LCR vs ESR Meter

[requim]

@alm - I didn't check the prices for shipping other than what they posted.  I've ordered much heavier stuff from Amazon UK and had it shipped here for significantly less (like $16 for 4-5x the weight).  Who knows, maybe it is a typo.  They say $44 euro to ship to US and Canada for one unit and also for ten units.  If the shipping is 4-5 euros, which would be inline with my expectations then it becomes much more reasonable.

As for gear I've already got what I believe to be the essentials excepting a DSO hence the reason why I am looking at ESR/LCR meters.. But based on your response I will further consider the pulse generator as a possible project.  I'm still tossing around the 3455a, but that's a harder sell because of the space requirements.  In reality I'd like to finish my current projects before I take on another.  At the moment my project list consists of implementing the service note changes on my HP e3610a power supply, figuring out what is wrong with my Agilent 34401a DMM which just won't settle on a value (ideas are welcome), and fixing an Agilent 6611C DC power supply which is on its way here.  In addition to that I have a KitchenAid Mixer that is brand new for all intents and purposes (used like 3 times before it died) that I plan on tackling.  I also have several computer projects I've got spinning which hopefully will come to completion here in the next few days.

What I'm most intrigued by at this point is doing one of the ESR meter projects I found on kripton2035's site so I could customize the code.  My two biggest holdbacks on pursing that project are that I would want the end product to have a really nice industrial design enclosure, which I have no way of fabricating and second that I lack any sort of understanding to improve or modify the circuity to implement good input protection which Dave is always ranting on about or improving accuracy or adding 100khz functionality to the meter.  Oh and I'd like to swap out the two line lcd display with a multiline LCD, and while I'm sure I could handle the software side of it, I have no idea what would be required on the hardware side.

[grenert]

@grenert - I looked at the ESR meter you recommend, but the 44 euros they charge to ship that to the US kills any sort of deal.  I imagine that thing probably less than two blu-ray's in size and maybe slightly more in weight.  No justification for the astronomical shipping costs on it.

Requim, that's the shipping for 10 meters!!! 
As I said, for one kit, it's 49 Euros shipped:
http://clientes.netvisao.pt/greenpal/kits.htm

[requim]

@grenert - Ah, my apologies.  Apparently I need to go back and retake reading comprehension 101.  I obviously wasn't paying close enough attention.  The two gripes I have about it are one that it needs a good industrial designer to make a nice looking case for it and two, that I would like an ESR meter that shows both ESR and capacitance simultaneously, I don't want to have to double probe everything.  Is that wrong of me?

[alm]

Why would you want to measure capacitance and ESR at the same time? In my mind they're two different answers to two different types of questions: is this cap still good vs. what's the value of this cap within 1% or what's the value of this unmarked cap. Of course having them in one instrument would be nicer, but I wouldn't worry about having to measure twice.

[Conrad Hoffman]

requim- There are two numbers that more or less describe a capacitor. They can be in many units and combinations of units. The nice $16,000 Agilent LCR meter at work lets you choose, so having it read out capacitance and ESR (which is really Rs) is no problem. Actually, many cheaper meters can do that as well. Traditional bridges will typically give you capacitance and dissipation factor, but Rs and thus ESR is just a short calculation away. One simply learns to interpret dissipation factor and be done with it. Interestingly, you can also predict the high frequency Rs (ESR) from the low frequency numbers with a great deal of accuracy. Thus, any LCR meter that gives loss (Rs or ESR) tells you everything you need to know, though not always in the units you may be used to or have been told are important. Now, an ESR meter using a high frequency is very handy, and I'd take one of those over a C-only meter any time, particularly for service work where you're looking for the injured and deceased, rather than matching and measuring values. Most of this is part of that not-quite-done link I gave above.

[kripton2035]

Why would you want to measure capacitance and ESR at the same time? In my mind they're two different answers to two different types of questions: is this cap still good vs. what's the value of this cap within 1% or what's the value of this unmarked cap. Of course having them in one instrument would be nicer, but I wouldn't worry about having to measure twice.

I have right here on my desk a "nice" chemical capacitor that once was in a smps
it is marked as a 1000µF 35V and has a good esr of some 0.5 ohms
but in reality the capacity is only 1000nF !!!!!
if you check only the esr, you will not detect the capacitor...
but it's quite difficult to measure capacity in circuit...

[alm]

0.5 ohms is about ten times of what I would expect for 1000uF/35V, so I would label this cap as suspect. Especially in a SMPS which typically calls for low ESR caps. I wonder if the 1uF figure is correct, or if the meter is confused by series resistance or leakage.

Desoldering every cap to measure capacitance is not very practical, although ESR measurements aren't perfectly reliable in-circuit either.

[Conrad Hoffman]

Hey, I finally got my hands on one of the new Agilent handheld LCR meters today. It was only for a few minutes, but it seemed really nice. Does everything a Digibridge can do. This one went to 100 kHz. According to my standards it was dead accurate for capacitance and dissipation factor (does ESR and a bunch of other stuff too). A bit pricey for the hobbyist, a few hundred bucks, but a lot of capability in a small package.

[requim]

Jealous.  Which model were you using?  I assume it was the U1733C.  If only I could find one for less than MSRP.

Conrad - I read the article you wrote regarding caps.  Question - what value is there in reading caps at 100 kHz instead of 10kHz or 1kHz.  Your article didn't seem to address that.

[Conrad Hoffman]

I'm pretty sure that was the model, as it had the 100 kHz capability plus a computer interface- they get you for even more $$ for the non standard USB cable that isn't included    It was still a rather nice meter.

