Passive Attenuator: builder renditions

Simple passive preamp circuits, or more accurately passive attenuators have been the subject of a few threads in the last few months. It started when LDP got us interested in a 2002 thread over at DIY Audio. There, a Mr Harry Haller (a pseudonym) proposed an extremely inexpensive, simple way to produce a passive attenuator with minimal parts that were virtually the cheapest available. In essence, the circuit is a variable shunt to ground.

Our AKer ashok in at least two threads produced a cogent analysis and explanation of the circuit, as well as detailed instructions for wiring it up. Other knowledgeable contributions followed. If you can fry an egg and burn toast, you can make this circuit.

As for getting your head around how it works ... well it tests your basic knowledge of Ohm's Law and the very basics of electrical theory. I freely admit to being challenged; such kind of abstract concepts have never been my forte, despite my passionate attempts over many years to master them. Nevertheless, according to my predilection that simple can be elegant, I've become an enthusiast of this device because it has put so much more music into my speakers!

And so, I thought that it was time we stepped away from the introductions and did a new thread where we showed each other our own renditions and variations of the passive attenuator. Of course, we can all go further; the challenges are nearly unlimited. And so a thread like this can quickly get into biased circuits, active transformers blah blah blah. Nothing wrong with that, and some of us may quickly move there. But can I ask that this thread stay on track? — passive circuits. If some of us move on, perhaps we can refer to sister threads and keep us all in touch.

A passive attenuator may be no more than a pot — a resistive device. After that, the variations seem to grow. Values for the parts can vary hugely. Our discussions so far have been around the Haller shunt design. But Rod Elliott's ESP site has shown us how we can fake the law of the pot using a linear or 'B' designated pot and some resistors. It is a different approach than the Haller design (and one I intend to try).

Apparently, there is no slam dunk about some of this. Many of these passive circuits will have to be tweaked and tuned to suit and marry what is both upstream and downstream in the chain. IMHO, all sorts of complaints and criticisms of the passive attenuator stem from a failure to adjust the values and implementation. This may not be everyone's cup of tea, nor is it a solution to many applications, but the passive approach to attenuation should not be disqualified from the running on that account. It is possible that with enough understanding and footling around, this humble device can be made to serve in all sorts of systems. It sure is attractive to those of us who can't afford $X-K or $XX-K preamps. Moreover, the savings may leave something in the kitty for an EQ device to play vinyl, or a DAC to squeeze more out of our digi source.

I'm now into my third rendition. If the first had been a bummer, I might have stopped there. But it was so successful that I fiddled and footled with the basic Haller shunt, all the while using cheap parts. Each step so far has been rewarding. And ... it's cheap enough!

See below for my Passive Attenuator V-2.

Lorne PA-V2

Passive Attenuator V.2 ... to give it a reference. This was my second version of the Haller circuit. I used two Cosmos 100K 'B' (linear) pots so I can balance the volume. More: 1% generic metal oxide R's, Teflon coated 22 and 24 gauge multi-strand wire and standard 60/40 solder. The unit has its own lifted ground plane. The shunt resistors slot into cinch pins that are mounted on, but insulated from the ground plane. This allows trial and error change-outs for resistor value without having to desolder. Yeah — I know — don't flame me over that issue. I got tired of changing out values using the iron on V.1. I look upon these these versions as test beds. The current value is 10K Ohms — my initial, and as yet, only tried-out choice in V.2.

V.1 used alternately 10K and 20K Ohm pots as well as a number of different R values under 10K Ohms. I don't know where the difference lies, but V.2's 10K R and the 100K Ohm (still cheap Cosmos) pots have hit the magic spot. Any version of V.1 was still an improvement over any active preamp I'd used to date, but the V.2 is a transformation. Every recording I play convinces me that I am hearing more music from my Kenwood/Trio L-05M mono amps and much modded Sansui SP-G88's than ever before. This is simply some of the best sound I've got stuck into during decades of this, that and the other — the reason that inspired me to start this thread.

