Lissen/Ever Ready 5214

[Amie]

What's your feeling about NiMH cells for LT on the D series filaments Mike?

[Mike Watterson]

They are fine for parallel 1.4V filaments. Better than Zinc, Alkaline or NiCd. Eastern Europe and USSR actually did do B7G D9* valves with a slightly lower filament rating for Ni and Ag based rechargeable systems. But actually the NiMH have a slightly higher average voltage than NiCd, though about the same end point. The end point of 1.0 to 1.1 rather than 0.9V better suits the DK91, DK40, 1L6, DK92, DK96 and locktal/octal mixer/osc as they stop oscillating at 1V when new and 1.1+ as they wear.

Also they keep well left flat, which is the opposite to Lead Acid and Lithium.

Many of the sets in the 1940s using Telefunken low profile Y8 battery valves used NiCd. (DEAC) as did many continental mains/battery sets using those and D4 series and D9 series after the war. I've made replica DEAC packs with a C or D NiMH soldered in. They also act as the regulator for LT when on mains.

Why bother with the expense and awkwardness though unless you are using the set every day? Often they will go flat in a month. The highest capacity ones often faster. A few kinds will hold charge for 6 months. But decent Alkaline cells have a 5 year or more shelf life and about x2 to x3 the running time of a good NiMH.

They are not so good in the long term for 7.5V to 12V LT systems, just on cell life as one can be flat first, radio will still go and cell will be reverse charged. Fine for that if carefully checked, or you use one cell less so that radio stops before any one cell is low. So a LT pack with 9V in it (radios with an RF preamp) might use 6 x C cells, or D Cells rather than the 7 to give closer to a "Fresh" Zinc.

Don't be tempted to use them for HT as obviously the risk of a destroyed cell is x6 higher than an NiMH PP3 pack. The NiMH PP3 and other series NiMH in gear without battery condition monitoring have a much shorter life due to one cell being flat before the others and getting effectively reverse charged.

I use 2032 lithium coins, carbon layer PP3, Alkaline AAA or Alkaline AA in HT packs. The AA work out cheapest. The Alkaline PP3 are not worth the extra cost. I avoid Duracell (poor base seals and overpriced), Energiser (overpriced) and Panasonic (overpriced unless on offer). Varta, Aldi, Lidl are fine as are most random or badge brands. Totally avoid Zinc carbon and Zinc Chloride round cells. They should be banned.
I've not made the PP3 styled (longer) 45V and 67V packs yet for the pocket sized valve radios. Likely I'd use Alkaline button or Lithium coins.
I use paralleled stacks of 2032 (bought at 12c each from China) using 1N4148 diodes) with a wired fuse on each parallel stack to completely fill a 67V B101. I use PP3s for B126. The next size up packs (HT only and combo) I usually use AAA, I do use PP3 in the B114.

I've investigated why Alkaline cells leak. There are two reasons:
1) Poor construction of the bottom seal (Duracell & Polaroid) . Unlike Zinc Carbon and Chloride the case is positive and the -V is a spike attached to the base plate.
2) Compartment or holder that's too tight, or spring is too strong. It breaks the bottom seal with time.

Another issue in older sets, especially valve sets is metal battery holders, typically for D cells (there may be two in parallel). The Alkaline cells often only have a thin film. I add a printed sleeve (sometimes in thin card) to help protect the cell from getting shorted. Some holders also can short to the metal top surrounding the + terminal, so a card or plastic washer is needed.

I've made sockets and plugs from cut up coffee tin. The last one I did was for a PP10 for a transistor Sky Prince. I found that a cheap nail emery board made into two layers duplicates the paxolin and holds the pins. The PP10 fills nicely with D cells in a simple wooden holder.

[Kalee20]

Likely that amount of metal degrades a frame aerial, so hardly a fair test. The Zinc of the B cells and F cells in an AD3 would have been differently distributed.

Yes I can see that - though TBH the space occupied by the SLA and chunk of aluminium is probably less than the B136 battery. But its a good point, comparison might not be fair.

Basically, on series or parallel sets operated on mains the simplest thing is to check volts on the DAF91 or DAF96 as it will be the end of the chain. Should be 1.35V +/- 5%. Most decent series sets have taps or a rheostat to set the LT current, usually stated. The 7.5V nominal battery sets typically at 6.8V.  Any purely with either HT or LT via a dropper off the mains want to be run off a shaver transformer. 1930s to 1960s concepts of safety and 70 year old insulators between metal panels and live chassis are not good.

I really do not like the idea of series-connected filament chains, with bypass resistors as more electrons, going up the chain, leap off into space. Too variable and flakey IMO. It's parallel every time for me.

I don't use my battery sets for everyday listening when a mains radio is available, and no way would I make a mains adaptor and use as a mains set. Though not eschewing the odd modification, I've always kept to the original design intent which is mains-independent.

However, my Vidor CN420a has still had a lot of use in last 10 years, with stabilised battery-derived power supply. LT is set at 1.38V and varies less than a millivolt till the battery (NiCad in this radio) is at its last gasp. It's clocked-up well over 1,000 hours in that time, with 3 valves the same (DK96 I zapped with a moment's carelessness; replacement S/H 'pull' has run 9 years). It goes on overnight travels with me. It's been used on the roof as I clear leaves from gutters.

I'm confident that the long life is due to LT care, after all, 1.5V from a new dry battery is over 7% above the nominal 1.4V rating which must have a disproportionate impact on valve life and cathode depletion. Also, not being forced to operate at 1.1V again means there's been very little possible damage by filament bombardment.

