostly really stable. Even the Fluke 720A is "Self Calibrating" since there is no other way to meet its accuracy spec.
A well designed autocal routine can work very well. its dependent on a few very stable standards and then well worked out ratioing. The internal references need to be checked but those are mostly really stable. Even the Fluke 720A is "Self Calibrating" since there is no other way to meet its accuracy spec.
I see great Impedance specs on the Hp3478A >10(10)ohm 30mV,300mV, & 3V ranges but in the 30V,300V ranges it listed at only 10Mohm....+/-1%
I read that to imply that at, say 50V its comparable to a the 269's 200Kohm/volt spec.........
Not implying the sweep meter is better in the least....just starting to get my head around not thinking ohm/volt and relative readings to calibrate,,,,
My habits have been to use my much cheaper digital for low voltage accuracy but the analogs for higher voltages . One + of a VTVM is it'll handle the occasional 'welding torch' event and live another day.
Don't see a digital as a 'test' meter.
Looking forward to the increased accuracy and utility of the HP3478A......
Just at that 1st step of ...."a good autocal routine" ...impedance of the meter.
Side Note:
I've experienced lately that Metrologist are not so free with their techniques since so many labs have shut down...but believe any bench'er should be able calibrate units he expects results from.....Haven't made that leap yet on most digitals..
The old Ham compliant about stickers all over his damn meters !
Thanks
Charles (Way off thread topic)
I read that to imply that at, say 50V its comparable to a the 269's 200Kohm/volt spec.........
Not implying the sweep meter is better in the least....just starting to get my head around not thinking ohm/volt and relative readings to calibrate,,,,
My habits have been to use my much cheaper digital for low voltage accuracy but the analogs for higher voltages . One + of a VTVM is it'll handle the occasional 'welding torch' event and live another day.
Don't see a digital as a 'test' meter.
Looking forward to the increased accuracy and utility of the HP3478A......
Just at that 1st step of ...."a good autocal routine" ...impedance of the meter.
Side Note:
I've experienced lately that Metrologist are not so free with their techniques since so many labs have shut down...but believe any bench'er should be able calibrate units he expects results from.....Haven't made that leap yet on most digitals..
The old Ham compliant about stickers all over his damn meters !
Thanks
Charles (Way off thread topic)
The older concept of Ohms/Volt came from current operated indicators (the 50 uA meter movement for example). With voltage sensitive devices (any electronic meter) the concept doesn't really make sense.
A 10 Mohm input will cause a 1% error with a 100K source. If you are measuring a 100K source at high voltages you probably don't need more precision (unless its a bias on an electrostatic transducer??). You can get higher impedance but its not real simple.
A 10 Mohm input will cause a 1% error with a 100K source. If you are measuring a 100K source at high voltages you probably don't need more precision (unless its a bias on an electrostatic transducer??). You can get higher impedance but its not real simple.
334
Hey Chris,
I managed to pick up a 344. Based on this thread, I figured it would a nice analyzer to start out with. I know the specs are not as nice as the 339. But, it seems to made really well. I only checked it out on voltage so far. Highest range was a little twitchy. I guess I need to think about cleaning the contacts. But, so far I like it. I do have 2 questions for you however:
1. Since these units are pretty old, do I need to replace any electrolytics? Or, should they still have some decent life to them?
2. I do not have any BNC cables yet. I need at least one from my signal generator to the 334. What do you recommend for BNC's? Do you make them yourself? Or, is there a particular cable that you like (impedance, make)? To make them yourself - most of the BNC cable connectors I have seen require crimping of the pin and the shield with special crimpers and dies.
Thanks,
Steve
Hey Chris,
I managed to pick up a 344. Based on this thread, I figured it would a nice analyzer to start out with. I know the specs are not as nice as the 339. But, it seems to made really well. I only checked it out on voltage so far. Highest range was a little twitchy. I guess I need to think about cleaning the contacts. But, so far I like it. I do have 2 questions for you however:
1. Since these units are pretty old, do I need to replace any electrolytics? Or, should they still have some decent life to them?
2. I do not have any BNC cables yet. I need at least one from my signal generator to the 334. What do you recommend for BNC's? Do you make them yourself? Or, is there a particular cable that you like (impedance, make)? To make them yourself - most of the BNC cable connectors I have seen require crimping of the pin and the shield with special crimpers and dies.
