P6015/Repairs

the original filling

Freon-114 aka Fluorocarbon 114 (R-114, CFC-114) has an absolute permittivity/dielectric constant of ≈2 as a liquid and near 1 as a gas.

Possible replacement dielectric filling

For the P6015 non-A version, a replacement for the CFC-114 filling gas that is no longer available is needed.

Several options have been suggested, see below.

Links

https://www.eevblog.com/forum/testgear/high-voltage-oscilloscope-probes-make-or-buy/

https://www.eevblog.com/forum/testgear/high-voltage-oscilloscope-probe/

Mineral oil

Compensation range may be insufficient - see silicone solution below; tbd

Ed Breya says:

I happened to notice I have an old bottle of white mineral oil, and decided to see just how much effect it has. I still have the probe #1 with butane fill for testing, so I washed out probe #2 with IPA, then did some before and after capacitance measurements for air vs oil fill. These aren't highly precise considering the variables involved, but it gets in the ballpark at least.
The C from the tip to the corona ring (and hence the whole capacitor structure) measured about 1.83 pF, without the outer shield, in air. The C from tip to everything is the total shunt capacitance, has measured a bit less than 3 pF in air, for all the probes, which is the spec.
From the tip to the ring only, with no shield, with oil fill, is about 2.43 pF, while total shunt C with it fully assembled is about 4.2 pF. So, going to oil fill increased all the Cs by about 30-50 percent, which isn't all that bad, considering.

Low-viscosity silicone

Dow Corning Xiameter PMX-561 has been suggested (R.E.Beverly III and associates AN-108, Dielectric Replacement in Tektronix P6015 HV Probes). The compensation circuit needs to be adapted to allow for the higher capacitance.

Butane (lighter gas)

Dielectric strength may not be sufficient, and obviously the gas is flammable.

HFC236-FA

https://groups.io/g/TekScopes/topic/30956868:

HFC236-FA has almost precisely the same characteristics as R114 (boiling point and vapour pressure), is environmentally friendly, and is used both as a refrigerant (a R114 replacement) and in fire extinguishers. So it should be a perfect solution to keeping the P6015 up and running.

P6015 refill with Diala oil

R114 has an absolute permittivity/dielectric constant of ≈2 as a liquid and near 1 as a gas.

As a substitude for the R114 I used Shell-diala-s4-zx-i with a dielectric constant of 2.

The diala oil has a high break down voltage of 70 kV, is enviroment friendly, not corrosive and cheap.
Drinking the oil is deadly, because it will creap into the lung!

Disasembling the probe body:

The resistor should never be touched
the resistor is plugged in and can be pulled out
The "golden cap" (resistor end cap) is screwed in. Use a big phillips screwdriver, put some cloth around it so no edges poke out, press it gently into the cap and turn. Be carefull!

The Probe body is modified as follow:

cleaned in ultrasonic bath (every printing on the device will disapear!)
cleaned with isopropanol
airfree filled with oil

Asembling the probe body:

put everything together (gold cap, resistor)
hold the probe body upright
put the sealing ring on top of the probe body
fill the probe body until the oil spills over the sealing ring
put the probe togehter
clean up the mess

The compensation box is modified as follow:

C1: 47 pF ceramic/polystyrene capacitor parallel
C2: 220 pF ceramic/polystyrene capacitor parallel
C3: 100 pF polystyrene capacitor parallel

Compensation:

follow the rules in the manual
the probe cable should be stretched out, because it has an importand influence on the first 100 ns

As a rule:

if you can't tight C1/C2/C3 not as hard as you need for compensation: you need more capacity.
if you can't loose C1/C2/C3 not as much as you need for compensation: you have too much capacity.

Using only ceramic capacitor seems to increase the ringing.

For soldering the capacitors into the compensation box, a hi-power soldering iron is needed as some pins have to be soldered directly to the chassis.

The modified P6015 is tested at 20 kV DC for 5 min.

oszillograms: