Analog parts notes for REV9

Development
1

Today I looked a bit at replacement analog parts for the comparator and opamps.

Possible op-amp replacement with standard parts:
LMV321/A. The gainsil is very nice, the A has better specs than the 6001, but is twice the price of the 321.

LMV331. Standard cheap comparator. There are alternate pinout Y parts that match our board, but I think that is asking for trouble and we should change the PCB to use the more common pinout.

2

Current sense opamp: Gainsil GS321A
Buffering opamps: Gainsil LMV321

Comparator: GSV311/HGV311, whichever is cheaper. They clearly have the same die from the same fab.

Other parts:
All PFETS: CBI SI2301 in sot-523
Diodes: CBI 1n4148 in do-523

Potential substitutions:
AIP replacements for all 7400 logic
Hopefully the 4066 is removed in this revision

3

Hi Ian

The current purchasable version appears to be revision 8.
I read on this forum that you are already working on rev9 which changes some analog components.
When will Rev9 be released? Do we have a release roadmap?

Diego.

4

Hey Diego,

Welcome to the forum.

Yes rev 8 is the one on sale. It’s an assembled board, no enclosure. The enclosure is due soon after the national holiday, according to the tooling contract. Then they have to make samples for fit testing and then final production. I’d say that’s a couple weeks away.
​
​My plan was to have rev 9 FCC tested before the holiday this week. However, I botched the board and didn’t update the opamp footprints. I rolled another and had it assembled last week. It passed self test in the office, and should arrive to me today. I’d like rev 9 to go in the enclosure and probably won’t FCC certify it as a bare PCB.
​
​The difference is mostly bom cleanup. And that’s a fancy way of saying cheaper and better available parts. For example rev 8 has a fancy low power opamp, rev 9 has a bog standard opamp, and an ‘A’ grade standard opamp (sorted by machine) in the current sense. I also used a cheaper standard comparator instead of the super low power RunIC part in rev 8. It’s not a battery powered device, the extra cost isn’t justified.
​
​rev 9 has a cheap common pfet instead of the expensive ultra low vgsth winsok part (both have same max vgs, but different typical, so I don’t see how it makes much difference).
​
​im on a quest to eliminate 7400 logic chips. Rev 9 uses pfets instead of 74hc4066 for the pull up resistors. That’s the big change I need to test when I receive the board. This is actually kind of a down grade, but not in any way that impacts usability (lowest pull up is 1v vs 0v, buffer is out of spec <1.65v so…).

I’m hoping to test rev 9 today. If it’s good we’ll send it for FCC testing asap (next week) if we get the sample enclosure.
​

1 Like
5

Finally talked to a Gainsil factory rep. He said they sell (design too?) the dies used in HGSemi’s parts. I knew it was something like that.

Currently we use HGSemi LMV324 and LMV331.

  • HGSemi LMV331 is sold out everywhere. Gainsil is in good supply, cheaper, and is the same die (we’ll verify with a decap). That’s an allowable change under FCC, and we have to do a batch CE again anyways (for the cable kits).
  • HGSemi LMV324 is in good supply, but Gainsil LMV324 is about half the price. I’m going to have some test boards made for certification before we use it in production, but I want to switch to this chip.

Hopefully we can consolidate on Gainsil parts for all the analog stuff, and get it direct from the factory. We’ve tried to talk to HGSemi, but they told us to go away (ask a distributor).