Precision vs plain capacitors???

[Tubejunke]

I am getting closer and closer to going through with restorations on two of my most beloved diagnostic pieces, the Heathkit C-3 cap tester and a Hickok Model 209A V.T.V.M., milliammeter and capacitor tester.

In a small amount of research I have done I have learned of the use of "precision" type capacitors which up to this point I have never had to consider in vintage electronics. I'm just wondering if the superiority of (some brands) of new components like caps just how essential this low tolerance application really is.

Has anyone ever reworked say a capacitor tester with run of the mill caps and had good results?

[bob91343]

In general, the only capacitor(s) that require close tolerance are in the bridge circuit, and are used for comparison with the unknown. And you can measure a bunch of 'ordinary' capacitors to find one close enough.

Modern parts are very precise, and may be good enough without selection. But the main control of the unit isn't that good anyway, and probably can only be exact at one setting.

It boils down to the accuracy you want. These bridges were not high grade, and if you need good results you need to move up to a laboratory quality unit.

[VTVM]

I just used yellow poly caps (and regular aluminum electrolytics) on my Heathkit...I forgot the model number, but it's one of the later ones from the 1970s. It's fine.
I also replaced any resistors that had drifted out of spec. There were about 8-9 resistors a little out of spec. Then went through the calibration procedure.

[Tubejunke]

I think what I will do for the time being is use the capacitor function on my DMM for capacitor value and use the load test on the Heathkit C3 too see what really goes on under more real world conditions. That is the main reason that I wanted one of the older units in the first place. I do sort of wish that I didn't jump on the first thing I saw now realizing that there are much better capacitor testers out there. Of course they would need work too most likely. I used to have a rule in that I don't want my restoration equipment to need restoring, but sometimes they just don't make things the way that they used to, so you have to adapt. Really, I like the old equipment better in certain cases; I just don't have time to rebuild everything I find, or need.

[NoPegs]

Originally Posted by Tubejunke

I think what I will do for the time being is use the capacitor function on my DMM for capacitor value and use the load test on the Heathkit C3 too see what really goes on under more real world conditions. That is the main reason that I wanted one of the older units in the first place. I do sort of wish that I didn't jump on the first thing I saw now realizing that there are much better capacitor testers out there. Of course they would need work too most likely. I used to have a rule in that I don't want my restoration equipment to need restoring, but sometimes they just don't make things the way that they used to, so you have to adapt. Really, I like the old equipment better in certain cases; I just don't have time to rebuild everything I find, or need.

"Better" units? It is a wein-bridge. If you use tight tolerance components and construction it will be very accurate, Heathkit or not. Visualize it as an old style weighing-scale (like the statues of Justice are holding.) You're putting a known value in one pan (precision component) and comparing it to your unknown. If your fulcrum is precise, your measurement will also be precise.

Granted, I agree that it isn't good for testing modern-application caps to precision measurements, but for anything from the discrete analog era(which is less likely to complain about exact values anyway), a 0.2% error is nothing to worry about. LCR/ESR meters have their applications, but sometimes you just need to slap 40v onto a cap and observe what happens.

[bob91343]

The reference capacitor is, of course, important in determining the unknown value, but so is the panel calibration of the bridge control. If the control has an error compared to the linear rotation, it will result in errors in reading.

One important thing is the knob-to-shaft alignment. If it's off at all, that will be reflected in all the readings in terms of an angular error. Many manufacturing tolerances come into play, and an inexpensive bridge can't deal with them all with much confidence compared, say, to a GR 1650, etc.

[Tubejunke]

Originally Posted by NoPegs

"Better" units? It is a wein-bridge. If you use tight tolerance components and construction it will be very accurate, Heathkit or not. Visualize it as an old style weighing-scale (like the statues of Justice are holding.) You're putting a known value in one pan (precision component) and comparing it to your unknown. If your fulcrum is precise, your measurement will also be precise.

Granted, I agree that it isn't good for testing modern-application caps to precision measurements, but for anything from the discrete analog era(which is less likely to complain about exact values anyway), a 0.2% error is nothing to worry about. LCR/ESR meters have their applications, but sometimes you just need to slap 40v onto a cap and observe what happens.

When I said better units, I just meant that there were much higher quality units based on the familiar wein-bridge. A Heathkit is what is is; a kit for someone (anyone) to assemble as far as I know. I don't know if you could buy them pre-assembled or not. There is a guy on a YouTube video who shows capacitor testing on a tester with all the functions mine has, but with about two more voltage selections (mine only goes 450VDC; it went 600VDC which is the rated working voltage a lot of the caps we run into happen to be. Also, his tester had a volt meter and an amp meter to boot to view leakage current.

