[Hammarlund] cap testing

[N4ue@aol.com]

I have had a lot of requests about testing caps, so I thought I'd share it 
here......

This is long, so don't let me put you to sleep. ha ha 

History: I collect boatanchors of all types. I have quite a few radios and 
LOTS of test equipment. When I got my R-390A and Collins 75A-4, I read with 
great interest, a LOT of articles on 'deadly caps', the 'black beauties of death', 
etc. On the R-390A in particular, if a certain cap in the IF deck shorts 
(common problem) , it takes out the IF filter in use at the time. The owner, 
unaware what caused his radio to be deaf as a rock, tries another selectivity 
setting. Another filter bites the dust, and so on.... I did a lot of research into 
these old paper caps and noted that the Collins Collector Association's 
Technical Advisor, (Dr. Gerald Johnson, PE) has long been a staunch advocate of 
wholesale replacement of these types of caps. I got my first exposure in cap 
failure, when I bought a URM-25D Signal Generator that had no modulation. Opening 
the case, I noticed a puddle of oil inside. Several of the large mica caps had 
split wide open. Next, I found some split "Black Beauty" caps inside several 
HP 608 series Signal Generators.

Equipment: I have an extensive number of cap testers in my collection. I have 
several 'lab grade' testers / bridges made by Boonton, Leader, etc. Several 
have current NIST  Calibration stickers. Still, they are NOT the tester I use 
almost daily.
I am going to make a blanket statement here:

"NO cap tester powered by a battery, will indicate a capacitor's leakage"

VALUE, yes. In fact, my NIST Calibrated, Leader LCR Bridge, (9 VDC) is 
excellent for identifying the capacity of an unknown part, but is useless for 
determining the leakage of a cap used in our beloved boatanchors.

Reading Dr. Jerry's postings and subsequent e-mails with him, have lead to an 
excellent, foolproof, method of determining the leakage of bypass and 
coupling caps.

Limitations: The good Dr.s method will NOT indicate the VALUE of a cap, but 
it will tell you if it's leaking. Please note: EVERY cap I have ever found to 
be a leaker (not shorted!), will still indicate almost it's exact marked 
capacity, when using any NIST Bridge. So, if you use only a cap checker (like those 
built into a DVM, or other battery operated device), you will only be able to 
check the capacity. In terms of leakage, this is a different, non correlating 
value.

How to test: You will need:
1, a VTVM (you cannot use a VOM) You can use an 11 Megohm input DVM, IF it 
has an analog bargraph. Trying to follow moving digits, just doesn't work.
2, a source of DC from 10 to 600 VDC.

E-Pay (er, E-Bay), is an excellent source of these, cheap. I have purchased, 
well.... you wouldn't believe me if I told you how many VTVMs I have collected 
from there...
I love to collect these, as there were so many different models, made by so 
many companies... all interesting. As far as the VDC source, you can find 
Heathkit cap checkers for cheap. I don't remember the model # (IT-11, IT-28, 
???etc). These have the good 'ole "EYE" tube. If you find one without the 'eye' 
tube, or with a weak eye tube, it doesn't matter. I only use these testers for the 
convenient switchable VDC output anyway. 
Note, since I have described this method to other boatanchor groups, the 
price of these cap testers has risen all the way up to the $10 level! You used to 
be able to get them for $2-$3......

Connections and use: Simplicity in itself.....

Caution, high voltages here, so use common sense!

1. connect the (-) of the P/S to the VTVM (-) or common
2. connect the (+) of the P/S to the center of a piece of coax (or one test 
lead)
3. connect the shield of the coax (or other test lead) to DCV on the VTVM

That's it! 

Say you wanted to test a .1 uf / 600 VDC "black beauty"....

Set the P/S to 600 VDC, the VTVM to 500 VDC (or whatever) and using INSULATED 
clips or probes, connect the cap. NOTE: The cap must have one leg lifted out 
of the circuit! One nice feature about most VTVMs is that you cannot burn out 
the meter, by having the VDC range on a low setting and hitting it with 
voltages 20X times higher. I say MOST of them are that way. Considering that MOST 
VTVMs use the same circuit, it's a pretty safe bet. ......
If the cap is good, the VTVM needle won't move. If you see any movement, just 
reduce the 'range' setting of the VTVM. You want ZERO voltage on the VTVM, 
even on it's lowest setting. ANY voltage indication, indicates some leakage. If 
you measure 40 VDC with 600 VDC applied, the cap is leaking THAT voltage, 40 
Volts.
I went one step further, and placed a N/O microswitch at the VTVM input 
terminals. Why? Pressing the switch will put the full P/S voltage to the cap, 
without having to wait for the R/C time constant. However, if I see ANY leakage, 
the cap goes into my very large cap scrap bucket. Leakage won't get any better 
with time, I can assure you! It is very common to see the above "BB" cap leak 
400+ VDC of the 600 VDC applied.

