don't try this at home....

I need to test for 100M of insulation resistance at 500V.

Rigged up 4 transformers to make 4x120V

I used a 10M build out resistance, so 100M load should drop voltage roughly 10%....... BUT  :'( my cheap rat shack VOM has a roughly 10M input impedance so my rig drops to around 220V before I expose the device I want to test...

Now my source impedance is 5M (10M in parallel with meter 10M), so drop from 100M load is even smaller.

I can measure a difference but not very accurate, and driving voltage should be 500V.

If I get brave I guess is could use my VOM in series and measure current.  The VOM has a uA scale.

JR

I think I'd be brave and check for current.

I probably will try current before I tear this down...It should be in the range of single digit uA.  I'm not going to leave it laying around with a plug still attached.

A friend on another forum that has been following my progress developing the outlet tester  has offered to lend me his Fluke insulation tester so I will then be able to do a proper (accurate) measurement (up to 1000V but I only need to pass 500V).

I just want to make sure this is right before I submit it to UL where failing any test gets expensive to repeat the whole submission process..

JR 

[edit- using current range I measure between 7 and 11 uA so calculates to well less than 100M... I think I need a real tester [/edit]

JohnRoberts said:

I need to test for 100M of insulation resistance at 500V.

Click to expand...

The last thing I should be doing is giving you advice, but have you considered a bridge?

A resistor in series with the DUT, another resistor(for better feel)  in series with a pot (preferably with a really long nylon shaft at 500V), both strings across the 500V. Measure voltage between the wiper and the node between resistor and DUT,  balance the pot for 0V,. Calculate ratios and current. Meter Z no longer matters once nulled.

Or you can go old school, but this one is probably out of cert:

http://www.ebay.com/itm/Vintage-Megger-Biddle-Instruments-Insulation-Major-Tester-/272123668560

Gene

Gene Pink said:

JohnRoberts said:

I need to test for 100M of insulation resistance at 500V.

Click to expand...

The last thing I should be doing is giving you advice, but have you considered a bridge?

A resistor in series with the DUT, another resistor(for better feel)  in series with a pot (preferably with a really long nylon shaft at 500V), both strings across the 500V. Measure voltage between the wiper and the node between resistor and DUT,  balance the pot for 0V,. Calculate ratios and current. Meter Z no longer matters once nulled.

Or you can go old school, but this one is probably out of cert:

http://www.ebay.com/itm/Vintage-Megger-Biddle-Instruments-Insulation-Major-Tester-/272123668560

Gene

Click to expand...

Thanx that's more like my price range.. ;D

I have a modern fluke insulation tester in the mail (on loan) coming to me, so I'll be able to confirm the design. 

With 100M input resistor not sure I believe the measured 40M  I got with my rube rig... :'(

JR

I received my loaner "real" insulation tester and it confirmed my < 100M measurements. In fact measuring a  100M resistor by itself resulted in 60M.

Looking at the resistor data sheet suggested a normal operating voltage of <100V  :'( :'(

Back to the design bench...  The good news is that the proper insulation tester confirmed my DIY results.  I was aware of noise issues with small SMD resistors, apparently they have voltage issues too...

JR

[edit- using two 100M resistors in series still works and deliver over 150M @ 500V so confirm that my design approach is valid.. I just need to revisit my resistor choices.  [/edit]