By the way, Scott Wurcer and Walt Jung wrote a really great application note 27 years ago which is must reading for anyone listening to this thread -- first appeared in EDN -- AN-245 "Instrumentation Amplifiers Solve Unusual Design Problems" -- page 10-15 describes a prenulling technique which can extend the measurement range to 140dB below fundamental.
Also directly from AD (see Fig. 4), plus a design from Cordell :
http://www.diyaudio.com/forums/soli...ordell-distortion-analyser-2.html#post2010617
http://www.diyaudio.com/forums/soli...ordell-distortion-analyser-2.html#post2003882
Patrick
http://www.diyaudio.com/forums/soli...ordell-distortion-analyser-2.html#post2010617
http://www.diyaudio.com/forums/soli...ordell-distortion-analyser-2.html#post2003882
Patrick
Last edited:
I found these on eBay and thought you guys might be interested. One item in this bunch is a shielded Banana to BNC adapater:
Banana Plug to BNC Adapters 3 pcs. #1 - eBay (item 380210857280 end time May-28-10 14:24:52 PDT)
it has a metal shield that surrounds the banana plug and connectors. perfect for use with a spectrum anyl. input. this should help keep the input signal completely shielded.
Zc
Banana Plug to BNC Adapters 3 pcs. #1 - eBay (item 380210857280 end time May-28-10 14:24:52 PDT)
it has a metal shield that surrounds the banana plug and connectors. perfect for use with a spectrum anyl. input. this should help keep the input signal completely shielded.
Zc
Hi ZC,
Those are great for all instrument using banana jack inputs. From oscillator outs to voltmeters, spec ans or anything else where low level signals ride on banana jacks.
I'm going to build some, long time in the plans. Shipping from the US into Canada is often a deal-breaker on Ebay things. I'm looking to see if that vendor has those shielded ones on their own.
-Chris
Those are great for all instrument using banana jack inputs. From oscillator outs to voltmeters, spec ans or anything else where low level signals ride on banana jacks.
I'm going to build some, long time in the plans. Shipping from the US into Canada is often a deal-breaker on Ebay things. I'm looking to see if that vendor has those shielded ones on their own.
-Chris
Hi Jack,
Those papers are really good. That's one of the best sources for continuing education - papers and app notes. Certainly, everyone interested really should be reading things like that.
I trust Kiethley, especially for nV stuff. I'm not so sure about Fluke bench meters (I have 3-8200A meters also!), I've had too much trouble with them, and I guess working in a cal lab exposed me to the more unfriendly aspects of them. Having said that, their calibrators are excellent, as are their voltage references.
I gotta know. What are you measuring with that Jack? EMF gradients across connectors or jacks to PCB? DC potential from ground to system shields? I'm starting to run out of ideas now.
-Chris
Those papers are really good. That's one of the best sources for continuing education - papers and app notes. Certainly, everyone interested really should be reading things like that.
I trust Kiethley, especially for nV stuff. I'm not so sure about Fluke bench meters (I have 3-8200A meters also!), I've had too much trouble with them, and I guess working in a cal lab exposed me to the more unfriendly aspects of them. Having said that, their calibrators are excellent, as are their voltage references.
Excellent deal! You're a lucky lad to also get the leads, it's a system and normal leads do not cut it at all. All copper? Low thermal EMF type contacts for sure.
I gotta know. What are you measuring with that Jack? EMF gradients across connectors or jacks to PCB? DC potential from ground to system shields? I'm starting to run out of ideas now.
-Chris
The shielded dual banana adapter is from GR (I have one). Its great and unavailable new. I just checked and then bought the one on fleabay since they are unobtanium. Hopefully some China vendor will sniff this out and make some. I don't need the other adapters but they'll go in the bin with the rest.
I was just looking at Keithley electrometers since I have a specific task but discovered they are limited to low voltage. I need to measure 200V with no loading (bias on a condenser mike) and it looks like I need to use an electrostatic voltmeter.
The 8500 series Fluke meters seem to be well made and hold up for a long time. The earlier ones had a reputation around here for being pretty flakey. The 8800's also seem to hold up well. I have one sick 8860 but another that works fine.
I was just looking at Keithley electrometers since I have a specific task but discovered they are limited to low voltage. I need to measure 200V with no loading (bias on a condenser mike) and it looks like I need to use an electrostatic voltmeter.
The 8500 series Fluke meters seem to be well made and hold up for a long time. The earlier ones had a reputation around here for being pretty flakey. The 8800's also seem to hold up well. I have one sick 8860 but another that works fine.
