An Improved VTVM Battery Eliminator

One of the disheartening things about buying an older (…well, they’re ALL older, now…) VTVM is opening it up and finding either an old, leaky battery, or evidence of one. This usually damages or destroys the battery contacts and sometimes also the circuit board. The battery, usually a standard 1.5V C-cell, is necessary for measuring resistances with the ‘ohms’ scale.

A great improvement has been suggested over the past couple of years to replace the battery entirely, using a modern voltage regulator, drawing power from the filament circuit. A version of this has been available on several Heathkit lists and I’ve used it in my V-7A and IM-18. This replaces the battery and removes the risk of leakage.

More recently, Peter Bertini, Pop Comm Magazine (and others) have pointed out that the circuit used an inefficient half-wave rectifier, probably adding stress to the (already old) VTVM’s transformer.

If, in ‘ohms’ mode, the probes are touched together (0.0 ohms resistance), the entire 1.55 Volts is placed across a 9.1Ω resistor, resulting in the maximum current draw of (I=V/R), or 1.55/9.1 = 170ma. This doesn’t count diode loss and heat loss in the voltage regulator.

The following circuit attempts to repair the inefficiencies and addresses ripple filtering in the regulator, so as to provide an efficient and accurate VTVM Battery Eliminator.

An Improved VTVM Battery Eliminator

The input is from the 6.3VAC filament circuit. The output goes to the same locations as the original dry-cell battery.

In the case of a Heathkit V-7A, is grounded, with +1.55VDC at the ‘free end’ of the 9.1Ω resistor (see a schematic at http://www.heathkit.nu/heathkit_nu_V-7A.html).

VTVM Battery Eliminator, Grounded
VTVM Battery Eliminator, Grounded

Overcurrent Issues

The transformers in these VTVM’s are barely capable of lighting the two vacuum tubes. Is it possible that the addition of a ‘battery eliminator’ for the Ohms measurement could cause a problem?

While converting a Knight-Kit KG-620 for battery-free operation, I was able to measure the current draw. Instead of an LM317 (1.5A) regulator, I substituted an LM317L – 100mz, current-limited regulator.

Idle current is measured around 5 ma and peak current at around 95 ma with shorted terminals, and range switch to Rx1. Current limiting would probably curtail measurements on the low-end (below 100 ohms).

A second way to address current draw is as follows: the KG-620 has a Type #47 pilot lamp, drawing 150ma. Replacing this with a LED will reduce the lamp current by 10x from 150ma to around 15ma. This effectively ‘recovers’ any current used by the battery eliminator.

This would also work for the Heathkit V-7A and IM-18 (and variants) as the pilot lamp is also a 6.3V, Type #47 lamp across the filament supply. Added benefits: a) reduced heat, b) reduced transformer load, c) never have to replace the pilot lamp again.

 

References

  1. Peter Bertini, “Fixing Up A Vintage Heath IM-13 VTVM”, Popular Communications, March 2010.
  2. AntiqueRadios.com, forums – Search “VTVM battery draw…”
  3. LM117/LM317A/LM317 3-Terminal Adjustable Regulator (Datasheet), National Semiconductor
  4. 3-Terminal Regulator is Adjustable (Application Note 181), National Semiconductor, Figure 2. Adjustable Regulator with Improved Ripple Rejection.
  5. The V.T.V.M.: How it Works, How to Use it , Rhys Samuel, Gernsback Library 1956.
  6. Servicing Radio and Television with a Vacuum-Tube Voltmeter, an excellent, 1951 document from Sylvania Electric Products.
  7. AD5X’s method of turning Type 47 Lamps into LED Lamps, saving heat & current

My First Project

This is probably the first project I’d done and it was when I worked at Mostek. I used a piece of scrap perfboard, some TTL ICs and sockets, and a few Old Style LEDs – big, old, current-consuming, 1982, RED LEDs. And some wire-wrap wire – it was all I could find, so that’s what I worked with.

1982 TTL Dual Dice - Top View
1982 TTL Dual Dice - Top View

The circuit is from Don Lancaster’s “TTL Cookbook”. It has worked for years and years, and uses almost any 6 to 9 volt Wall Wart. Unfortunatly, about 10 years ago a wire or two came loose from underneath (a rats-nest of misused wire-wrap). I finally got around to fixing it and actually replacing a rare burned-out LED.

A Rats Nest of Wire-wrap Wire - not the best choice, but it's what I had.
A Rats Nest of Wire-wrap Wire

I have no idea why I chose Wire-Wrap wire, other than it was something we had laying around or scrapped. It’s some of my earliest, gloppiest examples of soldering that I have. Wish I had that old Radio Shack P-Box that I build back when I was about 12.

ATX Desktop Supply Adapter

Built one of the ubiquitous ‘Desktop Power Supply’ from a recycled, ATX-form, PC power Supply. Actually used it occasionally to power a 12V charger in the garage. Unfortunately, it fizzled an electrolytic capacitor and it’s really not worth reparing.

So, I rescued all my hardware and built a reusable adapter for ANY standard ATX power supply.

Parts List:
1 ATX Power extender – this is an extension to the wide plug that goes to the motherboard, usually about 9 inches or so (like this one: http://www.directron.com/atxextension.html).
All of the other hardware bits needed to alter an ATX supply for desktop use (binding posts, a switch, a 10 ohm / 15 watt resistor (I used a Dale, metal-cased).
And an enclosure – a wide, thin, Radio Shack enclosure I had on hand.

The 10 ohm 15 watt resistor goes from the +5V rail to Ground. This is needed so that the supply can sense a load – otherwise, it will not start.

Excellent how-to’s everywhere on the net, but this one is great: http://web2.murraystate.edu/andy.batts/ps/powersupply.htm

Great current ratings here: http://web2.murraystate.edu/andy.batts/ps/PowerRatings.htm. The +5 and +12 rails can supply a good deal of power and these ATX switcher supplies are very efficient.

This adapter will provide desktop power supply access for a recycled ATX Supply.
ATX Power Supply Adapter

“Most Accurate” SB-630 – An Update of an Old “Timer”

Or, “Most Accurate” SB-630: Retro Style Desk Accessory, Updated on a Budget

A while back, I was given an SB-630 Station Console by an old timer acquaintance.

The SB-630 is a nice, but not especially necessary station accessory. Some hams built them just to have the complete SB-line. The console consists of a passive SWR meter, a phone patch, a motorized Digital Clock and the unique feature: a 10-minute Identification Timer. Better versions of the SWR Meter and Phone Patch were sold separately; the clock-timer was unique, so the SB-630 was merely an opportunity to wrap them all up in a single desktop accessory. Since there’s nothing unique about the SWR Meter or the Phone Patch, for my purposes, I shall focus on the clock-timer combination.

My plan was to build a new clock display, keeping some of the old style (albeit 1970’s style, not 60’s), and drive them with an Arduino micro controller. The real-time clock is provided by a Maxim DS1307 (formerly Dallas Semiconductor) 8-pin IC. The chip is tiny, uses very little current when it’s ‘on’, and is backed up (according to the datasheet for 10 years!) by a single CR2032 3-volt lithium battery.

The LCD is from SparkFun, a Red-on-Black, Backlit LCD which fits well.

Construction

All Tubes and the Plate-Filament transformer were removed (and saved, for future tube projects). The clock was given to a local collector who prefers to keep his Heathkits original. The functionality replaced – and enhanced – by adding a Real-Time Clock (RTC) chip, a WWVB receiver and Arduino code to interpret the 1950’s era clock signal.

The WWVB receiver reads each ‘pulse’ of the signal and interrupts the Arduino (INT1) to add the ‘tick’ to the buffer. Once the whole signal is received, it can be interpreted as a date and time. The RTC pulses (INT0) each second in order to drive the display clock.

In addition, an LM35 sensor provides the current room temperature.

Original WWVB decode source from http://duinolab.blogspot.com/2009/06/arduino-cmmr-6p-60-almost-accurate.html (Capt Tagon) and all others who’ve improved this code. Website seems abandoned, but the source code is good. My alteration is to remove the timer interrupt (1000 times a second) which operates the 1-second tick and replace it with the square wave output (SQWE) signal from the RTC chip, a Maxim DS1307.

The LCD is 16×2, and the layout needs to accommodate Date, Time, and Day, along with an ID Timer. A room temperature indicator is ‘extra’.

SB-630 LCD Layout
Character Display is laid out on the 16×2 LCD Display.

It’s a bit cramped, but I’m able to show everything I wanted to display, and I have plenty of Arduino pins left to trigger the ‘IDENTIFY’ lamps and ‘audio tone’, and a few pins left for future expansion.