[Mike Watterson]

The battery cell voltage of a nominal 1.5V Zinc Carbon falls pretty fast from 1.6V to 1.45V and then under 1.4V. Hence the 1.35 recommendation and the 1.4V spec.  In normal usage there would be little time above 1.4V and replacement around the 1.05V to 1.1V, volume gradually degrades with failing HT, but Local Oscillator stops dead at some point with falling LT. Vidor actually explain the symptoms as to when to replace HT or LT pack on the sets with separate packs. On the 250mA sets the small LT packs needed replaced twice as often as HT.  Larger LT packs and combo packs were designed that the LT and HT would exhaust at the same time. Since the o/p was class A on most sets that was fairly accurate. The sets with push pull etc might have also had an RF preamp, so the main concern on combo packs for them was the LT.

The Zinc Chloride is not much different to Zinc Carbon except at higher currents there are lower internal losses. The x5 life of Alkaline over Zinc is only applicable at the sorts of massive currents of a floodlamp or motor that ought to use NiMH or Lead Acid historically. It's more like x2. Thus a AA Alkaline is similar to Zinc B cell and an D Alkaline can replace a F cell. An F cell is 50% bigger (longer) than a D, but the x4 F in parallel in an AD4 pack gives much lower resistance than the two smaller cells in parallel in an AD35 (which changed the the size of cells from a cell larger than D to D during lifetime). A C Alkaline can replace a Zinc D, or one D Alkaline replace a pair of D, except not quite the larger cells in older AD35, so I put 2x Alkaline D in a AD35 box. The 250mA sets certainly used the original higher capacity cells in an AD35, but still would have used that twice as often as the HT pack.
The B114 would have had a short life, but the USA 1941 to 1946 models often used a single D cell in a holder for maybe less than 18 hours life. Like DAB Radio though some early DAB sets only last 6 hrs on a charge.

The 5214 AD3 pack had EIGHT F cells in parallel. It was intended for long service. Some USA Farm Radios (what we would call Table models) from 1938 when All Dry came in there had an LT pack with SIXTEEN F cells!

As an even more aside, Sylvania did button base versions of the 1938 1.4V dry battery valves they developed. Loctal base. The Loctal button base pins fit an octal plug and thus into an octal socket. The 1940 RCA B7G valves (indirect for mains and the 1.4V alldry) are essentially just miniature versions of the Sylvania loctal & Octal types. The first step in Philips versions that weren't a copy of the RCA came before Philips / Mullard did the B7G, they were the Rimlock /  B8A. The DK40 was superior to the DK92 and the RCA original. The DK92 and not quite compatible 1L6 replaced the DK40 and USA 1LC6 for "decent" shortwave. The DK96 is poor for SW.

The original RCA B7G battery valves used nickel filaments because such small filaments couldn't be done accurately enough for series. The Telefunken 1940s series had all different currents to minimise consumption, so used the NiCd as a regulator and LT for mains. Philips not only cut the current in half for the B7G series but with improved diamond dies also used tungsten. So the 25mA series have more fragile but longer life tungsten rather than the nickel of the original 50 mA series RCA design.
Note a 10% overrun about halves the life of a tungsten lamp.

Edit forgot to click on Add!
   

[Amie]

Thanks for the battery information Mike. As I can get pj996 Eveready brand (4×F zinc carbon) and 1604 alkaline, for little money that's what I'll use. I am tempted to make a small modification of using a couple of small lithium cells to provide fixed op valve bias, to save the power lost in the series bias resistor.... However the set will only be run occasionally, its @ 80 years old and deserves a little rest!


Amie

[Mike Watterson]

The HT voltage is not critical. Similar valves do work with as low as 45V HT packs, the main limitation is the volume.
So you'd notice no difference at all in HT battery life.
Let me know via PM it you need accurate images of the AD3 pack, or actual colour laser prints to stick to a card box. The socket is a paxolin B4 valve base! A B5 also works as it simply has an extra socket. I've made my own sockets for most packs out of plastic cut from marge tubs and cut up coffee tin.

The real F Cells do leak eventually so I use a big screwdriver to pop the top of the case and rewire using stubs of existing wires in parallel. Just check no packs have dud cells. I leave the cells in the casing as that fits inside the proper size AD3 pack, without even trimming the top.
Save the top springs to make D cell battery holders to fill things like a PP10. I used two battens, two pieces of plywood for the ends, two 996 pack springs and homemade socket to build a 6 x D cell holder (2 stacks of 3 on one diagonal) that slides into a PP10 box.

[Amie]

Yes Mike, thanks. I've given the socket a bit of thought and as I have some paxolin sheet I am thinking that I could use that and some sheet brass to make the socket, the pins have a tension spring on them, so a simple tube contact would do for the socket part, I've some vintage B5 sockets, but am loth to use them....

Amie.

[Mike Watterson]

Coffee tin may be better than sheet brass:
1) Easy to cut with scissors.
2) cut 3 flaps at one end of rectangle
3) Wind on a screwdriver with help of pliers.
4) Most originals are tinned steel, none are brass and some are sheet with one open join folded to a triangular cross-section.
The flaps then go between two layers with 4 holes and one slot in rear. One flap is bent into the slot to be a solder tag. The box is in two almost identical parts, almost the same size each missing one wall. Outer has a hole at the connector and 5 of the sides of printing. Inner has 5 small holes for connector which is glued to it and only print on the base.