Viewing Noise with a Measurement Amplifier

RE: Viewing Noise with a Measurement Amplifier

phoenixcomm — Thu, 26 Aug 2021 13:41:52 GMT

[mention:b0bc65b9ecdc4307bd967592f00e340a:e9ed411860ed4f2ba0265705b8793d05] oh, my great stuff. the analog stuff is over my head.

looks professional!

~Cris

RE: Viewing Noise with a Measurement Amplifier

shabaz — Fri, 20 Aug 2021 21:19:36 GMT

Quick photo showing the front panel assembled! The enclosure could do with a rubber surround : )

I finally found some suitable button caps on Aliexpress and have ordered them, currently I'm having to push the buttons with a pencil! but I needed to use it so I couldn't wait for the button caps to arrive.

The photo shows it measuring the noise from a homemade battery-operated supply (there is a Lithium camera battery inside there with a charging circuit). This is a very crude supply that was assembled in the past for just one experiment when I needed a lower noise supply than a bench supply. It contains a REG101 voltage regulator, and only outputs 5V up to 100 mA.

[View:/resized-image/__size/939x548/__key/commentfiles/f7d226abd59f475c9d224a79e3f0ec07-e4b59325-e032-46a7-81c7-892e4beb029b/contentimage_5F00_208526.jpg:939:548]

With the 10 kHz filter enabled and set to 100X gain, in the screenshot below the automated measurement at the bottom shows the noise to be 1.3mV RMS, so divided by 100 that's 13uV, which is quite low noise! The datasheet says it should be 35 uV RMS, but that's specified for 100 kHz bandwidth.

[View:/resized-image/__size/620x372/__key/commentfiles/f7d226abd59f475c9d224a79e3f0ec07-e4b59325-e032-46a7-81c7-892e4beb029b/contentimage_5F00_208527.png:620:372]

With the 10 kHz filter disabled (i.e. a bandwidth of over 500 kHz), I see this:

[View:/resized-image/__size/620x372/__key/commentfiles/f7d226abd59f475c9d224a79e3f0ec07-e4b59325-e032-46a7-81c7-892e4beb029b/contentimage_5F00_208528.png:620:372]

The measurement there is 3.4mV, which is 34 uV RMS, and that's almost spot-on compared to the 35 uV RMS at 100 kHz bandwidth value in the datasheet! I've not taken the noise of the amplifier into account. So everything is definitely in the right ballpark, and reasonable enough for relative measurements too.

I'm thinking the measurement amplifier should have a 100 kHz filter setting (i.e. settings of 1k, 10k and 100k rather than 100Hz, 1k and 10k), since 100k is so popular in datasheets, and would mean the measurement amplifier noise would barely need to be taken into account for most measurements, since it is low (about 3.7 uV RMS) up to around that ballpark. Anyway, for now I'm happy it is functioning good enough to use, and maybe at a later date the RC filter components could be recalculated. I've attached modified front panel files in export-alu-100k.zip in case anyone wishes to make use of that.

[View:/resized-image/__size/620x324/__key/commentfiles/f7d226abd59f475c9d224a79e3f0ec07-e4b59325-e032-46a7-81c7-892e4beb029b/contentimage_5F00_208529.png:620:324]

Aliexpress buttons; they need a 2.8 x 2.8 mm hole:

[View:/resized-image/__size/620x408/__key/commentfiles/f7d226abd59f475c9d224a79e3f0ec07-e4b59325-e032-46a7-81c7-892e4beb029b/contentimage_5F00_208530.jpg:620:408]

RE: Viewing Noise with a Measurement Amplifier

shabaz — Fri, 13 Aug 2021 21:12:57 GMT

Regarding alu, I was reading this today, it seems quite relevant, maybe not to this project, but to other physically smaller projects.

https://www.sculpteo.com/blog/2017/03/15/metal-laser-cutting-how-to-fold-sheets-of-metal-easily/

I'm not confident in getting it right (I know so little about metalworking) but I think it would be worth a shot, to try to devise such a thing for (say) 80x50mm boards (i.e. about credit card size) and be able to easily modify it to customize the holes or silkscreening for each project.

