Yet another rectifier question...

[Paul-in-KC]

All,

I have another question about a rectifier circuit.

Apologies for asking here again, but I couldn't find this referenced anywhere.

So... looking at different ODS schematics, I have seen a couple of different rectifier circuits (not counting the full-wave voltage multiplier used on the relay power).

One is the ubiquitous full-wave bridge rectifier.
The other I see in the "generic 3rd Gen low-plate classic" schematic. It has each of the high voltage secondary legs running through three diodes in series and then merging together and then on to the first filter caps.

Pretty easy to intuit that the diodes are going to prevent any negative voltage through. But why use three diodes in series on each leg? Is it because actual components are not "ideal"? Is it to provide some redundancy in the circuit? So, as long as a diode doesn't fail "open" - you could have up to two diode failures in each leg? Although small, there would be additional voltage drop with the three diodes in series (right?).

Is there a name for this design? Is there a reason to use it - vs a full-wave bridge?

Thanks.
-Paul

[pompeiisneaks]

The six diodes is a protection thing yes, if a diode fails shorted, the others remain in operation, and it also handles more current roughly across the three so they share the load overall. the name for that type of rectification is just full wave rectification, not full wave bridged. Here's the differences between those two:
https://www.electronics-tutorials.ws/diode/diode_6.html

and this page that discusses more detail and math behind the full wave bridge:

http://www.circuitstoday.com/full-wave-bridge-rectifier

full wave bridged increases the output voltage compared to a full wave rectifier. I'm not sure I recall the exact difference but if I'm remembering correctly the full wave rectifier is usually 0.9 X VAC for VDC, but the full wave bridge rectifier is 1.4 X VAC for VDC. I'll see if I can find that for sure, but that's one of the bigger differences, it grabs voltage from every single part of the cycle more efficiently.

~Phil

[pompeiisneaks]

Sorry I got it confused a bit. the CURRENT allowed through is different. The 1.4 vs .9 I was remembering is common to all rectification for peak vs avg, and based upon the resistive, capacitor or inductor based load.

I'm attaching a pdf from hammond that talks about some of these things.

Basically to compare apples to apples, a full wave has more current throughput if with a capacitor than a full wave bridged, according to this document.

HammondRectifier.pdf

~Phil

[Paul-in-KC]

The six diodes is a protection thing yes, if a diode fails shorted, the others remain in operation, and it also handles more current roughly across the three so they share the load overall. the name for that type of rectification is just full wave rectification, not full wave bridged. Here's the differences between those two:
https://www.electronics-tutorials.ws/diode/diode_6.html

and this page that discusses more detail and math behind the full wave bridge:

http://www.circuitstoday.com/full-wave-bridge-rectifier

full wave bridged increases the output voltage compared to a full wave rectifier. I'm not sure I recall the exact difference but if I'm remembering correctly the full wave rectifier is usually 0.9 X VAC for VDC, but the full wave bridge rectifier is 1.4 X VAC for VDC. I'll see if I can find that for sure, but that's one of the bigger differences, it grabs voltage from every single part of the cycle more efficiently.

~Phil

Thanks Phil!

Much appreciated.

[Paul-in-KC]

Sorry I got it confused a bit. the CURRENT allowed through is different. The 1.4 vs .9 I was remembering is common to all rectification for peak vs avg, and based upon the resistive, capacitor or inductor based load.

I'm attaching a pdf from hammond that talks about some of these things.

Basically to compare apples to apples, a full wave has more current throughput if with a capacitor than a full wave bridged, according to this document.

HammondRectifier.pdf


~Phil

Thanks again!

[Mr. dB]

Three diodes in series gives you three times the PIV, so three 1N4007s, which each have 1000V PIV, gives you the equivalent of a 3000 PIV diode (with slightly more voltage drop).

[Paul-in-KC]

Three diodes in series gives you three times the PIV, so three 1N4007s, which each have 1000V PIV, gives you the equivalent of a 3000 PIV diode (with slightly more voltage drop).

Ah, good point.

[sluckey]

But why use three diodes in series on each leg? Is it because actual components are not "ideal"?

Yes, that's the short answer. But a little more info should completely explain why 3 diodes were used. The long answer is part history and part technical...

Back in the late '50s and early '60s silicon diodes were just breaking into the consumer market. But there were no affordable 1000PIV diodes. 400PIV @ 500mA diodes were common and just becoming cheap enough to use in consumer electronics.

Also at that time, musicians were hungry for more powerful amps. Fender's answer was the Showman and Twin Reverb. These amps needed a lot of current to achieve the power goal. Dual rectifier tubes could provide the power that a quad of 6L6s needed, but that also requires a more expensive PT. Fender chose to switch to the more economical/reliable silicon diode. But a single 400PIV diode would fail with the voltages that were required to meet the power goal. The easy answer was to put three diodes in series which provided a 1200PIV rating. Other manufacturers were making similar changes in their high power amps. I chose the Fender AB763 TR as my example just because their six diode rectifier is pretty well known and the schematic has voltages listed which makes further understanding easier.

So why six diodes? One 400V diode is obviously not suitable for a B+ of 460VDC. But why not just use two diodes? That would give an 800PIV rating. Here's why...

Refer to the attached schematic. The PT puts out 340-0-340VAC RMS. That's equal to 480-0-480 VAC Peak. I'm gonna talk about peak voltage because that's what the diodes have to work with. The diodes pass the 480V positive half cycle of the AC to the filter caps. The caps charge up to this peak pulse and hold it. (The schematic shows 460V but I'm talking about the theoretical value of 480V.) Meanwhile, the AC voltage has flipped polarity and there is now a 480V negative pulse applied to the diodes. The diodes will block this negative pulse, but consider this... THERE IS A NEGATIVE 480V PULSE ON ONE SIDE OF THE DIODE AND THERE IS 480VDC ON THE OTHER SIDE OF THE DIODE! THAT'S 960V PEAK DIFFERENCE ACROSS THE DIODE!!!

And there's the answer. Two series 400PIV diodes would also fail. The minimum number of 400PIV diodes that could reliably be used is three. Fender continued to use six diodes up until they switched over to a full wave bridge rectifier.

So, that's why Fender used six diodes. I'm sure that if 1000PIV 1N4007s that only cost 4 cents apiece were available back in the '60s, Fender would have only used four diodes. Silvertone would probably just use two!

[norburybrook]

thanks for that Steve. I've learned something new. I've never questioned the diode arrangement so I'm gald someone has and we've had a good explanation.

M

[pompeiisneaks]

And again we see the massive brain and storage of information Steve's head holds I knew generally I'd had this explained to me before but forgot enough that my description was mediocre at best lol. Thanks!

~Phil

[dorrisant]

I'll pre-order his memoirs...

[Paul-in-KC]

But why use three diodes in series on each leg? Is it because actual components are not "ideal"?

Yes, that's the short answer. But a little more info should completely explain why 3 diodes were used. The long answer is part history and part technical...

Steve,

Fantastic - thanks for taking the extra time to provide the history lesson. I'm glad some other TAG members also learned something.

I feel like Paul Harvey should step in and say - "...and now you know the rest of the story"

-Paul