Good news. It sounds like these parts are popular enough to add new
packages, which suggests they may be available for a while. Several
other sources of phemts have discontinued them.
Less inductance is good.
https://www.dropbox.com/s/qttp09vpw3kcee3/SAV541_Xray.jpg?raw=1
We have a new xray machine that can resolve wirebonds. I'll get some
pics of the new versions, the entire sample kit.
Reply by Phil Hobbs●October 29, 20232023-10-29
On 2023-10-26 18:54, Joe Gwinn wrote:
> On Fri, 27 Oct 2023 01:19:18 +0300, Dimiter_Popoff <dp@tgi-sci.com>
> wrote:
>
>> On 10/26/2023 23:13, john larkin wrote:
>>> On 26 Oct 2023 19:54:19 GMT, Uwe Bonnes
>>> <bon@hertz.ikp.physik.tu-darmstadt.de> wrote:
>>>
>>>> john larkin <jl@650pot.com> wrote:
>>>>> https://www.minicircuits.com/WebStore/dashboard.html?model=TAV1-541NM%2B&utm_campaign=Product%20Focus%20Emails%202023%20%26%20Onward&utm_medium=email&_hsmi=279821766&_hsenc=p2ANqtz-9MbcHJHjYg6_2uzVDHSL13EsHrk1Q6ZhSHQEwOQSif6Y1l55GNDwI_YEWiJGP8nVJdlmIf1bwj6kW8ZaQVj8XtIGwlLgVa4tPz5ednrtuuRQiRkNY&utm_content=279665588&utm_source=hs_email
>>>>>
>>>>> MiniCircuits calls these both MMICs and Transistors, but they sure
>>>>> look like bare transistors.
>>>>>
>>>>> The 45 MHz lower frequency limit is strange.
>>>>>
>>>>> THey do have DC curves, which is refreshing in an RF part.
>>>>>
>>>>> Rds-on, seldom specified for an RF part, looks like about 5 ohms.
>>>>>
>>>>> I wish they had Spice models.
>>>>>
>>>>
>>>> I do not find the DC curves in the datasheet. Can you give a hint?
>>>
>>> Data sheet page 3. That suggets operation down to DC.
>>>
>>> The 45 MHz might be caused by coupling caps on the eval board or
>>> something. I want to use parts like these as switches.
>>>
>>> Some MMICs have internal bias circuits that indeed wreck the low
>>> frequency response.
>>>
>>
>> They do look interesting as switches indeed.
>> Are they JFET-s? That gate-source +0.7V max and the max gate
>> current suggest sort of that, although 0.7 looks too high.
>> I have biased (during wrestling this or that, not in an end
>> product IIRC) JFETS slightly above 0 G-S to get even less
>> channel resistance but they must have brought that to a
>> new level somehow.
>
> They are not JFETS. MiniCircuits has a good summary.
>
> .<https://blog.minicircuits.com/mmic-technologies-pseudomorphic-high-electron-mobility-transistor-phemt/>
>
> People have observed that pHEMTs seem to have relatively low flicker
> noise for a GHz amplifier. My guess is that this is because very
> clean materials and processes are needed to make a pHEMT.
>
> Joe Gwinn
>
The ones I've measured have 1/f corners of 10-50 MHz. With a flatband
noise of 0.3 nV in 1 Hz, that gets you about 1-nV noise at 1 MHz, not
awful at all.
They're great for bootstraps, as long as you bootstrap their drains as
well--otherwise they make very disappointing followers.
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.nethttp://hobbs-eo.com
Those are the same as the SAV series, but in a different package. The
$70 sample kit has both kinds.
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.nethttp://hobbs-eo.com
Reply by John Larkin●October 27, 20232023-10-27
On Fri, 27 Oct 2023 08:04:35 -0400, legg <legg@nospam.magma.ca> wrote:
PIN diodes. Interestingly also specified to 6 GHz max. The phemt has a
funny round 5th pin of unknown function.
I've tried using PIN diodes in time domain with expensively bad
results. Some RF parts are usable in time domain, some aren't. The RF
boys just ignore that market.
