Sovtek 5Y3

[labb]

I am having problems with Sovtek 5Y3's having a mechanical noise when used in a combo type amp. Anyone else run into this issue?

[tubeswell]

So you swapped the tube already?

[jeff]

I've had problems with Sovtek 5Y3s. Every one I've tried has raised my B+. If you just replaced a 5Y3 with a Sovtek make sure you check your voltages.

 The way I hear it is that Sovtek rebrands other rectifiers tubes as 5Y3s eventhough they're not 5Y3s. It's been my experience that a real 5Y3s have a drop of about 1.1 and the Sovtek about 1.2.
 The first amp I ever built was supposed to run at 360V but I built it using a Sovtek and was getting 420V. I didn't know about this and tried rewiring everything, replacing almost every part, trying to figure out why my amp wasn't to spec, what I did wrong. I tried another 5Y3 but bought a Sovtek again, and again 420V. It was almost enough to make me give up amp building because I couldn't figure out what I was doing wrong, I did everything right yet it was wrong. It wasn't till I fianally got a NOS 5Y3 that my voltages were where they should be.

 Sovtek is lying to you and ought to be ashemed of themselves! They should not be allowed to do this! At the very least they should label them "5Y3-like replacement" and tell you what the tube really is. Yeah it's a rectifier but it's not the same as a 5Y3. I don't know if all Sovteks are like this but I have two that are, and don't buy them anymore. I use mine to raise voltages where a 5Y3 is called for when building with a smaller transformer than called for.(300-0-300 with real 5Y3 ~ 275-0-275 with Sovtek)

 I don't know if the Sovtek is causing your mechanical noise/rattle but you should check your voltages to be sure you're not exceeding limits. Espically if it's a Princeton where the voltages are already way high. (420V on 6v6s is already close to the limit, use a Sovtek and you may be getting up to 460V!) If so get yourself a REAL 5Y3. Like I said I don't know if all Sovteks 5Y3s are like this but the two I have are.

Just something I'd check.
 Jeff

[labb]

As a follow up on the mechanical noise problem in the Sovtek 5y3. I put a couple of viton o-rings on the tube and the noise went away..

Jeff, you are correct in that the Sovtek 5y3 will give you a voltage of a 1.2 factor instead of the 1.1 that a "real" nos 5y3 gives. JJ is suppose to have a 5y3 on the market that is a drop in replacement voltage wise for the nos 5y3. I have not tried them myself but the reviews that I have read are very good.

[jeff]

I just hope your amp can take the extra voltage.
What's your B+?
What tubes are you using?

[labb]

I design around the elevated voltage of the Sovtek. I am using a PT that is 300-0-300, putting about 355 vdc on 6V6 tubes.

[HotBluePlates]

The way I hear it is that Sovtek rebrands other rectifiers tubes as 5Y3s ...

They're a russian tube type, rather than a true 5Y3. If I wasn't lazy, I could find you the actual russian tube number (or FYL will post it shortly...).

I am having problems with Sovtek 5Y3's having a mechanical noise when used in a combo type amp. Anyone else run into this issue?

I've had rectifiers that were microphonic, and not just Sovtek. It's a strange rattle that almost sounds like a speaker problem. It can potentially occur with any rectifier tube.

[tubeswell]

This is what I was meant to be looking for earlier.  http://www.newsensor.com/pdf/sovtek/5y3gt-sovtek.pdf

I still don't know the Russian tube equivalent name. (The 5C3S is a 5U4GA equivalent)

[FYL]

They're a russian tube type, rather than a true 5Y3. If I wasn't lazy, I could find you the actual russian tube number (or FYL will post it shortly...).

