iRIAA filter PCBs

[simon]

Okay, I'm not buying or selling but this seemed the best place to put this. Now I doubt anyone's interested in those funny old black disk things but if anyone would like a inverse RIAA filter I'm going to send gerbers off for a PCB based on the Hagerman iRIAA2 filter.

www.hagtech.com/pdf/iriaa2.pdf

The fab houses send 5 boards and I can't imagine I'll ever need more than one (unless I lose it) so there are spares for any regulars on the forum in the UK. The cost to you is zero, I'll even bung it in the post at my cost (I'm not expecting many takers ). Disclaimer - I haven't built this yet so can't be sure it works, but it's a very simple circuit so I'm confident it will.

This is the board: https://www.audio-talk.co.uk/phpBB2/vie ... 44#p213644
.

Hagerman iRIAA.png (117.61 KiB) Viewed 4955 times

.
Yes it's SMD, which allows 1% caps and 0.1% resistors and keeps board size to 43mm x 40mm. I haven't priced up the resistors and caps in detail but it looks like about £7 at Mouser. Plus phonos, chassis, standoffs etc. If anyone does want a PCB, and would like the SMD bits too I'll add them to my Mouser order and charge to you at cost.

If anyone does fancy one let me know. I'll upload the gerbers in a few days.

[Nick]

When its built, if you want to measure it for reference, you know where I am.

[simon]

That's a good idea, thanks.

I'll probably buy a few of each component to match. It'll only be as accurate as my DMM likely isn't, but at least I can match the channels.

[Andrew]

Be curious to see how surface mount compares to thru hole?

[simon]

I suppose SMT will be a tighter layout, but presumably it will largely be dependant on the accuracy of the parts used?

[Nick]

The fact there are 1% caps are a big win for SM. I assume they are C0G so distortion will be low enough and no one builds them to add a extra CD input to a integrated amp any more

[simon]

TBF I haven't fully scoped the parts, I looked at a couple of the caps and they were available in 1206 C0G 1%. I should probably do that, though if they're not available in 1% C0G there's probably not much choice in iRIAA anyway.
.
I've added two more output GND pads as I'll probably just use two pairs of output phonos rather than a switch and it's always easier soldering one wire per pad. Tightened up some of the spacing too which also allowed me to shave a whole 3mm off the board .
.

Hagerman iRIAA rev 1.png (128.09 KiB) Viewed 4174 times

.
Only one taker so far (one more than I expected ) I'll probably upload the gerbers in the next day or two to avoid Chinese new year.

[simon]

I've just checked the parts, which was worth doing. C2, 6n2, is only available in 1210 as C0G 1%. It is available as 1206 1% in X8G or 5% C0G. Better to resize the pad for 1210 or is X8G adequate here?

R1, 1k92, isn't available, only 1k91. I imagine the extra 10 ohms doesn't make much difference here?

[Nick]

I would say it depends if you plan any distortion measurements. But you don't need the inverse Riaan for that just set to constant output from the stage.

[simon]

That sounds like use a X8G then. That said I might revisit with all 1210 parts as even 1206 are a bit skinny.

[Andrew]

I don't believe X8G are as stable over time as C0G.

[simon]

Very interesting. I thought about 1210 and even 1812, but there's almost no availability in larger sizes if I want to stick with 1% C0G caps and 0.1% resistors. So I'll alter C2 to 1210 and leave the rest as 1206 and enjoy the challenge of soldering tiny parts.

[simon]

Updated with the larger 1210 footprint for C2
.

Hagerman iRIAA rev 2.png (140.26 KiB) Viewed 3745 times

.
Component spacing is hellish tight though, CAD software gives a false perspective of space I think. There's barely 2mm between them.

[Morgan Jones]

I would say it depends if you plan any distortion measurements. But you don't need the inverse RIAA for that just set to constant output from the stage.

Actually, you do. One of the specifications that is often missing from capacitor data sheets is how capacitance varies with frequency. C0G/NP0 is supposed to be the best of the ceramics (although I've never measured it). The ceramics we commonly use for bypassing power rails can vary by 40%. I prefer to use polystyrene or polypropylene for filters.

Incidentally, the Hagerman iRIAA is slightly limited because he elected to use 1% parts everywhere. If you use 0.1% resistors it is possible to do slightly better, and there's a modified version on p449 of BVA.

[simon]

Interesting. I dug out my copy of BVA but it's an earlier version I think as it only goes up to page 354. However I presume it's the same as fig. 6.10 on page 326 of the first edition? The second (b) fig looks the most appropriate accepting I won't be using a very short lead from the sig gen. Just another 4 resistors to add.