Rectifier spec explanation

[jordan86]

I popped in my newly acquired 5Z4 into a 15w (2 el84) Dr Z amp. It calls for a 5Y3 but Dr Z has given the go ahead on any of the standard 2A rectifiers. So I have tried them all at some point.

It sounds interesting. To my ear it is sounds and feels louder/punchier than a 5Y3 but cleaner(less dirt/drive) than a 5V4. I thought it was a straight 5Y3 but I now see it’s categorized as a low drop 5Y3. I like the tone a lot. Just puzzled as to how it’s effectively different than a 5v4 or 5AR4. The voltage drop is actually less than a 5V4. Very close to a 5AR4. But it’s a lower power rectifier I guess? I’ve not opened it up to check voltages on the EL84’s. But I presume it’s closer to what it would be with a 5v4.

What I don’t understand is how the maxPma and maxPV matter for an amp. Especially a smaller amp that doesn’t approach the max values of the 5AR4.

Here’s the specs I found on an old tube radio forum. I do understand the filament values and voltage drop and how that relates to sag.

Just trying to figure out if these specs would explains my tonal perception.

Tube# -
Fvolt-Famp- Vdrop- MaxPmA-MaxPv
5AR4/GZ34-5DA
5.0 - 1.9 - 17 - 250 - 450

5V4-G/GA
5.0 - 2.0 - 25 - 175 - 375

5Y3-G/GT-
5.0 - 2.0 - 60 - 125 - 350

5Z4 -
5.0 - 2.0 - 20 - 125 - 350

[trobbins]

There would be a variation in B+ level at idle, which would also affect cathode bias (assuming it is a cathode biased output stage), which would affect the sound for most of the volume range.  And yes as output signal gets high for some of what you hear then the sag characteristic would become noticeable.

If you look at any of the datasheets for those valves (eg. go to https://frank.pocnet.net/sheets5.html), not just a listing of some parameters, then locate the average plate current characteristic plot and compare how the plotted curve starts at 0V drop for 0mA current, and rises up with a bit of a curved characteristic.  Your amp will make each plate cyclically rise up from the origin (0V,0mA) to some peak point along that curve and then fall back to the origin.  For idle and up to medium sound output, the peak point is likely well under the max end point of a curve, and may only extend up to the end point of the curve for cranked output.  So you could choose (for comparison) a peak current point and see what the voltage drop gets to - although that voltage drop is not exactly what you would observe as a difference in B+.  Also the ripple voltage waveform is affected by how the current rise and fall cyclically changes and to what level it gets to and to what curve it follows, and the ripple harmonics therefore change too and that can affect how a note sounds as the mains harmonics modulate the note (and sometimes to a noticeable degree).

Note that any one sample of a rectifier valve could have noticeably different performance than any other sample of the same make/model/age etc, especially as those rectifier tubes have two valves in them that can each be a bit different than the other.

Note that it's best to get someone else to swap the rectifier valves when doing your kind of assessment, so you don't know what valve is in there.

[pdf64]

I popped in my newly acquired 5Z4 into a 15w (2 el84) Dr Z amp. It calls for a 5Y3 but Dr Z has given the go ahead on any of the standard 2A rectifiers. So I have tried them all at some point...

There must be quite an increase  in HT voltage, going from the stock intended 5Y3 to a GZ34. As such amps are typically fairly hot biased, I wonder the power valves can accommodate that, ie without being overstressed to some degree?
Did you check the HT voltage with each rectifier type, ideally at idle and also when cranking out full overdriven power?

[tubeswell]

Here’s the specs I found on an old tube radio forum.

Frank’s is a good resource for data sheets https://frank.pocnet.net/

[jordan86]

I popped in my newly acquired 5Z4 into a 15w (2 el84) Dr Z amp. It calls for a 5Y3 but Dr Z has given the go ahead on any of the standard 2A rectifiers. So I have tried them all at some point...

There must be quite an increase  in HT voltage, going from the stock intended 5Y3 to a GZ34. As such amps are typically fairly hot biased, I wonder the power valves can accommodate that, ie without being overstressed to some degree?
Did you check the HT voltage with each rectifier type, ideally at idle and also when cranking out full overdriven power?

I did at one time. I recall at idle they were roughly...

300v with 5Y3GT
340 with 5v4
350 with 5AR4

[HotBluePlates]

... my newly acquired 5Z4 ... Just puzzled as to how it’s effectively different than a 5v4 or 5AR4. ...

The 5Z4 data sheet says DC Output Current is 125mA max.  That's on-par with typical operation of a 5Y3GT (page 2).

The Peak Inverse Voltage rating of the 5Z4 (1100 volts) is lower than the 5Y3's rating of 1400 volts, so it's less-capable in this area.  Better have a lower voltage coming out of your power transformer: (1100v / 2) / √2 = ~389vac (and preferably much less).  This is why the 5Y3's data sheet (page 4) shows higher "Volts (RMS) Per Plate."

Comparing charts between the 5Z4 and 5Y3, they do seem to imply the 5Z4 has slightly less internal resistance than the 5Y3, but that's a tricky comparison.  A lot of times, the resistance of the power transformer supplying the rectifier, and the amount of capacitance used to generate the chart influence the graphed results (just as they will in an actual amp).

RDH4 Page 1171 comes to the rescue showing that at 100mA the 5Z4 is ~190Ω while the 5Y3 is ~500Ω.  So yeah, the 5Z4 is lower-resistance than the 5Y3.  But the 5V4 is even lower-resistance than the 5Z4 (I won't guess where the 5AR4/GZ34 falls on this chart, which apparently pre-dated the 5AR4).

... 5Z4 ... sounds and feels louder/punchier than a 5Y3 but cleaner(less dirt/drive) than a 5V4. ... trying to figure out if these specs would explains my tonal perception. ...

Before getting to specs, you don't know for sure how amp-voltages changed with each rectifier.  Sonic impressions probably have a lot to do with voltage delivered to the rest of the amp via the power supply.

I'm guessing you also don't know how any of these rectifiers measure on a tube tester, to know if one is more-worn than another (and less-able to deliver power demanded by the power supply & output tubes).

[jordan86]

Sounds like I need to pull the chassis and do some real data collection again. Thanks guys. Always learn more from you guys here.

And I really like the tone and feel of that 5Z4. All my tube rectified amps are less than 20 watts, so I think I may snag a few up. Dirt cheap too. If it is supplying B+ similar to a 5AR4 it could be a great sub for that in a small amp that doesn’t need the full power the 5AR4 provides. Thinking my Princeton may be the next candidate.

[jordan86]

Couldn't help myself. Had a free hour and pulled the M12 chassis and the RXJR to measure voltages with different rectifiers. Dr Z has talked about the stock M12 being biased cool from the shop, hence the flexibility to with go bigger rectifiers. I'm pleased to report the 5Z4 weighs in at a near perfect 11.2w plate dissipation. Attached my spreadsheet if anyone is curious how the modern JJ rectifiers measure against the NOS variants. All measurements taken at idle.

Doesn't totally answer my original question, but feels good to know where my voltages are at with future swapping and tinkering.

Side note: The NOS 5Y3GT's get WICKED hot. I doubt it's a concern but certainly took me by surprise when swapping them out.

[tubeswell]

The NOS 5Y3GT's get WICKED hot.

So the getter hasn't gone all powdery white?

Other than that (assuming they're conducting properly) each 5Y3 plate can be assumed to be seeing the VAC x [the forward current (which equals the HT load)]/2*.

*It's 'divided by 2' because the forward current only gets pulled through each plate during the positive part of the VAC cycle (for each plate).


Also (in terms of Rob's chart**), the 5Y3 cathode is connected to the reservoir cap, and will be sitting at B+ VDC, with the positive peak of each VAC pulse from each alternate plate cycle making the plate-to-cathode voltage difference.
 
**So the way it is recorded on the chart that you have attached from Rob is 'wrong' for rectifier tubes.


What's more, the rectifier cathode only sees the full forward current from both plates during the 'charging current' part of the reservoir cap's charging cycle. (This isn't over the whole DC cycle, because the charging current doesn't occur constantly throughout the rectified positive pulse cycle. The charging current stops after the reservoir cap's peak charge has passed, and only kicks in again after the reservoir of charge in the cap has discharged to the point where it falls below the next rising VAC pulse from each rectifier plate (at which point, another burst of charging current occurs as the rise in positive voltage pulls more electrons from the positive pole of the reservoir cap, until the next rectifier plate VAC pulse peaks again, and so on). So each burst of charging current is shorter than the full VAC cycle, and therefore is more intense than the 'average forward current', because it has a shorter time to do the amount of work needed to get the reservoir cap back up to peak charge. The more worn out the rectifier tube is, the less robust it is likely to be in being able to meet these demanding pulses of charging current. Also, the bigger you make the reservoir capacitance, the harder the rectifier tube has to work during the charging current part of the cycle to get the cap up to peak charge. Hence we have an upper limit for reservoir capacitance on tube datasheets.

[HotBluePlates]

... Hence we have an upper limit for reservoir capacitance on tube datasheets.

Implied, more than stated.

The is a Peak Current Limit that is clearly stated.  More µFs move the tube towards & past that limit, as you noted.  But Series-R (either from the transformer's windings, or from actual resistors in series each plate) will back the tube away from that limit.