Fixer Upper

Discussion of rare things not covered in other fora.
Forum rules
Only taste in good post.
Post Reply
User avatar
Fonotec
Site Admin
Posts: 105
Joined: 03 Apr 2013, 03:13

Fixer Upper

Post by Fonotec »

Just took delivery from an eBay vendor of a good condition B&K 1760 triple output laboratory power supply. I like this model a lot because it has a logic power supply with more than 5V available, variably. Most supplies with a spare section for powering logic offer only 5.0 V (fixed) for that output. Having a little more potential than normal is crucial for the design of the LS-76 (THE LATHE) logic circuit that powers many chips that are used to control three motors (turntable, feedscrew, and depth solenoid) and provide metering and automation manual override. In addition to motion control, there are audio sampling circuits which use 8 bit / 20 kHz Fs converters (bass loudness requires coarser feed than that of treble).

Although the circuit feeds the chips their Vcc signal and grounds in parallel, so voltage should remain constant, there is a lot of cable and a slow-blow 4A fuse which create a significant resistance under load which drops the regulated output by about 600 mV. (Civilian TTL logic chip Vcc potential has to be between +4.75 and +5.25 V for reliable use. But 5.0 V - 0.6 V = 4.4 VDC, which would be under the device's minimum Vcc by more than 300 mV. So, boosted regulation is the simplest way to achieve post-drop compliance to design tolerance.) Since the circuit chips are mostly obsolete, the inventory of spare parts might last longer under tightest regulation, so the B&K lab supply's "5 V" section will be inserted for the lathe brains power and that of the tt servo, via the main A and B sections of the 1760.

The 600 mV drop in the Vcc between regulation and destination is why L. J. Scully put an A14F diode between pin 2 and ground of the 7805 regulator. The original circuit forces the 7805 output into measuring 5.6 V above ground, which is necessary to deliver 5.0 V (after the drops) to the Vcc pin of the many chips (over 70 pieces). The silicon diode on the regulator's ground pin trick is like putting the regulator's output pin on platform shoes. It stands taller without being different from any other 7805 output... A later version of the power supply used a 7806 regulator, but the 6.0 V output only drops to +5.4 V, which is 150 mV over specification (since +5.25 V is the maximum recommendation for Vcc in civilian chips, and many of the chips used in the card cage are not available in the military 5400 series). I have a new version of this supply, but I replaced its 7806 regulator with a 7805 plus A14F diode on pin 2, just like the original psu.

Also, the B&K 1760 only has 2 Amperes for each of the main supplies (which is plenty for the +18 and -18 VDC used by the tt servo drive board), but the B&K 1760 complimentary "5 V" section, which is adjustable from +4 to +6.5 V, provides up to 5 Amperes*. The L. J. Scully logic supply connects a 2N5871 (PNP) power transistor transverse to the regulator (the 5.6 V output connected to both the 4 Ampere slow-blow fuse and the collector of the pnp - its base and emitter connected with resistors to the same bridge rectifier and caps as for the regulator). This boosts current on demand above what the 7805 could provide on its own.

Anyway, this nice lab supply had a little bug when I fired it up for the first time. Although the Fluke DMM showed stable voltages from all three supplies, the B supply's 7-segment numerical display was "hunting." Its reading kept changing, even though the voltage output was steady and corresponded normally to the deflection of the volts pot. I opened the lid and noticed that the 470 uF electrolytic capacitor on that display's daughterboard had exploded and was revealing its inner fur...


Image


I heated up my desoldering station and looked for a replacement cap. I'm glad I keep a bunch of components in my shop - the new one with over twice the voltage tolerance and a compatible form, with radial leads.

Image

Works.


* One Ampere is the equivalent of one Coulomb per second. For conventional flow analysis, one Coulomb is equivalent to 6.24 x 10/\18 electron holes. One Coulomb flowed through one Ohm during one Second = 1 Joule.
Post Reply