driving Pass A40

I have a question about the Pass A40. Can I use a passive pre amp to drive it. At this point I really only have one source my Cd player. If so what value of POT/Attenuator do I use? (There seems to be alot of different values to choose from)In the article NP states the A40 have a nominal input impedance of 40K ohms. This is my first amp and any help would be great.
Ward
 
There's not going to be a hard and fast value for your pot. I'd say to try a 47k log (aka audio taper) pot (pull one out of your junk box or buy a cheap one). For solid state stuff (i.e. the CD player and the amp) you don't generally run high values like, say 100k. For tubes, however, values like that are normal.
If you like, you can use a lower value, say, 20k (but still log taper). It won't hurt a thing.
Try it and see if your CD player has enough output to drive the amp to a: a loud enough volume to suit you, or b: into clipping, whichever comes first. If the volume is too low, you're only out the time (and possibly the money if you bought a pot) it took to put it together. Then you'll know you'll need a line stage, although it's not likely that you'll need much gain. 6 to 12 dB should be more than enough. If your CD player has sufficient output, then you can consider splurging on a fancy pot and all the other nice stuff.
A linear pot (the most common kind) won't hurt anything either, but you'll get all your volume in the first quarter-turn; no fine control. It's a cheap trick that stereo manufacturers use to impress the unwary--salesman turns the volume a little bit, shows customer the knob, customer thinks: Wow, if it's that loud with the knob at 9:00, just think how loud it will be wide open! Ain't so. Caveat emptor. You can use a linear pot, should you have one on hand, to determine if your CD player has enough output, but I wouldn't think you'd want to live with one.
Good luck.

Grey
 
Having used an A40 for several years with a passive PA, the short answer is YES!

However (there is always a catch!), it depends upon what you are running into the amp. Most modern CD players have outputs around the 2V mark and this will *easily* drive the A40. If you are using a RIAA section or other device, then it will depend upon the output from this.

IMHO, if you are listening to CDs, then the use of an active PA is totally irrational.

I use an ALPS 20KLog "high quality audio" pot. These cost a little (actually a LOT) more, however one of the really salient lessons from my youth was the noise from standard pots - don't use one of these except for testing purposes!

BEST option is to buy a good quality (ie. gold contact) multi-position rotary switch and build a stepped control using metal film resistors. It was in fact doing this which showed-up the gross noise we were hearing from standard pots!

My current project is a passive PA based on the volume control section from Pass' Aleph O PA. This uses an 8bit-ADC and a series of relays through metal film resistors - have a look at the service manual ;-)

cheers, mark
 
I just ran a simulation of the A40 through Circuitmaker and You will hit the rails at about 2.5V P-2-P.

Assuming the rating from the Adcom is RMS-AC, the unit has output peak-2-peak at 4V.

As stated previously, if you are using a CD player, you are *highly unlikely* to require a pre-amp .... the only thing it is useful for is adding 0.01% THD to your signal ;-)

cheers, mark
 
doktor,
Duh! I'm slapping myself on the forehead. I just remembered why everyone doesn't use pots instead of preamps on high-output sources. The pot reacts with the capacitance and inductance of the interconnect going to the amp. Depending on the level of the volume control (hence the source resistance seen by the interconnect), you will get variations in frequency response. How much will depend on the reactance of the cable.
A line stage, setting aside questions of expense and gain, also buffers the volume control, usually by placing a cathode/emitter/source follower as the final stage. This gives you low output impedance from the preamp and minimizes interactions with cables.
In theory, I suppose you could make a case that it might oscillate. In practice, I've never heard of it happening, so I doubt that you will face any problems in the practical sense. It's just something that you might want to be aware of from a performance standpoint.
I seem to recall that Nelson has some discussion on this same issue in the Bride of Zen writeup. I think he decided to put the pot at the back, even with cable-dependent things going on.
Sorry to be so slow getting around to this, but I haven't had more than about four hours sleep the last several nights running, and my brain is in molasses mode...

Grey
 
Above comment is quite true. Potentiometer + C_cable will form a RC roll-off filter.

Before we condemn this, recall that most pre-amps also have high roll-off filters. So, how big a deal is this?

For fun, I just measured the C_cable of an "average" cable from my box = 0.128nF. Lets assume our pot is at 50%R and it is a 10K or 20k unit. Using Circuitmaker to run a simulation, driving this with 2V input (out should be 1V), we see at the following frequencies:

For a 20K pot:
1kHz = 999.26mV
10kHz = 998.34
20kHz = 995.59
100kHz = 929.55

For a 10K pot:
20kHz = 998.56
100kHz = 978.87 (considerably better!)

If we take a 10K input impaedance for our power amp, this equates to less than 1dB at 100kHz (someone please check this, I'm no EE and could easily have stuffed the math bit up!!)

