Capacitance Meter II

[Jdansti]

I just used my capacitance meter kit to check a motor capacitor and thought I'd post this to show that the meter works on various forms of capacitors. Another reason for listing this is that the original post about the kit was flushed down the black hole earlier in the week, and I want to provide the information about the meter for those who would like one.

The meter kit is a Jyetech 06001, and they sell for $11 to $15 around the web.



The instructions aren't that great, for example, they don't tell you the polarity of the DC jack (center positive) and they don't tell you how to install a switchable 9V battery connector. Fortunately the do provide a schematic.

Here's a photo of the motor capacitor under test. It's marked as 35 μF, and reads 30 μF.

[armdnrdy]

Nice John,

I love me the test equipment. 

Do me a favor and test various pico farad capacitors. It is generally in the low range where capacitance meters fail.

I have a Fluke 87 that doesn't read small capacitors. I just metered a 150pf ceramic that registered 2.5 nano!

If this thing works good in the lowest ranges......I might have to have one!

I'll have to modify it with tone controls, germanium trannys, and carbon comp resistors just so it's cool! 

[Jdansti]

I'd like to see that information too. I'll try to do it this evening and I'll post the results. The manufacturer claims the meter is accurate to within ±2%.  I'm not sure that I buy that because they're basing the claim on the fact that two critical resistors are rated at ±1%.  There's a lot more that goes into the error than a couple of resistors. 

I can start at 10pF work up to 470 μF. I'll build a table showing the results of say 5 caps of each value and list the tolerances. 

If you want coolness, set it up to communicate with your computer and log all of your measurements.  The board and chip are designed to do this, but they don't give you the resistors or connectors to make it happen.

[armdnrdy]

That would be cool!

I would definitely want to box it up. Some thoughtful use of sockets, pin headers, and the like would be in order.

[Jdansti]

I got a bunch of readings from 1pF to 470μF, but haven't had a chance to tabulate them. I'll try to do that tomorrow.  I believe that when I do, I'll find that almost all values from are going to be within the specified tolerance ratings. I measured five different caps for each value when I had enough to do so.  One cap was way off but after measuring several of the same value, which looked fine, and then measuring the wacky cap again, I got the same high results.

The only problem that I ran into was that the meter would occassionally freeze on a reading.  When I pressed the zero switch to try to clear it, it would constantly display "C0" instead of displaying C0 for a couple of seconds and then displaying "0.00P". I was able to get around this by switching the power off and then back on. It appears to be some kind of bug in the software. Fortunately it's more of a nuisance than a show stopper.

[LucifersTrip]

http://www.diystompboxes.com/smfforum/index.php?topic=94116.0
http://www.diystompboxes.com/smfforum/index.php?topic=94116.msg812844#msg812844

[Jdansti]

Seems to be a popular meter!

[LucifersTrip]

You really can't beat it for the price, especially since it's a fun, quick project, too...and it really is a good idea to have if you build alot, so you can see how inaccurate caps can be, especially those cheap ceramic ones.

[Jdansti]

Here are the results in chart form. I split the results into four charts for better resolution.  All of the units for the charts are in pF to make the plotting easier and in order. The table is pretty long, so I'm not posting it.  If someone wants it in excel or PDF format, I can provide a link. Only a few caps had readings outside of their stated tolerance.  One 220 μF cap was way off. I measured 5 caps for most of the points on the charts. Some of the measurements were so close that they appear as one point.

[PRR]

> motor capacitor ....marked as 35 μF, and reads 30 μF

Motor caps don't have to be exact. Ideally they would vary all over the place as motor load changes; in practice they use a compromise value and it isn't too fussy. (And a sharp maker facing a 33u +/-10% spec will strive for 30uFd to save a penny of foil.)

Why is the cap loose? If new, it may need an hour at voltage to fully form. If old, it may have zapped-off some foil in start-stops and line-surges.

Electronic caps are more likely to try for the "exact" value, but 5% tolerance is pretty good and 20% tolerance is not uncommon. Most cap jobs, 20% off is plenty close enough. So I would not trust general electronic caps to any 2% spec without some cross-verification.

EDIT: yes, checking a w-i-d-e range of assorted caps does give some verification through diversity. Each one may be 5% off but they won't all be 5% the same way.

To get 2% or 1% caps you must either over-pay or luck-out on some costly-caps let-go cheap at surplus. And since the electronic industry moved to Asia, most of us will never see the deals we had around 1970. (Whole 18"x16" PCBs with thick-Gold connectors and loads of 0.1% R and 1% C.)

EDIT: your 310u "220u" is 40% high. Many large electros are rated -20%/+50%. Electro capacity is hard to control (you must roll it and form it before you can measure it exactly, and you don't un-roll over-value caps), and for most e-cap chores 'more is better' with really wide margin before the customer cries "too much!". The Japanese got very good at getting pretty darn close, but they've left the general cap racket to lower-price producers.

[armdnrdy]

Great work John!

Thanks for taking the time.....but now you're going to need to get another capacitance meter to check the actual capacitor values to see if your capacitance meter is accurate! 

[Jdansti]

Thanks but no thanks armdnrdy. I've had my share of dealing with NIST traceable standards from working in labs and with meteorological stations. NIST traceable weights, chemical standards, pressure standards, volume standards, voltmeters, etc. For home stuff, I'll just measure a bunch of non-traceable components and see how the results fall as Paul noted.

Thanks for the cap info, Paul.

[Jdansti]

>
Motor caps don't have to be exact. Ideally they would vary all over the place as motor load changes; in practice they use a compromise value and it isn't too fussy. (And a sharp maker facing a 33u +/-10% spec will strive for 30uFd to save a penny of foil.)

Why is the cap loose? If new, it may need an hour at voltage to fully form. If old, it may have zapped-off some foil in start-stops and line-surges.

Here's the cap from the replacement motor. It's rated at 124-149μF and it measured at 136μF on the meter, which is right in the middle of 124-149.



This motor can be wired for 115 or 230VAC. Why would the cap be rated at 165VAC?  Here's the motor plate:

[armdnrdy]

Nice pump dude!

How bout some gut shots!

[Seljer]

Because the capacitor is wired in series with one of the motor windings, a portion of the voltage gets dropped oover the winding (its either that or they're assuming you're not going to use it at 230V)

[Jdansti]

^ I hope it's not because they don't expect me to use 230V, because I do!

^^ I'll be able to show you my own guts if I continue to procrastinate about getting the pool water un-green. 

[PRR]

> This motor can be wired for 115 or 230VAC. Why would the cap be rated at 165VAC?

The cap may not be across-line, or in-line, but connected to *taps* which feel the same voltage for either 120V or 240V.

The cap voltage may have no relation to line voltage. Ferroresonant transformers put the cap on an arbitrary winding. The goal here is to minimize cap-COST. In general film-caps below 200V cost more per hunk of energy, electro-caps over 450V must be stacked at higher cost. Motor-caps started as wet electrolytic, were film-caps for decades, and I read that now electrolytics are hip again. So I don't know what you got, or what the motor does to it, but I assume 165V is a sweet-spot on the energy/cost curve.

[Jdansti]

Thanks, Paul.