Think of the time saved, not having to cut an enclosure on one or more sides : )

It may not look as pretty as a ready-made enclosure, but still it could be worthwhile perhaps.

RE: Viewing Noise with a Measurement Amplifier

shabaz — Fri, 13 Aug 2021 08:50:25 GMT

These arrived earlier (within 7 days of ordering!), so I propped it up against the enclosure to quickly test for fit.

Also in the foreground in the photo below, is the underside of the alu front panel, and a conventional FR4 fibreglass one underneath it.

There is not much in it, although the alu PCB has a smooth surface, unlike the faint regular woven pattern of fibreglass boards, and of course the edge will be visibly aluminium when it will be screwed onto this extruded enclosure.

Also alu won't be very forgiving if screws are over-tightened and it bends. So in summary for this particular use-case perhaps the alu isn't totally necessary, but still glad to have tested it.

One test I wish I'd done, but didn't, was to see what happens if there is no solder mask, and if the silkscreen still makes its way onto the bare aluminium. It is probably not supported I guess. But at least some older proto PCB services used to do the screen printing without checking if there was solder mask underneath in the past. It's a shame that is likely to not occur any more : (

[View:/resized-image/__size/1266x843/__key/commentfiles/f7d226abd59f475c9d224a79e3f0ec07-e4b59325-e032-46a7-81c7-892e4beb029b/contentimage_5F00_208524.jpg:1266:843]

RE: Viewing Noise with a Measurement Amplifier

jc2048 — Sat, 07 Aug 2021 12:09:17 GMT

I enjoyed reading this. Thanks for blogging it.

Your op amp is much more precise than the three-transistor cicrcuit I was playing with. The OPA1611 looks like a very nice device.

If it's of any interest, this is my transistor circuit response in more detail. My test equipment and measurements by hand aren't really quite up to this, but it gives an idea for comparison.

[View:/resized-image/__size/620x392/__key/commentfiles/f7d226abd59f475c9d224a79e3f0ec07-e4b59325-e032-46a7-81c7-892e4beb029b/contentimage_5F00_208523.png:620:392]

Perhaps I had better have a break from the 'Summer of FPGAs' and build the other two transistor circuits I wanted to try and see how they all compare.

RE: Viewing Noise with a Measurement Amplifier

dubbie — Thu, 05 Aug 2021 15:43:50 GMT

Shabaz,

Very impressive. I do find anything analogue electronics impressive. I do like the idea of putting all the switches and connectors onto the PCB to avoid wiring. I do find that the wires I use flex too much and just break off. I have even resorted to purchasing some special 'bendy' wire to see if I can overcome this problem. It does have the problem of needing a case with the holes in exactly the right places - something else I also find difficult to achieve.

The 'case' at the bottom, in your first photograph looks remarkably like a biscuit tin. Biscuits I understand.

Dubbie

RE: Viewing Noise with a Measurement Amplifier

DAB — Tue, 03 Aug 2021 18:31:06 GMT

Very good post.

I like the design, I might try it out for some thunder measurements during lightning events.

DAB

RE: Viewing Noise with a Measurement Amplifier

genebren — Fri, 16 Jul 2021 20:27:00 GMT

Very cool project. Your write and explanations were really great. This is yet another project that I need to think about doing.

Well done!

RE: Viewing Noise with a Measurement Amplifier

fmilburn — Fri, 16 Jul 2021 05:49:13 GMT

Hi Shabaz,

That is a useful instrument and the project is very well done. Thanks for the great writeup and documentation as I am going to look into building one but need to finish off my current backlog. I agree a PCB front plate could work well and liked the way they worked out on my AIS project.

Frank