Since 1993 Copyright Notice

E61 Some Notes and safety advice

Back to the EE61 Board.

1. Grounding / Sound. It pays off, to find out which of the two ends of the secondary is best suited for grounding, giving best high frequency performance. Just try this out, and you probably notice an audible difference, but this functionality is part of the transformer, not the PCB. You can quickly achieve the same be reversing the banana plugs of the primary winding.
2. Grounding / Safety / Part1. The transformer secondary should be grounded with one end, to have no risky situation in case the transformer shorts primary to secondary. Such transformer faults are rare, but extremely dangerous. Or perhaps you make such a short yourself by mistake, with a loose wire or a probe. Specially with a head phone amplifier, this is not funny! With grounding however, If such a short happens, and one end of the secondary is grounded, the secondary winding can not have HV on the other both ends, because copper resistance is very low. This would blow the power supply fuse, or burn out the transformer, but that is still better as high DC voltage on the output.
3. Grounding / Safety / Part2. Also the board itself should be grounded. The day will come when you appreciate it. Good engineering says, grounding should not be done with a banana plug, because plugs can be removed unintended. Using a soldered wire for this prevents accidental disconnection. There is a free ground pad on the bottom left side of the board. This pad should be attached to ground of the work bench. .
4. Frequency Testing. Tubes and transformers may not be used at full power at one single audio frequency. The worst is, to sweep the circuit at full power through the whole audio range, because then for sure, you will hit one or more resonance frequencies, and instant damage may occur. Already at 10% of maximum power, you can hear the tube and transformer begins to "sing" at specific frequencies in the kHz range. This are movements of some of the internal parts, and these can break off, or short, if you go to far with it or do this too long.
5. Grid damage. This should be avoided, because it occurs mostly unnoticed, but then give an unexpected increase of the plate current later on, at strange moments. This increase will out glow the anode, and each time it happens, the grid damages a little further. So to avoid this, make sure you always know at what current the tube would over heat. You can and should look up these limits in the tube data sheets. For auto bias it is 90% of the absolute possible maximum. For instance, the absolute (damaging) limit for 300B is 40 Watt, and this is no recommended use conditional. Maximum use is appr 36 Watt. So that is roughly 90%. For adjustable bias, if no other values are mentioned the data sheet, use 75% of the absolute maximum. For a 300B we talk then about 30 Watt. It may be tempting to go over that limit, and if you do so, it works, but this proves no safe design. If too close at maximum power, the older the tube gets, the more it may become instable, and suddenly drift away with an avalanche effect. That will self destroy the tube. This is one of the best arguments for auto bias, because you can bias the tube higher, and yet it thermal runaway will not happen.
6. First use. With a new build circuit, initially use the fuse, instead of the SE transformer. Only if the circuit works as it should, remove the fuse and insert the SE transformer. Take care, you may never plug or unplug the SE transformer, or the tube, with a working circuit. That may cause damage to transformer or speaker.
7. Rs = 1k Grid series resistor. This prevents oscillation or radio voices caused by long wires. What happens, Rg and the Input capacitance of the tube, together form an RC filter, blocking signals above 150kHz. The value of 1k is "general" and seen often, but by use of an oscillator and oscilloscope, you can indeed find out set your own 3dB point to a desired value somewhere far above 20kHz..
8. Gas test - The circuit. First, a true gas test as such does not exist. We can only measure grid current, and from the behavior, try to say if this is caused by gas, or by something else. All tube testers with a 'gas test' work this way. Since this test is not a real gas test, you will never get the same result from two testers. There are two ways to measure the grid current. One is, to insert a 470k resistor in the grid, which is normally shorted via a switch. Then, pressing the switch will open the contacts, and now any grid current will result in a small DC voltage across this resistor. The current comes OUT of the grid, and lifts up the grid voltage a little bit. Plate current shows a small increase. The tube gain plays a role here, but test circuit does not know the tube gain. This is why this method is not accurate.
   
   Another way is, directly measure grid current itself. We can use the R-C filter of the Mixed Bias circuit conveniently for this, as it is already present, and this can be used nicely for measuring gird current as well. So the only thing we need to do is, connect a sensitive DC voltmeter to the VAR Input, and pick up the small voltage which the grid current generates across R2. Grid current in uA = U(R2) *10. So, 100mV at the VAR Terminal equals 1uA.
9. Gas test - Causes. Grid current has four sources. 1) Bad socket. Remains of solder flux or other dirt, combined with decades of storage causes leakage from anode voltage to the grid. This obstructs every measurement of course. Such tubes will have crackeling noise, indicating this problem. 2) Space charge. To see this, use only the heater, and apply no anode voltage. Free electrons will land on the grid, and charge it negative. This results in a small negative voltage coming from the grid, so current goes INTO the grid. This is also a sign of a good working cathode, and some curious historic tube testers were made, using this principle, and it works indeed. This current can be measured at the VAR Terminal. As soon as anode voltage is supplied, this current disappears. 3) Grid Emission. This is a very bad form of grid damage, and it results from tube over heating. Some cathode material evaporates and finds its way to the grid. At very high anode temperature, the grid become so hot, this small remains of cathode material becomes emissive. A few uA will result with a bad tube. If the tube has grid emission, this can be concluded, when the grid current is only present at very hot anode. At levels of 70...80% of maximum dissipation, grid current begins to develop, and then rises sharply the closer we get to 100% of maximum dissipation, For a bad tube 3...6uA is possible. A good tube stays far below 1uA. 4) Gassy tubes. This is the case when the previous conditions can be excluded. A gassy tube will produce grid current too, but this does not accelerate so much at high temperature. Also, a gassy tube will always have lower plate current as normal.
10. Gas Test - Values. This depends totally on the tube of course. But here are some numbers.
    300B <1uA or <100mV
    Historic 45 <3uA, or below 300mV
    EML45 <0,5uA, or below 50mV