Pages: [1] 2   Go Down

Author Topic: DIY stepped attenuators  (Read 1941 times)

2E0DED

  • Member
  • Posts: 14
DIY stepped attenuators
« on: November 11, 2018, 10:53:20 PM »

looking at these, whilst cheap enough to buy i did want to ask are they an easy diy project?

the tuning and calibration is the issue for me?
Logged

N1CX

  • Member
  • Posts: 201
RE: DIY stepped attenuators
« Reply #1 on: November 12, 2018, 02:39:06 AM »

I've built a bunch. Theres alot of things that can go wrong. I've gotten 31db return loss out of several I've done, some ~24-26 when I didn't pay extra attention to shielding sections. Calibration has been spot on or very very close for values used when measured on a network analyzer.

Leads need to be kept short, good quality switches, special attention to shielding is most important. Above all biggest thing is to plan plan plan. To make a small one with say 6 switches requires ALOT of small dexterity work and pre planning is critical otherwise you may not end up with a nice job and get frustrated. There are things like preassembling resistors to switches that are critical to doing it right.

Im sure others can give other tips here too.
Logged

G3RZP

  • Member
  • Posts: 2254
RE: DIY stepped attenuators
« Reply #2 on: November 12, 2018, 05:42:26 AM »

Another difficulty can be getting the right value of resistors if you want the steps to be whole numbers of dB and the return loss to be reasonably high. I obtained for about £5-00 the two attenuators - 10dB  and 1 dB steps - from a scrapped Telonic 2GHz sweeper. They needed putting in a die cast box with semi rigid 50 cable to the connectors and that produced a far better job than I could have made
Logged

LA9XNA

  • Member
  • Posts: 428
RE: DIY stepped attenuators
« Reply #3 on: November 12, 2018, 07:53:40 AM »

There used to be a build description in the ARRL handbook.
Logged

VK2TIL

  • Member
  • Posts: 695
RE: DIY stepped attenuators
« Reply #4 on: November 12, 2018, 12:35:05 PM »

They do require quite a bit of mechanical work.

This is the classic ARRL design;

https://www.arrl.org/files/file/Technology/tis/info/pdf/9506033.pdf

(the schematics do not appear well in the above; Google may find a better copy).

and a version of it;

www.orarc.net/SteppedAttenuator.pdf

This basic model should be OK over HF; getting higher up the frequency scale requires a lot of work and some test equipment.

I built one;



There was a good deal of mechanical work;



The layout used SMD resistors and the switches were soldered "SM fashion";





I don't have a photo of the shields between and around each section but here is the final shielding made from copper foil;



The final version works well to over 500 MHz.

I enjoy mechanical work but I don't think I'd make another one; if I did, I would put about 3 or 4 sections in diecast boxes and interconnect them as required in use.

A collection of attenuators is very useful;



The rotary units were, as G3RZP described, bought separately and housed in aluminium cases.

Fixed units are also useful; the cheaper ones from ebay and elsewhere work fine for HF and VHF.






Logged

G3RZP

  • Member
  • Posts: 2254
RE: DIY stepped attenuators
« Reply #5 on: November 12, 2018, 01:17:49 PM »

When you look at the effort that people like HP and Telonic and Marconi Instruments went to in order that they could make precision attenuators, surplus is really attractive. Although I did make a 'special to type' test fixture in my home workshop for work, where the attenuator was a 'waveguide beyond cut off' type to test the effectiveness of a shield to go over a radio chip that was to go into an implantable medical device - pacemakers and the like.

Waveguide beyond cut off attenuators can be quite complex mechanically....
Logged

KM1H

  • Member
  • Posts: 11155
RE: DIY stepped attenuators
« Reply #6 on: November 12, 2018, 02:44:35 PM »

I made a pair in the 80's, for 50 and 75 ohms using slide switches which tested better on the VNA at work and were readily available....cheap. Material was double sided PC board left over from work related projects.

The steps duplicated the Narda, KAY, ?? ones on the bench at 20, 10, 5, 3, 2, 1 dB

The resistors were 5% 1/2W carbon film selected to <1% tolerance, another benefit of working in a large RF lab. I forget how high I tested them but they are still used to around 150 MHz but were used much higher.

Carl
Logged

N3DT

  • Posts: 2465
    • HomeURL
RE: DIY stepped attenuators
« Reply #7 on: November 12, 2018, 03:17:53 PM »

I built one with slide switches and common resistor values dead bug with a double sided PC board case a long time ago. It didn't work too bad through maybe 15M. I found a couple RLC Electronics (?) at a hamfest that looked a lot like the rotary HP models so I got them, a 1dB and a 10 dB step. The 10 worked but the 1 did not, it had some burned resistors in it plus the switches (micro) were dirty and hard to clean. I never did get it working very good even though I bought 1% resistors from Mouser. Be a lot easier if you can find a working one and pay a few bucks. I've seen them on ebay not expensive, but better get a guarantee it works. I've got a bunch of in-line pads that work about as good if not better.
Logged

KE0ZU

  • Member
  • Posts: 581
RE: DIY stepped attenuators
« Reply #8 on: November 14, 2018, 06:36:19 PM »

Older ARRL handbooks from the 50s thru 70s had step attenuator projects in them.    The ones I remember used slide switches.   

