1kW Dummy Load project

[BriansBrain]

My new 1kW Dummy Load project

8 x 100Ω - 250W RF resistors - Aliexpress.
188x120x37mm Aluminium Box - ebay.
Heatsink 300x190x50mm - DutchShop.
Copper Strip 120x75x3mm - DutchShop.
Total parts inc post = €155

1kW Dummy Project-1.JPG

1kW Dummy Project-2.JPG

1kW Dummy Project-3.JPG

Heatsink 300 x 190 x 50 mm.JPG

[BriansBrain]

This is the Plan of mounting the resistors on the Copper Plate

Dummy Plan.jpg

Dummy Circuit.JPG

[Marcel]

Dear Brian,

I don't want to discourage you, but this only works well at low frequencies, you will measure 50 ohms neatly, but you will see that the SWR is not good as the frequency increases.

Look inside the Alpha power HF dummy load to the internal construction and compensation.

[Marcel]

Dear Brian,

I don't want to discourage you, but this only works well at low frequencies, you will measure 50 ohms neatly, but you will see that the SWR is not good as the frequency increases.

Look inside the Alpha 2100 HF dummy load to the internal construction and compensation (Teflon coax C en Coils).

[Albert H]

We always used to mount the resistors in a radial pattern around the input socket. The SWR of those loads was never much worse than about 1.15 : 1

[BriansBrain]

I know that the SWR is not gonna be perfect

Finished putting it together today, gonna test it tomorrow

1kW Dummy Load Finished.JPG

[Mersey100]

Is the copper strip mounted directly to the heatsink or stood off it, cant really tell.

[rigmo]

I know that the SWR is not gonna be perfect

Finished putting it together today, gonna test it tomorrow

1kW Dummy Load Finished.JPG

nice work but i do not like it.. why end of coax is not grounded and closer to meddle point od loads.. . need to be wavelength coax velocity.
use panel mount soldering coax directly on connector ...

[BriansBrain]

Is the copper strip mounted directly to the heatsink or stood off it, cant really tell.

It's directly on the heatsink, if you look caefully you can see the cutout

[BriansBrain]

why end of coax is not grounded and closer to meddle point od loads.. .

Thanks
OK understand that one

need to be wavelength coax velocity.

Will that help ?
I am using RG402 ?

RG402.JPG

Capture.JPG

use panel mount soldering coax directly on connector ...

Thanks
I will have to get some of them

[BriansBrain]

OK..... I have just found the spec of RG402
Velocity of Propagation 69.5%
So... https://k7mem.com/Ant_Freq_Wavelength.html

Capture.JPG

[BriansBrain]

OK....... for starters I have now grounded the end of RG402

RG402 End now grounded.jpg

[BriansBrain]

OK.... you guys are correct...... as usual
I have learn't a lot on about RF this forum
I measured it on the spectrum analyzer and the load does change at higher frequencies
This is the proof on the BIRD

Forward = 100W

1K Dummy Load @ 100W Forward.jpg

Reflected = 3.5W

1K Dummy Load @ 100W Reflected.jpg

[jvok]

I bet if you shrunk everything down you could get it a lot better than that. Put the resistors right next to each other so you can solder straight between the pins and get rid of the wires in between. Also see if you can solder the pins at the end of each chain straight down to the copper instead of wiring to a solder tag. You might need a torch or at least a really big iron to get the copper hot enough.

Every piece of wire in there is an inductor which is where your problems coming from.

You might find it runs a bit hotter with the resistors close together but honestly doubt itll make much difference with that big copper block to spread the heat.

3.5w reflected is 1.5:1 SWR which isn't far off. I wouldn't expect to get much better than 1.1:1 maybe 1.2:1 for such a big load.

[BriansBrain]

I bet if you shrunk everything down you could get it a lot better than that. Put the resistors right next to each other so you can solder straight between the pins and get rid of the wires in between. Also see if you can solder the pins at the end of each chain straight down to the copper instead of wiring to a solder tag. You might need a torch or at least a really big iron to get the copper hot enough.

Thanks for the suggestions

Every piece of wire in there is an inductor which is where your problems coming from.

I understand that now

3.5w reflected is 1.5:1 SWR which isn't far off.

Well considering the cost.......... and that I made it........ it's not that bad then

[yellowbeard]

You should have told them you were building a George Foreman grill powered by Elenos...

[rigmo]

1kw dummy load

[BriansBrain]

You should have told them you were building a George Foreman grill powered by Elenos...

Good one

[Marcel]

If you have a (nano) Vector Network Analyzer, you can immediately see all changes, only measuring resistance with a universal meter or Bird Wattmeter (reflected power) will not help you any further.

Unfortunately, I can inform you in advance that this method of construction will never work properly, no matter how you are going to place the resistors / internal coaxial cable(ground)and wire.

The Alpha 2100 HF dummy load construction and compensation (Teflon coax C and Coils) goes up to about 60 MHz with these types of resistors and that is not for nothing.

Take a look at the inside of a Bird dummyload on the internet.

They use a funnel there to get things straight down to a few GHZ. (good VSWR over a wide range).

Marcel

[BriansBrain]

Unfortunately, I can inform you in advance that this method of construction will never work properly, no matter how you are going to place the resistors / internal coaxial cable(ground)and wire.

I knew I was not gonna have a perfect dummy with this design from the start

All I needed was something I could test repaired the ETG1000 at 500W without melting or blowing up
And I succeeded for a very low price