Talk:Regency TR-1

Germanium Transistors

Can anybody reading this shed light on why the TR-1 used germanium NPN transistors but a few years later virtually all germanium transistors were PNPs? BrianDGregory (talk) 14:02, 23 May 2010 (UTC)

That is an interesting question and 15 minutes of searching has not come up with an answer. I suspect it has something to do with grown junction transistors being replaced by alloy junction transistors - the Regency web site mentions in passing that the TI transistors were grown-junction type, which were expensive to produce. But i don't know enough about germanium transistor history to explain why the polarity switched. --Wtshymanski (talk) 15:04, 23 May 2010 (UTC)

And this [1] indicates the above is on the right track, thoguh not conclusive. --Wtshymanski (talk) 15:14, 23 May 2010 (UTC)

Thanks. You're a star. BrianDGregory (talk) 22:38, 27 May 2010 (UTC)

Hand-picked Transistors?

An article (http://www2.electronicproducts.com/What_s_It_Worth_Vintage_transistor_radios-article-fasj_worth_feb2012-html.aspx) says:

"Introduced in October of 1954, the Regency model TR-1 was the first widely marketed transistor radio. Developed by both Regency and Texas Instruments, the Regency TR-1 used four germanium transistors, which allowed the small 22.5-V battery to last for more than 20 hours. It differed from other prototypes in that due to its circuit design, hand selection of transistors for each individual radio was not necessary. This lead to higher production levels and lower production costs and made the price to the consumer affordable."

This contradicts the Wiki article.194.72.120.131 (talk) 08:32, 4 May 2012 (UTC)

As well as I know, grown-junction transistors are always hand-picked as otherwise the quality is too low. But yes, the TR-1 required them to be less hand picked. Getting them up to that frequency was barely possible, only with the high supply voltage and neutralization. They needed matching neutralizing capacitor, which was at least hand picked. Gah4 (talk) 16:19, 19 October 2020 (UTC)

Numbers

It would be interesting to add the total number produced, and when manufacture finished, if they are known.

Baska436 (talk) 23:30, 18 October 2012 (UTC)

 I was wondering the same thing myself.  — Preceding unsigned comment added by 98.82.127.122 (talk) 23:57, 18 October 2012 (UTC) 

Schematic

I've not seen this on WP, but for a device this old (and with performance issues), why not show the entire circuit schematic? Is this still patented? Seems like somebody out there knows exactly how it was constructed. There are no integrated circuits at the time. so why not? --2600:6C48:7006:200:B056:6066:1296:EF0B (talk) 00:53, 18 October 2020 (UTC)

The schematic is there, which is from the patent. The service manual is on the http://regencytr1.com/ web site. Gah4 (talk) 03:51, 18 October 2020 (UTC)

Operation voltage

Why used this radio such a high voltage and why use later-built transistor radios lower voltage levels? — Preceding unsigned comment added by 2003:DF:1F1A:A715:537:139:56BA:95C3 (talk) 14:11, 18 October 2020 (UTC)

The info is in the article. The transistors had a poor frequency response. By operating them at a high voltage (near to the max voltage), the depletion layer is larger and hence the parasitic (Miller) capacitance is lower. This reduces the Miller feedback and gives a better frequency response. As the battery ran down, the depletion layer would shrink, the capacitance would rise and the radio would oscillate with the increased feedback. This occurred at around 15 volts. If it was not for this, the transistors would operate at lower voltages. The IF transistors had to have their Miller capacitance neutralised (the two capacitors at the top of the schematic marked as 100-200μμfd).

More modern radios use better transistors and thus can operate from much lower voltages. 86.164.169.96 (talk) 17:51, 18 October 2020 (UTC)

Grown-junction transistors are not good at higher frequencies. Higher voltage helps, but they still need negative feedback for neutralization. (Neutralization was common on vacuum tube amplifiers, and brought over to transistors.) Note the range of feedback capacitors, which are chosen to go with each transistor. Gah4 (talk) 20:24, 18 October 2020 (UTC)

Er No! Neutralisation is positive feedback. The Miller effect is a negative feedback capacitor that reduced the gain of the stage at increasing frequencies, to the point that the gain could be less than unity at the desired frequency. The big problem is that the effective feedback capacitance is the actual capacitance multiplied by the transistor's physical gain making it a significant problem. The neutralising capacitor was a capacitor that fed back a signal in phase with the signal on the base (positive feedback, though you might not work this out from the schematic because the secondary is phase inverted w.r.t. the primary) and was designed such that the positive feedback more or less cancelled out the negative Miller capacitance feedback (this is why the capacitor had to be chosen to operate with its transistor - the collector to base capacitance varied between transistors). The technique worked reasonably well, but only up to a point.

Further, the first transistor (the frequency changer) was not neutralised. It relied solely on the large collector to base voltage to reduce the Miller capacitance to get the desired frequency response (up to around just 1.5&nbspMHz), though it did not do that well as the set had a relatively poor sensitivity compared to its contemporary vacuum tube receivers. 86.164.169.96 (talk) 12:25, 19 October 2020 (UTC)

Well described in Neutrodyne, it is negative feedback to reduce oscillation caused by positive feedback. But negative feedback at one frequency is positive feedback at another. The Miller effect is a positive feedback, capacitance from plate to grid in a triode. But then as described in Miller effect, it increases the input capacitance, which is negative feedback as seen by input signal. Gah4 (talk) 16:13, 19 October 2020 (UTC)

Wikipedia is never a reliable source of information on anything. If the Miller Effect was positive feedback at any point in the transistor's pass band it would increase the gain of the stage not reduce it and we wouldn't need to bother. The tetrode vacuum tube would never have been invented. In a circuit without a following transformer, the collector circuit always has to include a choke so that an in-phase signal (i.e. positive feedback) can be fed back to the base to counter the negative feedback. The neutralising capacitor is connected to the junction of the choke and a resistor (the choke being connected to the collector and the resistor to V_CC. Where there is a following transformer, the secondary can be phased to provide the positive feedback. Have a look at some older UHF tuner designs. These always required neutralisation and relied entirely on an inductor to provide the in-phase signal. 86.164.169.96 (talk) 17:30, 19 October 2020 (UTC)

Not disagreeing with the description, but the term neutralization implies undoing something, which in this case means negative feedback. The Miller effect, when added to the delay of electrons from the grid to the plate, gives positive feedback at some frequencies, causing oscillation. Neutralization undoes that. Or it cancels out the plate-grid capacitance with an opposing voltage on the grid. Gah4 (talk) 17:57, 19 October 2020 (UTC)

Check the schematic of the Regency TR-1 (the subject of this discussion), and point out to me any vacuum tubes it contains (Clue: it is a transistor radio). At the frequencies of interest (around 1.6 MHz and 262 kHz) the signal on the collector is in anti-phase to the signal of the base. The Miller capacitance feeds back this anti-phase signal. An in-phase signal is required to cancel it out - that is the neutralisation for this application.

I am aware at that at much higher frequencies various factors can make the signal in-phase (vacuum tubes being far worse than transistors in this regard) and a different form of neutralisation is required, but the Regency TR-1 contains no such devices, so your misdirection is irrelevant for the purpose of this discussion. 86.164.169.96 (talk) 14:36, 20 October 2020 (UTC)

not a forum

Yes this is not a forum for general discussion, but I try to give some benefit of the doubt, with some discussions leading to article improvements. Questions about transistor technology and how easy or hard it was to design and produce the TR-1 don't seem so far off to me. Also, they might lead to improvements in other pages. (Yes I know that isn't the idea, but it does happen.) Gah4 (talk) 20:30, 22 October 2020 (UTC)