Bendix 6106/5Y3WGT vs. generic 5Y3GT

[Brent Hutto]

I've seen comments online, including on this forum, that the fancy-pants Bendix 6106 with its indirectly heated filament and bulletproof construction should be expected to exhibit less voltage drop than a generic NOS 5Y3. I have one of those rectifiers on order from KCA and decided to research the matter to see what I should expect in my amp.

I've spent a few minutes comparing PDF's of ancient datasheets, one labeled as being from Bendix (Red Bank) and the other from Tung-Sol (NJ). Under standard operating conditions, the "rectifier efficiency" on those two datasheets appear very similar with if anything the Bendix (6106) one being lower by a miniscule amount in the range of 40, 50, 60mA per plate.

So out of curiosity I'm going to measure up my little single-ended amp with the NOS 5Y3 in it and then again with the 6106 and post on here with my findings. I think that will be a fun educational experience for me and perhaps someone somewhere will find it useful.

[phsyconoodler]

Yup.Put it in and measure the voltage differences.If less voltage drop is what you are looking for,a modern 5Y3,which is not really a 5Y3,may be what you are looking for.A 5V4 can be good too.
Let us know what the Bendix does for you.
According to the data sheets,it is a replacement for the 5Y3 where extra durability are required.The bonus is you can mount it in any position,unlike a 5Y3 which needs to be mounted vertically.It has a 45 second heater warm-up time which may be a bother in this day and age of lightning fast everything.

[Brent Hutto]

My first choice would be for it not to change the voltage drop at all. I'm just swapping the 6106 in there because there's a microphonic ring (at about F#) somewhere in there and I've eliminated the preamp and power tube by swapping. I didn't have another 5Y3 so while I was ordering, what the heck I sprung for the heavy-duty Bendix.

So my concern was maybe it would have less resistance and increase my plate voltages noticably. But based on the datasheets that seems unlikely. I'm betting it doesn't make 10 volts difference one way or another but the only way to know is to measure.

[Blue Strat]

What Phsyco said. It was murder testing these with the extended warmup!

[redtoploader]

The Bendix is a great choice. Those things will last a lifetime.

[Brent Hutto]

I don't have the 6106 yet but I went ahead and practiced doing my measurements with the 5Y3 in there. I have two different power tubes and switchable cathode bias resistors so I did it with two different total current draws. The entire load on the rectifier is 41.1mA with the JJ 6V6S and 54.7mA with an NOS 6L6GC.

At the lower current draw, my measured eta=0.849 is a little worse than the datasheet efficiency for that current (0.875) but right in the ballpark. Upping the total load to 54.7mA I got an eta=0.816 which is better than the datasheet predicts (0.785) so the slope on my particular old 5Y3GT is a little shallower than spec'd.

My numbers were:

339.0VAC/338.8VAC input and 407VDC/407VDC output at 41.1mA load

337.3VAC/337.0VAC and 389.5VDC/388.4VDC output at 54.7mA load

and I derived the total current load by measuring the voltage drop across the first 1K/5W resistor in the high voltage chain (measured value 985ohms).

I also have a question about the operation of my amplifier. Not sure about the 6L6GC but with my JJ 6V6S and the 680ohm cathode bias resistor selected my screen voltage is slightly higher than the plate (362.4VDC screen, 361.8V plate) which seems odd. There's a voltage drop (correct polarity) of 0.8VDC across the 470ohm screen resistor so the grid is drawing current. I think that's what it's supposed to do but I thought it was a bad thing to ahve the screen go positive w.r.t the plate...

[Blue Strat]

You might want to ask the amp "designer" why he's running a 12/14 watt tube, the 6V6, at 16.7 watts. You're not supposed to exceed the max dissipation of the tube.

Note that I don't warranty tubes that are operated beyond their maximum specified limits.

[Brent Hutto]

I think my presentation of the numbers wasn't clear, sorry I typed in a hurry.