So, what do we learn from this:

1. Choose a lower value for the pot, 10K is better than 20K, but you can likely go slightly lower than this if you are building with metal film Rs.

2. Keep lead length to a minimum. If we halve the cable length, then we have half the C_cable (0.064nF) and our worst case for 20k is not 929.55 but 979.28mV. If you are going 20cm then use 25cm, not the standard 1m! (This *IS* a DIY list, just make one)

3. Cable quality *does* matter. You can do much better than the 0.128nF per meter of the cable I measured. Good quality cable should actually list "pF/meter".

AND ..... I still recon this is less of a sin than would be committed by most pre-amps ;-)

cheers, mark

[Edited by mefinnis on 03-07-2001 at 01:25 AM]
 
making attenuator

Now you guys have me thinking about making one. I found the following link that calculates the values.

http://margo.student.utwente.nl/klaas/audio_step.htm

now is 3.33... dB step size ok? I know it probally depends on listening preferences but is this a good starting point. Also note the highest/lowest level dB. Any ideas? I also read that a make before break switch should be used? All the info so far has been great thanks.
Ward
 
doktor, Mark, everybody else...
Got sleep?
(Mark, do they have an ad campaign down there for milk with various celebrities with milk on their upper lips? I believe the ads are sponsored by some national milk industry board. This may not translate well to other countries.)
At any rate, I'm firing on more cylinders today. Back with an easier & cheaper solution.
Duh! (Accompanied by the sound of me slapping my forehead again.)
Put the volume pot into the amp itself.
Advantages:
-no pesky capacitances or inductances to worry about
-fewer boxes hanging around; why go to the trouble to have a box with nothing in it but a pot?
-cheaper! subtract 1 pair nice cables, 2 pair nice jacks (in and out), 1 enclosure, and of course the misc. rubber feet and hookup wire that always creep in the back door
The only disadvantage that I can come up with is that it's a little 'old fashioned,' as it is not current practice to put level controls on power amps. So what! It's your amp, your project...do what you want with it. Incorporate a switch and some extra input jacks if you want so you can hang a tape deck off the side.

Grey
 
doktor,
3.33 dB sounds a little wide to me. Let everybody else weigh in with their opinions, then do something scientific...like flip a coin.
Two observations:
-I'd still check (using a cheap pot) to make sure you are happy with the volume attainable doing the direct drive thing. Some of these switches and the attendant resistors can get damned expensive. It'd be a pity to put out a huge amount of money, only to find that you're not satisfied.
-Keep in mind that there's no law that says that the steps on the switch *have* to be the same increment. You could have, say, six dB steps starting from the low end up to somewhere just south of a 'normal' listening volume, then drop back to 1 dB steps for fine control in that general region. This will save a lot of positions on the switch. If you haven't already checked, you'll probably find that 24 to 36 position switches are the most common, but that doesn't leave much elbow room for fine control if you try to divvy it up equally. Yeah, I know...this complicates the math, but you'll only have to do it once.

Grey
 
I hinted earlier that we have made a number of stepped volume controls and I would have to say these are my preferred option. You have much greater flexibility.

Cost *IS* an issue and the comment about checking first seems wise, just ignore the "white noise" in the background ;-)

For example, my preferred switch would be a Blore-Edwards 29 position gold contact job, which goes for the small sum of $AUS400 !!!! **If anyone has a spare, I'll pay the freight ;-)

I would propably settle for the ELMA 12 position, low-profile PCB mounting unit, its only $AUS100.

You get the idea ..... good switches cost real money :-(

Anyhow, doktor, Grey is absolutely on-track when he suggests making the steps different. This is what we have always done. Think about the way you listen ..... for me, I want fine control at low volumes (background use, multiple people in the room etc.), then as I start to get louder, less control and bigger steps is acceptable.

Suggestion: Aim for at least 6 x 2dB steps at the bottom end - 3.3dB is actually slightly greater than "doubling" power, ie. 2W, 4W, 8W, 16W etc.

So, if we realise that good quality, large No. contact switches cost lots, the design of the volume control in the Aleph P starts to make more sense: 8 DPDT good quality relays (not cheap), 8 switching transistors, 2 x 8 metal film resistors (peanuts) .... and 8bit AD-converter (not much), and a "cheap and nasty" pot (peanuts).

The end result, if you do it right is a 255 position rotary switch controlling balanced metal film resistors ...... starting to make more sense gents !

NP does his at quite high voltage for reasons that escape me. You can get 5-12V relays and the ADC0800 will run off 5V, so my plan was to have one of those plug-in DC supplies (like for your mobile phone) @ 12V which would be totally external and regulate it accordingly.

Comments welcomed! mark