Schematics

The DS1307 has it’s own battery-backup, which retains the time on power-off.

DS1307 schematic
Detail of the DS1307 Real-Time Clock connections.
Schematic of the CMMR-6P-60
Schematic of the CMMR-6P-60 (now unavailable)
Overall Schematic of the Microcontroller
Overall Schematic of the ATMEGA168 IC and it’s connections.

 

The WWVB Receiver IC is no longer available. A better choice today would be a GPS receiver, which would provide the same (or better) accuracy and is in the same price range now, with some GPS modules selling for as little as US$14.

Finally, the ATMEGA168,  LCD, ID Lamp Relay, Temperature Sensor, and the Tie-in with Original Switches.

Code

Source Code contains classes for the DS1307 and the CMMR-6 WWVB receiver.

 

Garage Parking Light

I really liked the Nuts & Volts Magazine project “Garage Parking Assistant” in January 2010, but I don’t ‘do’ Basic Stamp – no reason really, just not my favored platform and N& N & V seems to base a lot of projects on that platform.

So, I translated it to Arduino and replaced the Basic Stamp with a minimalist Arduino (Atmega 168, a clock resonator, a couple of resistors and capacitors, and an LM7805 voltage regulator). From there, code translation (BASIC to  “C++”) was easy. I make no originality claims and I’ve only provided a code translation.

Garage Parking Light – Translation to Arduino

This is a project I’d been thinking about for some time, but was too lazy to do. Finally a version came out in Nuts & Volts, January 2010. Their version was based upon a Parallax Basic Stamp module. Not a huge problem: I translated it for Arduino and built the whole thing on a small circuit board. This project is also well documented over on
Savage Circuits. I have nothing against the Basic Stamp modules, it’s just that I had parts on hand to build an Arduino version.

Parts

The Parts List is similar to the magazine defined project.

Qty Description Source Notes
1 Project Enclosure, approximately 3″ x 2″ x 1″ Anywhere For the Sensor
1 Project Enclosure, approximately 5″ x 3″ x 1.5″ Anywhere For the Controller & LEDs
1 PerfBoard to fit the Larger Box Anywhere
1 10mm Red LED Red is brighter; use a larger resistor
1 10mm Yellow LED
1 10mm Green LED
1 PING))) Ultrasonic Sensor Radio Shack
1 470 ohm, 1/4W Resistor Anywhere Red LED is Brighter
2 220 ohm, 1/4W Resistor Anywhere
1 Minimal Arduino or RBBB Kit www.moderndevice.com RBBB Kit or build a Minimal Version from parts

Arduino Code

The original Basic code for the Stamp was translated to Arduino’s familiar  C++. I’ve retained the original author’s comments and I’ve used the constants provided therein for evaluating distances to the arriving car. I didn’t provide the “Setup” mechanism, including the momentary press button, but I’ve included

PING Ultrasonic Sensor
PING))) Ultrasonic Sensor, mounted on the wall at bumper height.

it in the translated code. I used the original vehicle distances, as they suited my needs.

The Arduino platform used can be whatever is on hand. I usually use an RBBB from Modern Device, but in this case, I used a minimal Arduino built from a couple of resistors, a resonator for clock (or a crystal if you have one). The minimal Arduino, and a programming
header was constructed on the same board as the LEDs are placed, but off to the side.

One feature I did add was a ‘darkness’ sensor — I didn’t feel that the PING))) Ultrasonic Sensor should be working when it’s not needed, so I sense darkness with a simple analog read of the voltage drop across a Cadmium Sulfide (CdS) resistor.
Since the garage is dark except for:

  • Daylight – bright light enters from the open garage door
  • Night – headlamps will turn the Parking Light on
if ( analogRead(LightSensor) > 600 ) {		
	// Lights are on, or door is open and it's daylight.. Start to Work...
		
	// establish variables for duration of the PING))), 
	// and the distance result in inches and centimeters:
	long duration, inches, cm;
	:
	: (rest of PING))) Loop here...)
}
else { 
	// DARK! Sleep Now...
	digitalWrite(RedLED, LOW);
	digitalWrite(YellowLED, LOW);
	digitalWrite(GreenLED, LOW);
	delay(1000);
}

The CdS sensor peeks through the enclosure via a small hole above and to the left of the LEDs.

