From: Bil Herd <>
Date: Fri, 17 Feb 2012 01:56:07 -0500
Message-ID: <>
"sloppy" is a sloppy way of saying you have to take into account all of
the mins and maxes even when they aren't all printed in the spec sheet,
including gains, voltage drops, currents, temperatures, and operating
voltage, and then watch that you're not close to the tolerance of a
resistor getting in your way times the probability that more parts means
more parts stuffed wrong statistically plus higher failure rates and more
cost with more parts. All parts cost three times not including failures;
initial cost, insertion cost, pcb area cost. Failure modes may also be
semi-nefarious if parts are in linear region instead of all the way
switched (outside of SOA) , etc so harder to test for all conditions
including extended use by user, it could be barely working and still past
the test.  Also the testers were mostly digital so there were several
parts being tested by the one cassette output instead of being probed

In my case it was run a wire from an existing package, drive from a
testable/probable logic level, and pretty much know it's either on or off.


-----Original Message-----
[] On Behalf Of Jim Brain
Sent: Friday, February 17, 2012 1:38 AM
Subject: Re: CASSETTE MOTOR DRIVE Questions

On 2/17/2012 12:30 AM, Bil Herd wrote:
> Jim Brain asked:
> Questions:
>    * Why was the darlington pair needed with the 2n3904/2n2222 setups
>      not with the 2sc1815/7406/7416 setups?  Is the TIP29 that hard to
>      toggle?
> The TIP29 has the gain(hFE) of a small brick typically, as low as 15
> at current so I assume if I run the numbers that the 4401 is used to
> source more current when the TIP29 was on and to keep it in saturation
> at full current. If a small value resistor (higher current) was used
> to bias the TIP when on then the 2N3904 would have to sink too much
> current when the
> TIP29 was off to be practical. So it is hard to drive because it
> doesn't help that much by adding DC gain.
I can understand that if the 2n3904 was used to drive the TIP29, but the
TIP29 was biased on with a 1.5K resistor.  The 2n3904 merely drove the
base to ground (.6V or so), thus removing current from the base.
Obviously, it would need to sink the 7.5mA through the resistor
(11.4/1500), but where was the resit of the current required?
>    * The 2n3904/2n2222 and TIP29 are commodity.  Why did the VIC go with
>      more specific transistors?
> They were overseas sourced, if they used US transistors they would
> have had to ship them from US to Japan for stuffing, plus more
> expensive in general.  This was back when the Japanese were kings of
> the .5 cent transistor.
Ah, that makes sense.
> I didn't like the interaction of two transistors by the time you run
> min/max gains on both it felt a little sloppy, I designed it out on
> TED/Plus4 to use my favorite transistor... the output stage of an
> '06/'07
Yep, I noticed that, and when I saw it first on the C128, I went to look
at the 264 series, as I assumed you brought that over from the +4 design.
Can you expand on your "sloppy" note, though?

       Message was sent through the cbm-hackers mailing list

       Message was sent through the cbm-hackers mailing list
Received on 2012-02-17 07:00:31

Archive generated by hypermail 2.2.0.