The only draw-back is in very low level listening. As I turn towards the full shunt, the knob movement becomes hypercritical and balance matching is very difficult. An audio taper pot may be a solution, but it may present compromises too. It strikes me that for low level listening, the ESP faked law pot may be the ticket. It continues to use a linear pot. The Haller design may be better off with a high quality taper (logarithmic) pot. But I am not finished with the shunt device yet ... as you may see in my planned $10 V.3 version. You see, I don't mind replugging four RCA connectors to switch from a low-level-listening version to a boogey version, and vice versa. Not everyone would wanna do this, especially the people who want remote volume control! If you are physically challenged the choice is a no-brainer. Now I am not saying that V.2 is a 'boom-or-off' device, but very low level listening is what I often do in the dead of night, and it is at the limit of what I can do with V.2.

In the pic, a U.S. 1922 Columbia silver dollar is used to show scale.

Passive Attenuator.

Hi.. Where do you hookup this device, between cd player and preamp, or, preamp and power amp?

Thanks.

Hi Zekk,

Let us agree to call this device — and similar devices — a passive attenuator. That is, it attenuates volume in the system. It is passive because no part of the circuit is biased, and it relies completely on the current that arrives at its input. Such devises are often misnomered as passive preamps. There are passive preamps, but this device is not one of them.

The object of using this devise, as well as others like it, is to make the preamp redundant. Most line sources such as Redbook CD, DVD, cassette tape, and tuner have sufficient output to drive a power amplifier. The exception is the TT — or vinyl turntable. There, the output is insufficient and has to be amplified. Moreover, it must go through a special EQ or equalizing circuit in order to reclaim specific frequencies that have been suppressed in the recording process. There may be case-by-case exceptions, but generally speaking, we can eliminate the preamp. So the chain goes like this: source > attenuator > power amp > speakers. You may choose as I did to have: source > selector > attenuator > power amp > speakers. The source may be a TT and phono amp combo: TT > phono amp > selector > attenuator > power amp > speakers. Or if you use a DAC: CD> DAC > selector > attenuator > power amp > speakers. Another option is to combine the selector circuit and the attenuator so that they are in the same box. (Duh) One enthusiast has put up a page that is entirely devoted to such a project.

There are advantages and disadvantages to this approach. The advantage is that by deleting the preamp, you skip many, many electronic components, plugs and switches. All these can be thought of as incrementally yet cumulatively detractive — or at best a buffer (unwanted in most cases). The disadvantages of such a device are: matching the impedance of the source/attenuator/amp. This may become become more of an issue than with a preamp. In many cases, your attenuator may be well-matched to some particular suite, but not to another. Versatility may not be its virtue unless you build in a facility for switching the shunt resistance. But if you are going down that path, there may be other, more attractive designs. The beauty of this one is its simplicity. If you really get into this, the issue of impedance matching will become more of a consideration for you in regards to audio generally. Prior to my building these simple devices, I just sort of winged it. Now I am really interested.

Also, the user has to work out how to select sources. If you only use CD it's a no-brainer. I built a switch box that includes a recording loop (pics and so on in another thread). I know one guy who simply unplugs and plugs what he wants in a rat's nest of cable — inelegant but it works.

You may want to know if the idea works to an advantage. My answer is an unqualified yes. I have had trouble with very low-level listening as I said above. But I have taken care of this with an experimental device I built for $8. Instead of using the regular, linear, 100K Ohm volume pots, I used an 18 turn Bourne's type pot as used on circuit boards. This experiment has provided great improvements, but at the expense of some inconvenience. I will explain that in another post to follow.

By the way, you don't have to use a shunt type attenuator. There are other designs such as Project 01 over at Rod Elliott's ESP site. And some people just box up a ganged pot or dual pots of the tapered or A curve logrithmic type — usually referred to as an audio pot. Some people think that B-non-curve linear pots track better, but you either have to compensate the linear characteristic to suite the way we hear sound, or simply live with the effects. The upcoming version is a compromise that retains the shunt design and the linear pot.

