Resurrection of my WD88A XL

[NickW]

Originally Posted by petercom

NickW, there's something odd going on, but neither Dr John nor I can work it out at a distance.

Even allowing for the voltage drop across the diodes of around 0.3V each (x2) you should be seeing 8v DC on the output of the bridge. So your DC output of 5.85V is clearly crazy.

Are you sure the diodes are all the correct way round? One of them the wrong way round will part cancel the rectification.

That's the only thing I can think of at the moment.

Hi Peter, John

Ok, I've opened the amp up and I am pretty well 100% sure the diodes are wired correctly. I've taken a photo, but I don't know how to add it into this reply. I can e-mail it to you if you give me an address.

regards, Nick

The wire link that links the negative ends of the diodes together is made under the tag board with the heavy duty single-core black wire, and suitably stood away from the board to avoid shorting out the two tags between them. You can see the two blobs of solder that fill the central holes in the two tags.

[Richard]

Hi Nick,

Yes, it looks the same as before but with one red AC connection made via a fuse.

An outside possibility is that if one of the new diodes is faulty open circuit then the rectifier will be working as a half wave (with 2 diode voltage drops) and that would give lower DC.


Rich

[petercom]

Please, anyone, bear in mind that if you decide, for whatever reason, to replace the diodes in the heater circuit with types other than the ones we supply you may need to adjust the series resistor in the PI filter between the DC heater capacitors.

The value of this resistor is given for a specific voltage output from the bridge. Because of the differences in voltage drop across different types of diodes the resulting voltage may increase or decrease from the design values.

So if you are replacing the diodes please check that the heater voltage fed to your valves (with the valves in place, of course) is between 6.1 and 6.3 volts DC.

[systemx]

Peter.
Thank you, understood.

Terry.

[petercom]

Originally Posted by Richard

Hi Nick,

Yes, it looks the same as before but with one red AC connection made via a fuse.

An outside possibility is that if one of the new diodes is faulty open circuit then the rectifier will be working as a half wave (with 2 diode voltage drops) and that would give lower DC.


Rich

It's beginning to look as though it is something like that. NickW is talking to Dr John about it and between us we'll sort it out.

[NickW]

Hi Peter,

John suggested I measured the diodes with my AVO on "diode setting" which I did early this morning. It looks like they all drop ~ 0.44V FB which rather suggests they are all working ok. I've talked to John, and we have concluded we are "puzzled" - it all seems to be ok, but is clearly behaving differently to intended. I have to be away next week with work, but John has very kindly suggested that I take the amp over to his when I get back and that we try and work out what is going on inside (it turns out that we live relatively close to each other).

regards, Nick

[systemx]

Received my 88XL kit on Thursday and have completed the LT side of things using 35v 105 deg caps in the filter and STPS1045D diodes in the bridge. With the supplied 0.47 ohm resistor and the circuit loaded with the valves, the voltage across the valves is 6.25, which is close enough. Time for bed.

Terry.

[Richard]

Originally Posted by systemx

Received my 88XL kit on Thursday and have completed the LT side of things using 35v 105 deg caps in the filter and STPS1045D diodes in the bridge. With the supplied 0.47 ohm resistor and the circuit loaded with the valves, the voltage across the valves is 6.25, which is close enough. Time for bed.

Terry.

Hi Terry,

That's good news.

Trying schottkies in Nick's amp might be a worthwhile excercise if only to eliminate the diodes as the cause of his particular problem.

Here's a good article which explains diode testing and a circuit to check Vf at higher than test currents,

http://www.allaboutcircuits.com/vol_3/chpt_3/2.html

After a bit of thought, I came up with an easier method using a constant current bench supply; stick it across the supply, wind up the current and read the corresponding voltage drop,




UF5408 1000V 3A fast recovery type showing 1.29V drop at 2A




1N5822 40V 3A Schottky showing 0.58V drop at 2A

Put these 2 types in a bridge and the resulting DC is the difference x2 = 1.42V. Add that to Nick's existing 4.8V = 6.22V which is quite close to your result. Just my thoughts anyway, there may be something else going on which John will find no doubt. Time for the pub!

Rich

[systemx]

Rich.
Wow, that,s a posh power supply!. At this end it's just me and a meter, so the only way to be sure was to build the circuit. Also as I have fitted physicaly larger caps, I wanted to make sure it would all fit. Today will be the relay panel and sig wiring from it. I intend to supply this post filter, so I will pick up the supply from the valve close to the selecter switch, the 6.25v was measured with a resistor across to draw an extra 30mA from the supply, so all should be o.k. I can then get back to following the build instructions.

Terry.

[Richard]

Hi Terry,

Yes, I rescued it from a mate's garage, he was using it for charging batteries! Nice bits of test kit come up quite cheaply from time to time if you're not in a hurry. A wander round the larger outdoor radio rallies in the summer is very pleasant and usually turns up something or other,

http://www.hamradio-badarc.co.uk/radio_rallies.htm

I have most of the regular stuff now at a total cost well under that of a single amp. It does add an extra dimension to the hobby.

Likewise adding bits of knowledge is interesting. 6 years ago I noticed the voltage difference between schottkies and standard silicon diodes, yet here we are now adding to that. It was discussed here at the time but no-one knew the reason or the spec to account for it. Paulo suggested the forward voltage drop but a search on the web was inconclusive. Even now if you search the web you'll find very little about it. The oft-quoted figures of 0.3V for germanium and 0.7V for silicon abound and past that not a lot.

However it clearly is a spec that's on all the data sheets. It's there on those graphs in post 28. The figure of 0.7V in reality just seems to be a typical figure for silicon at typical lowish current levels. Most of the time it will have little bearing but in low voltage apps (where the drop is a significant proportion of the output) it may be important.

Even at mA current levels the difference will be nearly a volt between schottkies and silicon used in a bridge. Here are some more pics from that session taken at the lowest current the ps would give, and they too agree with the graphs,



UF5408 1000V 3A fast recovery type showing 0.55V drop at 16mA




1N5822 40V 3A Schottky showing 0.16V drop at 16mA

Oh, getting carried away here good luck with the build, I can't see you getting a lot of sleep til it's done!

Rich