R3000 Diodes, output tube plate to ground.

[Enzo]

Look at various brands of amps. ANy with flyback diodes. The diodes in each amp are doing the exact same thing, so if Fender uses R3000, Peavey SR2873, Bugera R2000, and so on, it's all OK.

[Leo_Gnardo]

Look at various brands of amps. ANy with flyback diodes. The diodes in each amp are doing the exact same thing, so if Fender uses R3000, Peavey SR2873, Bugera R2000, and so on, it's all OK.

And so on: I'll add MusicMan, and strings of diodes used in Ampeg SVT, V4, V2 & similar. Plus the highly prized Trainwrecks. Ken Fisher gives 'em a big recommendation in his Trainwreck Pages. No harm in pinching off the extension of hi voltage to just a tad beyond B+, cut the stress on output transformer insulation as well as tubes & sockets.

[g1]

Pulled out the micrometer... and the diode falls in specs to the datasheet - the diode was 5.25mm x 2.5mm

PY brand is Pingwei. Their datasheet for R3000 here: http://www.alldatasheet.com/datashee...ING/R3000.html

[Gtr0]

are these the ones you're looking for ?? GP02-30

https://www.mouser.com/ProductDetail...qzDBUV3A%3D%3D

I believe the way I ended up buying these was I ordered "R3000" from tube depot in the past, got it in the mail, read that it says GP02-30 on the actual diode, then looked it up on Mouser and ordered it from them, haha

edit: yes, zoom in on R3000 in the tube depot picture and you can see it says GP02-30

Those are the ones. However they are not in stock. And my experience with mouser is that even though they have a site in my country of residence (Belgium), they seem to ship everything from the US. I tried to order some KOA Speer resistors... the resistors were $12, shipping was $30! And that was the cheapest option. I would rather give that kind of money to a Mom & Pops retailer from the US than Mouser.

And so on: I'll add MusicMan, and strings of diodes used in Ampeg SVT, V4, V2 & similar. Plus the highly prized Trainwrecks. Ken Fisher gives 'em a big recommendation in his Trainwreck Pages. No harm in pinching off the extension of hi voltage to just a tad beyond B+, cut the stress on output transformer insulation as well as tubes & sockets.

Yep, I first saw one of those Trainwreck pages where there were 3 IN4007 in series. And that is also a viable option. Just figure a single R3000 would simplify things.

PY brand is Pingwei. Their datasheet for R3000 here: http://www.alldatasheet.com/datashee...ING/R3000.html

Thank you!!! I searched and searched. The diode falls in specs according to their datasheet. And I did buy a bag full... perhaps I will give them a go whilst looking for another supply.

[J M Fahey]

Since you have a bagful, nothing stops you from using 3 of them in series

So even if they are rebranded 1N4007 sold as theb real thing, you are still safe.

My main concern is not PIV but as mentioned above, creep voltage.

Dry air is an excellent insulator ... humid air or much worse: condensation on amp parts on a humid day is not, I would like to have 3 plastic bodies in series instead of just one.

Just curious about the Austrian ad: 38 Euro was the 1000 piece bag price or for a single one?

Just curious 2: please measure forward voltage drop with meter set to the diode scale ... and post it here of course.

[R.G.]

It's worth worrying a little about why those diodes are put there, and what they do. The general idea (I think) is that it prevents a sudden secondary-open or discontinuous-signal transient from making the OT primary think it's a flyback-ing inductor and flying its primary leads apart with high voltages that might arc a tube, but might also puncture the insulating film on the wires inside the trannie. If it punctures the insulating film, the trannie will either be toast immediately, or die more easily next time.

The OT primary in a P-P amp has two half-primaries held by a CT to B+. The diodes are arranged so that in normal operation they don't conduct, so the diodes have to have a reverse breakover voltage that is higher than 2X B+. Otherwise, the voltage on the turned-off output tube(s) would reverse-break and conduct on normal operation.

So the diodes that are active when a flyback/transient event occurs are always the ones which are on the half of the primary that happens to be headed towards ground, while the other half flies up above B+ by an equal amount. When the side headed towards ground gets below ground by enough to forward-bias the diodes there, the diodes conduct and clamp that half-primary to just a bit more than B+ across it. By transformer action, the other half-primary is also clamped to a bit more than B+, stacked on top of the real B+. That is, not hugely more than normal operation.

This is a great theory. Like most great theories, it's a bit too simple. Well, it probably works OK in many situations, but not all. The issue is that there is always leakage inductance on every transformer lead. This is inductance that by definition is not coupled to the rest of the transformer, and it still can cause voltage spikes on the high-side half of the PT primary that plate diodes won't clamp.

My preference has been to use some kind of clamping device across the whole primary or both half-primaries that will conduct at more than B+ across these diodes. Now both half primaries AND any leakage is clamped to just a bit more than the magnitude of B+, and only a bit more voltage than normal operation. But it does require some thought and application to select clamping devices of the right rating. And diodes are so very easy, no thought needed.

[trobbins]

For those protection diodes to conduct, there must either be some leakage inductance at play, or some dI/dt and then the valve stopping conduction rapidly. The other half-winding may not have any current flow if in class B action.

However it plays out, its better if practical to loop the protection current directly back to the winding (as per a shunt device across the winding), rather than through the diodes and back through the B+ supply to the winding mid-point. A MOV across each half-winding is a simple configuration.

[R.G.]

For those protection diodes to conduct, there must either be some leakage inductance at play, or some dI/dt and then the valve stopping conduction rapidly. The other half-winding may not have any current flow if in class B action.

I'm still on my first cup of coffee, so my parsing of this is definitely suspect. You're right, the catch/protection diodes only conduct when something tries to stop current flow in one of the windings on the transformer. There are two currents that could be interrupted, those being the magnetizing current that sets up the flux in the core so that transforming works, and the transformed current that ultimately is transferred to the secondary.

Opening the secondary pretty much affects only the current through the secondary leakage inductance, not the magnetizing current. This would create a spike that is transformed back through to the primary by transformer action. This transformation may be imperfectly carried to the primary windings due to the normal frequency response limits of the windings.

Opening a plate lead on the conducting tube will generate a spike from both the primary leakage and the magnetizing inductance. Lots of energy stored in the magnetizing inductance! But we're still dealing with imponderables about how much voltage gets generated from these different events.

The non-conducting half primary gets just as much voltage on a transient as the conducting one, as transformer action requires that all turns experience the same volts per turn (minus fuzzinesses like leakage and interwinding capacitances).

However it plays out, its better if practical to loop the protection current directly back to the winding (as per a shunt device across the winding), rather than through the diodes and back through the B+ supply to the winding mid-point. A MOV across each half-winding is a simple configuration.

I agree. I used MOVs in the Workhorse amps, but Steve ought to pop in here in a minute to remind me that MOVs have capacitance, while TVS devices have much less capacitance, and hence less chance of messing with the frequency response of the OT.

[Gtr0]

Just curious about the Austrian ad: 38 Euro was the 1000 piece bag price or for a single one?

Just curious 2: please measure forward voltage drop with meter set to the diode scale ... and post it here of course.

they were actually 6 or 7 Euros for a bag of 100


They all (the several that I have tested) measure at 1.5x Volts. Where In4007 and UF5804 measure at about .5V. Does that sound right??

[J M Fahey]

For those protection diodes to conduct, there must either be some leakage inductance at play,

Oh, the main primary inductance is there all the time and is way larger than any leakage one, by definition, so that is the most dangerous one.

or some dI/dt and then the valve stopping conduction rapidly.

Which happens twice every cycle on an overdriven amp.

However it plays out, its better if practical to loop the protection current directly back to the winding (as per a shunt device across the winding), rather than through the diodes and back through the B+ supply to the winding mid-point. A MOV across each half-winding is a simple configuration.

yes.
That said, MOVs must b e designed and rated, both for trigger/clamping voltage and dissipation ; while diodes take care of both points, automatically.
A nice feature I might add.

[Tom Phillips]

...They all (the several that I have tested) measure at 1.5x Volts. Where In4007 and UF5804 measure at about .5V. Does that sound right??

Attached is a copy of the common 1N400X data sheet and one for the R3000.
Note that the R3000 data sheet lists a 5V Maximum Instantaneous Forward voltage drop spec at 0.2A DC.
The 1N400X data lists Maximum Instantaneous Forward voltage drop as 0.93V typ and 1.1V Max.
The 1N400X data also lists Maximum Full−Cycle Average Forward Voltage Drop = 0.8V whereas the R3000 data does not include a similar spec.
R2500F-R5000F Data Sheet.pdf
1N4001-7 Data Sheet.pdf
We don't have an apples to apples comparison here. However, the data indicates that R3000 diodes can exhibit higher forward voltage drop than our familiar 1N400X rectifier diode. Therefore, based on the manufacturer's data, I don't think there is any problem with your diodes.

Cheers,
Tom

[Mick Bailey]

The forward drop of an R3000 is definitely 1.5v when tested with a DMM.

[Tom Phillips]

Makes me suspect that there are two diodes in series within the R3000 package.

[Enzo]

Years ago working on monochrome video monitors, the flyback transformer spits out 19kV for hte third anode cup on teh CRT, but it needs to be rectified. They used these "stick" rectifiers. About 1/4" diameter, and almost as long as a business card, they rectified all those kilovolts. Inside they were a stack of multiple little discs of diode.



Here is a nice article about the topic.
http://www.kronjaeger.com/hv/hv/comp/rect/index.html

[J M Fahey]

Oh, I saw the "gleich_gross.jpg" label on screen (sometimes my ISP connection is ssslllloooooowwwwwwwww and text loads before actual pictures, by a few seconds) and imagined the image might have referred to something like:



Then I remembered "gleichrichter" is German for "rectifier"