The bookend attenuator ...

I think that this experiment shows that the volume pot is just as crucial and critical in this design as it is in any other. Some of us may have been thinking that the shunt could cheat the quality of the pot to at least some degree.

The pics show a bookend I bought for 100 Yen, some inexpensive RCA plugs and two 18-turn Bourne's type pots that normally would be on a circuit board. Not very convenient. Any listening session means holding onto a jeweler's screw driver. But, you know what? ... the sound is so good that I don't even mind. It has to be the nearest thing there is to a straight wire from source to power — and yet I can control both the balance and volume to within a scintilla of what I want.

Try it yourself. The whole kit will cost you $8.00 or less. It taught me that precise balance is more elusive than I thought, and yet it means a great deal in the production of the stereo image. I have had some decent preamps, but this ludicrous device tops them all.

What next? Well, I could build the next one with some really good — $$$ — linear pots in good faith that it will be an excellent attenuator. And/or I can move on and build the Project 01 over at Rod Elliott's ESP site. From there ... well there is another ESP project using op-amps, but that is a different country.

BTW, the interconnect is DIY as well — in two different flavors.

Lorne,

You have done quite a bit of experimentation with the passive attenuator. Good work.

The bookend attenuator with the Bourn's pots - is that a law faking design? I imagine the pots are linear, and you have a shunt resistor at the output jack.

I have hesitated to post details and pictures about my passive device because it no longer is passive. I have an opamp buffer right after the source selector that drops the source output impedance to around 40 ohms. Makes it very easy to drive any type of load.

On the output of the buffer is a 1:1 transformer (one each for Left and Right). On the secondary of the transformer is a 10K PEC stereo pot, with audio taper. The attenuated output of the pot drives the power amplifier.

Your experiments have given me the impetus to try linear pots with law-faking resistors. I will probably get mono pots to allow for proper balance.

Only issue to watch out for is that the input impedance of the pot will now change with the volume setting. Using a 100K linear pot with a shunt of 15K means that the input impedance is 100K with the pot turned all the way down (total attenuation). Turned all the way up, the input impedance is 100K in parallel with 15K - approximately 13 KOhms.

ashok and lorne,

You will find, I think, that using a stereo, or 2 pole ganged pot and varying the shunt resister to match (or attempt to match) variations in the two 'sides' will make a more convenient attenuator.

If it becomes that big an issue, what I've taken to is 'simply' building a 24 step attenuator and hand matching the resisters. Pain in the arse but a very satisfying result. I know of one guy who matched his resisters to .001%. He's nuts, but you have to admire the ingenuity...


See also, on the web somewhere a monograph by 'VoltSecond' on various attenuator designs. Just look for VoltSecond and attenuator, you should find it.



Good luck guys. I'll be following these threads with great fascination.....

Hey lorne, I'm in the process of putting together a shopping list for a basic passive. Actually, its the one redboy built and not sure how similar it is to your versions. It's calling for a 100K ohm pot but you have noted using a 10K ohm pot. Would it be just the resistor values that would change?

I'm new to all this mumbo jumbo but am trying to follow along.

Sounds like you have tried a few combinations. Any suggestions?

The reason I bring up the 10k ohm pot is because I have a Proton preamp that has a nice Alps 10k ohm pot in it. I hate to pull the volume control just for that but the thought has crossed my mind.

Thanks, RC

I put together TWO passive pre-amps within the last year.
One uses dual 10K mono stepped ladder-design attenuators (generic MF resistors), the other is built around the Haller circuit using a 10K Alps Blue pot and a pair of 16.9K Dale RN-60 resistors.
So which one do I like better?
Surprising so (to me), the Haller design.
For less money too than the Hong-Kong purchased mono stepped attenuators.
Don't get me wrong, they BOTH sound very, very good!!
And both sound quite similar to each other, so the "perceived" differences are quite small.
It's just that with the Haller design, the system seems to have a little more "snap" (try to define snap, right?).
Maybe it's the Dale RN-60 resistor-vs-the generic MF resistors in the HK attenuators?
Who knows??
I had NO system matching problems with either unit.
Kind of hard to justify NOT sticking with these passive-pre's versus an active pre, when (IMO) they sound so much better than ANY active pre-amp that I have owned.....for far less money!?