My Keithley goes to 100V, just used it last month to verify the bias on a Rode NT01A. I have a few electrostatic voltmeters that I pulled out of the dumpster, unfortunately the extras are not accessible right now. The one I have is 1500V but little or no resolution below 400V.
Last edited:
I used it to measure the leakage in some electrolytics -- I found some vintage HP gear -- a couple of 465A amplifiers that were burned out in the innards -- turns out that the coupling caps were almost as low leakage as the wet slug tantalum that Williams used in his application note on the 775nV reference.
Sad to say -- I've purchased several dud 465A's -- and they were advertised as parts units -- all suffering the same issue with the power supply regulator and charred circuit boards.
I can't get his amplifier circuit to work, however. Just oscillates around 100kHz. Same thing with the circuit which Scott showed . I stick with the SSM2019.
Lithium Batteries
Chris,
Just wanted to say thanks. Worked out great. Just completed the installation of the new lithium batteries. I have 2 of these meters - one is the newer version and one is the older one.
I did use a 10K resistor as you recommended. But, I used my variable power supply instead of an extra battery as the temporary supply. The nice thing about using a variable supply is that I dialed in the voltage to exactly match that of the mew battery going in. This way, there would be no draining of the new battery while the temp supply was tacked in since the voltages were identical (for all practical purposes). Even though, I'm sure that this wouldn't be much of an issue anyway since both batteries would have been new and their voltages close. Plus, the resistor would have slowed the draining down. But, since I already have the supply, I figured why not. This supply also allows you to float the ground from the chassis. So, I wasn't worried about shorting out the positive side while soldering.
The other thing I was wondering while I was in there is - maybe I should replace the electrolytics being that they must be getting old as well.
Anyway - thanks again - you were very helpful - much appreciated,
Steve
Chris,
Just wanted to say thanks. Worked out great. Just completed the installation of the new lithium batteries. I have 2 of these meters - one is the newer version and one is the older one.
I did use a 10K resistor as you recommended. But, I used my variable power supply instead of an extra battery as the temporary supply. The nice thing about using a variable supply is that I dialed in the voltage to exactly match that of the mew battery going in. This way, there would be no draining of the new battery while the temp supply was tacked in since the voltages were identical (for all practical purposes). Even though, I'm sure that this wouldn't be much of an issue anyway since both batteries would have been new and their voltages close. Plus, the resistor would have slowed the draining down. But, since I already have the supply, I figured why not. This supply also allows you to float the ground from the chassis. So, I wasn't worried about shorting out the positive side while soldering.
The other thing I was wondering while I was in there is - maybe I should replace the electrolytics being that they must be getting old as well.
Anyway - thanks again - you were very helpful - much appreciated,
Steve
Hi Demian,
The solution is pretty simple, and you probably have everything you need already. All you need is some resistors, a stable power supply and something to use as a null meter. The impedance is actually not important since you are adjusting your power supply for a null between it's output and the condenser supply. By definition, no current flowing due to zero potential difference.
They use this same method (more refined divider) to measure standard cells at NRC and NIST. Use a normal DVM for your rough measurement, use a 'scope probe to present a 10 M impedance to start. Not accurate, but this doesn't matter for this. Once you are within the safe voltage range for your meter. You can also use a 50-0-50 uA analog null meter. It's like doing an experiment in school.
Anyone have some leads for a Wiltron 360A Scalor Network Analyzer? I got one of these (low freq option) that I'm trying to repair. Of course, no probes came with and I can't find any. I suspect an HP 8510 may possibly use similar probes. I don't know if I can fix the thing yet, so I'm not into spending a ton trying to find out.
Now, if anyone has a 3577A or similar they want to get rid of .... Didn't think so, not surprising.
-Chris
The solution is pretty simple, and you probably have everything you need already. All you need is some resistors, a stable power supply and something to use as a null meter. The impedance is actually not important since you are adjusting your power supply for a null between it's output and the condenser supply. By definition, no current flowing due to zero potential difference.
They use this same method (more refined divider) to measure standard cells at NRC and NIST. Use a normal DVM for your rough measurement, use a 'scope probe to present a 10 M impedance to start. Not accurate, but this doesn't matter for this. Once you are within the safe voltage range for your meter. You can also use a 50-0-50 uA analog null meter. It's like doing an experiment in school.
So darn true. And people throw them out when they come across them in a lab situation I've just missed several throughout the years. It looks like the only alternative will be to construct some new. Banana plugs can be had new. Cut out some insulator material to hold the plugs, make a copper top plate with clearance for your hot connection, then sheet stock wraps around to form the skirt. Solder the metal together, then mount the insulator and terminals. I'm planning a BNC connection on mine, and may even include a 600 ohm load on some (may as well). If you want pretty, dipping the outside in paint or that tool grip stuff would do the trick.