The 41.1mA is the total load of the entire amp (power tube plus 12AX7) with the 6V6. The higher load of 54.6mA is for the entire amp with a 6L6GC. I wanted to look at the entire load on the rectifier to figure what eta it was running at.

My cathode currents are 37mA with the 6V6 and 53mA with the 6L6 at plate voltages around 360 and 330 respectively (and around 25-26V of cathode bias). Making the dissipations 12.5 and 17.5 watts or so...

[Blue Strat]

Ah! That sounds a lot better!

[Brent Hutto]

I've not read up on 6L6's as much as 6V6's but it's my impression a 6L6GC is just loafing along at 17-18W plate dissipation. IIRC it is spec'd for 30W or 25W or something ridiculous like that. But at 12.something watts a 6V6 ought to be cooking along at a pretty good boil.

[Blue Strat]

Yep, these amps that run both 6V6s and 6L6s are incapable of running both types at their full potential. There's always a compromise. It's unusual and refreshing to see that the max spec of one type isn't being sacrificed.

[Brent Hutto]

I exchanged several E-mails back and forth with Sara during the shopping phase. She said the goal was to get it operating with a little less tube abuse than a vintage Blackface in 6V6 mode and then to have a really conservative, clean Class A sound with the 6L6.

I took that to mean she didn't intend to get the same kind of tone with each and just allow tube-rolling for the heck of it. I think the amp has two quite different sounds, all considered, and it's pretty darned obvious which tube you have in there.

Of course it always sounds like a 12" Jensen Neo in a good-sized cabinet, I guess. That probably determines the basic tonal range no matter what's going on in the electronics. Fortunately, I seem to like the sound of a 12" Jensen Neo!

[Trout]

I've not read up on 6L6's as much as 6V6's but it's my impression a 6L6GC is just loafing along at 17-18W plate dissipation. IIRC it is spec'd for 30W or 25W or something ridiculous like that. But at 12.something watts a 6V6 ought to be cooking along at a pretty good boil.


It might be interesting to try a 6L6G in that amp. The older ones have a lower dissapation rating around 19W? Or maybe an RCA 6L6GB might be the hot ticket.

I'll have to dig around in my garage and see If I have any, I still have a lot of unsorted boxes of tubes from a TV shop.

Trout

[Brent Hutto]

It's an interesting question whether an older tube running hard might have a little edgier or lively sound than a higher-spec one that's loafing. Especially given the miniscule amount of NFB on this particular amp a tube wanting to break up a bit ought to have a Tweed-y good time of it.

Any advice concerning my screen-voltage query? Isn't it unusual to have the screen a couple volts positive w.r.t. the plate? Or does that not matter as long as the screen current is in the proper direction?

[Trout]

I beleive her screen voltage is far enough under the anode voltage that it should not be an issue. There is also a screen resistor not common to the original BF champ in mine.

Rule of thumb is lower is better than over, but like all things guitar amp, margins vary.

Soon as I get a couple hours to get in the garage attic, I am about 99% sure I have some G's & GB's in a box up there, I will test a couple out and see what happens.

[Brent Hutto]

There is also a screen resistor not common to the original BF champ in mine.

In mine it's a 470/2W connecting the screeen to the low side of that second 1K step-down resistor in the high voltage chain. I need to check my voltages more carefully this evening, when they're only like one volt different it would be easy to write them down backward.

Hopefully, my new Bendix rectifier will be waiting in the mailbox when I get home. And a "spare" (a.k.a. another to play with) 6L6GC too. I can measure all kinds of configurations.

[Groovey Records]

Besides the never ending hunt for the EF86 tube sound I crave, (you can never have too many) The final touch on my TopHat Supreme 16 has been finding an EZ81 Mullard NOS from 1959.

The tube does its job and it does it noticbly better then the current production
Electro Harmonix- which is no slouch- Until Mike Mathews put these on the market there was no reliable source for EZ81 tubes so tube rectification was handeled other ways in EL84 Amps.

Now my TopHat has more touch sensitivity along with the sag and bloom.
The Cut control is much more effective in shaping tone, instead of have ing effect in only the last 25% of adjusment the cut control shapes the sound with greater effect and an increased frequency range.

Visually now all the tubes have settled into a more uniform appearance. The glow of the tubes (except the EF86) is evenly matched.

This latest tube puts the S16 together with a full set of late 50's and early 60's Mullards.

For the adjective police out there I simply state that my distortion generators sound like I like them to sound. The amp now generates distorted tones even better then it already did.

Old and Improved Sound, cool

EnJoY ThE MuSic
GrooVey RecOrds
listening to Houses Of The Holy

[Brent Hutto]

My Bendix tube didn't come today (doggone that postman) but I did take a few minutes to re-measure. With the 6L6GC I get what I'd think is a normal differential with the screen voltage a little more than two volts negative w.r.t. the plate. And there's about 4.4mA coming through the screen resistor so the 6L6 screen is dissipating a watt and a half or so which shouldn't be a problem.

But with the 6V6S the screen was seventh-tenths of a volt positive vs. the plate. Not surprisingly, only a fraction of a milliampere is flowing through the screen resistor to the 6V6S screen which basically isn't having to dissipate anything. I guess as long as the screen doesn't have current flowing the opposite direction it's no biggie.

Once again with this evening's measurements (and with my line voltage running about a volt higher than this morning) the 6V6 plate dissipation works out to right at 12W, running with 360-odd volts on the plate and a 680 ohm bias resistor. And it sounds pretty darn good. Oh, and the microphonic ringing is gone once I tightened up the tube retainer on the rectifier tube. So all is well with the world but I still want to try the Bendix and see how its operating parameters compare...

[Brent Hutto]

Well, regardless of what the datasheets have to say the reports of a Bendix 6106 dropping less voltage than a generic NOS 5Y3GT. I dropped one of the Bendix rectifiers into my Richter single-ended Class A amps today and measured it at two different current draws (by switching between 6L6 and 6V6 output tubes). The line voltage was constant at 123VAC give or take half a volt so differences are almost entirely due to the different behavior of the two tubes...

G.E. 5Y3GT

39.8mA load
342/343VAC to the rectifier
409VDC out from the rectifier
Efficiency eta=0.845

54.9mA load
341/342VAC to the rectifier
392VDC from the rectifier
Efficiency eta=0.813

Bendix 6106

40.5mA load
342/342VAC to the rectifier
418VDC out from the rectifier
Efficiency eta=0.865

57.0mA load
339/339VAC to the rectifier
400VDC from the rectifier
Efficiency eta=0.835

So you can see that for moderate loads of 40-57mA the Bendix will produce about 8 volts more DC output for a similar AC input voltage. The result at the plate of my 6V6 is an increase from 364-371V and with the 6L6 a similar increase from 328-336V while idling in Class A, cathode-biased operation. Not sure how that would equate in "sag" terms on a push-pull amp but it sure sounds like you'd get slightly less with the Bendix 6106 than most 5Y3GT rectifiers.

P.S. Through four sets of measures today I get the same results as earlier concerning my screen voltage. It was consistently around one volt positive w.r.t. the plate with all combinations of 6V6/6L6 and 6106/5Y3GT that I tried. But the 6V6 always draws around 1mA of current from the high voltage supply and the 6L6 a little over 2mA draw so that sounds about right.

[Trout]

I think she pretty much took the design as far as possible without abusing the lesser tube (6V6) which is actually a good thing. I imagine that the screen is perfectly safe when considering the original had no real protection (screen resistor). I feel it is a more stable amp than my scratch built monsterized 5F2 running an old NOS 807.

I have logged no less the 250 hours on my RS 5.8 and it has been flawless in operation. I run it dimed at least 75% of the time. I have flip flopped on the 6L6 several times, Some days I really dig the jazzy tones, other days I just want grit and pop in the 6V6.

Fun flexible amp, I lent it to my brother for several recording projects, he loves the fact he can crank it pretty good and not have to hide in the hall LOL