Reply by legg●October 27, 20232023-10-27
On Thu, 26 Oct 2023 10:19:50 -0700, john larkin <jl@650pot.com> wrote:
Package size and pinout is roughly equivalent to HP 'minipak'
used for HMPP-3890/92/95.
RL
Reply by John Larkin●October 27, 20232023-10-27
On Fri, 27 Oct 2023 03:06:22 -0000 (UTC), Phil Hobbs
<pcdhSpamMeSenseless@electrooptical.net> wrote:
>john larkin <jl@650pot.com> wrote:
>> On Fri, 27 Oct 2023 01:19:18 +0300, Dimiter_Popoff <dp@tgi-sci.com>
>> wrote:
>>
>>> On 10/26/2023 23:13, john larkin wrote:
>>>> On 26 Oct 2023 19:54:19 GMT, Uwe Bonnes
>>>> <bon@hertz.ikp.physik.tu-darmstadt.de> wrote:
>>>>
>>>>> john larkin <jl@650pot.com> wrote:
>>>>>> https://www.minicircuits.com/WebStore/dashboard.html?model=TAV1-541NM%2B&utm_campaign=Product%20Focus%20Emails%202023%20%26%20Onward&utm_medium=email&_hsmi=279821766&_hsenc=p2ANqtz-9MbcHJHjYg6_2uzVDHSL13EsHrk1Q6ZhSHQEwOQSif6Y1l55GNDwI_YEWiJGP8nVJdlmIf1bwj6kW8ZaQVj8XtIGwlLgVa4tPz5ednrtuuRQiRkNY&utm_content=279665588&utm_source=hs_email
>>>>>>
>>>>>> MiniCircuits calls these both MMICs and Transistors, but they sure
>>>>>> look like bare transistors.
>>>>>>
>>>>>> The 45 MHz lower frequency limit is strange.
>>>>>>
>>>>>> THey do have DC curves, which is refreshing in an RF part.
>>>>>>
>>>>>> Rds-on, seldom specified for an RF part, looks like about 5 ohms.
>>>>>>
>>>>>> I wish they had Spice models.
>>>>>>
>>>>>
>>>>> I do not find the DC curves in the datasheet. Can you give a hint?
>>>>
>>>> Data sheet page 3. That suggets operation down to DC.
>>>>
>>>> The 45 MHz might be caused by coupling caps on the eval board or
>>>> something. I want to use parts like these as switches.
>>>>
>>>> Some MMICs have internal bias circuits that indeed wreck the low
>>>> frequency response.
>>>>
>>>
>>> They do look interesting as switches indeed.
>>> Are they JFET-s? That gate-source +0.7V max and the max gate
>>> current suggest sort of that, although 0.7 looks too high.
>>> I have biased (during wrestling this or that, not in an end
>>> product IIRC) JFETS slightly above 0 G-S to get even less
>>> channel resistance but they must have brought that to a
>>> new level somehow.
>>
>> These are phemts, which are enhancement mode parts that turn on at a
>> few tenths of a volt. Capacitances are crazy low. We use them as
>> switches, typically to discharge a timing ramp capacitor or some such.
>>
>> I have taken a lot of data on their SAV541 and SAV551 parts, which
>> look very similar. I can post that if anyone is interested. The SAVs
>> have wire bonds inside; maybe these don't.
>>
>> I did find remarkable effects if they are driven past suggested data
>> sheet limits. They just keep enhancing!
>>
>> The Supertex depletion-mode mosfets enhance some too.
>>
>>
>
>Any and all measurements welcome!
>
>Thanks
>
>Phil Hobbs
Here's some random measurements.
https://www.dropbox.com/scl/fo/944svmozmwh4cfploqhyv/h?rlkey=gjueh508twp483vq20dlkeaq1&dl=0
In another case, I pulsed the gate positive at a low duty cycle, and
it just kept conducting harder. I went to +1.2 volts. That is probably
long-term bad news, but they sure switched hard.
The models are based on your models, tweaked a bit per measurements.
The xray shows the geometry and wire bonds.