Anyone have some leads for a Wiltron 360A Scalor Network Analyzer? I got one of these (low freq option) that I'm trying to repair. Of course, no probes came with and I can't find any. I suspect an HP 8510 may possibly use similar probes. I don't know if I can fix the thing yet, so I'm not into spending a ton trying to find out.
Now, if anyone has a 3577A or similar they want to get rid of ....
Didn't think so, not surprising.-Chris
Hi Steve,
If you see some old electrolytic caps and can get similar replacements, then go ahead. The new ones will be rated at higher voltage to keep the can size similar, just work carefully. If you can't get the same axial (?) mount, using a radial might work. I hate seeing that, so make an extra effort to get caps that have the same mounting.
I use a battery because having the leads slip off is too easy with a power supply. The additional battery is small and allows you to move the unit around if you need. It worked for you, so thats great.
Hi Jack,
Some of those older higher grade components are really best left in there. I measure them to be sure, but otherwise I'd be using a 'scope and such to see if they work in an operating unit. To many new techs like to jump in and tear everything out. A heart breaker when you get one pre-distressed. So unnecessary.
Sorry to hear about your experiences Jack. There are a lot of scammers on fleabay and other auction sites. I even bought scrap from a HAM once. So much for that image of trustworthy people. I shouldn't really say that, because HAM radio operators are generally of excellent moral character, and are helpful people typically.
-Chris
If you see some old electrolytic caps and can get similar replacements, then go ahead. The new ones will be rated at higher voltage to keep the can size similar, just work carefully. If you can't get the same axial (?) mount, using a radial might work. I hate seeing that, so make an extra effort to get caps that have the same mounting.
I use a battery because having the leads slip off is too easy with a power supply. The additional battery is small and allows you to move the unit around if you need. It worked for you, so thats great.
Hi Jack,
You mean those small caps that are worth about $30 each to replace? Those are everything that solid tantalum caps are not. Anyone removing these need a good slap. Alert, Marantz 240 and 250 power amps use these.
Some of those older higher grade components are really best left in there. I measure them to be sure, but otherwise I'd be using a 'scope and such to see if they work in an operating unit. To many new techs like to jump in and tear everything out. A heart breaker when you get one pre-distressed. So unnecessary.
Well, if I save up some money, can I buy one or two from you? I'm used to charred PCBs. I bought a "working" HP 3586A that I needed to use. The board holding the backup battery is baked, corroded with a cap where the battery is supposed to be. No, it didn't work. If anyone knows what battery is supposed to go in there, I'd love to know. Agilent can't help me, nor can some other people I have contacted. A picture of an existing board would be wonderful in fact.
Sorry to hear about your experiences Jack. There are a lot of scammers on fleabay and other auction sites. I even bought scrap from a HAM once. So much for that image of trustworthy people. I shouldn't really say that, because HAM radio operators are generally of excellent moral character, and are helpful people typically.
-Chris
Hey Chris,
I used a power supply. But, I still tacked the positive and negative leads from the supply to the board. There was not going to be any slipping off. I made sure of that.
Thanks,
Steve
I know about the null meter solution but the source is in the G-Ohms so leakage is a real issue with the null meter (I have a Keithley and a Fluke null meter somewhere, but I switched to using my 8506 for that when I read that NIST had switched)
The Wiltron 360A is a vector analyzer, which means no probes in the sense you are mentioning. It has two RF receiver channels that are operated in sync/phae locked to it can measure the phase difference. The inputs will probably fry with more that a volt or two. Here is the link for info: Vector Network Analyzer - 360 - Anritsu This is a realm where everything is really expensive. The test cables for it could easily run $2,000. Some idiot fried one input to the Agilent 8753 VNA and the part from Agilent was $2,000 plus $900 to recalibrate it. We found a spare on fleabay for $500 and got it working OK.
Do you have the whole instrument with the S Parameter test set? I believe it has the source internal but I'm not sure. This page seems to have all the manuals for download: Anritsu 360 - 40 MHz to 26.5GHz Coaxial Vector Network Analyzer (888-683-2872) (National Test Equipment, Inc.) It seems to have millimeter wave capability. LF may mean 10 MHz.
The Wiltron 360A is a vector analyzer, which means no probes in the sense you are mentioning. It has two RF receiver channels that are operated in sync/phae locked to it can measure the phase difference. The inputs will probably fry with more that a volt or two. Here is the link for info: Vector Network Analyzer - 360 - Anritsu This is a realm where everything is really expensive. The test cables for it could easily run $2,000. Some idiot fried one input to the Agilent 8753 VNA and the part from Agilent was $2,000 plus $900 to recalibrate it. We found a spare on fleabay for $500 and got it working OK.
