power supply trick

THE LATHE, aka: LS-76, and the LJ-10 and LJ-12 tape machines
Post Reply
User avatar
Site Admin
Posts: 86
Joined: 03 Apr 2013, 03:13

power supply trick

Post by Fonotec » 04 Apr 2013, 05:18

LS-76 power supplies take in 120 VAC from the wall and use a maximum of 600W in action. There's a modest multi-winding transformer which provides fractions of the line voltage for various regulation sections for the various circuits used at once. The brains of The Lathe are the cage cards with Logic chips (mostly Fairchild 74xx series gates, timers, and dividers) and they mostly require a very stable +5 V DC for their power source. The datasheets for these civilian chips specify a range of +4.75 through +5.25 VDC. So, for the first version of The Lathe, which is the main one I'm using (although with the updated tt servo control system and hidden vacuum tower) the logic supply section of the psu employs a 7805 regulator, which normally puts out a steady +5 VDC, pretty much on the nose, dissipating unusable energy as heat. And there's a heat sink bolted onto the tab of this regulator in the Scully psu to facilitate that, since it's being used to capacity and with a high input voltage, about midway within its tolerance (i.e., nearly +16 VDC).

There are seven 22-pin cage cards with more than 10 logic chips each, so more than 70 chips, in parallel, running on the same regulation. There is a fair amount of cable, but, when not energized, the circuit still measures very low resistance, from fuse to Vcc, or ground to ground. However, when "on," there's a 600 mV, diode-drop effect, caused by the parallel TTL loads. Scully therefore provided +5.6 V by putting a silicon diode on the ground pin of the +5.0 V regulator. So, unloaded the power supply "+5 V" test point measures, +5.6 V, but, as soon as the supply is turned on with the card cage snake (already) attached, the fuse in the test point measures +5.0 V, as it does, up in the cage.
Diode_Boost.jpg (48.09 KiB) Viewed 1037 times

On the updated LS-76 power supply, Scully selected, instead, a 7806 regulator. This doesn't need a silicon diode on pin 2 to make it higher, since it's already really too high in voltage for the 5V chips, even with the expected drop of 0.6 volts in the circuit. That leaves the voltage at Vcc, +5.4 V, which is fine for the military versions of these parts (e.g., 5400, instead of 7400), but not every part used has a military counterpart (anymore?), according to my internet searching so far.

I tried a Germanium diode in series with the output of pin 3 of the 7806, but the circuit would not allow the voltage to drop. The regulator just compensated, boosting the output back to 6. So, I removed the 7806 from the power supply of #660 and installed a new 7805 in the regulator position with an A14F silicon diode on pin 2 to ground, making the desired + 5.6 VDC, unloaded.

* The diode trick reminds me that the big island called, Hawai'i, has a volcano called, Mauna Kea, which would be the tallest mountain in the world if we were to measure it from the bottom, at the sea floor...

User avatar
Dr3d Sl3d
Site Admin
Posts: 43
Joined: 03 Apr 2013, 01:11
Location: All over the world...

Re: power supply trick

Post by Dr3d Sl3d » 20 Oct 2016, 03:19

The detail of the PSU drawing, below, shows the way in which a power transistor can boost a 1-Amp regulator's current, while maintaining the stable voltage. The transistor is a 2N 5871 (TO-3 package), with 7 Amps capability. Its base and emitter are connected to the diode-rectified and capacitor-smoothed current going to the regulator's input, and the collector is connected to the regulator's output, going to the fuse, and on to the card cage...

No diode voltage boost needed, here (unlike in the 7805 version of this supply), but the transistor current boost is a nice power supply trick, too.

(Slide pane all the way to the right in order to see the PNP transistor in the drawing.)
Bridge_To_Base_Amp.jpg (70.59 KiB) Viewed 1041 times

Dr3d Sl3d
Cappin' the Rap in the M&M's...

Post Reply