Steve

gearhound said:

I put together TWO passive pre-amps within the last year.
One uses dual 10K mono stepped ladder-design attenuators (generic MF resistors), the other is built around the Haller circuit using a 10K Alps Blue pot and a pair of 16.9K Dale RN-60 resistors.
So which one do I like better?
Surprising so (to me), the Haller design.
For less money too than the Hong-Kong purchased mono stepped attenuators.
Don't get me wrong, they BOTH sound very, very good!!
And both sound quite similar to each other, so the "perceived" differences are quite small.
It's just that with the Haller design, the system seems to have a little more "snap" (try to define snap, right?).
Maybe it's the Dale RN-60 resistor-vs-the generic MF resistors in the HK attenuators?
Who knows??
I had NO system matching problems with either unit.
Kind of hard to justify NOT sticking with these passive-pre's versus an active pre, when (IMO) they sound so much better than ANY active pre-amp that I have owned.....for far less money!?

Steve

Click to expand...

Thanks Steve,
I'm going to take that and run with it. Two will be built, one with the 10K and the other with the 100k pot. Should be a fun project!

RC

doucanoe said:

Thanks Steve,
I'm going to take that and run with it. Two will be built, one with the 10K and the other with the 100k pot. Should be a fun project!

RC

Click to expand...

Which design are you going to build - the standard one, or the series-shunt?

Here is the standard (both R1 and R2 change as the pot is turned):



And the series-shunt (R1 is fixed, R2 changes as the pot is turned):

I believe that it's the first one. Nate is going to help me through it and like I said, it's all mumbo jumbo to me at this point. My attention span for the details is limited to more about how it sounds than what circuit exactly. I listened to his second version of the build and was quite impressed. Not necessarily better than the Mod Squad and Channel Islands passives I have owned in the past. Same goes for my current Promitheus passive, but very solid.

He is the more informed guy at this point than I am. I'm sure that he will read this and have something to say about the selected circuit.

RC

doucanoe said:

He is the more informed guy at this point than I am. I'm sure that he will read this and have something to say about the selected circuit.

RC

Click to expand...

RC, I think that may be the nicest thing you've said about me!

We'll be building the first circuit - the standard circuit - but doucanoe's making this interesting for me by throwing the 10K pot into the mix. I can build these things, but I have no real understanding of how to calculate the values of the resistors in them. gearhound used 16.9K resistors in his - is that the "optimum" value for the 10K pot?

I think these little passives are pretty system dependent. So far, they've worked well in my setups but I know ejfud has had mixed success with them, depending on the amps he's used. Someone else can explain why that would be, I'm sure.

Can't speak for anyone but myself, but I see using a 50k or 100k ohm pot for use with tube amps which require higher impedance.

10k pot would load down a tube amp too much for my liking.

Ron

Redboy said:

RC, I think that may be the nicest thing you've said about me!

We'll be building the first circuit - the standard circuit - but doucanoe's making this interesting for me by throwing the 10K pot into the mix. I can build these things, but I have no real understanding of how to calculate the values of the resistors in them. gearhound used 16.9K resistors in his - is that the "optimum" value for the 10K pot?

I think these little passives are pretty system dependent. So far, they've worked well in my setups but I know ejfud has had mixed success with them, depending on the amps he's used. Someone else can explain why that would be, I'm sure.

Click to expand...

I aim to please!


I'm sure it's an impedance matching thing. Gary said that it worked well with the "magnavox" tube amp but sounded rolled off and not as musicaly involving when used with that big DIy Pass class A clone amp of his. A switch to the Promitheus changed things for the better.

Not sure of why, but it seemed to be a better match. My experience with all passives had been about the same. When it's not "right" you can hear it pretty quickly. When it is, it's very, very good.

RC

spkrman57 said:

Can't speak for anyone but myself, but I see using a 50k or 100k ohm pot for use with tube amps which require higher impedance.

10k pot would load down a tube amp too much for my liking.

Ron

Click to expand...

Thus, two builds are in the works so I can listen for myself.

I will send you the 10K Ohm pot version because I know you don't care for tube amps much

RC

Passives are highly system dependent. The input impedance of the passive must be high enough compared to the output impedance of the source driving it.

The output impedance of the passive must be low enough to drive the next chain in the device, usually the power amplifier.

A rule of thumb is 1:10 - the output impedance of the device doing the driving should be about 1/10th of the input impedance of the device being driven.

I would think that the Pass clone amp has an input impedance that is lower than the input impedance of the Magnavox tube amp. The passive volume control would not have had a low enough output impedance to drive the Pass clone satisfactorily.

A different (lower) value for the passive device could have driven the Pass clone well, but that will also lower the input impedance that the passive device will present to the source upstream of it.

The output impedance of the source and input impedance of the power amp are good starting points for determining the value of the passive control that will go in between the two.

There are many designs for active buffers which can be used just after the volume control. These have a very high input impedance, and very low output impedance, making it easy to drive both solid state and tube gear.

revelation!

I just built a Haller passive shunt attenuator. I used the cheapest parts from Radio Shack -- 100k stereo pot and clunky knob, 5% 10k resistors, nickel rca's, 18 gauge solid copper wire, and a plastic 5.5"x2.5"x2" box.
I burned the hell out of the case a couple times.
I ham-fisted the solder joints.
Cooked one resistor because I forgot the alligator clip.
Mistakenly drilled the first two rca holes along the internal bracing so I couldn't get a flush mount on the inside -- had to drill two more for that side.
It took me 2.5 hours and it looks like it should be taken out of its misery.







But. Oh. My. Gosh. That little box of junk just opened up my music like nothing I've used before. Wow. My first DIY creation is a success! I advise everyone to at least try this out.

:thmbsp:

[edit: How's that for my 250th post? cool. ]

Sorry guys ... I've been out of the loop

I started this thread and referred to others who had a similar interests. But for one or two lads it remained a quiet corner. And — for some strange reason — I stopped getting notification of any activity in my Email. So, I've missed questions people have asked me and have been out of the discussion. Looks like there is some interesting stuff to get into, so let me digest the posts and I'll come back. I've built and am using Rod Elliot Project 01 fake law design. Best ... Lorne

In answer to Ashok:

You have done quite a bit of experimentation with the passive attenuator. Good work.

Click to expand...

Thanks a lot. The parts are so cheap that I can't resist.

The bookend attenuator with the Bourn's pots - is that a law faking design? I imagine the pots are linear, and you have a shunt resistor at the output jack.

Click to expand...

This is the Haller design again. I used type B, or linear pots. As I said earlier, volume control was very sensitive at the lower levels. So I began to think that the next one should be a Haller shunt device with type A or log taper pots. I ended up shifting my idea and built Rod Elliot's Project 01 fake law design using cheap Cosmos linear B pots instead.

Your experiments have given me the impetus to try linear pots with law-faking resistors. I will probably get mono pots to allow for proper balance. .... Only issue to watch out for is that the input impedance of the pot will now change with the volume setting.

Click to expand...

That issue concerned me too, but so far I've not had any problem for the most part. I say for the most part because I suspect that my cassette machine is not happy with it.

I have hesitated to post details and pictures about my passive device because it no longer is passive. I have an opamp buffer right after the source selector that drops the source output impedance to around 40 ohms. Makes it very easy to drive any type of load.

Click to expand...

Well Asok, I'm so intrigued that if no one else minds I'll suspend my earlier request for passive only devices. We can get by on the fact that yer buffer may be considered an outboard add-in:yes: Anyway, I am dying to see a schematic, some pics as well as some tips on how to build the device.