My Colpotts oscillator kept doing weird things, probably GHz
oscillations, which was fixed with a *500 ohm* gate resisor.
Reply by Phil Hobbs●October 27, 20232023-10-27
john larkin <jl@650pot.com> wrote:
> On Fri, 27 Oct 2023 01:19:18 +0300, Dimiter_Popoff <dp@tgi-sci.com>
> wrote:
>
>> On 10/26/2023 23:13, john larkin wrote:
>>> On 26 Oct 2023 19:54:19 GMT, Uwe Bonnes
>>> <bon@hertz.ikp.physik.tu-darmstadt.de> wrote:
>>>
>>>> john larkin <jl@650pot.com> wrote:
>>>>> https://www.minicircuits.com/WebStore/dashboard.html?model=TAV1-541NM%2B&utm_campaign=Product%20Focus%20Emails%202023%20%26%20Onward&utm_medium=email&_hsmi=279821766&_hsenc=p2ANqtz-9MbcHJHjYg6_2uzVDHSL13EsHrk1Q6ZhSHQEwOQSif6Y1l55GNDwI_YEWiJGP8nVJdlmIf1bwj6kW8ZaQVj8XtIGwlLgVa4tPz5ednrtuuRQiRkNY&utm_content=279665588&utm_source=hs_email
>>>>>
>>>>> MiniCircuits calls these both MMICs and Transistors, but they sure
>>>>> look like bare transistors.
>>>>>
>>>>> The 45 MHz lower frequency limit is strange.
>>>>>
>>>>> THey do have DC curves, which is refreshing in an RF part.
>>>>>
>>>>> Rds-on, seldom specified for an RF part, looks like about 5 ohms.
>>>>>
>>>>> I wish they had Spice models.
>>>>>
>>>>
>>>> I do not find the DC curves in the datasheet. Can you give a hint?
>>>
>>> Data sheet page 3. That suggets operation down to DC.
>>>
>>> The 45 MHz might be caused by coupling caps on the eval board or
>>> something. I want to use parts like these as switches.
>>>
>>> Some MMICs have internal bias circuits that indeed wreck the low
>>> frequency response.
>>>
>>
>> They do look interesting as switches indeed.
>> Are they JFET-s? That gate-source +0.7V max and the max gate
>> current suggest sort of that, although 0.7 looks too high.
>> I have biased (during wrestling this or that, not in an end
>> product IIRC) JFETS slightly above 0 G-S to get even less
>> channel resistance but they must have brought that to a
>> new level somehow.
>
> These are phemts, which are enhancement mode parts that turn on at a
> few tenths of a volt. Capacitances are crazy low. We use them as
> switches, typically to discharge a timing ramp capacitor or some such.
>
> I have taken a lot of data on their SAV541 and SAV551 parts, which
> look very similar. I can post that if anyone is interested. The SAVs
> have wire bonds inside; maybe these don't.
>
> I did find remarkable effects if they are driven past suggested data
> sheet limits. They just keep enhancing!
>
> The Supertex depletion-mode mosfets enhance some too.
>
>
Any and all measurements welcome!
Thanks
Phil Hobbs
--
Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC /
Hobbs ElectroOptics Optics, Electro-optics, Photonics, Analog Electronics
Reply by Joe Gwinn●October 26, 20232023-10-26
On Fri, 27 Oct 2023 01:19:18 +0300, Dimiter_Popoff <dp@tgi-sci.com>
wrote:
>On 10/26/2023 23:13, john larkin wrote:
>> On 26 Oct 2023 19:54:19 GMT, Uwe Bonnes
>> <bon@hertz.ikp.physik.tu-darmstadt.de> wrote:
>>
>>> john larkin <jl@650pot.com> wrote:
>>>> https://www.minicircuits.com/WebStore/dashboard.html?model=TAV1-541NM%2B&utm_campaign=Product%20Focus%20Emails%202023%20%26%20Onward&utm_medium=email&_hsmi=279821766&_hsenc=p2ANqtz-9MbcHJHjYg6_2uzVDHSL13EsHrk1Q6ZhSHQEwOQSif6Y1l55GNDwI_YEWiJGP8nVJdlmIf1bwj6kW8ZaQVj8XtIGwlLgVa4tPz5ednrtuuRQiRkNY&utm_content=279665588&utm_source=hs_email
>>>>
>>>> MiniCircuits calls these both MMICs and Transistors, but they sure
>>>> look like bare transistors.