Hardware Differences

Since the controller changes from a Basic Stamp to an Arduino, there are a few hardware changes. The Arduino Digital and Analog Pins selected are:

  • Ping Pin – In/Out, Digital Pin 8; uses the PulseIn Function to detect the reflected Ping
  • Red, Yellow and Green LED Driving Transistor Bases: Digital Pins 11, 10 and 9
  • Program Button (implemented in code, but not used): Digital Pin 12
  • CdS Voltage Sensor: Analog Pin 5 –> inputs from a voltage dividor including the CdS sensor

The original article uses the Basic Stamp to drive the LEDs directly. I used a transistor to drive them, as the Arduino pins cannot source about 40ma of current. The Digital Pins might be able to supply enough current, but I’ve used the pin to drive the transistor base and let the transistor switch the current as a conservative design choice.

Source Sketch Files

  • Version 1 – works with Arduino-0018, uses a CdS light to turn off when garage is dark.

Yet another Part 15, AM Transmitter

Part 15, Low Power, AM Transmitter, with an Audio Pre-Amp

I’m using this schematic from user “35Z5” at the ARF.
Many thanks to user “Norm Leal” (for the original 6888 Transmitter), Bill Hamre (Parts Kit), and 35Z5 (6GY6 Version) and many more who know more about tubes than I’ve learned since I was 12 years old.

Schematic: Two-Tube AM Broadcaster - 6GY6 Version
Schematic: Two-Tube AM Broadcaster – 6GY6 Version

I decided to use a Blonder-Tongue BTC-99 UHF Converter as a cabinet. Somewhere on the ARF forums, someone is weeping, I know, but I needed an enclosure and it was unlikely to ever be used again in it’s original incarnation.

6GY6-Version Chassis is Prepped
6GY6-Version Chassis is Prepped

Besides, the lil’ fellah will live on in a new project. Everything will be used, except the tube (6AF4) and the transformer, which will be saved for future projects.

Some additional prep work was done cleaning the chassis up (thank you, steel wool!), and recycling for reuse the original switch, 7-pin tube socket, and terminal strips. The original transformer was only 10 watts — too small for the new load, so it will be saved for some future project.

6GY6-Version Chassis Prep Detail
6GY6-Version Chassis Prep Detail

I stripped the chassis the evening of the final Analog to Digital TV conversion: June 12, 2009.

I think chassis prep, whether it’s recycling an old project or re-purposing an unusual container for an enclosure (like a mint-tin, or an old UHF converter) takes about 50% of the project and maybe 60% of the physical work (metal work, cleaning), but when done correctly will result in something worth looking at.

  • An additional 7-pin socket is fitted.
  • A fuse is added, retaining the original cord, but clipping the ‘neutral’ side to assure one-way insertion.
  • The longer rear terminal is removed and saved and replaced with a piece of perf-board. The crystal oscillator and R3 will be mounted here.
  • Holes are drilled and grommetted for the new transformer leads.
6GY6 Version power supply complete bottom
6GY6 Version power supply complete bottom

The HV Supply is complete (blue electrolytics), as well as the filament supply (pins 3 and 4, both tubes). Also the crystal oscillator is wired and supplied via the 5.1V Zener (top-most).

6GY6 - Audio Output complete
6GY6 – Audio Output complete

End of the first evening, about 1 1/2 hours of actual build-out, excluding cabinet prep. Note the marking, in red of tube placement. I changed this later to place the 6AB4 in back, closest to where the input jack would be to keep audio signal paths short.

6AB4 Output, Sine wave
6AB4 Output, Sine wave
6AB4 Output, Saw wave
6AB4 Output, Saw wave
6AB4 Output, Square wave
6AB4 Output, Square wave

Some testing of the Audio Output from the 6AB4 – nice! Input signal on bottom\; output of Pin 1, after C4. About an 8:1 amplification. Also tested Saw and Square. Square Wave photo is fuzzy. Testing is done with a Heathkit IG-1271 Function Generator.

6GY6 Version Wiring complete
6GY6 Version Wiring complete

One more evening’s work and we’re complete. Did some preliminary testing to make sure there were no High Voltage wiring disasters awaiting, and also some spot-checks to make sure the audio and RF wiring is correct. A final check with a DVM *and* a VTVM (belt and suspenders approach) to make sure there’s no large DC or AC on the audio input path, or with reference to ground and only at this stage are we willing to risk connecting a $135 iPod nano!

