Bridge rectifier vs. two-diode rectifier?????????

[anson]

Hey guys,

I've got a 120 watt homebrew amp and am using a bridge rectifier for the HT voltage. My transformer has a center tap, but I used a bridge rectifier anyway just out of cenvention. Is there any advantage to using a bridge rectifier over the simpler two-diode rectifier seen in many amps???????

Is it more effecient, or safer? does it exhibit less sag? thanks for your help!!!
Anson

[R.G.]

Hey guys,

I've got a 120 watt homebrew amp and am using a bridge rectifier for the HT voltage. My transformer has a center tap, but I used a bridge rectifier anyway just out of cenvention. Is there any advantage to using a bridge rectifier over the simpler two-diode rectifier seen in many amps???????

Is it more effecient, or safer? does it exhibit less sag? thanks for your help!!!
Anson

If you left the center tap on ground and hooked the ends of the HV winding to the AC posts on the bridge, took B+ off the + output of the bridge and left the - output of the bridge open, it is EXACTLY the same in an electrical sense as a two diode setup. There is no electrical good or bad about it.

The only difference is that there are two additional diodes that take turns connecting the - bridge terminal to the then-unused half of the HV winding. As long as the - bridge terminal is unused, no good or harm happens.

If you connect something to the - bridge terminal, then the fun starts. That point is sitting at the negative mirror image of B+ but without a filter cap. Your finger touching that point would be very unhappy. So you now have a new dangerous point in your amp.

[Enzo]

I interpret the question different from the way RG did. I might be wrong, but...

What RG said was right or course, but it depends what you do.

If you ground the center tap and use two diodes, then you are rectifying half the transformer winding voltage. If you used a bridge the same way, the voltage would be identical, but you cannot ground the negative corner of the bridge.

Most amps that use a bridge are rectifying the whole transformer winding. The negative bridge leg is grounded and the + leg is the B+.

So if you have a 350-0-350AC transformer, grounding the CT and using two diodes - or half a bridge - provides 350VAC rectified. If you put a bridge on the winding and leave the CT not connected, you wind up with 700VAC rectified.

[Gregg]

I stopped using center tapped PTs long time ago. If you want a tube rectifier, then yes. Rest of the time I use toroidal PTs with a single HV winding + bridge rectifier.

[Enzo]

Sneaky little spammer digs up a 3-1/2 year old thread.

[jpfamps]

I stopped using center tapped PTs long time ago. If you want a tube rectifier, then yes. Rest of the time I use toroidal PTs with a single HV winding + bridge rectifier.

You can still use a valve rectifier with a bridge rectifier by using two silicon diodes for the other half of the bridge.

This has the added advantage of reducing the PIV on the valve rectifier.

[Steve Conner]

Correct. Center tapped PTs are a throwback from the days when adding two more diode valves was more expensive than just adding more iron and copper to the PT.

Now, with a bridge rectifier you can use a smaller PT for the same rating, and you can still have some tube mojo by using two vacuum diodes and two silicon ones.

That is, if you can find a PT that has a HT winding designed for a bridge rectifier, but also a 5V heater winding. Good luck on that one.

[jpfamps]

Correct. Center tapped PTs are a throwback from the days when adding two more diode valves was more expensive than just adding more iron and copper to the PT.

Now, with a bridge rectifier you can use a smaller PT for the same rating, and you can still have some tube mojo by using two vacuum diodes and two silicon ones.

That is, if you can find a PT that has a HT winding designed for a bridge rectifier, but also a 5V heater winding. Good luck on that one.

Tube Town in Germany do a range of toroidals for valve amps with 5V windings.

Although it looks like they are centre tapped, they actually have two HT windings which can be wired in parallel using a bridge.

[Wilder Amplification]

One thing I will add is that the transformer has to be made for the rectifier you're using whether it be full wave grounded center tap or full wave bridge.

What I mean by this is that typically on a transformer that is made to be used with a FWCT rectifier, the winding is made for DOUBLE the required AC voltage at 1/2 the required current. This is due to the fact that with a grounded center tap rectifier, you're rectifying 1/2 the full winding voltage. Typically the way it works is that each 1/2 of the secondary operates at a 50% duty cycle (i.e. 1/2 the winding is only passing current for 1/2 the full AC cycle). This allows you to pull DOUBLE the full winding current from 1/2 of the full winding at 1/2 the voltage. Because of the 50% duty cycle of each 1/2 winding, this makes the AVERAGE winding current equal the full winding current.

With a transformer that's made to use a full wave bridge rectifier, the full winding is rated for the required voltage and required current. The entire winding operates on a 100% duty cycle (i.e. the full winding is passing current 100% of the time). If this transformer is center tapped and you try to operate it with a full wave grounded center tap rectifier, you'll only have 1/2 the required AC voltage, which won't be enough voltage to operate the amplifier.

What you CAN do with that center tap on a transformer that is made to use a full wave bridge rectifier is wire up a full wave bridge voltage doubler just like the older 100 watt Marshalls use. Before I go on, I will say that this circuit isn't really a "doubler" at all unless you look at it from the perspective that you're "doubling" the center tapped voltage. However, having the bridge across the full winding does that anyway.

