I scored a box full of lead-acid batteries today. They are 12V/3.3Ah maintenance free types, 32 pieces. Hooking up 16 in series would give me 192V, wow!!! I kind of bought them with this in mind, maybe try battery powered B+ as well. Many people may call me crazy for trying this, but most people already do anyway, so...
The only trick is the recharging. I can run the 26 with it's 6-7mA for about 20 days, theoretically, but I have to be able to recharge them within one night. If I were to parallel the 16 batteries I would need about 1/10th the total capacity in current. That would be 16 x 3.5 = 56A... divide by 10 and I need a charger that can handle about 6A. This is one mean charger! The next idea is to just use four small chargers that each handle four of the lead-acid batteries. This way they only need to deliver 1.0-1.2A each and I can use the cheaper LM317T as opposed to the three ampere TO3 type regulators that usually cost many times the price of the little fellows. The charger supply does not have to be anything fancy. Just a rectifier, capacitor and some other parts to complete the circuit. It is in no way hooked up to the tube or signal so there are no demands on component quality and supply stability.
I have added a relay circuit to switch the batteries from series to parallel connected. In series they will be connected to the 26 circuit, while connected in parallel they will be hooked up to the charger. This way, when I switch off my amps the batteries will be charged. This let's the amp do everything with a single flick of a switch. The battery filament supply works in a similar fashion.
The 26 is loaded by a 200H choke. This gives a gain that approaches the mu of the tube (8.3) and also has a relatively low voltage drop due to lower Rdc (400-600 ohms) compared to RC loading (>22000 ohms). This is just what I need for my battery B+ supply that gives a maximum of 208V. The filament is also battery supplied, a lesson I learned in previous experiments with the 26.
The new supply is against all my former beliefs. It uses semi-conductors, I mean regulators !!! I have found batteries to be the best method of heating the filament of the 26. One drawback is that it pulls 1 amp to do the job. I need some hefty 2V batteries to get my preamp to play for a full day. The problem I have encountered these last few months was that the battery of one of the two channels always drops more quickly in voltage than the other. The result is that one channel starts playing softer than the other. In the beginning this is hardly audible, but after a while you start to think there is a phase problem. Then it hits you that one of the filaments is dying out. This is not good for your batteries and not good for the tubes either. Something had to be done.
Because I am using 2V batteries, I need to drop about 0.5V by means of a resistor (actually 2x 0.25R/2W). This works very nicely; for a full battery. Pulling 1A for several hours is a hefty load for my two 8Ah batteries. I have recently switched to a single 25Ah battery, but these don't seem to like cyclic loads (drain, charge, drain, charge, etc etc). They tend to drain very quickly and take a long time to recharge again.
So what can I do to keep the voltage to the filaments stable. Voltage stabilizing was the best option, I need to have a higher supply voltage so that there is a margin of voltage drop before the filaments start to notice. I mean to say: a 4V battery can drop 2V and still supply 2V output with a stabilizer, sort of, just as an example. A normal battery will not go below 80% of it's voltage since it loses any means to supply current when it is this far drained.
I set up a voltage reg on a 6V battery, and as expected I don't like the way the regulator influences the sound. Damn!!!! What else can I do. I don't like the batteries drain being audible, and I don't like voltage regulators. Why not use a current regulation like Tom Ronan has on his 26's. I've tried this before my battery epoch and liked it a whole lot better than just a voltage regulator, but less than a battery alone. Since the battery alone also has problems of it's own, this might be the way to go.
Because of the 'high' voltage drop over a standard LM317 regulator, I am no longer able to use the 2V batteries. 8V is just enough to get 1.5V/1A out of them. I had to shovel out about $9 for a low-drop regulator, LM1085T. This type allows me to use a 6V battery instead of a 12V type. The extra voltage that a 6V battery now offers will see to it that I always have the correct voltage on the filaments, even if the battery starts to get really drained.
© Jim de Kort