Mini-circuits pHEMT designer's kit

One very nice piece of info--the kit includes the new SAV-331+, a _depletion_ pHEMT that they claim has a typical NF (unmatched!) of 0.5 dB in the flatband and 0.9 dB down to 40 MHz.

Cheers

Phil Hobbs

makes no sense. Idss is 228 uA and operating point is 60 mA at -0.69 on the gate. The 228 uA is a funny font too.

Of course, being an RF part, we get no DC curves.

There should be a profession, Data Sheet Checker Consultant. The state of data sheets is a disgrace.

Yeah, it isn't the biasing that's cool--it's the hint at a much lower

If we ignore current noise (probably okay for a pHEMT at low frequency), in an unmatched system (as shown in the datasheet), the flatband 0.5 dB noise figure corresponds to a voltage noise 'density' of

e_N = 0.91 nV/sqrt(Hz) * sqrt(10**0.05 - 1) = 0.32 nV/sqrt(Hz)

and the 40-MHz value of 0.9 dB corresponds to 0.44 nV/sqrt(Hz), just about exactly 3 dB above the flatband value. So we expect these guys to be better than the CPH3910 (0.9 nV) down to about 10 MHz.

I use pHEMT bootstraps quite a lot, because their behaviour is a good match--their low frequency noise gets killed by the increasing reactance of the source capacitance, and their high-frequency noise is 10 dB better.

The old Avago parts (RIP, *sniff sniff*) were about like that, but the Skyworks SK65050 is around 4 dB noisier at low frequency.

Cheers

Phil Hobbs

Another graph opportunity! RF people can't seem to imagine their parts being used for anything but small-signal high-freq amplifiers.

The data sheet says the usual "adjust the gate voltage until you get the drain xx mA that we tested it at." We have to test these parts ourselves and trust they are repeatable. In the case of some MiniCircuits mmics, they weren't.

I suppose if you want to keep changing fabs, it's better to not commit much. "min" and "max" are so confining.

We might have an opportunity to collaborate on something like that...

It amazes me that a transistor can have so much less noise than a 50 ohm resistor.

Hi John, IOW: why 'loosely documented' parts cost lower than one might otherwise expect. They pass the qualification testing onto the customer. Oh, of course, the select class of major customers, they care most about - those who buy 1M+ units. For them MC will hold it to a tight spec, do the selection, etc. regards, RS

The John Doe troll stated the following in message-id <sdhn7c$pkp$ snipped-for-privacy@dont-email.me:

And the John Doe troll stated the following in message-id <sg3kr7$qt5$ snipped-for-privacy@dont-email.me:

And yet, the clueless John Doe troll has itself posted yet another incorrectly formatted USENET posting on Fri, 12 Nov 2021 14:59:10 -0000 (UTC) in message-id <smlvfu$7ls$ snipped-for-privacy@dont-email.me.

KGyA/A3mpGHx

You're making that up. The MCL parts cost 2x-3x more than the parts they're replacing.

Cheers

Phil Hobbs

Microdyne used at least 25 different amplifiers from MCL. I have a handful that had to be replaced on new boards, when they didn't perform like the prototype. I snagged them on their way to the scrap container for parts. I waas also given a couple pounds of SMD that were spilled from bagged inventory. I recently bought a pair of LCR tweezers, so iyt might be fun to see what was in there I just took them to use as dummy components for people to use t practice soldering by hand.

We're using an HMC659, at $330 each, and the data sheet is terrible.

Even expensive parts have dreadful data sheets.

I agree that high volume users usually get better support and no doubt have access to data that we peons don't. They probably get design assistance and free reviews too.

It looks to me like when the high volume uses go away, lots of parts get dropped, regardless of how much it hurts most users. TI is unusually good about keeping stuff available. Rochester has an interesting function here.

Hi Phil, I wasn't addressing you original topic, just the general state of spec sheets and MC. Indeed, the first 2 sentences were suppositions. The remainder was based on my 'emotional long-term commitment' (what we call a job) at that converted bowling alley. Re HK's pricing philosophy, and creation of part variants, etc. True, this could have changed since RK took over-- I haven't tried sourcing new parts from them for many years. cheers. RS