PSU Design

Seriously? I'm surprised there's no mention of this in any of the Keighley books. What sort of cap, though?

Polystyrene, polycarb, or some other I forget. Mylar's DA would be a nuisance.

Somebody, maybe they posted here, measured the self-discharge of a good film cap. He got numbers like a couple per cent per year. And even that could be mathed out.

1 pA into 1 nF is a millivolt per second. There are 86K seconds in a day.

Tek designed and built their own transformers. And CRTs.

Tek's schematics ca 1971 were brilliant and beautiful and fun.

Imagine a draftsperson at Philips or Siemens or Oxford trying something like that. Or even HP.

Sloman is a uniformly sour and sad old git. Ignore him.

Oxford Instruments came a bit later. Philips were Dutch and Siemens was German - they took themselves seriously, at least in their documentation. The Philips brother were nephews of Karl Marx. though he wasn't much of an uncle to them.

And I don't have any enthusiasm for flattering John Larkin. That makes me a very sour git in his book.

Pity he doesn't post anything that might deserve positive comment.

You need a sufficiently-good readout amp and reset circuit. We talked about that a bit in the 'femtoampere' thread just lately.

Cheers

Phil Hobbs

Polystyrene, polyprop, or (especially) Teflon.

"Okay guys, we have a thousands of these transformers we designed for the previous series and never used. We need to get rid of 'em on the

7623A, but they don't have enough secondary windings. See if you can come up with an overly complex and convoluted design to overcome this limitation. The designer who comes up with the most confusing and tortuous configuration using the most hard-to-find components will get a raise."

Bill and I are around the same age. He seems to have lost his sense of humour over the years. I've clung on to mine as a survival mechanism to cope with the f***ed-up world we live in today. If I ever became humourless, I'd be just like Bill. Only nothing like as good at electronics. ;-)

The setup would be a metal box with the parts inside some mechanical means to connect the readout circuit to the cap. That shouldn't be hard. The readout thing could be grounded or maybe precharged when it's not connected to the cap.

A reed relay might work, but that has hazards. Reeds are nasty.

Cursitor Doom prefers not to get the jokes I make at his expense.

He doesn't realise quite how comical his gullible devotion to climate change denial propaganda really is.

But much dimmer. And since he doesn't get my jokes at his expense, I don't find his sense of humour all that impressive.

I know a few people who are at least as good at electronics as I am, but lots more who are merely pretty good.

Not in my experience. Dry reeds do bounce, but that's about it .

This Carey Vibrating Reed Electrometer only seems to have been made from 1975 to 1980. We had one in Melbourne when I was a graduate student (1963-1969) so it must have been an earlier model.

I don't think you have anything to worry about, Bill. what you lack in humility you more than make up for in conceit. ;-)

I charitably assume that's an actual quote from a Tek engineer, and that you didn't make that up.

Did you ever work with any ex-Tek engineeers? I have worked with two.

The world is a pretty good place and by most indications keeps getting better.

Not sure I understand why you think there aren't enough secondary windings? There is one for -/+50, one for -/+15 one for +5 and one for

80-90 which added to +50 made 130V. Seems like plenty, how would more windings make it simpler? The 130V regulation occurring at 50V is economical, other HV supplies of that era used that trick too.

piglet

Here is the service manual extract...

LOW-VOLTAGE POWER SUPPLY The Low-Voltage Power Supply circuit provides the operating power for this instrument from six regulated supplies. Electronic regulation is used to provide stable, low-ripple output voltages. Each supply (except the -i 130 V supply, which is fused) contains a short-protection circuit to prevent instrument damage if a supply is inadvertently over-loaded or shorted to ground. Fig. 3-22 shows a detailed block diagram of the Low-Voltage Power Supply circuit. A schematic of this circuit is shown on diagram 8 at the rear of this manual.

Power Input Power is applied to the primary of transformer T801 through line fuse F1000, thermal cutout S1000, and POWER switch S1001. The Voltage-Selector Jumper, P1001, connects the two halves of the primary of T801 in parallel for 110-volt (nominal) operation. Voltage-Selector Jumper P1002 connects the two halves of the primary in series for 220-volt (nominal) operation. The line fuse, F1000, must be changed to provide the correct protection for 220-volt nominal operation.

Each half of the primary of T801 has taps above and below the 110-volt (220-volt) nominal point. When the Voltage Selector Jumper is moved from LOW to MED to HI, more turns are effectively added to the primary winding and the turns ratio is decreased to compensate for the increased primary voltage. This configuration extends the regulating range of the 7623.

For the R7623, a fan provides forced-air cooling. The fan is connected in parallel with one half of the primary winding of T801. Therefore, it always has the same voltage applied regardless of the position of the Voltage-Selector Jumper.

