Using a silicon diode rectifier instead of an AZ31 valve

[Nowhere-Man]

1N4000x aren't made the way they used to be.  Some years ago I measured prv of a few with a Tek curve tracer, and found they were all around 1600v no matter if they were labeled 4005 or 4007 and maybe even some 4001 (I don't recall).  In the old days they would have been graded after manufacture but now they're all alike.

Resistors typically aren't rated for the voltages involved, and add one more failure point.

I just made a circuit that uses two silicon diodes and the capacitors soldered across each diode. The cathodes were joined and the anodes connected to the transformer secondary. I started with a 1 K resistor between cathode DC ripple and the large H.T. can. This was a very simple design that was forwarded to me. I now conclude the resistor needs to come down somewhat, maybe 0.5 K.

Out of interest I tested voltage from the joined cathodes and then again at the other end of the series resistor. This equated to a drop of 30 volts. Some of this had been a bit vague as I'd read rectified ripple D.C, tends to be at the peak of the RMS so RMS * 1.414. In actual fact, the ripple wasn't as high as expected, except the odd spike on my meter at close to 300 D.C.

They were using this system in the Ham field back in 1960 (silicon rectifiers). Some people have their own preference in design but, for now, I think I can lessen my stopper resistor somewhat to a lower number.

[Nick]

FWIW, 100nF MKT or MKP capacitors are the standard supressor caps to use. No need for an X or Y rated cap here...

[Radio Fixer]

This was interesting to me in Alan D's ARF posting.

"The originals were often wet electrolytics, which would purposely leak above their rated voltage and hold the voltage down. They could absorb tens of milliamps without harm, being "liquid-cooled" so to speak. In the 1940s there were "wet replacements" available that were dry but could still handle the initial surge."

Its one of the reasons we fit 450 to 500V reservoir capacitors for, as with a directly heated rectifier the HT soars skywards (420V in one measurement) whilst the indirectly heated other valves, and particularly the output take there slow warm up. In some cases an indirect sub. can be used for the original directly heated type. HT only climbs slowly (to 340V in the same tests) and is lower by say 10%, (as the indirect will have a larger drop across it)which can be useful.

Gary

[pwdrive]

I always used to fit the series limiting resistors in the anode feeds, it simplifies the understanding of whats going on voltage wise and gives the opportunity to spread out heat dissipation from the said.

Aim for values that will give the same HT voltage at the reservoir capacitor that's specified in the manual as some local osc designs in receivers can be quite critical, eg: if the HT is down by 30 volts the local osc might boot up on MW but not on LW.

Lawrence.

[Nowhere-Man]

I always used to fit the series limiting resistors in the anode feeds, it simplifies the understanding of whats going on voltage wise and gives the opportunity to spread out  heat dissipation from the said.

Aim for values that will give the same HT voltage at the reservoir capacitor that's specified in the manual as some local osc designs in receivers can be quite critical, eg: if the HT is down by 30 volts the local osc might boot up on MW but not on LW.

Lawrence.

I wasn't too sure where to solder in my diodes. In the end, I opted to solder straight onto the pins of the valve where the transformer leads go. I was a bit concerned about how safe that would be.

I may have said before in the HAM radio Amateur Handbook they were using silicon way back in 1960. I quite like this book because back then the consensus was you should use semiconductors and tubes together with the emphasis on tubes. It stated anyone who chose to use a silicon rectifier should go ahead as they were becoming popular in the HAM field. Who would have thought that all that would end with the tubes being dumped as semiconductors took over?

Should I wish to revert to a tube I'd need a U50 or U78 but, for sure, silicon does provide a cheap alternative.

[Murphyv310]

I have never been a fan of either directly heated valve rectifiers or silicon rectifiers. This is a personal thing as neither are particularly elegant. With these types of rectifiers the off load voltage rises very quickly before the rest of the valves conduct. It's sore on the smoothing caps and even old paxolin valve bases.
Adding series resistors can reduce the the switch on surge but they are a compromise, insufficient current can be a problem with a warm set and of course series resistors like this get hot!
If at all possible I will swap a directly heated reccy for an indirect one, eg 5U4 - 5Z4. I've even replaced 4 volt reccys with a 5Z4 with no problem. My 1936 Pilot has an EZ80 in an adapter base and it's heater is 2v under run. It works 100%

[Nowhere-Man]

I have never been a fan of either directly heated valve rectifiers or silicon rectifiers. This is a personal thing as neither are particularly elegant. With these types of rectifiers the off load voltage rises very quickly before the rest of the valves conduct. It's sore on the smoothing caps and even old paxolin valve bases.
Adding series resistors can reduce the the switch on surge but they are a compromise, insufficient current can be a problem with a warm set and of course series resistors like this get hot!
If at all possible I will swap a directly heated reccy for an indirect one, eg 5U4 - 5Z4. I've even replaced 4 volt reccys with a 5Z4 with no problem. My 1936 Pilot has an EZ80 in an adapter base and it's heater is 2v under run. It works 100%

Any idea how hot the series resistor gets? Today I decided to test the anode on my output valve for voltage. At the moment I'm finding it to be about 245 DC. The manual gives around 290 and points out the test was with an Avo 8 on Medium waves and at full volume. Of course, I expected the voltage to be a bit low as I started with a 1K resistor. However, I should point out this wasn't resistance in series with each cathode but just one resistor at the bonded cathode ripple output. Plus, my grid bias resistor on the output tube is 50 ohms out on the high side.

I can say, I'm really not so keen having the diodes soldered to the valve pins as it's a bit "rickety". The soldered joints seem really O.K. but you never know if it takes a knock and something comes off. Maybe one more reason why it seems a good idea to put a 1 amp fuse in the supply hot line (something I have to do). I guess you could say now I know I have a working set, maybe it's worth considering putting the tube rectifier in should one turn up at an affordable price.

However the case may be, I have a feeling my next project may involve either a metal rectifier or the silicon variant. I have a thirties TRF radio that actually came with a metal rectifier. My theory is someone at some time modified the radio around the fifties in order to do without the battery. Whichever way I go, it will involve a 120 volt HT supply to the Smoothing cap. All the valves are present this time and they're as big as pepper pots.

[pwdrive]

How hot the resistor gets depends on the current flowing through it and the resistors wattage rating.

Lawrence.

[Nowhere-Man]

How hot the resistor gets depends on the current flowing through it and the resistors wattage rating.

Lawrence.

Mine is a 7 watt. I've heard small ACDC sets get very hot and I've found this to be the case. Someone advised not to run an ACDC set more than one hour due to the heat. This is when the tubes are all crammed together in a small space.

I'll be changing my resistor to a lower value of Ohms but at the same time seeking to allow for the initial current on start-up. It does say in the manual also to take into account the fact some meters won't read identically to the AVO 8.

[Refugee]

Have you replaced the grid coupling capacitor on the output valve?
If it is a bit old and leaky it will make the anode voltage go low and make the radio get hot.