>>>>
>>>> The 45 MHz lower frequency limit is strange.
>>>>
>>>> THey do have DC curves, which is refreshing in an RF part.
>>>>
>>>> Rds-on, seldom specified for an RF part, looks like about 5 ohms.
>>>>
>>>> I wish they had Spice models.
>>>>
>>>
>>> I do not find the DC curves in the datasheet. Can you give a hint?
>>
>> Data sheet page 3. That suggets operation down to DC.
>>
>> The 45 MHz might be caused by coupling caps on the eval board or
>> something. I want to use parts like these as switches.
>>
>> Some MMICs have internal bias circuits that indeed wreck the low
>> frequency response.
>>
>
>They do look interesting as switches indeed.
>Are they JFET-s? That gate-source +0.7V max and the max gate
>current suggest sort of that, although 0.7 looks too high.
>I have biased (during wrestling this or that, not in an end
>product IIRC) JFETS slightly above 0 G-S to get even less
>channel resistance but they must have brought that to a
>new level somehow.
On Fri, 27 Oct 2023 01:19:18 +0300, Dimiter_Popoff <dp@tgi-sci.com>
wrote:
>On 10/26/2023 23:13, john larkin wrote:
>> On 26 Oct 2023 19:54:19 GMT, Uwe Bonnes
>> <bon@hertz.ikp.physik.tu-darmstadt.de> wrote:
>>
>>> john larkin <jl@650pot.com> wrote:
>>>> https://www.minicircuits.com/WebStore/dashboard.html?model=TAV1-541NM%2B&utm_campaign=Product%20Focus%20Emails%202023%20%26%20Onward&utm_medium=email&_hsmi=279821766&_hsenc=p2ANqtz-9MbcHJHjYg6_2uzVDHSL13EsHrk1Q6ZhSHQEwOQSif6Y1l55GNDwI_YEWiJGP8nVJdlmIf1bwj6kW8ZaQVj8XtIGwlLgVa4tPz5ednrtuuRQiRkNY&utm_content=279665588&utm_source=hs_email
>>>>
>>>> MiniCircuits calls these both MMICs and Transistors, but they sure
>>>> look like bare transistors.
>>>>
>>>> The 45 MHz lower frequency limit is strange.
>>>>
>>>> THey do have DC curves, which is refreshing in an RF part.
>>>>
>>>> Rds-on, seldom specified for an RF part, looks like about 5 ohms.
>>>>
>>>> I wish they had Spice models.
>>>>
>>>
>>> I do not find the DC curves in the datasheet. Can you give a hint?
>>
>> Data sheet page 3. That suggets operation down to DC.
>>
>> The 45 MHz might be caused by coupling caps on the eval board or
>> something. I want to use parts like these as switches.
>>
>> Some MMICs have internal bias circuits that indeed wreck the low
>> frequency response.
>>
>
>They do look interesting as switches indeed.
>Are they JFET-s? That gate-source +0.7V max and the max gate
>current suggest sort of that, although 0.7 looks too high.
>I have biased (during wrestling this or that, not in an end
>product IIRC) JFETS slightly above 0 G-S to get even less
>channel resistance but they must have brought that to a
>new level somehow.
These are phemts, which are enhancement mode parts that turn on at a
few tenths of a volt. Capacitances are crazy low. We use them as
switches, typically to discharge a timing ramp capacitor or some such.
I have taken a lot of data on their SAV541 and SAV551 parts, which
look very similar. I can post that if anyone is interested. The SAVs
have wire bonds inside; maybe these don't.
I did find remarkable effects if they are driven past suggested data
sheet limits. They just keep enhancing!
The Supertex depletion-mode mosfets enhance some too.