Do you have the whole instrument with the S Parameter test set? I believe it has the source internal but I'm not sure. This page seems to have all the manuals for download: Anritsu 360 - 40 MHz to 26.5GHz Coaxial Vector Network Analyzer (888-683-2872) (National Test Equipment, Inc.) It seems to have millimeter wave capability. LF may mean 10 MHz.
Hi Steve,
Well, you thought of that. Excellent, no one can really complain about the way you did it.
I look at it this way. A calibration and cert for a DVM will probably run $150 or so. The cost of a battery in comparison really isn't visible against the other possible costs. I should really check with Agilent (Remove all Doubt!) service to see what various calibration charges are. I do have a couple things I should have re-certified some year, once I can afford it again. The piece of mind and guarantied in tolerance conditions can be used to whip the rest of my stuff into shape. Yeah, only a cross-check, but that is better than doing nothing and hoping for the best. Agilent / HP gear (and many others) do tend to hold their calibration for years from what I've seen. Those Fluke 8200A that I have also do, but they tend to eat their Nixie display driver, I don't know why. The internals are far more accurate than their displays resolution. IT's not uncommon to see the display agree exactly with the applied voltage, or a 6.5 digit monitoring meter. That's pretty odd to see. I also have a Fluke 8000A that I should make a regulator for. They depend on the cells to smooth the power supply, and caps may not hold the voltage down in range.
-Chris
Well, you thought of that. Excellent, no one can really complain about the way you did it.
I look at it this way. A calibration and cert for a DVM will probably run $150 or so. The cost of a battery in comparison really isn't visible against the other possible costs. I should really check with Agilent (Remove all Doubt!) service to see what various calibration charges are. I do have a couple things I should have re-certified some year, once I can afford it again. The piece of mind and guarantied in tolerance conditions can be used to whip the rest of my stuff into shape. Yeah, only a cross-check, but that is better than doing nothing and hoping for the best. Agilent / HP gear (and many others) do tend to hold their calibration for years from what I've seen. Those Fluke 8200A that I have also do, but they tend to eat their Nixie display driver, I don't know why. The internals are far more accurate than their displays resolution. IT's not uncommon to see the display agree exactly with the applied voltage, or a 6.5 digit monitoring meter. That's pretty odd to see. I also have a Fluke 8000A that I should make a regulator for. They depend on the cells to smooth the power supply, and caps may not hold the voltage down in range.
-Chris
The conversation about a Wiltron 560A Scalar Network Analyzer has been moved to it's own thread, located here.
-Chris
-Chris
Anatech
Here is another ebay seller selling just the Banana to BNC adapter.
05-01192 BNC-Female Twin-Banana Plug Adapter Connector - eBay (item 360162020427 end time Jun-06-10 10:51:15 PDT)
Zc
Here is another ebay seller selling just the Banana to BNC adapter.
05-01192 BNC-Female Twin-Banana Plug Adapter Connector - eBay (item 360162020427 end time Jun-06-10 10:51:15 PDT)
Zc
Chris,
Thanks for the calibration thoughts surrounding an HP-3478A. I'm at that stage now. I've always tended to question the relative accuracy of a digital throughout a usable range when considering the impedance the meter may induce into the results.........Think it's foolish to reason that a 5 sig. digit readout is more accurate than even a 3 sig.digit unit.
Being 'old school' I've tended to approach things initially with a 'high impedance' VTVM then tandem my 100K-200K ohm/volt Simpson 269 (utra-precision in its day) and a Triplett 630NS (NS being the higher imp model) All Analog......Of course never claiming significant digit accuracy......but more comfortable with relative readings void of any meter impedance effect.
The 'Ham' influence from my father taught Metrology 'on the go'. The 3478a auto calibrate has seemed 'faith based' . Thoughts
Charles
Thanks for the calibration thoughts surrounding an HP-3478A. I'm at that stage now. I've always tended to question the relative accuracy of a digital throughout a usable range when considering the impedance the meter may induce into the results.........Think it's foolish to reason that a 5 sig. digit readout is more accurate than even a 3 sig.digit unit.
Being 'old school' I've tended to approach things initially with a 'high impedance' VTVM then tandem my 100K-200K ohm/volt Simpson 269 (utra-precision in its day) and a Triplett 630NS (NS being the higher imp model) All Analog......Of course never claiming significant digit accuracy......but more comfortable with relative readings void of any meter impedance effect.
The 'Ham' influence from my father taught Metrology 'on the go'. The 3478a auto calibrate has seemed 'faith based' . Thoughts
Charles
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.