At this point the only technically negative thing I’ve found is that my recycled transformer seems to output
a higher than 6.3 VAC for the tube filaments. For testing, I’ve kept the Iso-Variac at a low voltage output (about 114 VAC), but when I do the final checking at full house current (here, about 124 VAC), I’ll include a 2 ohm, 5-watt resistor in series with the filaments in order to drop the voltage from about 7.4 VAC to just above 6.3 VAC. This will make the tubes have a much longer life.

6GY6 - Final Cabinet Front
6GY6 – Final Cabinet Front

Finally Complete, and it looks as well as it sounds.
Transmitted audio is fantastic. It lacks the compressor of the SSTran solid-state Part 15 transmitter, so there’s a small amount of unevenness in volume. ‘Big Band’, or older mono music that was mixed originally for AM Radio sounds great, certainly, but I’m quite impressed with the bandwidth – more modern music sounds swell. Very good coverage of the highs. I did a quick test and found that the transmitted audio covers at least 8500 hz (that I can hear) and that would mean a bandwidth of 17 khz – quite a bit more than a standard (US) AM Broadcast channel spacing of 10 khz. See the excellent discussion of AM Broadcasting bandwidth on Wikipedia.

Distance: Testing with the “SSTran-style Antenna” yields about 75 feet (23 meters) distance from the little ham-shack / office / la-BOR-atory to a receiver in the front room, which is using a tuned Terk AM Advantage antenna, and shows full-scale and good audio.

The tuning knob is non-functional, but I may move the output filter’s VC1 to a larger capacitor and mount it there. The “On” know works!

Very fun to build and this one is turning out to have the best transmitted audio so far.

Accumulated Kit Comments / Suggestions

  • J1 (supplied, not listed) was a 3.5mm (1/8″) mono – suggest a stereo jack; schematic indicates a rudimentary mixer and most folks have some
    sort of stereo source (iPod, computer speakers, etc.)
  • R1 and R2 – 10K supplied as 1/2 Watt. Suggest 1/4 watt – they’re cheaper and fit the holes in the audio input jack better

Build Comments / Suggestions

  • I used some inexpensive RG-174 coax on both the audio paths and the signal paths. Only one end of each shield was attached to chassis ground. This seemed to remove any trace of hum incursion from signal paths passing by AC lines (Filament, transformer). Segments shielded with RG-174: AFTER the 6AB4 to the 6GY4, the crystal oscillator TO the 6GY4 and the RF after the PI network to the antenna terminals.
  • I used a take-out (recycled) transformer from an old Heathkit which had dual secondaries (150 / 6.3), but two transformers such as the PC-12-800.with a 12 V (CT) secondary may be used, as per Jon’s Electronics And More two-tube write up (see: Alternate Power Supply).
  • The Power Cord – make sure it can be inserted only ONE WAY, and that one way assures that the ‘HOT’ leg does NOT wind up on the chassis. I identify the ‘neutral’ side, and clip a bit out of the middle out of the prong and widen it a bit, making sure it fits in the LONG side (at least in the USA) of a power socket. If your house is older and has pre-war or non-standard wiring, use an isolation transformer.
  • As Always when working with house-current / mains voltages, please use extra care and caution. Use an isolation transformer if you have one available. Test before touching. Use a ‘one-hand’ rule – don’t get the HV (or any AC voltage) across your heart. Generally – do not get yourself between ground-reference high-voltage (your house current) and ground.

Updates

  • December 2009 – Since I used a 6AB4 instead of a 6C4, I changed R5 from 100K down to 82K – this improved the plate voltage to 92V and increased volume.
  • August 2010 – Finally cracked how to put an air variable capacitor in the blank spot behind the big dial. The original tuning mechanism rotated with a Bakelite shaft. I was able to cut that shaft and use a shaft coupler to put an air variable tuning capaciter there. This basically replaces the little trimmer capacitor VC1, and gives some tank tuning to handle differences in antenna. With the original VC1 trimmer, this could only be tuned with the cabinet removed and was fixed.
  • January 2011 – Replaced the Crystal Oscillator power supply with a regulator – replaced: D2 (1N200X), D3 (Zener), and R12 with: D2 – 1N5817 (Schottky), D3/R12 – LM2931 LDO 5V Regulator.
    The purpose is to provide better regulation and slightly more current in order to use a PLL Oscillator plug-in.
    Measured B+ is 177.5 VDC.

a personal website