Diode rectifier vs Weber Copper Cap

[Helmholtz]

There are 2 versions of the 1N4007: The "Plastic Rectifier" type and the glass passivated (GP) type, both plastic bodied.
In my former company we used millions of 1N4007s.
Our quality department on a regular basis analyzed some percentage of claimed/failed products and fed back the results to the R&D depts.
The "Plastic Rectifier" version was found to have a significantly higher statistical failure rate than the somewhat more expensive GP type.
As a consequence it was decided to only use the GP version in our premium/professional products.

[trobbins]

That stands to reason.

Also note that any vintage power diode from the late 1950s to even up to the early 70's has likely been exposed to stressful voltage events and environment (especially humidity coupled with higher than ambient temp). It's relatively easy to check ss diodes for reverse leakage current at 1kVdc using a megohmmeter, and I have found it not uncommon for 1960's power diodes to show leakage (even at 500V for 600V rated PIV diodes). Maybe not as common as vintage e-caps showing excess leakage current higher than modern e-caps, but I put vintage power diodes in the same basket as vintage e-caps (ie. I replace them).

[catalin gramada]

My reservations for using the copper cap would be:


5) The Weber Copper Cap is a diode rectifier with a series resistance for sag in a copper bottle. If you use a diode rectifier you can choose your own series resistance that suits the design as you see fit.

JM2C

It will sag anyhow. For 15% voltage "regulation" as many vintages get in respect with nominal current it may be expected some between 50-100 ohm secondary internal resistance. Is better to check first to get an ideea how stiff you supply is.
just my 2c

[Chuck H]

It will sag anyhow. For 15% voltage "regulation" as many vintages get in respect with nominal current it may be expected some between 50-100 ohm secondary internal resistance. Is better to check first to get an ideea how stiff you supply is.
just my 2c

+1 on this point. Sometimes when designing an amp, and I don't usually use "clone" transformers, I'll need to compensate for other power supply specs to get the right amount of sag reaction from the WHOLE circuit.

[Pedro Vecino]

The only copper cap that I have disassembled, not in a tubular format to be inserted into a socket but a rectangular box attached to the chassis (I think it was a Fargen amplifier) contained two diodes, a resistor and what I think was an NTC thermistor.

[Helmholtz]

Sometimes when designing an amp, and I don't usually use "clone" transformers, I'll need to compensate for other power supply specs to get the right amount of sag reaction from the WHOLE circuit.

In typical amps using a GZ34, only about 1/3 of the B+ sag is due to the rectifier. Main contributor is PT resistance (I think that's what Catalin meant).

I like the GZ34 for its delayed ramp-up feature.

[Chuck H]

In typical amps using a GZ34, only about 1/3 of the B+ sag is due to the rectifier. Main contributor is PT resistance.

Yes. For example, the Hammond PT I used in my personal amp has a secondary DCR of about 100 ohms while, say, a Princeton Reverb can be over 300 ohms. You can hear this in an amp that's overdriving hard. Further, my amp doesn't use a rectifier tube. I found a 150 ohm resistor off the rectifier gave the amp a good "feel" for my own tastes.

[Helmholtz]

For example, the Hammond PT I used in my personal amp has a secondary DCR of about 100 ohms while, say, a Princeton Reverb can be over 300 ohms.

The reflected PT primary resistance adds to effective PT resistance. It's typically around 40R with a 50W amp and higher with smaller PTs.

A GZ34 roughly corresponds to 60R series resistance.

[pdf64]

In typical amps using a GZ34, only about 1/3 of the B+ sag is due to the rectifier. Main contributor is PT resistance (I think that's what Catalin meant).

I like the GZ34 for its delayed ramp-up feature.

It may be a chicken or egg scenario, ie the use of a valve rectifier should necessitate the equivilant resistance of the PT's HT winding being high enough to comply with the requirements noted in the valve type info. Or external resistance added.
So even though per se the rectifier may not be the most significant source of sag, if a valve rectifier is not being used, it's feasible to greatly reduce HT sag, ie by using a PT with a lower equivilant HT winding resistance.
Unfortunately the 40-18005 PT info above doesn't cover that.

Yes, the slow ramp up of HT voltage provided by GZ34 is a wonderful thing

[Greg_L]

As a player though, sometimes "sag" isn't wanted. I appreciate some sag when it's there, but in general I prefer the tighter response of SS rectification, all other things being the same.

[Helmholtz]

It may be a chicken or egg scenario, ie the use of a valve rectifier should necessitate the equivilant resistance of the PT's HT winding being high enough to comply with the requirements noted in the valve type info. Or external resistance added.
So even though per se the rectifier may not be the most significant source of sag, if a valve rectifier is not being used, it's feasible to greatly reduce HT sag, ie by using a PT with a lower equivilant HT winding resistance.

Yes, but I don't think there's much margin to lower effective PT resistance for a given PT size.
Lower resistance means thicker wire and thus requires more winding space. (Ok, omitting the 5V heater winding saves a little space.)
I think the GZ34 was just designed/optimized for typical PT and reservoir cap values in that power range.

A stiffer supply is likely to spoil a vintage amp's tone/response.

I like a good amount of sag induced compression. That's something SS amps don't provide.
For the same reason I hate oversized filter caps, which kill dynamic sag.
Never liked the results when I replaced a GZ34 with SS diodes in a vintage amp.
This said, I never liked the increased voltage drop/sag with a 5U4 either.

[Helmholtz]

Unfortunately the 40-18005 PT info above doesn't cover that.

From the PT datasheet loaded B+ values (upper left), the source impedance calculates as 150R.