bass amp tube question

[dhuebert]

I'm thinking of another bass amp and would like to make it pretty much the same as the BFA. This one might go on the road and to that end I would like to use something more common than KT88. Would EL34 or 6550 survive 535 volts ultralinear? I'd like the amp to be fairly robust, of course.

Don

[EWBrown]

A set of 6550Cs should handle that kind of voltage just fine, I've seen 'em "beat on" a lot harder than that in some big VTL MB450 monoblocks.

I read up in the VTL book last nite, they do warn about pushing most new build EL34s much above 500V, though some of the old Siemens and Mullards could take a B+ of 800VDC and hold up just fine. (I forgot to add, that the screen voltages were held much lower, this was not UL operation).

VTL uses (and specify) the Svetlana 6550Cs, (operated either UL or speudo-triode modes) with 550VDC B+ and around 1900 uF of PSU capacitor, which offers up to 287 Joules when needed. Most new EL34s would go "pyrotechnic" sooner or later on 535VDC... (Especially the "continental slimmies" ) . They also set the individual tube biasing to 30 mA each, it sounds low, but at the elevated plate and SG voltages, it's necessary.

The J/J KT77s mentioned in the next reply are a good possibility

You may want to inquire over at Eurotubes, Bob has a lot of experience with J/J tubes and what they can and cannot do. He also has an interesting approach called "integrated quads" in which a PPP amp uses an EL84 and a 6L6GC on each side of the OPT primary, claims that this offers the best of both tube types.

HTH

/ed B in NH

[Ty_Bower]

I've been wondering about that... how hard can an EL34 be pushed, and does it make a suitable alternative in a circuit designed for a KT88. Clearly, the EL34's dissipation rating is lower - but can it handle the voltage if the bias current is turned down to an appropriate setting?

Considering you can buy a set of EL34 for about one third the cost of a set of KT88, it seems an economical route.

[sorenj07]

grab KT77's, they're a step inbetween (EL34-25W, KT77-32W, 6550-40W)

[dhuebert]

Mostly I worry about the potential difference between the control grid @ ~- 50 volts and the screen grid @ ~ +535 volts. We're talking about 600 volts across a millimeter of vacuum, seems precarious to me. If anyone would like to talk about this, I'd appreciate it.

Don

[TerrySmith]

Look at the Dyna MK-2, it runs EL34's at 500+ volts.

[Uncle Ned]

Look at the Dyna MK-2, it runs EL34's at 500+ volts.

Actually, what's listed on the Mk2 schematic is considerably less than 500V, like the EICO amps it takes advantage of the claimed 8W of screen dissipation Philips claimed for the EL34. However I 'm sure the EL34's in most Mk2's actually see more than what the schematic says. ;-)

It is possible to run EL34's at over 500V on plate and screen, but for a bass amp it's probably not the best or as good as 6550's.

A better way to push EL34's for lots of output would be to hold the screens to 400V or so then up the plates to 700+, but then you're talking about a different output transformer or running them PP parallel to match up with 3500-5000 ohms a pair of 6550's would like to see.

The best wattage output results I've seen with very high ultralinear voltages (like in the 550-650V range) were KT90's and the old British GEC KT88's (with ~625V 4 put out ~200W). GE, SED, Sovtek, JJ, etc just all seem to run out of gas (although newer ones may be better than what I was using) above 525-550V, pushing the voltage higher just doesn't seem to generate much more output.

1 mm = ~ 40 thousandths of an inch, that's a fair amount of gap for a vacuum, Jennings says contact spacing in a vacuum relay can be as close as 1000 volts per .001 inch, but there's more going on in a tube than a relay ;-)

[gerryc]

I'm thinking of another bass amp and would like to make it pretty much the same as the BFA. This one might go on the road and to that end I would like to use something more common than KT88. Would EL34 or 6550 survive 535 volts ultralinear? I'd like the amp to be fairly robust, of course.

Don

That's pretty coincidental, I was having a similar discussion on another forum about building a 200 watt bass amp UL with a quad of KT88. Seems like it would take upwards of 550 volts. However, over there the feeling is that the current KT88's would not fair very well at those voltages (although I've seen datasheets from JJ and Shuguang that show screen ratings of 600v). From that I would expect that 6550's wouldn't do well since their screen rating is 400, I think.

But then again, I've seen circuits where these ratings are exceeded all the time and so I'm game for trying something like this. I think that part of the solution is to use large enough screen stoppers, at least 1K. Kevin O'Connor has a discussion about this in his book The Ultimate Tone 3 (TUT3), and cites a number of commercial amps (Fender, Marshall, and Ampeg among others) that use too low of a value. Looking at your BFA, you could try increasing the 330 ohm screen resistors to 1K and trying a good set of 6550's in there. Right now, I have an Ampeg VT22 head that I converted to 6550's. Yes they are tetrode connected, (not UL), but they run the plate and screen voltages pretty high right around 535, and it's working fine. However, the conversion instructions say to change out the 470 ohm screen resistors for 1K 5 watt before using the 6550s. That may be the key here.

I actually had a question to ask you as well. Somwhere I think you said that you're getting 160 watts out of your BFA using the Hammond 1650T. Do you get good deep bass at full power? I was wondering if I could push one of those to 200 watts instead of using a 1650W (28 lb monster).

[Uncle Ned]

Right now, I have an Ampeg VT22 head that I converted to 6550's. Yes they are tetrode connected, (not UL), but they run the plate and screen voltages pretty high right around 535, and it's working fine. However, the conversion instructions say to change out the 470 ohm screen resistors for 1K 5 watt before using the 6550s. That may be the key here.

