
Another local Ebay find was this Hewlett Packard HP 403B AC Voltmeter in a protective HP vertical case. The case did an excellent job of protecting the meter over it’s now 48 year life.
As with most of my repair finds the first step was to clean it up, and look for potential problems before powering it up.

After removing the meter from the HP instrument case, I took off the top, bottom, and two side panels which gave me some access to the internals.

A quick check of the Nickel-Cadmium batteries showed that only one pack was able to take a charge, with the other three packs not accepting any charge.
That was pretty much what I expected since these cells had most likely not been charged in many years. There was also some signs of leakage, so it was time for them to go.
Unfortunately the meter uses the battery packs as a power supply filter and voltage divider for the +13V, ‑6.5V, and ‑13V power rails.

After a few quick measurements I ordered a pair of Exell 10 cell 12V NiCD battery packs to arrive in a few days. The existing packs were 5 cell 6V units that were connected in series with the ground connection connected in the middle of the packs.
I would have to tap the connection between the 5th and 6th cell on one of the new packs for the ‑6.5 supply, but the rest of the connections were all to the existing tabs.

While waiting for the batteries to arrive I spent some time cleaning up the circuit boards and checking capacitors. The only capacitor that was even slightly marginal was C21 a 65 uF at 60 VDC rated aluminum electrolytic in the power supply AC rectification section.


There was a small amount of corrosion on the bottom side of the 403B-65A main circuit board, which cleaned up easily. The board is positioned directly above the battery packs, so the corrosion was probably from leaking electrolyte.



After working on the circuit boards I opened up the shielded and chassis ground isolated enclosure for the attenuator 403B-65C board, and the range switch. The range and power switch were cleaned with DeoxIt D5 contact cleaner. I also cleaned the front jacks and BNC adapter while I was setup for making a mess.
The attenuator cover is held on with two screws above the side with the adjustment holes, one screw in the lower chassis near the meter, and one screw attaching the power supply board next to the AC power inlet connector.

The battery packs that I ordered arrived, so I used some 3M VHB double sided tape to attach the two packs to the tray. The pack layout matched the layout of the existing wiring harness, with only a few bends need to line up the wires with the terminals. The blue wire attaches to the ‑6.5 volt tap on the opposite side of the ‑13V pack on the strap between cells 5 and 6. Be sure to insulate the new connection and also not overheat the cells when soldering.

I was a little concerned when first turning on the meter, as it pegged the needle of the meter to maximum, with a return to near zero, and then another excursion to maximum before settling back to zero and staying there.
After some research, I found that sequence as a normal turn on as long as the needle returns to zero within around 20 seconds.
After a 20 minute warm up, I was able to complete all the calibration procedures and most of the calibration checks with the exception of the 100 and 300 VAC checks which I didn’t have the equipment to generate signals above 50 VAC accurately.

After 24 hours with the switch held momentarily in the battery check position the batteries had charged up to 27 volts total. Due to the extra capacity of the new 12 volt packs at 1,000 mAh versus the original pack capacity of 225 mAh, the meter should be able to operate continiously for 7 days before needing recharged. I did adjust R39 for a charging rate of 10 mA which doesn’t include an additional 5 mA current for meter electronics which is always on when charging.
A fully discharged battery would take around 120 hours to fully charge back up though.

This average responding AC voltmeter after it was calibrated, has proven to be very accurate on all ranges. It would be a great addition to an audio technicians workbench or anyone working with sub-megahertz AC signals. The only thing left to do now is to make some room for it on my workbench.
I did one not that long ago. The meter read 15% on all ranges, but was fine when not plugged into 120VAC. The C21 cap was dried out…30V ripple or more. Once changed out (60uF, 60V) the unit is now good, 3mV ripple on the battery versus about 50mV before that.
Batteries are good for about 10 years… YMMV.
These are great for wide band AC signals.
Can someone please inform me upon changing out the 6 germanium transistors for 6 silicon transistors if any other circuit mods are needed…maybe ill just have to wing it..thanks.
Hi Jon,
According to my service manual and datasheets Q1, Q2, Q3, and Q5 are 1853–0020 PNP Silicon TO-92 transistors, and Q4, Q6 are 1854–0071 NPN Silicon TO-92 transistors.
My service manual only covers Serial numbers prefixed: 0986A, so if your unit is an earlier prefix you might have germanium transistors.
If so you might want to compare schematics between manuals to see what changes were made if any.
Greg (Barbouri)
This was an excellent post and quite helpful. Since I didn’t want to use my meter as a portable and didn’t have the case anyway. I replaced the batteries with a diode stack consisting of 1n4735A 6.2 volt zeners. After replacing a few capacitors the meter fell right into calibration. I was careful to mark all the bias points when doing the rewire. The meter is much lighter and is a welcome replacement for my old 400D. The charge light stays on but no problem at all.
Hi. I’m a big fan of HP gear and have several pieces in my collection of test gear . A while back I started working of refurbishing one of two 403B’s in my inventory. I got side tracked and am just now getting back to it. My Nicads were shot. I have new Nicads. I can’t find my note on the color coded wiring for connecting the batteries.Mine has a red, blue and purple. the white with a black tracer is at ground. HELP!!!
Hi David,
Red (2) goes to + terminal on the positive pack.
Jumper goes from — terminal on positive pack to + terminal on negative pack.
White/Black (90) ground connects to pack jumper.
Violet (7) goes to — terminal on negative pack.
Blue (6) connects to the ‑6.5 volt tap on the opposite side of the negative pack, on the strap between cells 5 and 6 for a 10 cell pack.
Good luck with the repair,
Greg (Barbouri)