Sometimes I wish I were into the modern, low voltage stuff where I could just use an LCR/ESR meter and be pretty spot on with what is going on. AND I believe you can test in circuit which would be great although I don't understand how you can do that with parallel components in the picture. That doesn't change I don't think. LOL

[powerking]

Unfortunately, testing a cap with one of these "old school" Wein bridge testers requires it to be out of circuit and to do the leakage test(s). Since one lead is disconnected anyway, you're @ 25% of the labor to just replace it; many many folks have documented that almost every paper/mica/electrolytic out there on vintage gear repairs is just about guaranteed to be A) way off in capacitance or B) D.C. operating voltage leaky. The modern ESR cap testers use a ~~ 200mV 100kHZ test signal to effectively measure the power factor/leakage in circuit, which can be thrown out the door if the cap is seeing potential above that and up to its rating. This is a reason why the old school high voltage leakage selectable testers are still valuable.

Tom (PK)

[Tubejunke]

My take on capacitors in the past 3 decades is largely based on whether or not a unit was used either not at all, or way too much. I say this because I have radios and a couple of TVs that work fine on all original caps. I believe whole heartedly that it is because the set was used only periodically through the years but never set in a basement or attic for decades without use. Most of the time the ones chunked away in storage aren't just going to pop on like the day they were put there.

Then there are TVs and electrolytics which some people believe are the first thing to get rid of. I have found that they hold out pretty darned good if you bring them up slow. Its the non polarized stuff that needs to go. Particularly paper and the early brownish "drops" that need to go. It's good training to try to just repair a set (easier with a radio), but often you will get lost as when you find one thing that is wrong another is going wrong, or is wrong at the same time. A little different than normal troubleshooting.

[init4fun]

Hi Tubejunke ,

Great for you to have gotten a C3 ! I'm happy to hear you got one since I know you had been searching for one for a while . Tell me , in your question you mention "precision VS plain capacitors" and I wanted to expand on this a bit . If the precision aspect is referring to actual value only there is one more piece of the pie missing . There is the rating "temperature coeficient" to consider as this is actually far more important than raw capacitance and here's why ;

A cap who has a perfectly defined value is put into our bridge circuit and the indicator pointer is aligned properly . All is great and in use the dial / eye tube / cap under test indicate properly the value of the cap being tested .

BUT ! the cap with the perfectly defined value has a poor temp coeficient and in the cabinet being heated by the tubes the capacitance begins to drift a bit . Now , after it has sat running for a while the same cap we tested when we first powered it up* will test with a different reading . I only put this out there as a side discussion on the fourth major cap classification (the first three being rating , working volts , and environmental temperature expectations) since most usually mention the first two , some get into the third when dealing with hot tubes in tight cabinets , and it's usually in the realm of highly accurate measuring equipment that the fourth charactistic comes into play ....

* PS , it's generally considered good practice to warm up any tube containing test equipment for a period of time to stabilize temperatures (and thus component values leading to more accurate readings) . I usually let mine run 10 or 15 minutes before use , this seems time enough for them to reasonably stabilize .

Happy cap testing

[Tubejunke]

Originally Posted by init4fun

Hi Tubejunke ,

Great for you to have gotten a C3 ! I'm happy to hear you got one since I know you had been searching for one for a while . Tell me , in your question you mention "precision VS plain capacitors" and I wanted to expand on this a bit . If the precision aspect is referring to actual value only there is one more piece of the pie missing .

Thanks, and you are very well spoken on this subject. From what I gather looking around on the Web at forums where people are restoring vintage test equipment, it appears that what we are really talking about is simply a very tight tolerance for what is essentially laboratory grade equipment made to precisely measure a given component; here it is resistance and capacitance.

Back when the C3 and similar instruments were made most electronics found in the home could have components varying up to I think 30% with no negative effects. But I think; and it makes sense, that you wouldn't want a number, any number of components being off or drifting 15% much less 20-30% in a device that is designed to give you a precise measurement. Percentages add up too, so there's another potential issue.

That was then though and this is now as they say. Supposedly we have far superior capacitors compared to what they had back then. I believe this only to a certain degree. With small value caps I say give me the "film" or "orange drop" type over the old paper/foil. Just a better design. Now these days we have the stumbling block of being able to obtain well manufactured caps since everything is made in some foreign country, and from what I hear you really have to watch what you get and from what company. A penny saved is no longer a penny earned in this arena.

In the 50s you simply could get American made products like a Sprague "Atom" cap and KNOW for the most part that you put in a quality component. I don't even think Sprague is in business anymore. I hope someone will chime in with perhaps a list of good brands and where to get them. Actually, I think I have made a thread on that at some point, but a reminder for newer members won't hurt. Back to the topic and my original thoughts which I am still unsure of; I wonder if good modern, fresh standard tollerance (whatever that is) would be good enough for this application. I'm sure that replacing whatever I have that is bad would make things better, but I want the device to do what it was meant to do. Kirk out!

[NoPegs]

Originally Posted by Tubejunke

I hope someone will chime in with perhaps a list of good brands and where to get them.

CDE (Cornell Dublier Electronics.)

Vishay

NichiCon

Nippon Chemi-Con


That's my short-list for who makes good caps, in no particular order, although CDE units do cost a bit more.