Thermal cutout S1000 provides thermal protection for this instrument. If the internal temperature of the instrument exceeds a safe operating level, S1000 opens to interrupt the applied power. When the temperature returns to a safe level, S1000 automatically closes to re-apply the power.

-50-Volt Supply The following discussion includes the description of the

50 V Rectifier, 50 V Series Regulator, -50 V Feedback Amplifier, - 50 V Reference, and -50 V Current Limiting stages. Since these stages are closely related in the operation of the -50-volt regulated output, their performance is most easily understood when discussed as a unit.

The 50 V Rectifier assembly CR808 rectifies the output at the secondary of T801 to provide the unregulated voltage source for both the —50- and +50-volt supplies. CR808 is connected as a bridge rectifier and its output is filtered by C808-C809. Transistors Q886, Q896, Q900 operate as a feedback-stabilized regulator circuit to maintain a constant -50 volt output level. Q886 is connected as a differential amplifier to compare the feedback voltage at the base of Q886B against the reference voltage at the base of Q886A. The error output at the collector of Q886B reflects the difference, if any, between these two inputs. The change in error-output level at the collector of Q886B is always opposite in direction to the change in the feedback input at the base of Q886B (out of phase).

Zener diode VR890 sets a reference level of about —9 volts at the base of Q886A. A feedback sample of the output voltage from this supply is connected to the base of Q886B through divider R880-R881-R882. R881 in this divider is adjustable to set the output level of this supply. Notice that the feedback voltage to this divider is obtained from a line labeled -50 V Sense. Fig. 3-23 illustrates the reason for this configuration. The inherent resistance of the interconnecting wire between the output of the -50-Volt Supply and the load produces a voltage drop which is equal to the output current multiplied by the resistance of the interconnecting wire. Even though the resistance of the wire is small, it results in a substantial voltage drop due to the high output current of this supply. Therefore, if the feedback voltage were obtained ahead of this drop, the voltage at the load might not maintain close regulation. However, the -50 V Sense feedback configuration overcomes this problem since it obtains the feedback voltage from a point as close as practical to the load. Since the current in the —50 V Sense line is small and constant, the feedback voltage is an accurate sample of the voltage applied to the load.

Regulation occurs as follows: If the output level of this supply decreases (less negative) due to an increase in load, or a decrease in input voltage (as a result of line voltage changes or ripple), the voltage across divider R880- R881-R882 decreases also. This results in a more positive feedback level at the base of Q886B than that established by the —50 V Reference stage at the base of Q886A. Since the transistor with the more positive base controls the conduction of the differential amplifier, the output current at the collector of Q886B increases. This increase in output from Q886B allows more current to flow through Q896 and Q900 to result in increased conduction of - 50 V Series Regulator Q903. The load current increases and the output voltage of this supply also increases (more negative). As a result, the feedback voltage from the —50 V Sense line increases and the base of Q8868 returns to the same level as the base of Q886A. Similarly, if the output level of this supply increases (more negative), the output current of Q886B decreases. The feedback through Q896 and Q900 reduces the conduction of the -50 V Series Regulator to decrease the output voltage of this supply.

—50 Volts adjustment R881 determines the divider ratio to the base of Q886B and thereby determines the feedback voltage. This adjustment sets the output level of the supply in the following manner: If R881 is adjusted so the voltage at its variable arm goes less negative (closer to ground), this appears as an error signal at the base of Q886B. In the same manner as described previously, this positive-going change at the feedback input of the differential amplifier increases the conduction of the -50 V Series Regulator to produce more current to the load, and thereby increase the output voltage of this supply. This places more voltage across divider R880-R881-R882 and the divider action returns the base of Q886B to about - 9 volts. Notice that the feedback action of this supply forces a change in the output level which always returns the base of Q886B to the same level as the base of Q886A. In this manner, the output level of the —50-Volt Supply can be set to exactly —50 volts by correct adjustment of R881.

The —50 V Current Limiting stage Q908-Q909 Q910 protects the - 50-Volt Supply if excess current is demanded from this supply. All of the output current from the —50 Volt Supply flows through R903. Transistor Q908 senses the voltage at the collector of the -50 V Series Regulator Q903 and compares it against the -50 V output level at the base of Q909 which is obtained from the other side of R903. Under normal operation, Q908 is held in conduction and Q909 is off. However, when excess current is demanded from the -50 V Series Regulator due to a short circuit or similar malfunction at the output of this supply, the voltage drop across R903 increases until the base of 0908 goes more negative than the level at the base of Q909. Then Q909 takes over conduction of the comparator. The collector current of Q909 increases the voltage drop across R896 to reduce the conduction of Q896 in the —50 V Feedback Amplifier and limit the conduction of Q903. Q910 is connected as a constantcurrent source for Q908-Q909.