What it DOES do though is allow you to have the first filter caps wired in series, then use the center tap at the series junction of the caps to balance the voltage across each cap without having to use bleeder resistors on those two caps. What happens here is that each 1/2 of the secondary winding charges up each cap to 1/2 the DC voltage output. The full wave bridge is constantly switching which 1/2 of the winding is charging up which cap.

[Steve Conner]

Yes Jon, except the wire has to be sized for the RMS current, not the average. With every other half cycle missing, the average is cut in half as you point out, but the RMS only goes down to 0.7 of what it was. (0.7 is the square root of 0.5.)

So the FWCT transformer needs two windings each of 0.7 = 1.4 = 40% more copper in its secondary than a bridge one of the same rating. The primary is unchanged, so the transformer as a whole ends up 20% bigger than if it were designed for a bridge.

If you take a transformer that can be configured either way, like the tube town type ones as recommended by jpfamps, I guess they'll deliver 20% more juice without overheating when wired to a bridge rectifier.

[jpfamps]

If you take a transformer that can be configured either way, like the tube town type ones as recommended by jpfamps, I guess they'll deliver 20% more juice without overheating when wired to a bridge rectifier.

Just wired an amp using a TT135 (345-0-345 allegedly) and got a loaded HT of 505 VDC when wired with a bridge, which is fine for my application, but is a bit high for a 50W Marshall for which this is designed.

I'm sure with full wave rectification (a la Marshall) the HT voltage would be lower*.

Incidentally, in the absence of a bias winding (and my reluctance to try adding my own bias winding) and a separate transformer to derive a bias supply (it was somewhere in the Royal Mail system), I used a capacitively coupled bias supply, which I'm not overly impressed with as it doesn't seem that stable. The transformer has now arrive so this is being rectified.

* We're always told NOT to use full wave rectification with toroids (although several of the Tube Town toroids are designed so they can be wired for FWR), which I've always assumed is due to issues with DC offset. Is this correct? I did ask a transformer winder whether using FWR was a good idea, and got the slighty cryptic reply "I've never been asked to wind a transformer like this".

[Steve Conner]

I can't think of any reason not to use full wave rectification with a toroid.

Except that if you're paying extra for a toroid, it's presumably because you're interested in smaller size and higher efficiency. So why would you throw away 20% of the advantage you paid for by using a full wave?

When you change from FW to bridge by reconfiguring two windings as discussed here, to a first approximation the unloaded output voltage doesn't change at all. (there is still the same number of turns driving the rectifier.)

The only difference is that it will sag more under load. Again to a first approximation the sag will be 50% more than it was (the secondary resistance is doubled, because only one wire is used at a time, instead of both in parallel) but toroids tend to have quite low sag to start with.

Toroids are very sensitive to DC magnetization. In a low voltage power supply, if the two diodes weren't matched, that could generate a DC offset. But I think in the voltages used in tube amps, that must be a complete non-issue. The classic power supply for a SS power amp might look like a bridge, but it is two full-wave rectifiers, one feeding each rail.

You could argue that even if the diodes were matched, maybe the two halves of the secondary would have different DCR or leakage inductance, and there would be your DC offset. But as far as I know, the two secondaries on toroids are always wound bifilar in one go, so they are identical. Saves time on the winding machine.

Now if you try a HALF wave rectifier, the results are messy.

[jpfamps]

Toroids are very sensitive to DC magnetization. In a low voltage power supply, if the two diodes weren't matched, that could generate a DC offset. But I think in the voltages used in tube amps, that must be a complete non-issue. The classic power supply for a SS power amp might look like a bridge, but it is two full-wave rectifiers, one feeding each rail.

You could argue that even if the diodes were matched, maybe the two halves of the secondary would have different DCR or leakage inductance, and there would be your DC offset. But as far as I know, the two secondaries on toroids are always wound bifilar in one go, so they are identical. Saves time on the winding machine.

Now if you try a HALF wave rectifier, the results are messy.

I hadn't thought about the rectifier diodes not matching.

I expect you're right about this not mattering too much for a valve HT supply.

[Wilder Amplification]

I expect you're right about this not mattering too much for a valve HT supply.

Seeing as how in electronics we deal mostly with percentages rather than absolute values I would say Steve Conner is correct.

[vmazz]

Ive got a question related to this topic. I have a PT that has 2 HT windings one higher, about 350-0-350 and other 300-0-300, the lower one has a CT the higher one does not. I want to hook them up to a DPDT on-off-on switch for my standby and High and low voltage. This will be a marshall 50 watt style power section and id like to hook up the CT on the lower voltage winding to a junction between series main caps as mentioned above. But I am worried about what would happen when I switch over to the high voltage winding setting and the CT for the low is still hooked up. These will likely be feeding into a bridge rectifier configuration. What are the problems with keeping the CT hooked up ? I realize bias needs to be adjusted between the 2 settings, and I may eventually use a 3pdt with a resistor to adjust the low voltage bias as metroamp kits did in the past. Any ideas on this setup? Thanks