The Sovtek 5Y3GT is a rebadged 5Ц4М (5Z4M), a Russian indirectly heated recto quite close to a 5Z4GT or 6106.

http://www.tubes.ru/techinfo/HiFiAudio/5z4c.html (C = large bulb, M = compact version)

[alerich]

I have read of Sovtek re-branding other tubes that weren't really what they were labeled... like the 5881. I read that the Sovtek 5881 is really some sort of avionics servo tube that just happens to have the same internals. Fender has shipped a boatload of amps with those tubes in them. They are very robust and survive the warranty period easily. I bought a silverface Super Reverb with new Sovtek 5881 tubes once. It sounded pretty good - good enough to buy. When I re-tubed it with real 6L6 tubes the difference was remarkable. This was about ten years ago so they may have changed this practice since then.   

[tubeswell]

I have read of Sovtek re-branding other tubes that weren't really what they were labeled... like the 5881. I read that the Sovtek 5881 is really some sort of avionics servo tube that just happens to have the same internals.

Does 6P3Se sound right?

[HotBluePlates]

I have read of Sovtek re-branding other tubes that weren't really what they were labeled... like the 5881.

Sovtek made their name in the early-mid 90's. There were no American manufacturers of guitar amp tubes left (though you could still buy military surplus easily), and European manufacture was pretty much down to Tesla (maybe a couple of east-European tube companies were still around that I'm not aware of). Or you could buy Chinese.

For the longest, Russian tubes couldn't be imported. You still got them from some rebranding companies, often claiming an RCA 6L6GC "Made in England" or "Made in West Germany". These folks were selling you Russian tube types cheap with a bit of false advertising.

Then the Berlin Wall fell, and soon after the Soviet Union. Trading with Russia wasn't so much of a bad thing, and Sovtek saw an opportunity. But you don't own any amps calling for a 6П3C (cyrillic for 6P3S), but you do have amps that call for a 6L6. They rebranded with the label you're familiar with and sold them to you. LOT of other companies you think are reputable have done the same.

The Russian tubes aren't bad (some maybe even quite good) if you use them as they were designed to be used. Most folks don't run into problems with them. And most users don't poke at their amp with a meter, and never notice the Sovtek 5Y3 gives them higher B+ voltages.

But if you know the difference, or if your amp runs even the correct American type well beyond its published limits, you may find problems. I never had any failures or sonic issues with Russkie output tubes, but almost all the amps I've owned had feedback loops around the output stage (which might mask sonic differences).

Be an informed buyer. If the Russian type will work for you and save money, have at it. But they're not making any more 50's-70's American or Western European tubes. Eventually, they'll be gone, and what will you use then? It might be prudent to design around the true ratings of tubes which are available now and for the foreseeable future.

[jeff]

"The Russian tubes aren't bad (some maybe even quite good) if you use them as they were designed to be used..... It might be prudent to design around the true ratings of tubes which are available now and for the foreseeable future."

If we're building from schematics thinking we're using 5881s, but we're not, we're using 6P3Ss, how do we compensate for this?
(Aren't we trying to bake a cake using salt that's labeled "Sugar"? If we use the wrong ingredients how can we expect the cake to taste good?)

 Seeing as our amps aren't designed to use 6P3S, how do we bias them when used in an amp desinged for 5881s? I don't know what the difference between the 5881 and the 6P3s is(can't read russian) but how do you bias it?
 70% of 5881
 70% of 6P3S(?)
 or some point that makes the amp happy?
Will the 5881 and the 6P3S have the same bias point? If not, how do you bias them?

[FYL]

Just consider the 6П3C as a lower-rated version of the 5881 : 20W plate dissipation vs. 23W, 2.75W vs. 3W screen dissipation, 375V vs. 400V anode (design center ratings), etc. Curves et al. are more or less equivalent, so just run the russkies a tad cooler.

[tubeswell]

FYI 6P3S, 6P3Se tube data

http://www.tubes.ru/techinfo/HiFiAudio/6p3s.html

Essentially the 6P3S is a 20W tube and the 'e' version (with the wafer base) is the militarised version.

BTW I wouldn't run the 6P3S above 350V. Whereas in my humble experience the 6P3Se is much tougher and can easily handle 460V (but it is still ostensibly a 20W tube although I have heard reports of people running it as a 30W tube). I have a whole bunch of both types and I don't find much use for the 6P3S in guitar amps, whereas I happily use the 6P3Se in place of 5881s all the time.

