DC to 500 MHz RF step attenuator: 0 to 79 dB in 1 dB steps

[PeterMcMonty]

Hello,

I would like to contribute to this forum with this project I built long ago (1995).

It is a RF step attenuator for lab use, main features are:
- 50 ohm input/output impedance,
- very good performance from DC to 500 MHz,
- 8 attenuation cells: 1, 2, 4, 8, 16, 16, 16, 16 dB allows you to select attenuation from 0 to 79 dB in 1 dB steps,
- BNC input/output connectors,
- each cell is operated with a DPDT miniature slide switch.

I made it in three versions, two of them with PCB stripline technology in order to make it usable at high frequencies. I've tested and characterized them with:
- Anritsu mod.MS610B (10 KHz ÷ 2 GHz ; Zi = 50 ohm) spectrum analizer with built-in 50 ohm dummy load,
- Racal-Dana mod.9081 (5 MHz - 540 MHz ; -120 dBm ÷ +10 dBm; Zo = 50 ohm) syntesized signal generator.

You can see a photo gallery of this project on my facebook page:
http://www.facebook.com/media/set/?set=a.1934259518611.2115379.1305998984&type=3

A project description (unfortunately in italian, but if anybody is interested to I can traslate in english) in pdf is downloadable from: http://www.box.com/shared/yq460xfvnd

Another pdf file with some measures performed on prototypes (also pdf) is downloadable from: http://www.box.com/shared/3rr2ant105

Here you can download an excel spreadsheet with calculations for pi and T attenuation cells from 1 to 20 dB and their approximations using standard E24 resistors: https://www.box.com/s/txavd5irqg0t50bv3ze1

Finally, I have recently made a pcb with Proteus 7.10 (thanks to Sonsivri forum for this firmware !!! ): ISIS schematics, ARES pcb layout and gerber files are downloadable here: https://www.box.com/s/wdqeqlaqe8dv5qbvsnpv.

I hope this little contribution to the forum may be helpful to somebody. I'm glad to answer to any question about this project.

Greetings to everybody,
PeterMcMonty

[bobcat1]

Hi

Nice project but no schematic found - can you please add schematic in PDF

Thanks

Bobi

[PeterMcMonty]

sorry to read you so late, Bobi,

you can find schematics in my photo gallery, as I stated in my post:

You can see a photo gallery of this project on my facebook page:
http://www.facebook.com/media/set/?set=a.1934259518611.2115379.1305998984&type=3

If you don't have a Facebook account (my photos are public, but you need to access facebook to see them), I include here schematics in PDF.

Thank you for your interest and again sorry for the inconvenience,

PeterMcMonty

Posted on: November 17, 2012, 09:35:23 21:35 - Automerged
btw, I forgot... I guess you don't have Proteus 7.10 installed, otherwise you can download the entire project (schematics and pcb) from the link I gave in my post:

Finally, I have recently made a pcb with Proteus 7.10 (thanks to Sonsivri forum for this firmware !!! ): ISIS schematics, ARES pcb layout and gerber files are downloadable here: https://www.box.com/s/wdqeqlaqe8dv5qbvsnpv.

[ehnonymouse]

I've used the Peregrine Semiconductor chips to do something similar up to 5GHz. Two cascaded to get 60dB.

[PeterMcMonty]

Very interesting device indeed! I've checked that in Italy you can get from DigiKey at 3.7€ each: not a bad price!
( http://www.digikey.it/product-detail/it/4302-52/1046-1012-1-ND/2614516 )

What I can say is that in 1996 was hard to find such a device. Moreover, passive attenuators don't need power supply and works truly from DC.

By the way, I made some more measurements and I found that my prototypes on stripline are quite useable up to 1 GHz

Posted on: February 06, 2013, 02:06:20 14:06 - Automerged
here is the return loss of attenuator #2 from DC to 1GHz:
see img. DSCN9382_1 attached

and here the insertion loss of the same unit:
(see img. DSCN9393_1 attached)

[ehnonymouse]

Yes, time marches on and things seem to get easier. I had to have a computer controlled attenuator, so the chips were ideal. Looks like a great project you did though.

[PeterMcMonty]

thank you very much for your appreciation. The chips you used seems great for lots of applications.
Now I've got two samples of AD8307 Low Cost, DC to 500 MHz, 92 dB Logarithmic Amplifier from Analog Devices (you can get them for free, as samples: http://www.analog.com/en/rfif-components/detectors/ad8307/products/product.html ) and I'm thinking on how to use them. First idea is a VSWR bridge with input port and output port power measurement, performed by a PIC microcontroller, to get a very simplified network analyzer. Unfortunately I get no information about phase difference (the right device should be AD8302: http://www.analog.com/en/rfif-components/detectors/ad8302/products/product.html , but no samples are available for free), but I think it could be an useful tool yet and a good starting point for some more sophisticated.