transistor bridge

[Bernius1]

Have any of you ever seen 4 transistors used in a
bridge rectifier-type arrgt. for differentiation or use as a comparator? I know there's a use there somewhere, but the permutations are a blur.
I figure the E-B junction, 4-ways around. But C-B junctions reversed might allow differential switching. Maybe this path is trod?

[greg123]

I used to use a 4 transistor package (16 pin dip) for a bridge rectifier and diff amp. Their junctions were almost always well matched.<p>Greg

[Bernius1]

THE TRANSISTOR ARRAY.. # SOMETHING LIKE MPS2907 ?
I ONCE SAW A BJT ARRAY WITH 6 OR 8 NPN's WITH TIED EMITTERS. I'M JUST BEING LAZY, BECAUSE I SEE THE ARRGT OF TWO TRANS'S IN A DIFF.AMP, AND I READ OF RAIL-TO-RAIL OP-AMP'S, AND I CAN FORESEE A USEFUL CIRCUIT; ABLE TO DIFFERENTIATE/COMPARE INPUTS, SWITCH BETWEEN OUTPUTS, AND AMPLIFY THE SUM OR DIFFERENCE OF THE RESULT WITH MINIMAL PARTS. BUT WHICH SEQUENCE OF E-B OR C-B JUNCTIONS IS CONFUSING. I JUST DIDN'T WANT TO WASTE MY TIME BREADBOARDING IF SOMEONE ELSE HAS ALREADY SUCEEDED.

[chessman]

Not sure if this is what you're looking for, but I do have a schematic for a motor H-Bridge with 8 transistors (4 darlington, 4 small-signal drivers).<p>Truth table:<p>input | output
A | B | A | B
---------------
0 0 | float
1 0 | 1 0
0 1 | 0 1
1 1 | 1 1<p>There are two output terminals, A and B, and they are used to drive a motor in this circuit. But as you can see, the truth table shows the outputs can be used as logic with reference to ground.<p>All the transistors could be replaced with two '3904s, and two '3906s, if the circuit is only to be used for small-signal applications.

[Chris Foley]

Is it possible we're talking about the venerable CA3046? The part's still around (booted from RCA to several sources, now made by Intersil and refuses to die) and consists of five NPN transistors with a common substrate, two of which are connected as a differential pair. The datasheet's available at...<p>http://www.intersil.com/data/fn/fn341.pdf<p>and for a good description of what you're talking about, I think you want to look at the old RCA app note in the RCA Linear book. I have a reference to AN5296, but I'm not sure if that's RCA (they used the prefix ICAN on all their appnotes, I think) or somebody else. It doesn't come up on the Intersil website.<p>Good luck.

[Bernius1]

First off, THANK YOU VERY MUCH!! All very helpful. I'm fascinated with logic. Flip-flop based arithmetic units are like a chess game. Right sequencing yields desired results. So I see on every schematic that lines can't cross like a roadway intersection, and yet have all 4 inputs and outputs as viable options,INCLUDING Vcc & Gnd. But it seems to me that a bridge of transistors might not only allow electrical 'intersections' to be controlled, but analog signals routed through them could be amplified. And (4) transistors X (2)types (npn..) X (2) junctions X (2) junction app's (fwd.or rev.bias) X (2)input/output options = 64 arrangements, only 1 of which might be my 'Grail'. But part-count reduction and data multiplexing options should be worth it. Opinions????

[hlreed]

There is an algebra for what you want to do.
It is Shannon's switching algebra. The variables are bit flows. Give them names and you make a variable. Let Out be output and I1 and I 2 be inputs. Then the algebra is:
Out = I1 AND I2 ; If I1 = I2 then Out = I1 else 0
Out = I1 OR I2 ; If I1 > I2 then Out = I1
Out = NOT I1 ; If I1 = 0 then Out = 1 else 0<p>The combinations of these make a binary structure that you can do logic with. Then it maps to your switches.
Good stuff.

[Ron H]

Harold, could you take another look at the OR function? It doesn't make sense to me.<p>Ron

[hlreed]

Ron,
Sorry, I use that for data stream selection.
Here is truth table.<p>OR
I1 I2 Out
0 0 0
0 1 1
1 0 1
1 1 1<p>Also, you can use any chip on the market as a node with this algebra.<p>[ February 07, 2003: Message edited by: Harold ]</p>

[Ron H]

Well, the truth table is what I expected, but it doesn't agree with this:<p>"Out = I1 OR I2 ; If I1 > I2 then Out = I1"<p>Am I missing something?<p>Ron

[hlreed]

Ron,
It was a mistake. No hidden meaning.
I shoulda said: Out = I1 OR I2 ;
Out = 1 if I1 or I2 = 1 else Out = 0<p>What I said before was really for integers, not bits. More XOR than OR. I don't work with bits any more.
I should have left off the explaination.<p>XOR is: Out = I1 XOR I2 ;
It is also; Out = I1 - I2
Its truth table is:<p>I1|I2|Out
0 0 0
0 1 1
1 0 1
1 1 0<p>[ February 08, 2003: Message edited by: Harold ]</p>

[chessman]

Good ol' XOR, famous in the world of microcontrollers. Very useful if you are checking a register for a certain value!!<p>variable XOR constant<p>the constant is the value you're checking for. The function will equal 0 if the two values are the same. Simply check the MCU's zero value flag.<p>No_Vice:<p>Do an internet search for Boolean Algebra. You should find everything you're looking for.

[Bernius1]

Chessman : Already have it. Dover books. If you look at the circuit for simultaneous multiplication of (2) 3-digit numbers, there are
roughly 12 input lines, and an asst'ment of adders & half adders. But the logic flow is uni-
directional. And you must CHOOSE between pull-up,
or pull-down, or an inverter. Because the bridge
has an 'intersection' characteristic, I assume
( yes, that's right, ASSUME) that it would enable
either cross-circuit logic, or simplified memory
matrix access, or could mimic a bilateral switch.
I saw the stepper motor ctrl'r. Thanks for the
input. But I think there's more. Your* input is
greatly appreciated.<p>*this word is plural in the Koine Greek.<p>Harold: Shannon's switching algebra? Any suggested books? I hear Border's calling me!!!

[hlreed]

I don't think Shannon' switching algebra made it into a book. However it started a whole discipline of switching theory. There is a web site devoted to Claude Shannon.
Shannon wrote this paper in 1938 and it was widely distributed among colleges.
Shannon took Boolean algebra and made the functions switches. What this allowed the designers of computers to do was to write down the equations before making any hardware.
Of course algebra does not care what the switches are made of.
Search for Shannon. Boolean algebra. Switching algebra. etc.
You might like my book,
Brains for Machines/Machines for Brains by
Harold L. Reed, Nova Science, 1996
It is still at bn.com at the lowest price or if you ask, I will send you one.<p>[ February 11, 2003: Message edited by: Harold ]</p>