Observations and comments:
Several staunch lovers of Boatanchors insist that having leaky caps in the 
radio won't hurt the performance. They are partially right. IF (and a BIG IF, it 
is!!) the leaky caps were ONLY the cathode Bypass caps, this might be true. 
We are talking very low voltages and signals here. HOWEVER, if the cap is a 
COUPLING cap, having some B+ ride into the next stage, IS going to upset things.
One example: When I bought my 75A-4, I was told it needed to be recapped. I 
did some research and Chuck Rippel identified the usual suspects. Since this 
radio has quite a collector value, I would replace one or two caps and power the 
radio back up, checking to ensure it still worked. I was using the cal signal 
as a reference. AS I went, the radio begin to 'hear' better in increments. 
Once I was done (ALL Black Beauties were nfg, as well as a couple of micas), the 
radio needed a complete alignment. But that radio can rock and roll with any 
in my collection! Just a different radio. It is hard to believe that it worked 
at all, with the bad micas. One leaked 450 VDC of 500 VDC applied!

Back to the Hammarlund. I recently had the time to go 'back through it'. I 
had replaced the electrolytics and checked the tubes when I bought it. At that 
time, I noticed that the caps Hammarlund used, appeared (! very deceiving!) to 
be high quality caps. They looked like large Silver Micas. Brown in color and 
had a nice hard shell. For grins, I started checking them. Every .047 was OK. 
Every .1 was bad! Some of these appeared to have the radio built around them! 
In any case, the bad ones were replaced.

Electrolytics. Please note that the above method will NOT work on 
electrolytics. By their design, they have a certain amount of leakage. To test them, you 
can use the hand held cap checkers to check their value. Or do what I do, 
chuck em! Electrolytic caps do not age well. The best thing you can do for them is 
to use them. I fire up all my Boatanchors periodically. Please DO NOT buy NOS 
electrolytic caps! Sitting on a shelf unused for 20+ years, is much harder on 
them, than being used occasionally. If a piece of equipment has sat unused 
for years, start powering up with a Variac. If the cap feels hot, it ain't gonna 
make it. Compare a modern electrolytic with an old timer. You will see the 
modern one is much smaller and better sealed. No comparison. I either refit new 
caps inside the original can, or leave the can in place (for appearances) and 
hang new ones underneath.
Some suppliers have recently started making NEW replacement, multi-section 
metal can caps. They are great, but a little too $ for me....

Replacement caps:
A word is in order here. One hears much about "Orange Drop" caps, (to be 
referred to as "OD"  from here on). They are excellent! One time, I was testing 
some caps in an old radio and it appeared that every one was leaking close to 
100%. As a sanity check, I pulled out a new OD and tested it. Zero leakage, as 
expected. Went back to checking the caps. A couple of DAYS later, I returned to 
the test bench and wanted to ensure my Leader LCR Bridge was working. 
Connected the OD. POW! Took out all the FETs in the Leader. It went back for repair 
and recallibration. A $175 lesson. Take a regular old cap (non OD), charge it 
to it's rated voltage. I'll bet in 15 seconds, it's leakage will have bled off 
the charge. I have ZERO affiliation with any mfg. Just a VERY, VERY happy OD 
user!
I have also had good luck with the WIMA style. Although their leads are 
radial and short for PCB mounting, it's easy to hang a tinned wore on each lead....

Where to buy:
I like to order from Antique Electronic Supply and RF Parts. People bitch 
about the $, but you will find that their service is the stuff legends are made 
of. LOTS of on-line places have great prices (Digi Key, etc), but I have no 
experience with them. I really, really stay away from ebay for most parts. Don't 
be fooled by ads like:
"100 - .01 caps for $4.99". You'll get 100 caps sure, good for 25 VDC. I have 
bought some ODs there, and cheaply. They were pretty old, although that makes 
little difference. Look at the AES on-line catalog and read about some of the 
properties of ODs. Self healing dieletrics, etc....You can see WIMAs and 
other types there......

By the way, I DO NOT try to unsolder the old cap's leads from a tube socket, 
etc. It's easy to cause lots of damage that way. I just cut the old cap out, 
leaving just enough of the original leads to make the new connection (1/4" or 
so). Just put a small 'loop' on the new cap and the old lead, crimp and solder. 
Try to keep the overall lead length the same.....After practice, it's 
difficult to tell it's a replacement.
My Hallicrafter SX-101A had 20+ bad caps in it. It too, is now incredibly 
sensitive.

One last thing. When you get an old VTVM, replace the caps in it too! Leaky 
caps in a VTVM is one reason people hated them, they 'drifted'........

I'll be glad to assist anyone needing more help!

Ron

N4UE