ESR meters usually measure at a high frequency, I think between 10 kHz and 100 kHz, though I don't have real experience with them. It isn't in the cap page, but I'm finding you can predict almost everything about the losses with two measurements at two arbitrary frequencies. That's still too much trouble for troubleshooting, so I just make the 1 kHz measurement. That seems sufficient to locate any bad cap I've run across. Big caps will have too low an impedance for a 1 kHz measurement with most bridges, so one is forced to switch to a lower frequency. IMO, that's still sufficient. I'm still making changes on the cap page and may add a whole different page for some techniques that aren't well known.

[Ronnie]

Hey, I finally got my hands on one of the new Agilent handheld LCR meters today.

Where did you buy the Agilent U1733C? I plan to get one sometime next year at Allied Electronics for US $400 though it is out of stock as of this writing.  http://www.alliedelec.com/search/productdetail.aspx?SKU=8350760

[Rufus]

Hey, I finally got my hands on one of the new Agilent handheld LCR meters today.

Where did you buy the Agilent U1733C? I plan to get one sometime next year at Allied Electronics for US $400 though it is out of stock as of this writing.  http://www.alliedelec.com/search/productdetail.aspx?SKU=8350760

Having it for a few minutes I don't think he bought it. http://www.newark.com have stock at $400. Looks like they are in short supply. RS Components in the UK (who are grouped with Allied Electronics) are quoting mid Jan delivery.

After posting here how I have managed without an LCR meter for years I decided to get one and see what I was missing.

I am generally impressed with it apart from some dumb features which are mostly annoying because they are so dumb. ESR mode is stupid, don't attach any added value to it for having an ESR mode.

[ElektroQuark]

ESR mode is stupid, don't attach any added value to it for having an ESR mode.

Why?
I find the ESR mode of my BK Precision LCR meter an invaluable tool. I have repaired various PC motherboards changing some electrolytics that have the correct capacitance but big series resistance, making computers unstable, hanging or not booting units.
Or are you referring at the Agilent LCR meter ESR implementation only?

[Conrad Hoffman]

FWIW, I'm still making edits and corrections to my cap testing tutorial page- http://www.conradhoffman.com/capchecktut.htm

After running some actual circuit tests on my box of "caps removed from service due to poor measurements", I've realized that the page had some significant errors. The main one being that losses for low frequency PS caps can be quite high before it makes any difference at all. Ripple levels are controlled almost purely by Cs and nothing else. I can see an esr meter leading one astray concerning LF filter caps if it's measuring at the high frequencies they typically do. I'm still testing some things concerning higher frequency caps and will edit as I learn more. I've also discovered that high frequency performance can be predicted remarkably well by two low frequency measurements, but that's probably a subject for another less practical page.

Conrad

[amspire]

FWIW, I'm still making edits and corrections to my cap testing tutorial page- http://www.conradhoffman.com/capchecktut.htm

After running some actual circuit tests on my box of "caps removed from service due to poor measurements", I've realized that the page had some significant errors. The main one being that losses for low frequency PS caps can be quite high before it makes any difference at all. Ripple levels are controlled almost purely by Cs and nothing else. I can see an esr meter leading one astray concerning LF filter caps if it's measuring at the high frequencies they typically do. I'm still testing some things concerning higher frequency caps and will edit as I learn more. I've also discovered that high frequency performance can be predicted remarkably well by two low frequency measurements, but that's probably a subject for another less practical page.

Conrad

Not sure if this is right. Even though a capacitor can be doing useful work at low frequencies with a high ESR, if it is out of spec, then it means the capacitor has probably lost electrolyte and should be replaced. A loss of electrolyte can often have a bigger effect on ESR then it does on the capacitance value. Also even if the low frequency waveform looks alright at low frequencies, above spec ESR means heat generated at the capacitor core that leads to failure. The heat is rarely noticable on the outside case of the capacitor unless it is close to failure. So if a mains filter capacitor is spec'ed at 10mOhm ESR, but it has 100mOhm ESR, it is a very sick capacitor and should be replaced.

Richard.

[Conrad Hoffman]

I work on a lot of old gear. Though I agree in principle with everything you've said, as a practical matter it can take years or even decades for the Rs (esr) to increase to where it's a problem. It's not always loss of electrolyte that's the problem, though if that happens the caps will probably have a very short life. There were two I just pulled out of my Wavetek signal generator that had corroded leads at the seal. Measured OK, but I couldn't bear the thought of leaving them in there.

There were various things I was quite sure of until I started measuring my box of rogue caps that I've pulled over the years. Though it made me feel better to replace them, many of those caps, in spite of what I considered high Rs, were doing their job just fine and could have been left alone, probably for years. IMO, there's just too much cap replacement going on for no good reason. If there is a reason, I'm all for it, but I have a real aversion to "technician blight"- torn up non-original equipment that's no better off for it.

I'm just starting some tests at higher frequencies for coupling caps and such, so we'll see how many myths I have to forget on that one!

[Rufus]

ESR mode is stupid, don't attach any added value to it for having an ESR mode.

Or are you referring at the Agilent LCR meter ESR implementation only?

OK, I should have said its ESR mode. Its ESR mode is a primary display mode so it measures and displays just like R mode (series equivalent circuit). That makes the ESR mode pointless. In ESR mode (unlike R mode) you can't take it out of autorange, or use relative mode, or calibrate the range which makes ESR mode completely stupid. A firmware update could probably elevate ESR mode to just being pointless.

ESR as an alternate secondary display mode would at least be a little useful.