[Madison]

I am having problems with Sovtek 5Y3's having a mechanical noise when used in a combo type amp. Anyone else run into this issue?

Yupe.
Drove me nuts.
Stopped using them years ago.
I like their preamp tubes.

[jeff]

 When we say mechanical noise do we mean microphonics?
 I think I understand microphonic tubes in the signal chain, but I don't understand why a microphonic rect tube would cause problems? If the filter caps filter out the 120Hz AC(or rather pulsing DC), why wouldn't those filter caps also filter out any mechanical noise/microphonic signal the rect. tube "picks up"? Know what I mean, anything comming out the rect. tube should be smoothed to pure DC by the filter caps, right?

Is it because big caps filter low(120Hz) freq. good but have trouble filtering out the higher, microphonic induced, frequencies?  Would a small cap parallel to filter cap filter out any noise that the big caps don't(big cap to filter out the 120Hz/small cap to filter out microphonics, ...although we do use 20-50uF for pre amp stages)?

I'm having trouble understanding microphonic rectifier tubes.

Thanks
 Jeff

[HotBluePlates]

I think I understand microphonic tubes in the signal chain, but I don't understand why a microphonic rect tube would cause problems? If the filter caps filter out the 120Hz AC(or rather pulsing DC), why wouldn't the filter caps also filter out any microphonic noise/signals the rect. tube "picks up"?

The first filter cap works with the rectifier, and takes what is essentially a very large, pulsating d.c., and filters out some of the pulsations. You get essentially a large d.c. output, with an a.c. output superimposed on it, but the a.c. component is smaller than what it was with a rectifier and no filter cap.

The choke/resistor and the next filter cap do the same: reduce (but not eliminate) the a.c. component. And so on with the successive filter caps. At each filter cap, you can measure some amount of a.c., even if it is very small by comparison to the d.c. So you almost never have pure d.c. unless you have some form of regulator, though we do strike a balance with having "clean enough" d.c. at each stage.

Microphonic noises aren't hum, though. Especially in the case of a rectifier, the likely microphonic noise is a short, sharp impulse and can be very large.

Further, caps need time to charge/discharge. The bigger the cap or the impedance of the supply, the longer that charge/discharge takes. If you mess with the Duncan PSU Calculator, look at the plot of output voltage, paying particular attention to the beginning of the plot. You'll see the supply voltage ramp up as the first filter stage charges up.

So, sudden microphonic impulses (assuming they aren't as severe as an actual intermittent interruption) are really beyond the capability of the filter caps to filter out. That's why when you have a microphonic rectifier tube, it tends to sound like odd scratchiness or rattling that's akin to a speaker problem. Since it's output is no so much a sound that gets amplified, but a disruption of the supply voltage/current to every other stage in the amp, the sonic effect is different than, say, a microphonic preamp tube.

[jeff]

Further, caps need time to charge/discharge. The bigger the cap or the impedance of the supply, the longer that charge/discharge takes....So, sudden microphonic impulses (assuming they aren't as severe as an actual intermittent interruption) are really beyond the capability of the filter caps to filter out.

 OK I think I get you. We use big caps because the 120Hz ripple is low freq, but that can't handle the impulse because charge/discharge time.
 
Let's assume for a sec were just dealing with just that impulse. Ignore the 120Hz AC ripple for a moment, just focusing on the impulse. The big cap can't handle because charge/discharge time but is there a smaller cap that could?

And what would happen if we used the big cap and the small cap in parallel? Would each handle what it can(big cap filters 120Hz, small cap charges/discharges fast enough to filter the impulse) or would it just act as a very very slightly bigger cap because there in parallel?
In other words would a 20uF and a .001uF act excatally as one 20.001uF cap would or would the .001uF make up for the 20uFs shortcomings?(.001uF is just random value for example I don't know what exact value should be)
 
 I seem to remember seeing schematics with a big filter cap and a small cap in parallel, just adding a tiny capacatince to a huge one seems pointless. So what does using a big cap and a small cap in parallel do?