A little help with pads please?

[Led]

Hi knowledgable slutz. Could somebody give me a little help understanding T and H pads please? In the attached diagram I get the unbalanced T pad. At least I assume I do - the top line is the positive and the bottom line is ground. I don't really get which wires the ones in the H pad diagram are. Are they the + and - ? Where is ground in the H pad diagram?
Any calrification would be helpful and appreciated.
Thanks.

Led

[Francis Vaughan]

are they the + and -

Yes. A simple way to think of the relationship between T and H, draw a horizontal line through the middle of the H. Since you have + on one side, - on the other, the line is mid way between, and you can label it as a notional zero. Voila! You have two T pads (one upside down) connected together at their zero. Since you don't need the zero point you ignore it. There is no ground. Balanced lines don't have one. If you have a shield you simply keep it connected from one side to the other - it doesn't touch the pad.

[Led]

Thankyou Francis. Much appreciated.
Cheers

[tINY]

There is no ground in a balanced interface. That's the point.

The ground is at the transmitter and both outputs have equal output impedance. The "virtual ground" at the middle of H-Pad could be a DC or AC voltage compared to the real ground - it's called common mode signal.



-tINY

[Led]

Cheers Tiny, I got the -20dB pad happening. Next challenge is to make it switchable to varing attenuation but I'll post another thread about that.
Thanks,

Led

[ulysses]

For some reason the attached images aren't visible this week, so I can't see whether or not your drawing addresses the fact that in the modern voltage-transfer world, the "H" pad is typically more of a sideways "U" (or a backwards "C") and the "T" pad generally looks more like a "7" or an "L" rotated 180 degrees.

In the olden days of power-transfer, they used to match the source and load impedance. That's useful if you've got a generator powered by a human voice and you want it to travel across town without any help. Such a system would probably never require a resistive pad, but it set the standards that hung around for more than half a century and so it's what our electronic vocabulary is built on. In such a system, a constant-impedance pad needed build-out resistors on its output side to maintain the source impedance as seen by the load, just as it maintained the load impedance as seen by the source.

Today, we generally try to maintain the lowest possible source impedance and the highest possible load impedance. This is pretty much impossible for small amounts of attenuation, but we do the best we can. In any case, it's pretty rare today for us to seek out a higher source impedance, which is why the two resistors on the right-hand side of an "H" pad (or the one resistor on the right-hand side of a "T" pad) are rarely more than zero ohms. In other words, you generally omit them.

[tINY]

Good point. I guess the matched impedance helps if there are filters expecting to see a certain line impedanc (though those went out of vogue in the late 50's).




-tINY

[Led]

Hi, interesting point. I've been looking into pads for 2 bits of gear.
A pair of Federal AM864/U limiters and a pair of Siemens V72 pres. Both are 600 ohm input (and from the 50's), so I've been looking for ways to pad down the input without altering the frequency response. I was looking at some of those Eckmiller faders on ebay but I really only want a fixed attenuation on the federals, around 20dB, and with the V72's I'd be happy to have say 3 lots of switchable attenuation instead of a fader, so I was hoping to find a cheaper way out than the $160 US the faders have been going for on ebay.
Thanks again for the input.
Cheers.

[dubrichie]

alrighty, i have some more questions!

i need to build some pads because when i use condenser mics on drums with my Wunder preamps my AD converters clip at the lowest gain setting on the preamp.

may i assume that it is a better idea to use the pads between the preamps and the converters than between the mics and the preamps?

if so, what resistor values should i be using?

can i, as suggested above, simply omit the R2/2 resistors on the right hand side of the H-Pad?

also, can anyone refresh my memory of the model number of a suitable XLR barrel connector to build the pads inside of?

many thanks,

[dubrichie]

BUMP!

[tINY]

The first thing you have to do is figure out where it's clipping.

If the mics are clipping, you need a mic with a pad. If the pre-amp is clipping, then you need a mic pad in front of the preamp. If the input to the ADC is clipping, then you need a line level pad between the preamp and the converter.

Assuming that it's the ADC clipping, I'd use about 4.64k on each leg (preamp side) and a 1.00k shunting the ADC side. This should give you about a 20dB pad.

I wouldn't use barrels (they're hard to find). Use panel mount XLRs and put them on a rack panel or on a metal enclosure.




-tINY

[dubrichie]

thanks a million tiny.

i'd like to use the highest quality components that i can get my hands on, any suggestions of a make / model for the resistors?

[ulysses]

Use 1% (tolerance) metal film resistors. There are dozens of companies who make them and they're all the same. Some jackass is bound to come along and tell you how much better the old 10% carbon composition resistors sound, but they're full of shit. This is a resistive pad, and all you want is fixed resistance without extra noise.

Buy some extra resistors for the series legs (4K64 in Tiny's suggestion). Measure them all with an ohm meter. For each pad you build, pick two shunt resistors that both measure exactly the same on your meter. It doesn't matter how close they come to the nominal value (they'll be within 1% of it), but what's important is that they're equal to one another.

There. You're done. Now make some records.

[dubrichie]

thank you sir.

[dubrichie]

excellent, i've sourced 0.1% metal film resistors locally and Redco are supplying me with Switchcraft XLR Barrels for $7.98 each, which is about half the price of anywhere else i've found.

if i was to put the pads before the preamps, so i can crank them a bit more, what resistor values would you suggest?

thanks,

[ripple_fx1]

Here's a link to PAD101:
Uneeda Audio - Build your own attenuator pads

[dubrichie]

thanks, i have read all that and understand it.

i suppose what i meant to ask was "what is the average output impedance of modern microphones" or "what is a good choice of impedance for what the pad presents to the preamp".

the input impedance of my preamps is known and take into account.

[ripple_fx1]

Gotcha - 150 ohms is the nominal Z....

[dubrichie]

thanks again.

soldering time!