If you decide not to make them, look for an HP-355C ("0" to 12 db), and an HP-355D ("0" to 120 db) unit.   These are 1GHz rated units and can handle 1/2 Watt power levels.  These are 50 Ohm units



For audio the 600 Ohm HP-350 attenuators, shown flanking the VTVM below, are a good choice.

Logged
Regards, Mike
https://mikeharrison.smugmug.com/
Pics and bold print are usually links.

W1VT

  • Member
  • Posts: 6071
RE: DIY stepped attenuators
« Reply #9 on: November 15, 2018, 08:22:20 AM »

Slide switches typically work better than toggle switches but usually require more work (think square holes) in construction.
Logged

DL8OV

  • Member
  • Posts: 1061
RE: DIY stepped attenuators
« Reply #10 on: November 16, 2018, 12:31:05 PM »

One alternative that has been missed. There are some very nice Teledyne relays around that work well as far as 1 GHz. If you make an attenuator using these you can use whatever switch you want to just drive the coils.

Peter DL8OV
Logged

KM1H

  • Member
  • Posts: 11155
RE: DIY stepped attenuators
« Reply #11 on: November 16, 2018, 12:35:57 PM »

BUT now youre into using a VNA to design the most efficient PCB and overall layout to get to 1 GHz.  Certainly not for the casual homebrewer.

Carl
Logged

G0HZU

  • Member
  • Posts: 345
RE: DIY stepped attenuators
« Reply #12 on: November 16, 2018, 02:56:00 PM »

Quote
When you look at the effort that people like HP and Telonic and Marconi Instruments went to in order that they could make precision attenuators, surplus is really attractive.
Agreed.

However, I've got various step attenuators here and speaking from experience I would not recommend DIY or even commercial attenuators that have the classic series of toggle switches that set the overall attenuation. Usually this means a line of switches arranged as 1dB, 2dB, 2dB, 5dB, 10dB, 20dB etc. This type of attenuator is OK for very occasional use but if you want to do anything serious that involves changing a level continuously in 1dB steps then this type of attenuator quickly becomes a pain to use. The rotary ones are best for this type of use, especially if you can get one that has dual controls. i.e. 1dB steps and 10dB steps on the same attenuator.
JFW (eg 1GHz 50dB in 1dB steps) rotary attenuators like this do appear on ebay quite often but prices vary a lot. But bargains do appear now and again.

The other thing to be wary of is how good the isolation is at high attenuation settings at higher frequencies. Even some commercial attenuators perform poorly here at frequencies as low as 30MHz.  
Logged
Regards
Jeremy

WD4HXG

  • Member
  • Posts: 396
RE: DIY stepped attenuators
« Reply #13 on: November 18, 2018, 01:32:21 PM »

looking at these, whilst cheap enough to buy i did want to ask are they an easy diy project?

the tuning and calibration is the issue for me?

Check out www.chemandy.com particularly their page for calculators.
My favorite is the PI configuration attenuator at:   https://chemandy.com/calculators/matching-pi-attenuator-calculator.htm
but the tee configuration works as well. Just my particular fetish. If
the PI attenuator configuration yields obtuse values then I try the
tee configuration calculator.

You might also want to check Mini-Circuits website:            https://www.minicircuits.com/WebStore/Attenuators.html
for their surface mount and coaxial attenuators. For most works all
you will need is 1 dB, 2 dB, 3dB, 6 dB, 10 dB, 20, db attenuators to give
you a range of 1 dB to 39 dB. You simply cascade the attenuators to
the value you need. So if I wanted 25 dB attenuation then I would
cascade the 2 dB, 3 dB and 20 dB attenuators. I use the lowest
attenuation device first as if I apply to much power then there is a
better chance the lowest attention device will survive due to it dissi-
pating the least amount of power in dB in the cascade.

Surface mount resistors are sold in 10 piece quantities on this side of
the pond. I suspect you have vendors over on your side of the water
which provide similar pricing and quantities.

Please do not discount a used step attenuator but if you cannot test
it then you are at the mercy of the seller. Not a good place to be in
my opinion.

Regards

Chuck
Logged

G3RZP

  • Member
  • Posts: 2254
RE: DIY stepped attenuators
« Reply #14 on: November 20, 2018, 06:11:07 AM »

For an application such as a signal generator output, a 'waveguide beyond cut off' attenuator is fairly straightforward to make, and only needs calibrating at a couple of points if the mechanics are reasonably well done.
Logged
Pages: [1] 2   Go Up