—15-Volt Supply Basic operation of all stages in the —15-V Supply is the same as for the —50-Volt Supply. Reference level for this supply is established by divider R945-R946 between ground and the —50 V Sense voltage. The divider ratio of R945-R946 sets a level of -15 volts at the base of Q943A. The level on the 50 V Sense line is held stable by the

-50-Volt Supply as described previously. The -15 V Sense voltage is connected to the base of Q943B through R940. Any change at the output of the —15-Volt Supply appears at the base of Q943B as an error signal. The output voltage is regulated in the same manner as described for the

-50-Volt Supply.

+5-Volt Supply Basic operation of the +5-Volt Supply is the same as described for the previous supplies. The +5 V Current Limiting and +5 V Feedback Amplifier (except for Q985) is made up of a five-transistor array U973. Notice that both U973C and Q985 in the +5 V Feedback Amplifier are connected as emitter followers, since inversion is not necessary in the feedback path for positive output voltages. Reference voltage for the +5 V Feedback Amplifier stage is established by divider R970-R971 between the +5 V Sense and - 50 V Sense feedback voltages. This divider establishes a quiescent level of about 0 volt at the base of U973E.

• 15-Volt Supply

The + 15-Volt Supply operates in the same manner as described for the previous supplies. The unregulated

• 15 Volt Supply provides the source voltage for the High-Voltage Oscillator stage in the CRT circuit through fuse F814 and P870.

+50-Volt Supply Operation of the +50-Volt Supply is the same as described for the previous supplies. The unregulated +50 volts, from 50 V Rectifier CR808, is used to provide a positive starting voltage for the -50-Volt Supply.

• 130-Volt Supply The + 130-V Rectifier CR806 provides the rectified voltage for the + 130-Volt Supply. However, this secondary winding of T801 does not supply the full potential necessary to obtain the +130-volt output level. To provide the required output level, the +50-Volt Supply is connected in series with this supply through Q850. Basic regulation of the output voltage is provided by +130 V Feedback Amplifier Q852, and +130 V Series Regulator Q850.

The output voltage of this supply is connected across divider R855 R856. This divider provides a quiescent level of about +50 volts at the base of Q852. The reference level for this supply is provided by the +50-Volt Supply connected to the emitter of Q852. If the output of this supply changes, this change is sensed by Q852 and an amplified error signal is connected to the base of Q850. This error signal changes the conduction of the +130 V Series Regulator Q850 to correct the output error. Fuse F855 protects this supply if the output is shorted. However, since the response time of F855 is slow to a shorted condition, VR851 provides additional current to the base of Q850 to protect it from damage due to over voltage. Diode CR852 limits the reverse bias on Q852 to about 0.6 volt when F855 is blown.

Graticule Light Supply Power for the graticule lights is supplied by the Graticule Light Supply. Rectified voltage for this supply is provided by 5 V Rectifier CR820-CR821. Q835 operates as a series regulator transistor. Emitter follower Q829 determines the conduction of this series regulator as controlled by front panel GRATICULE ILLUM Control R1095. Currentlimiting to protect this supply is provided by Q827. Under normal operation, divider R830-R831-R833 sets the base of

0827 below its conduction level. However, if excess current is demanded from this supply, the voltage drop across R837-R838 increases until Q827 comes into conduction. The collector of Q827 then limits the conduction of this supply to limit its output current.

Divider R822 R823 provides a sample of the line voltage in the secondary of T801 to the plug-in unit. This provides a line-frequency reference to the plug-in units for internal triggering at line frequency or for other applications.

Sorry no idea.

piglet

External.

Well it's simple for you because you're an electronics wizzard, Erich. I'm just a humble hobbyist and "not particularly good at electronics" as I describe myself. I certainly did the right thing by not taking it up as a career, I'm perfectly certain of that.

[snip] Whereabouts in there did you find this, Erich?

Page 3-27 of the 7623 service manual at Tekwiki

piglet

Cursitor Doom does try to be satirical from time to time - he likes to pretend that he has a sense of humour, but what it does illustrate is his tin ear.

There's nothing charitable in that assumption - it is simply moronic.

Did they get fired after they'd worked with you, or was the sloppy habits that got them fired that let them work with you?

<snip>

Ask any Pollyanna who is a sucker for climate change denial propaganda.

Argh! Tekwiki - I'd forgotten all about that particular resource. Many thanks indeed.

Well, that Tek schamatic is kind of a tangle. Not at all obvious on first sight.

Maybe you're not autistic enough to be very good at electronics design.