if you want a constant voltage drop of 10 volts or 15 for a 83 mercury rect can you put a few 1n4007 in series and it will produce the same voltage drop ? i was thinking about this when i saw posts about 83 replacement in certain hickock testers
the mercury doesnt really change its voltage drop over load . 1n4007 a bounch of them wouldnt change over load to ?
the mercury doesnt really change its voltage drop over load . 1n4007 a bounch of them wouldnt change over load to ?
Well, with a 0.7 volt per silicone diode, you would need 21 diodes for a 15 volt drop. Make that 42 for full wave. Not very practical.
In tube tester service you might consider a high power zener (15V/5W or more) in series with the 1N4007 - noting that this is a no no if there is any significant amount of capacitance after the rectifier..
I was considering the series zener plan for a tube tester that I was about to calibrate. I figured the diodes would be better balanced with better long term stability than a tube. Of course, I am going to put some power resistors to load the filament winding as well.
Considering that a lot of tube testers don't use any electrolytic capacitors at all, except maybe one across the meter movement, a zener is a viable idea. Such testers, your Hickok 539B included, work from the haversine pulse peak like a transistor curve tracer does.
Capacitance causing oscillation is only an issue if the zeners are being used as a shunt regulator. If they are being used as a series dropping element then the capacitance is only limited by the peak current capability of the diode, just as it is with the regular rectifier diodes.
I would only run SS if you need a replacement rectifier. If the merc rectifier still works then there is no compelling reason to replace it IMO. The things seem to last darn near forever in tube testers.
I would only run SS if you need a replacement rectifier. If the merc rectifier still works then there is no compelling reason to replace it IMO. The things seem to last darn near forever in tube testers.
Last edited:
42 1n4007 in series is still cheaper than the power resistors you need to load the fillament whit
Just have to be careful to make sure that the ripple current does not exceed the zener current rating, and that the overall dissipation is acceptable as well. I think most 5W zeners may be OK with a reasonable sized input cap - just check the ratings of the proposed zener to make sure. (I have not tried it with a cap input filter, hence the caveat.)
FWIW I have a 539B and will continue to use the 83 rectifier until it fails - which may not be in my life time given the limited use it receives.
I don't think ripple is a problem. Since I haven't seen a datasheet with repetitive peak current ratings, I assume you're OK as long as you don't exceed dissipation, i.e. average current.
I'd be a little more worried about initial surge. A 15V 5W Zener will have somewhere around 6A surge capability. If that's not enough for the capacitance used, you can put two 6.8V 5W Zeners in series for roughly double the surge capability. Or use a series resistor to limit the surge.
I'd be a little more worried about initial surge. A 15V 5W Zener will have somewhere around 6A surge capability. If that's not enough for the capacitance used, you can put two 6.8V 5W Zeners in series for roughly double the surge capability. Or use a series resistor to limit the surge.
- Status
- Not open for further replies.