I am wandering if an insulation tester is a worthwhile investment. How do you actually use one? Can you use it to determine for example if there is a deadly leakage of B+ to the chassis? What is the value you should be reading on such a tester to determine there is not a leak? The videos on You tube are not that great.

I also note that it is useful for testing caps? I am new to all of this and just replace all paper caps anyway, but out of interest, what reading would you be looking for to determine if a cap is good/bad?

Basically it is an ohm meter. It measures high ohms. If you understand how your ohm meter works, it is the same except instead of using a 1.5V cell or 9V battery it generates 250vdc, 500vdc and 1000vdc to test the (non) circuit.
Most electronic (battery powered) units have a low ohms setting for measuring a fraction of an ohm.
It is called an insulation tester because it is for measuring two objects which are not supposed to be joined. Usually B+ is joined to the chassis through components, but I would use it to measure the 2 primary windings to the casing of a power transformer. The low ohms could measure the winding resistance.
Legally, all plug in devices have to be tested for earth leakage before they are returned to service. Clip the test lead to the earth pin and probe the metal case and measure low on ohms. this tests the earth wire. then turn the devices switches on, clip the test lead to the active and neutral pins, put the other lead on the earth pin and measure at double the working voltage. (500v) You should get a reading of many millions of ohms. Most plug in items are double insulated and are not earthed nowadays.
You could measure high voltage capacitors with it, but be aware that the voltage they put out can be 20% above what the switch says.
Some things measure a high resistance until a certain voltage is reached, then it 'breaks down'. This test should find these problems.

I hope this helps.

With capacitors note the polarity and do not exceed the voltage rating.
Filter capacitors will have leakage but coupling capacitors should be close to infinity.

What is it that you are trying to achieve?
Maybe you would be better off with a ring tester and esr tester?

Some things measure a high resistance until a certain voltage is reached, then it 'breaks down'. This test should find these problems.

I guess you're referring to "megger" testers, which test electrical insulation. Here are some examples:

http://www.megger.com/au/index.php

I have serviced a lot of radio's. The Valve tester I have has a "neon flash" test for non polarised caps.

That is basically a neon tube (with appropriate resistor) in series with the cap. Crude, but a very good pass fail indicator, as the neon extinguishes on a good cap.

It must be run on DC and is only useful above the striking voltage of the neon.

I do have a Tag & Test permit and I do have one of those Muti Voltage type Insulation testers (Megger is a name often used).

Distinguish absolutely ,the difference between Polarised & Non polarised capacitors. The Insulation tester is only good for Non Polarised caps & that is one thing I use it for, apart from Earth Leakage.

Even the VCT has a seperate setup for measuring Polarised caps. Polarised Electrolytics will always leak. If they don't they "depolarise" & fail.

The general formula for those I believe is "leakage in micro amps = 0.01CV" C is in mfd V is the Voltage rating.

If any one has built one of the "reformers" such as the one described by Rod Champness in "Silicon Chip"
Those will read leakage on a NP cap and you could easily adapt the series neon to these, within a lead avoiding the need for a meter. Do Note: HV DC bites.

Other than that, applying voltage and measuring current / resistance is really the only reliable easy way of leakage checking NP caps

Marc

I have an old chassis I purchased (a Fisk Radiola 52 - console) without the speaker. I had the complete table model and plugged the console version in to see if I could get any sound. I simply got hum (I hadn't yet swapped any caps - I know better now) and whilst trying to get a station, I got a zap from the station knob.

The chassis wasn't earthed which would have trigged the mains safety switch (I would hope).

Someone had obviously previously swapped some caps in it and the wiring just didn't look right. Eg. there seemed to be a wire from what appears to be from one of the secondary wires on the power transformer to a 0.05 paper cap which in turn appears to have been connected to part of the chassis.

Not sure if it was a series of really bad caps (a lot of them read 'open' on my fluke meter) or some other more serious problem.

Anyway, too scared to turn it back on and I wanted to know if there was a fail safe way (without me being electrocuted) of testing the chassis for live B+

A check with the ohmeter from the wire referred to above to the chassis gives me a reading of a few ohms - I would have thought that is not good but not proficient enough to know.

I may just scrap the chassis for parts.

That is why I thought maybe a megger would be the ideal way of testing a chassis in the future so I don't get zapped again.

Any tips appreciated.

Regards
Angelo

GTC, Megger is a brand name who used to make insulation testers.
Marcc, battery powered insulation testers put out dc. I am sure the the wind up megger testers (with the dynamometer) were dc also but the voltage varied with speed until the clutch slipped.
When measuring a polarized capacitor take note of the + & -. Make sure you know which lead is + & -. Also note the 500v test will be putting out more than 500v.
Wouldn't it be more economical to test the capacitor from the B+ with a sensitive meter or replace it with a new $2 capacitor?

GTC, Megger is a brand name who used to make insulation testers.

Yep, I had a (very) short term factory job in the early 70s which consisted of running a megger over pallet loads of relay sets for telephone exchanges all day. Most boring job imaginable!

Be carefull with $2 HV electrolytics, you often get what you payed for.. seen a few fail recently within 2 years.

Why I am pointing out the difference beween NP an Electrolytics is that two entirely different circumstances & constructions apply.

I will not Megger / Insulation test an Electrolytic.

Any result I consider misleading. I built a reformer, as for one; It is a good idea to reform caps that have been sitting for a while.

It is also usefull for (and has been used for recently) pinging electrolytics that get to a certain voltage then start leaking hyperbolicly, well before they get to their rated voltage.

The cheap & nasties in the last exercise were new 450's on a heater rectifier & 250V rail (table gram). No hope of exceeding voltage rating. You could not get them to 300V. They were tossed & Spragues put in (No more Hum) Fail time 18 months.

Extremely rarely, in a set I refurbish, do any old electrolytics anywhere in the set get left in circuit. They may get left topside, for aesthetics only.

Paper caps & "Micamold" anything are normally toss out on sight.

Mica's are rare fails, but do. The HV tester is handy for testing them and trimmers for shorts as you work through a radio. I make it a policy to HV test any mica that ends up with an end free, and I also check any resistor that can be, or has an end free (out of circuit) That actually saves re-work.

I tested the open type AWA Mica's, in a Bandmaster D184E recently. They're all still in there and it's working fine.

I replaced about nine resistors in the pass, on a Breville 730 recently. No mica's all Electrolytics & paper caps and the Pentagrid.

Mains & some other wire was replaced, (short on heater rail) No mains earth leakage.

No rework. First table set I have worked on with an 8" speaker albiet there will be a second Breville coming in soon.

Marc