Thanks,
Steve
You should use 50 ohm BNC cables. These are the same as old 10Base-2 ethernet used, so you can sometimes find old ones that way. You can also make them yourself from 50 ohm cable (e.g. RG58) and twist-on connectors. I buy most of my cables from Monoprice. They have a wide selection of 50 ohm BNC cables which will probably be cheaper than you could make yourself (make sure to get RG-58 not RG-59 ones!). They also have BNC-to-RCA (and everything else) adapters which can come in handy when making measurements with audio gear.
Last edited:
That's pretty close to the reading i got as well. I played around with my 339A tonight and got .004% of the internal oscillator. I had long cables, adapters and was not careful about cable placement etc so that's good enough for me.
I started working at An audio manufacturer and we have Tek AA501 Dist analyzers and SG505WD and WQ oscillators and i really like how fast the AA501 is and the direct reading THD in % is really nice. This has me considering selling my 339A and buying a AA501 setup. It seems to be smaller and easier to use and faster.
If I keep my 339A I should probably recap the power supply. Suggestions on brand, series of caps to use for best results?
If the 339 shows no signs of ripple, there's not much point digging around the innards.
The SG505 and AA501 are really nice instruments. The Thd of the SG505 should be 0.0006% or lower.
The AA5001 is has remote interface over GPIB in the appropriate chassis.
There was a govt spec for these that all the analyzer companies built instruments for. The AA5001, the HP8901, the Boonton 1120 and 1121, the Krohn Hite 6880 and some obscure Sound Technology models all have similar specs and seem to be targeted at the specific contracts. All of the Boontons on the surplus market I have found have mil tags on them.
Any of these would be very good on a bench. The Boonton is quick and easy to use and has synthesizer frequency accuracy as well as quite good performance. The HP is similar but you want the b version with its true balanced input. The Krohn Hite stuff seems to have a .002% distortion floor. However the auto tuning works real well. I saw the ST stuff at Sound Technology but only saw one sample on eBay in the last 10 years.
Many of these can be had for around $300 if you are patient.
There was a govt spec for these that all the analyzer companies built instruments for. The AA5001, the HP8901, the Boonton 1120 and 1121, the Krohn Hite 6880 and some obscure Sound Technology models all have similar specs and seem to be targeted at the specific contracts. All of the Boontons on the surplus market I have found have mil tags on them.
Any of these would be very good on a bench. The Boonton is quick and easy to use and has synthesizer frequency accuracy as well as quite good performance. The HP is similar but you want the b version with its true balanced input. The Krohn Hite stuff seems to have a .002% distortion floor. However the auto tuning works real well. I saw the ST stuff at Sound Technology but only saw one sample on eBay in the last 10 years.
Many of these can be had for around $300 if you are patient.
M-Audio Audiophile 192 anti-aliasing filters
I am trying to use an Audiophile 192 (AP192) in a software spectrum analyser using the Cordell Distortion Magnifier. The problem I observe is that the AP192 ADC and DAC anti-aliasing filters appear to always kick in at around 20 KHz at ALL sample rates. Attached is the power spectrum for loopback of white noise.
I would really like to run the AP192 at a 192 KHz sample rate, with anti-aliasing filters at around 96 KHz. Is there any way to control the cut-off frequency of the AP192 anti-aliasing filters? Perhaps this is a limitation of the AP192 Linux sound card driver.
My configuration is:
AMD Athlon dual-core X86-64 5600+
4 GB RAM
Fedora 14 Linux
running Audacity and other sound software
M-Audio Audiophile 192
I am trying to use an Audiophile 192 (AP192) in a software spectrum analyser using the Cordell Distortion Magnifier. The problem I observe is that the AP192 ADC and DAC anti-aliasing filters appear to always kick in at around 20 KHz at ALL sample rates. Attached is the power spectrum for loopback of white noise.
I would really like to run the AP192 at a 192 KHz sample rate, with anti-aliasing filters at around 96 KHz. Is there any way to control the cut-off frequency of the AP192 anti-aliasing filters? Perhaps this is a limitation of the AP192 Linux sound card driver.
My configuration is:
AMD Athlon dual-core X86-64 5600+
4 GB RAM
Fedora 14 Linux
running Audacity and other sound software
M-Audio Audiophile 192
Attachments
I realize this is an old thread, but at the moment I'm considering the Rigol DSA815 with tracking generator (DSA815TG) at $1400, new in box. Considered a beginners analyzer because of it's "limited bandwidth" of 900KHz - 1.5GHz as well is a not the best PLL, but if you're looking to cover the AM & FM bands, 1.5 GHz is plenty. Check the YouTube &EEVBLOG Videos.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.