To add this, because in a way - even the Test Bench stuff doesn't always show their cards either.
It deals with the power feeds into a test circuit...
Since the covers been blown on the fact that someone invented a part and die design that they feel is better than all the rest - but when placed in circuits prototyped with the various test jigs in mind - they fall flat. Because of direct trials in the real world they never did truly see forthcoming - we have empirical and physical as well as the FINANCIAL wasteland of evidence to prove they are not what they seem to be.
So dirty pool is played and suddenly eureka is shouted from the pulpits of a press meeting with some students and bystanders hired to say a few good things about this ...er - thing - and it works and works well (just don't ask them how - PLEASE!) Then they, in the Pubic Relations orifice - produce a PDF or two and offer free samples' of it - the test circuit, product support and ice cream to the first thousand vendors - whether they even sell transistors or not...nothing like free press...pressing on...
Not to bore you with repetitive diagrams, but this is something not too many people are really asking about...when it comes to the care and feeding of an Infant Transistor until they know how it will grow up and work out it's life as a functional block in society.
Power supply...
Yes, how do you feed a new part? Well, that requires a means to start out small and gain strength - usually with voltage and current steppers that allow the part to be examined and demonstrate it's stability in proper polarity of operation - including the die, wiring - stamping and epoxy along with the welds and leads being attached. Make sure it powers up and doesn't short right away...kinda' thing.
It may sound hilariously redundant but it is a quality of inception issue - you create the design and unless you see the physical model of it under test and passing all those trials. Variables exist, no matter what. Even the best model runs can't show every aspect - they just have to verify it won't blow up the table or create a quantum tunnel and suck all the air out of the room or introduce a gorilla of an unstable element from the place it contacted before we realized what it was doing and turned off the power to it.
Hi honey I'm home! What's for Dinner!?
GROWL! <SLASH>
"Next Week on Judge Judy...Paying the price for cheap groceries from a cheating cashier..."
Did anyone ever think about how to keep it happy in an environment it's not?
When I see those jigs, I can't help but wonder how the power supply itself is designed to isolate itself from the circuit but yet act as the provider?
Since we're looking at variables - I can pretty much guess they use a variable power supply and we too, should design these concepts using such a means to isolate the part from potential problems in power feed paths. Referring back to those diagrams - usually they use a bypass cap - you see that in each and every one of those test circuits. A feed thru cap of some value is used to filter out the externals for the benefit of providing clean pure clear power for the device to show it's true nature.
But, did anyone else notice the base region? How it not only receives power, but the amount of power it receives - like an expected, inherited, impedance - like a reactive component high enough in value to take it out of the equation.
So I am bringing this up because many a shortcoming in todays radios can be attributed to poorly designed power feeds in amplifier stages that added more of a source of noise and unfettered feedback paths from the power supply rail or the adjacent circuits - and they had to scramble to "band-aid" the problem.
We can make the part operate linearly and use an R1 and R2 ladder divider circuit in some fashion to provide the power source but then are we isolating the part and the power source from each other?
[ATTACH=full]24584[/ATTACH]
Yes, look back at the test circuit used - and in which configuration was it used in.
So, if we have an idea of the required power in voltage and current in a DC realm, we need to understand that we also, are dealing with an RF spectrum issue and we don't want anything but the RF we want going thru it - to be amplified. The part can amplify RF - but which RF do you want it to amplify, in the spectrum - is the study question...
This is very important, for in the realm of zero drain (extremely high-impedance to power flow) inputs where they are/appearing as, highly capacitive - we need to be prudent in the design of a power supply feeding system that maintains voltage and current to allow a current to flow thru and or a power level window and provide the input for sourcing but then remove the power not needed thru return - we don't want the part taking in too much power it is not designed to work with.
A good example of this is the MOSFET - might as well accept the fact that we are going to be dealing with these in several ways in both power supply and power RF delivery.
So we see these RF transistors as having a very streamlined die design and small surface to apply a signal to the central region of the die - we also have to remember that the speed of light and it's relativistic realm in RF is a finite speed and we can introduce problems and unwanted artifacts if we do not isolate the power supply with a level of isolation to reduce the interaction of the feed into the part from the parts' needed interaction with the signal it's trying to get in there to amplify.
Anyone whom has access to a Galaxy radio or even a CB radio that is using IRF MOSFET's for the Final - can see resistor divider networks that isolate the Gate as a divider from turning on and staying on - to being able to amplify but little else from the SOURCE as well as Input - but reflecting back (SIC) we also inject a SECONDARY problem of a power flow that can hinder the level, quality and ability of the signal to arrive at the destination. They added in resistors into the bias feed a SERIES sourcing but the voltage divider feeds this voltage into a higher impedance device - so the device sees power at a given impedance thru it's Gate region and the Signal itself sees this inclusively as a high impedance admittance.
In a basic sense we need to review, reexamine, our designs and then properly isolate the feeds from each other to keep unwanted feedback paths from appearing. So it looks like we'll need to reevaluate the expected single point impedance and yet use a divider to allow the injected signal a path to exit the region as well.
Someone designed a transistor - thought it was the best thing since sliced bread - but, in the real world - it sucked. So they then designed a test jig to make that part shine and look like it was the RF worlds savior - because if you add in all those extra parts, it functions like never before experienced in real life.
Oh really? Said the Secretary - as she busied herself with typing up yesterdays cliff notes about the bathroom stall issue...
:+> Andy <+:
