Support document for the 1802 Membership card. Last updated Mar 20 2015. Edited by Herb Johnson, (c) Herb Johnson, except for content written by Lee Hart and others. Contact Herb at www.retrotechnology.com, an email address is on that page.
![[dtr4307]](DTR4307.jpg)
Lee Hart selected transistor packages with internal resistors, to save
space on the Membership Card boards. In his design the transistors
have series 22K to base, 47K from base to emitter. the T0-92 transistor
100mA collector current, 50V E-C breakdown.
Here's the data sheet for the FJN3307 NPN and Here's the data sheet for the FJN4303 PNP
There's no single standard for these among manufacturers. here's examples.
Fairchild refers to these as "resistor biased" transistors
TO-92: FJN3307 is NPN, FJN4307 is PNP - these are what Lee Hart chose
alternatives: possibly FJN3207, FJN3307
ON semicondutor calls these "Bias Resistor Transistors" or BRTs
R1 is the base resistor, R2 the b/e resistor.
SC-59 is the largest surface-mount package, then SOT-23, sOT-323
SC-59: MUN2134 is PNP, MUN2234 is NPN
SOT-23: MMUN2134L is PNP, MMUN2234L is PNP
Old part number for TO-92 DTC114E series
DT1XXZ XX is 13, 23, 14, etc, Z is E or Y or T
None are R1=22k and R2=47K
Toshiba calls them "Bias Resistor Built-in Transistors (BRT)"
Resistors are described as "base series" or "base-emitter"
SC-59A or TO-236MOD are the largest packages, then SC-70, SC-75
SC-59: RN1405 is NPN, RN2405 is PNP
naming convention suggest RN1XXX is NPN, RN2XXX is PNP, XXX is package and resistors
NXP calls them Transistors Switching - resistor biased
no singles available, all "reel"
old data sheets describe PDTC123ES series as NPN and TO-92
Rohm calls them "pre-Biased transistor arrays"
DTB113ZS for TO-92, DTB113ZS for SC-59
- Herb Johnson
Lee Hart: Luckily, the specs on these transistors are not critical. None of them is being used at anything close to 50v or 100ma. If you read the specs closely, you'll find that those internal resistors have something like a 2:1 spread in resistance values. A "22k" resistor could actually be anywhere from 15k to 30k!
[Herb says: Rohm....for 1K resistor, range is .7 to 1.3. That's ±30%.
ON semi....same range for 1K
Toshiba...same range
On semi ....15.4K 22K 28.6K, again same percentage]
Thus in *this* application, most of the oddball variations will still work. For example, Mouser currently has a version of the PNP with two 22k resistors, Fairchild 512-FJN4303RTA in stock for $0.062 each in quantities of 100.
The 22k/47k was cheap and available. That's all.
Let's see... how critical are they? I'm using all these transistors as simple switches. None of them needs to carry more than a few milliamps of collector current. The transistors have a gain of 50-100, so they only need tens of microamps of base current to do this. The base is being driven by 5v (and maybe more, from the RS-232 input). So the base series resistor can be as high as 5v/50ua = 100k before the transistor wouldn't turn on.
The base-emitter resistor is equally tolerant. It shunts off part of the base current. As long as it isn't too low, there will still be plenty to turn the transistor on. The difference between 22k and 47k is only 0.6v/47k = 12 microamps. That's insignificant compared to 5v/22k = 230ua the 22k base resistor supplies.
None of these transistors are being used to provide any analog timing. You shouldn't be able to detect any difference in operation even with much difference resistor values. You would notice that the threshold at which they switch would change. For instance, DTR might have to go to below 3v instead of 3.5v to cause a reset. But that shouldn't matter, because DTR is driven by digital 0v/5v or -12v/+12v signals.
- Lee hart
This page and edited content is copyright Herb Johnson (c) 2015. Lee Hart and others who produced the content edited, hold copyright to their produced content. Contact Herb at www.retrotechnology.com, an email address is available on that page..