10
\$\begingroup\$

We are trying to decide between two different DACs in the same family that are pin compatible. One DAC is about 3x more expensive, but has higher resolution and will definitely work. The second is less expensive, but has lower resolution. We will have to do more extensive testing to verify that the lower resolution DAC works for our applications. We are a startup and would like to sell our product out as soon as possible. If we pass the FCC test with the higher resolution DAC, and switch to the lower resolution DAC later would we have to retest?

The rules I found on the FCC site said that if the part is electrically equivalent we would not necessarily need to re-test, but I am not sure if changing DAC resolution is considered "electrically equivalent".

Our device uses RFID, so it is an intentional radiator, but the DAC is not related to the RFID circuit.

\$\endgroup\$
  • \$\begingroup\$ Can you link to the datasheets for the two DACs? That will help determine whether they're electrically equivalent. \$\endgroup\$ – Kevin Vermeer Sep 23 '11 at 14:28
  • 1
    \$\begingroup\$ Both DACs are on the same datasheet. Only the resolution is different (same full-scale voltge, power-on reset behavior, package, etc.) \$\endgroup\$ – wrdieter Sep 23 '11 at 15:29
  • 1
    \$\begingroup\$ I bet the changes they consider electrically equivalent would be things like replacing a 1 uF cap with a 1 uF cap of the same footprint, but different manufacture. Your situation probably wouldn't fall into this since internally the components are not electrically equivalent. \$\endgroup\$ – Kellenjb Sep 23 '11 at 17:57
  • 1
    \$\begingroup\$ @Kellenjb - I appreciate (and agree with) your "bet", but the OP needs a concrete answer. Got any references? \$\endgroup\$ – Kevin Vermeer Sep 23 '11 at 20:11
  • \$\begingroup\$ @KevinVermeer I don't have any references, otherwise I would have written an answer :-) \$\endgroup\$ – Kellenjb Sep 23 '11 at 20:36
5
\$\begingroup\$

Since your device is a intentional radiator, you need to be very conservative. Common sense may say some changes don't matter, but that's a judgement call the paper pushers get to make and it's very expensive to argue with them. To be fair, some little seemingly inconsequetial changes can effect emissions in unexpected ways.

To be really safe, test your final product and then ship only that.

Get cracking on prototyping with the low resolution DAC so you know which one you want as soon as possible.

\$\endgroup\$
3
\$\begingroup\$

I'm a bit surprised that the FCC allows any changes at all. What's "electrically equivalent"? Even the same schematic on a different PCB layout may cause big changes in radiation. So two identical schematics aren't necessarily electrically equivalent. And, like Olin says, don't assume your change is only a minor one which won't matter much.

To put this in perspective: When I was a fresh design engineer I had to make a few minor changes in a timing lookup table. Of course I was sure the product didn't need retesting! (wasn't FCC.) Until my boss asked me if I was prepared to personally pay the 100 000 euros it would cost to fix it if it turned out to be a problem after RFP (Release For Production). Next thing I did was starting the testing procedure.

You never know how big a small change might be.

I agree with Olin: design for one solution and test that one. Regard the other variant as a different product, and expect new testing if you ever decide to replace your original product with that one.

\$\endgroup\$
  • \$\begingroup\$ +1 for the point about different PCBs. I am also surprised the FCC allows any changes, at least not without them being notified. Plus how exactly do they define "electrically equivalent"? \$\endgroup\$ – Oli Glaser Sep 23 '11 at 22:52
2
\$\begingroup\$

I often see that every part of the system that directly connect to power lines or telephone lines etc. -- and so might accidentally act as antennas and leak unintentional radio interference -- and also any part of the system that intentionally transmits radio signals -- designed as separate "modules" that plug into a motherboard. Those modules are the parts that the FCC is especially interested in.

Compared to putting everything one one big PCB, breaking up the system into several PCBs and buying connectors to attach them to each other has higher material and assembly costs. However, I hear rumors that such a division reduces total costs because it costs a lot less to FCC test and FCC certify new systems that use the same modules plugged into entirely new motherboards.

There's even a rumor that this allows "skipping FCC certification for those motherboards" -- Wikipedia: audio/modem riser (AMR slot), which doesn't sound likely since "all products containing electronics that oscillate above 9 kHz must be certified." -- Which products should have FCC certification and about how much does that cost? .

See also

Wikipedia: Communications and Networking Riser (CNR slot);

" 'Contains FCC ID' compare to just 'FCC'. Certificate of conformity ".

\$\endgroup\$
2
\$\begingroup\$

While pulling up FCC info for this question: Can anyone locate specs on this wifi module? I noticed that the module was listed as modular, valid for any printer that had one of two voltage regulators (The regulator was part of the printer, not on the wifi module).

Even more, when searching for this, found two things. The first is a Hermon Labs Powerpoint/PDF guide to FCC Certification, which includes a section on what does and does not require changes. And the second is the controlling FCC rule on permissible changes without recertification, CFR (2010) Title 47 (Vol 1) Section 2-1043(a)

§2.1043 Changes in certificated equipment.
(a) Except as provided in paragraph (b)(3) of this section, changes to the basic frequency determining and stabilizing circuitry (including clock or data rates), frequency multiplication stages, basic modulator circuit or maximum power or field strength ratings shall not be performed without application for and authorization of a new grant of certification. Variations in electrical or mechanical construction, other than these indicated items, are permitted provided the variations either do not affect the characteristics required to be reported to the Commission or the variations are made in compliance with the other provisions of this section. Changes to the software installed in a transmitter that do not affect the radio frequency emissions do not require a filing with the Commission and may be made by parties other than the holder of the grant of certification.

Furthermore, the FCC Permissive Change Policy 2(c) explains the test behind what is an equivalent chip for the Transmitter Portion, for Class 2 changes:

(i) The new chip component is pin-for-pin compatible.
(ii) The new chip has the same basic function as the old chip, from an external perspective (internal circuitry may differ).
(iii) No change in radio parameters has occurred.
(iv) The same conditions apply when a small area (approximately the same area as the chip) of the PCB is replaced with an equivalent chip.

Based on this, a change to a different version of the same chip family, without any other changes to the board, especially a part that's not related to the intentional radiating parts, would not require recertification nor notifying the FCC. It should plainly fall under a Class I Permissive Change. Class I permissive change includes those modifications which do not degrade the characteristics reported (2.1043(b-1)).

But, the main thing, you could have both tested under the same FCC ID in the first place. If you are worried about it, that is the best thing to do.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.