I recall that that particular noisy amp was actually a monolithic, fabbed by AT&T, in their brief foray into making their own high-voltage ICs.
I'm thinking that APEX sells to people with lots of money and not much space, like for satellites or something.
If the TCA0372 ever got unavailable, I could make my own mouse-bite board to drop in. There should be enough room, given the wide SO16 package.
-- John Larkin Highland Technology, Inc The best designs are necessarily accidental.
We just did the same thing last week, changed a product spec to bracket what we're actually seeing in production. Sorry.
No this was different. We were using some Apex opamp (PA124A or something I've forgotten the number) Apex stopped making it and PA123B was the replacement. I glanced (in retro-spec too quickly) at the spec sheet and OK'ed ordering the replacement. (I think this was the second time there had been an upgrade/change.)
In ~1 year we used up old stock and the new part got installed in the board. During testing; "Why is this so noisy?" I ask myself. After a day of poking around, I figure out it's the new Apex opamp.
LT6090 came to the rescue. I think I posted a pic of the most size mismatched plug-in pcb. soic8 to TO3-7 (to3 pac with seven pins)
George H. (who has no problem with drawing a circle around the spec sheet 'bullseye', after you've launched a few arrows. :^)
Turns out I misremembered. The PA84, PA85, and BB 3584 all had mildly-degenerated JFET inputs, and so their 1-Hz voltage noise came in between 3 and 10 nV. The super noisy one I was remembering in connection with STMs was the late unlamented LM11, which some folks tried to use as the STM preamp. Its offset current was typically 1 pA, but as it was (weirdly) _bipolar_, its 1-Hz voltage noise was north of
100 nV.The smart kids used OPA111s at the time. ;)
Cheers
Phil Hobbs
-- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC / Hobbs ElectroOptics Optics, Electro-optics, Photonics, Analog Electronics Briarcliff Manor NY 10510 http://electrooptical.net http://hobbs-eo.com
Bias current cancellation reduces DC current and increases noise current.
-- John Larkin Highland Technology, Inc The best designs are necessarily accidental.
Yeah, but it reduces DC current by one or two orders of magnitude, and boosts noise current by 0.15 orders of magnitude.
The LM11's selling point was that its bias current didn't go nuts at high temperatures. JFET amps at 120C are, um, disappointing that way.
Cheers
Phil Hobbs
-- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC / Hobbs ElectroOptics Optics, Electro-optics, Photonics, Analog Electronics Briarcliff Manor NY 10510 http://electrooptical.net http://hobbs-eo.com
Hi John,
What is R303 in the schematics in reference, please.
H
That's a
RIEDON PFC10-50RF1
or alternatly
CADDOCK MP725-50.0-1%
Both are excellent wideband power resistors, even with the tabs hard grounded.
The Riedon is a flat ceramic resistor that drops into a DPAK footprint. The thermistor is hung on that. It all becomes a sophisticated power overload computer.
-- John Larkin Highland Technology, Inc The best designs are necessarily accidental.
I see. I was just wondering why that metal tab is grounded. Thanks John.
H
Those resistors are rated for something insane like 25 watts. That must assume serious heat sinking. Something more like 5 or 8 watts is more reasonable on a PC board. I have vias to the layer 2 ground plane and copper pours out to the board edges, where we can heat sink to the box.
how did you manage that ? For me that is impossible.
H
I hacked some proto boards and tested some resistors for temp rise. Advanced thermodynamics.
Here's one example.
The 44 volt version can toast things if the user shorts the output and cranks up the duty cycle.
The big gold strips get bolted to the case, which is the main cooling path. I figure that resistor is good for 6 watts or so. R42 is the shutdown thermistor.
I have some capacitance and TDR data around here somewhere, on those resistors.
Polished PCBA as usual ..even for prototypes ;-) ref des of components on it is clearly readable. Fine
BTW I see you have been using a OPA power, Does some parts of this OPA family can be a replacement of the TCA ?
Best regards, H
That's actually a production board. There was no prototype. I did the design and layout at home, in a week during the initial hard virus lockdown. It works, with a minor ECO.
The box was an experiment in laser blasting artwork onto an anodized box.
We have sold.... TWO!
The TCA is only good for 40 volts, and the OPA547 is rated for 60... at about 12x the price. The OPA also has a resistor-settable current limit, which helped my overall power scheme.
-- John Larkin Highland Technology, Inc The best designs are necessarily accidental.
Bravo !
It's usually bad when an engineer thinks that some product would be cool, with nothing like market research. But sometimes a year or two later, people start buying. Or someone sees the web page and asks for an OEM product. So something dinky that makes search targets but doesn't itself sell may be worth it.
-- John Larkin Highland Technology, Inc The best designs are necessarily accidental.
John,
Nowadays market research is everything for marketers. Sometimes best seller OEMs products are designed and build outside that process but mainly from talented engineers ideas. Agreed.
Best regards, H
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.