Mylar-Aluminum Grav-Cap Experiments and Observations by John Pietrasik

The following was an attempt to make a working grav-cap, also known as an electric rocket, by John Pietrasik. Thank you to John for sharing this work. All of the following text and photos are from him.

Report dated Feb. 5, 2001


A multi-layer Mylar-aluminum foil "grav-cap" is tested. Results and conclusions are noted.


The original documents describing the construction of the "gravity-warp capacitor" (or "grav- cap") state that a functional device could be constructed using tin for the plate material and wax paper for the dielectric, although "better" (i.e., more modern dielectrics) could possibly improve performance.

The report of success from "Cliff L." (from Bill Beaty's website) seemed to verify this statement. Cliff L.'s device was constructed using Mylar layers for the dielectric and aluminum foil for the plates. The device described here is a variation of that design.

Device construction

Acrylic endcaps (Figure 1) were used in place of wood, and Delrin threaded rods (Figure 2) were used in place of steel. This "all plastic" design yielded a completed unit that was both small in size (6" O.D.) and arc-resistant (i.e., no arcs would accidentally jump to the threaded rods).

Figure 1. Acrylic base (endcap)
FIGURE 2. Threaded delrin rods

Each "capacitor section" was composed of professionally die-stamped pieces in the following configuration (see Figure 3 and Figure 4):

Figure 3. Stamped aluminum plate
Figure 4. Stamped mylar layer

Layer 1: 0.0015" Aluminum foil (+ Polarity)

Layer 2: 0.002" Mylar

Layer 3: 0.0015" Aluminum foil (- Polarity)

Layer 4: 0.002" Mylar

Layer 5: 0.002" Mylar

Originally, the device was to have 0.001" Mylar between the plates and one layer of 0.002" Mylar separating the sections. However, the 0.002" film was found to be mechanically easier to work with and provided a higher breakdown voltage. Therefore, the grav-cap was constructed using only this film. Figure 5 shows the device with the first "capacitor section" installed.

Figure 5. Completed first section

Power to the device was provided by an EMCO 4100 series HV module (10kVDC, 10W max output). In order to adjust the high voltage and monitor the actual value, a small interface box was constructed. This device connects directly to the EMCO 4100's 7-pin Molex header and contains a power switch, a 10-turn adjustment potentiometer, a pair of banana jacks for the +24V and GND inputs, and a pair of banana jacks for the voltage monitor output (1VDC per 1kVDC actual output). A digital multimeter was plugged into these jacks (see Figure 6).

Figure 6. Power supply module and interface box

Construction of the device was completed on 01/23/2001. It contained 120 isolated sections (i.e., 240 aluminum foil layers). The device (without power leads attached) is shown in Figure 7. Using a FLUKE 87 DMM the total device capacitance was found to be 0.212 uF as shown in Figures 8-10.

Figure 7. Completed device (no connecting leads)
Figure 8. Capacitance-measurement connections
Figure 9. DMM showing capacitance
Figure 10. Close-up of DMM capacitance measurement

The "beam-balance" test setup is shown in Figures 11-14. The device was attached to one end of a 3/8" diameter dowel rod while a plastic vitamin bottle was attached to the other. The bottle was filled with coins until the weight balanced. I did not have a scale for weight measurements but estimate the device to weigh approximately 3 lb (1.36 kg).

Figure 11. Beam-balance setup
Figure 12. Close-up of grav-cap on beam balance
Figure 13. Close-up of +24VDC supply and interface box
Figure 14. Close-up of vitamin bottle counter-balance


The first tests were performed the same evening that the device was completed (01/23/2001). Figures 15-20 show the device unpowered then powered at 1kV, 2kV, 3kV, 4kV, and 5kV. Adjustments were made slowly. In all cases, NO beam-balance deflection occurred (i.e., no force was observed).

Figure 15. Position of device when unpowered
Figure 16. Position of device at 1kV
Figure 17. Position of device at 2kV
Figure 18. Position of device at 3kV
Figure 19. Position of device at 4kV
Figure 20. Position of device at 5kV

On 01/25/2001 a "pulse" test was conducted. While in the OFF state, the power supply module was adjusted to 4kV. Power was then turned on, forcing the grav-cap to charge quickly. The 4kV setting was reached in approximately 10 seconds. Again, NO force was observed.


Though similar in construction to the "grav-cap" of Cliff L., this device appears to have differences significant enough to prevent it from functioning.

Additional experiments will be conducted in the near future to determine these differences and show what is necessary to construct a working device.


I would like to thank the following individuals for their contributions and insights on this project (whether directly or indirectly): Bill Beaty, Tony Dekelaita, Steven Dufresne, and Cliff L.

Do you have a project you'd like to share on too? You're more than welcome to. Click here for details.
Liked this? Share it with: