• The move to the new server is done. There are some software and database maintenance updates in process. This has us passing the hat around to help out. We appreciate any donations. Seriously, even a dollar helps. The payment page may be found here - https://www.audiokarma.org/support.html

full wave rectifier vs rectifier as a time delay

Pio1980 said:
Consider the VTA ST 120 on one conventionally connected 5AR4 attempting full power thru both channels.
Click to expand...

I wonder why it is designed that way, it clearly exceeds the ratings of the GZ34. Actually, even the original ST-70 pushed it to the limit, two in parallel would have been better. Maybe the designer counts on the fact the amp is seldom called on to deliver full power. This connection seems like a quick fix for those amplifiers who ask too much from a single rectifier.
 
maxhifi said:
Nice - this is a cheap way to cut the rectifier loss in half!!
Click to expand...

The voltage drop VS current is not a linear function. According to the charts, the reduction in voltage drop will be considerably less than half when the two sections are in parallel. Also, the parallel connection will create diode switching noise unless Schottky types are used, just as in a conventional SS supply. Whether this is a concern depends on other factors, but it should be recognized as a potentially negative characteristic of the modification.

Jack
 
maxhifi said:
I wonder why it is designed that way, it clearly exceeds the ratings of the GZ34. Actually, even the original ST-70 pushed it to the limit, two in parallel would have been better. Maybe the designer counts on the fact the amp is seldom called on to deliver full power. This connection seems like a quick fix for those amplifiers who ask too much from a single rectifier.
Click to expand...

The ST-120 should never have been produced with only one 5AR4 in the PS. It was clearly an attempt to upsell the product without incurring the cost of a different chassis. As for the supposed improvement in 5AR4 current capability from this mod (parallel anodes), the only gain lies in the reduction in dissipation that results from the decreased voltage drop. Each section of the tube otherwise carries the same average current as when the anodes are separate.

Jack
 
dcgillespie said:
Great job Gadget -- Were you using a single set of diodes to drive both rectifiers -- or double sets, as in one set for each tube?

The 22 watts obtained at the onset of clipping with both channels driven is definitely low. For reference, operating from a 121 vac power line, my early version SA260 would produce 30 watts RMS in both channels with both channels driven from 20 Hz to 20 kHz -- although THD at the frequency extremes rose considerably.

Dave
Click to expand...


Just one set of diodes.


I'm pretty sure I've clocked this at close to 30 watts output before. I'm also reasonably sure when I checked the power supply voltages after the initial overhaul they were pretty close to schematic. Power output was measured on the Sound Tech 1700a so its not just one multimeter giving me bad info. There is a thermistor in there, not impossible it has gone high resistance and is causing me problems.
 
So, if the amp is idling at recommended bias, it's already drawing ca 200mA without output product. Add output draw to bias and get?
 
TriodeLuvr said:
The ST-120 should never have been produced with only one 5AR4 in the PS. It was clearly an attempt to upsell the product without incurring the cost of a different chassis. As for the supposed improvement in 5AR4 current capability from this mod (parallel anodes), the only gain lies in the reduction in dissipation that results from the decreased voltage drop. Each section of the tube otherwise carries the same average current as when the anodes are separate.

Jack
Click to expand...
Using a PT rated for silicon rectification would have made great sense, and splitting the PS with another quad cap in the rectifier socket position and another choke.
 
TriodeLuvr said:
The ST-120 should never have been produced with only one 5AR4 in the PS. It was clearly an attempt to upsell the product without incurring the cost of a different chassis. As for the supposed improvement in 5AR4 current capability from this mod (parallel anodes), the only gain lies in the reduction in dissipation that results from the decreased voltage drop. Each section of the tube otherwise carries the same average current as when the anodes are separate.

Jack
Click to expand...
I found it odd that the optional tube set for the ST-120 includes a GZ34 on the VTA website, when Latino recommends a coppertop SS rectifier. One picture of the ST-120 on his site shows a GZ34, another shows a coppertop SS rectifier. :dunno: A GZ37 can be used if you want to pay $$ for one.
 
just to be clear here, none of this was about trying to squeak more power or voltage out of an amp. Whether it does that is entirely unrelated. The only question I was trying to answer is what, if anything, does bridging the plates on a rectifier do. The fact that it drops less voltage seems to imply that the tube operates with lower internal impedance. Indirectly this makes me suspect it should be able to handle more current when operated in this manner vs the "usual" way but I don't really have a way of proving this with any certainty.
 
6DZ7 said:
I can't speak for how a VTA120's power claim is figured. But I know for the ST-70 I have no worries about my 5AR4 being able to handle it. I just keep in mind that to get 3dB of sound power you have to double the wattage so when you wring your hands over a few % difference in B+ by paralleling the plates, .... I ain't bovered. And I think there's more to it than is being considered here, such as taking the cathode into consideraton, and if you lower the internal resistance which lowers the voltage drop and makes it easier for current to flow through it, what happens if you increase the 1st filter cap and make it pull more to charge it with a plate that is in constant conduction instead of having 1/2 cycle off. Will that also increase the risk of a hot start arc-over?
Click to expand...

Check this out, from the KT88 data sheet - this is for 70W output, which is about the same as the VTA ST120's 60W rating when you take into account output transformer losses.
gec KT88 DATA.png

A mono 60W KT88/6550 amplifier should need about 280mA of current available, and stereo is twice that, or 560mA.

As a single GZ34 is rated for only 250mA, it is quite clear why it struggles when it is trying to supply power to two channels! In fact it is nearly fully loaded just supplying the idle current of around 200mA, let alone full load current!
 
If I'm not mistaken, the 'pins 4 and 6 jumped together' configuration ends up deploying all four diodes in the two 5U4 (or 5AR4) valves in parallel full time (sort of akin to using ss full-wave rectifier with a single following damper diode between the rectifier and the filter).

With 4 valve diodes in parallel for each rectifier current pulse (rather than the original configuration where only two diodes are in parallel) then the rectified current pulse has a lower net resistance path to the first filter. WRT each of the valve diodes, its current pulse doesn't reach the same peak level and so the peak voltage drop is lower. A valve diode doesn't present a constant resistance to current (like a Coppertop), as the incremental resistance falls as the current level increases towards the peak level being passed. If one was looking at a datasheet plate voltage-current characteristic, then each rectifier pulse would not follow the plate curve as far along the V-I curve.

Similar to using a damper diode after ss rectification, the rectified current passed to the filter is still 'valve rectified' in characteristic in that the current pulse being passed to the filter has a softer start and finish, and lower peak level, and doesn't exhibit the rapid turn-off transient that just using a ss diode by itself would - (and also avoids any slight reverse recovery influence that may or may not be significant). Hence the first filter ripple waveform has lower harmonic content and is easier to then suppress hum than just an ss alone rectified configuration.

WRT to the first filter cap, that jumper mod would cause an increase in ripple current - that increase can be estimated by using say PSUD2.

If so inclined, and the amp can 'manage' a further increase in B+ and ripple current, then the 33R series resistor could by bypassed. The 33R value would likely be related to the OEM power transformer effective resistances and the 5U4 rating.
 
Last edited:
After reading this thread I think I can see why Bob retired. It appears he was building circuits that he had no clue about. He should have fixed this rectifier problem before retiring thou.
 
tonyg88 said:
After reading this thread I think I can see why Bob retired. It appears he was building circuits that he had no clue about. He should have fixed this rectifier problem before retiring thou.
Click to expand...
I think he went with Laurent and Tucker's notion that the amp wouldn't be stressed in normal domestic use with the single 5AR4, even tho it isn't the Laurent driver.
 
Pio1980 said:
I think he went with Laurent and Tucker's notion that the amp wouldn't be stressed in normal domestic use with the single 5AR4, even tho it isn't the Laurent driver.
Click to expand...

I totally understand that, but I like equipment which can put out full power "all day long" without adverse effect. I think it's poor practice to do otherwise. Once I used a Dynaco ST-70 in mono with an audio oscillator and a step up transformer to supply 50Hz power to a European record player. It consumed about 50W and didn't stress the Dynaco after hours of operation.
 
I did loose a 5AR4 in my ST-70. I don't crank the volume much at all and when I do it's not for long periods. I chalked it up to the JJ rectifier I used. I've had zero problems using a TAD 5AR4.
 
trobbins said:
If I'm not mistaken, the 'pins 4 and 6 jumped together' configuration ends up deploying all four diodes in the two 5U4 (or 5AR4) valves in parallel full time (sort of akin to using ss full-wave rectifier with a single following damper diode between the rectifier and the filter).
Click to expand...

yup, thats how its functioning.
 
maxhifi said:
I totally understand that, but I like equipment which can put out full power "all day long" without adverse effect. I think it's poor practice to do otherwise. Once I used a Dynaco ST-70 in mono with an audio oscillator and a step up transformer to supply 50Hz power to a European record player. It consumed about 50W and didn't stress the Dynaco after hours of operation.
Click to expand...
Agreed
 
Last edited:
maxhifi said:
I totally understand that, but I like equipment which can put out full power "all day long" without adverse effect. I think it's poor practice to do otherwise. Once I used a Dynaco ST-70 in mono with an audio oscillator and a step up transformer to supply 50Hz power to a European record player. It consumed about 50W and didn't stress the Dynaco after hours of operation.
Click to expand...
That's probably within the safe operating envelope of the OEM with the OEM 5AR4.
 
maxhifi said:
I totally understand that, but I like equipment which can put out full power "all day long" without adverse effect. I think it's poor practice to do otherwise.........
Click to expand...

In hind sight, it's hard to argue with the Hafler/Laurent/Tucker logic. They designed arguably the most popular stereo basic power amplifier in the world, that was hardly known for eating rectifier tubes. Their call to use but one rectifier -- based on the use patterns they envisioned their purchasers as having -- kept the price of the kit at the intended price point, which no doubt helped to rocket its sales. But for my own purposes, I absolutely agree with this comment: Coming from a band background in the 60's, this is a design approach I like to call "Band Rated". Like a Timex watch, it can take a lickin' and keep on tickin'. From a durability standpoint, the approach serves any design concept very well.

Dave
 
You must log in or register to reply here.
Back
Top Bottom