SIL resistor question

[gez]

There are some tasty SIL resistor packages in the new ESR catalogue (though they've been selling them for quite a long time).  I was just wondering what they're like in terms of noise?

The blurb describes them as 'metal glaze elements on a ceramic substrate'.  Will they be noisy?  The 4 isolated 8 pin packages would be ideal for providing a couple of dividers to bias op-amps, when more than one is needed - it would save on board space!

Anyone ever use them?

[Paul Perry (Frostwave)]

I've only used them for logic pullups, but i believe they are as good as (indeed identical) normal "metal film" resistors. But, you usually only need one source of 1/2V, so..??

[gez]

But, you usually only need one source of 1/2V, so..??

Thanks for the feedback Paul.  One divider for audio (op-amps etc) and one for an LFO (I can think of at least two LFOs offhand where it would be unwise to share a divider with audio).

The way these packages are arranged it would be easy to keep traces out of the way of each other, but (and here comes question No.2) would I be playing with fire by using one of these rather than discrete resistors?  I should think that all these resistors are formed on the same substrate, but surely they're isolated within the device or is there anyway signals could interact (capacitance etc)?

[Paul Perry (Frostwave)]

The main problem as I see it, is that if it is all that critical to have separate bias sources (and personally I doubt it, since so little current flows in the bias inputs), then you will potentially create problems by having the various bias lines run near to each other , as they approach the same SIP resistor array. Just put resistors on end, and you save a lot of space anyway.

[sir_modulus]

good Idea though Gez, To save some space, I've also seen people use resistor arrays (vertically mounted) as they only come with the # of resistors you need (you can get nice ones with only two for small voltage dividers).

[gez]

The main problem as I see it, is that if it is all that critical to have separate bias sources (and personally I doubt it, since so little current flows in the bias inputs), then you will potentially create problems by having the various bias lines run near to each other , as they approach the same SIP resistor array. Just put resistors on end, and you save a lot of space anyway.

With the type of LFO that has positive feedback to the divider to pull the centre of it towards whatever rail the output slams into (thereby setting the thresholds), it would be unwise to use a single divider.

I don't use this type (just an example), but I often run LFOs from regulators at lower voltages, hence the need for more than one divider.  

As I said in my previous post I don't think layout will be an issue, not the way I have things in my mind's eye (though how it pans out in practise is another thing altogether - I always was a chancer though!).  By using the extreem pins for the separate ground paths and the middle two pins tied to V+ (or V+ via a regulator for one half), the 'centre taps' would be taken left and right to their respective amps.  There should be plenty of room to avoid traces getting in each others way and this should take up less room.

So long as they're quiet for audio - thanks for this snippet Paul, though God help you if you're wrong!  :twisted: ( ) - I think I'll get a few and see what results.