I've done that, and put EL34's in the same kind of amp. (The general consensus was that 6550's sounded better) The reasons they got away with the high screen voltages on 7027's/6L6's that were originally in that amp, is the diode power supply, damping diodes on the plate, a big honking power transformer, and the static bias being set very low (like ~25 ma IIRC).

I don't see any problem with running 6550's at those voltages (~525-535V), there's no shortage of amps that already do.

BTW, which forum were you having the KT88 discussion on?

[gerryc]

Right now, I have an Ampeg VT22 head that I converted to 6550's. Yes they are tetrode connected, (not UL), but they run the plate and screen voltages pretty high right around 535, and it's working fine. However, the conversion instructions say to change out the 470 ohm screen resistors for 1K 5 watt before using the 6550s. That may be the key here.

I've done that, and put EL34's in the same kind of amp. (The general consensus was that 6550's sounded better) The reasons they got away with the high screen voltages on 7027's/6L6's that were originally in that amp, is the diode power supply, damping diodes on the plate, a big honking power transformer, and the static bias being set very low (like ~25 ma IIRC).

I don't see any problem with running 6550's at those voltages (~525-535V), there's no shortage of amps that already do.

BTW, which forum were you having the KT88 discussion on?

I was discussing it on the www.diyaudio.com tubes forum. I think there are two camps on the design philosopy. One being conservative and probably targeted towards home stereo applications, and the other being more liberal targeted towards musical instrument amps where "watts per pound" is more important (who wants to carry around a 100 lb amp that only puts out 50 watts?). As you said, a lot of designs run these tubes at high voltages but bias them on the colder side to help keep within dissipation ratings. I'm tending to lean towards the liberal side myself.

[dhuebert]

Do you get good deep bass at full power?

I hardly know what to say. In its current venue you can feel the bass through the floor at the back of the room 50 feet away. This is through an Ampeg 8x10. I have had many complimentary comments on the sound quality, personally I love it. I often hear solid state amps at other venues and I can always tell, and I hate them, yuck! OTOH, the other night I heard an old Ampeg SVT through what looked like a half an 8x10, YEOW that was sweet! Made me think of giving up on building bass amps.

you could try increasing the 330 ohm screen resistors to 1K and trying a good set of 6550's in the

Ya know, I used 1K5s in some of my other designs but ran out when I built the BFA so used what I had on hand. No question in my mind TUT is right here.

I was having a similar discussion on another forum about building a 200 watt bass amp UL with a quad of KT88. Seems like it would take upwards of 550 volts.

This brings me to the next thing I want to discuss: What about 6 x 6550? Can I get more power by adding another pair but lowering B+? I would like to get B+ around 500 volts to make it easier to find caps for the power supply.

I've never really understood how adding pairs makes more power but it works so...

Don

[gerryc]

This brings me to the next thing I want to discuss: What about 6 x 6550? Can I get more power by adding another pair but lowering B+? I would like to get B+ around 500 volts to make it easier to find caps for the power supply.

I've never really understood how adding pairs makes more power but it works so...

Don

Take a look at the discussion over there (I don't know if it's kosher to cross link to another forum). Check out the tubes forum and search for "parallel output transformers". There are some numerical examples that should be enlightening. Instead of looking at the tubes first, you need to look at the output transformer and power supply requirements to achieve what you want. Then look at what it requires of the output tubes to decide what to use. Paralleling output tubes only increases power if you can reduce the P-to-P impedance. The 1900 ohm of the 1650T or W is a bit limiting to give you more power without raising the B+ voltage. If you're at the limit at how much power a quad of tubes will give you at a given P-P impedance and B+, adding another pair won't buy you anything. With the same B+ voltage you could parallel 10 pairs of tubes and you won't get any more power. Either B+ must be raised or OT impedance must be decreased. Then you can add more tubes.

[EWBrown]

If you add parallel pairs, it is a good idea to reduce the OPT primary impedance proportionally , to get the full benefit of any resulting power increase.

The old-timers' trick for > 500V B+ is to connect the PSU filter caps in series, with equalizing resistors across each, typicaly 100K-330K, 1-2W is a good range. I'd use at least 300WVDC caps , for 500V, and allow some safety margin of around 20-30%. Here is a good use for those "big and fat" computer power supply HV caps...

/ed B in NH

[dhuebert]

Either B+ must be raised or OT impedance must be decreased

I have been hooking an 8 ohm load to the 16 ohm tap to decrease OT impedance. It works quite well to a point but its no free lunch. I had the BFA set up this way until I discovered the 8x10 was 4 ohms. I took the amp home and rewired the output and increased the power by 15 watts from 145 to 160. I assume that with a sextet I want the impedance to be 1/3 what it would be for a pair.

I don't know if it's kosher to cross link to another forum

I say go for it!

Don

[gerryc]

Either B+ must be raised or OT impedance must be decreased

I have been hooking an 8 ohm load to the 16 ohm tap to decrease OT impedance. It works quite well to a point but its no free lunch. I had the BFA set up this way until I discovered the 8x10 was 4 ohms. I took the amp home and rewired the output and increased the power by 15 watts from 145 to 160. I assume that with a sextet I want the impedance to be 1/3 what it would be for a pair.

I don't know if it's kosher to cross link to another forum

I say go for it!

Don

Ahhh. Actually, I started thinking about the same thing. Theoretically, connecting a 4 ohm load to an 8 ohm secondary would reflect an impedance back to the primary of half it's design value, in this case the 1900 ohms P-to-P of the 1650T would look like 950. Now that's theoretical. I'd guess that in reality it would reduce the bass response somewhat since the primary inductance wouldn't be high enough. But, this is not science, it amounts to engineering tradeoffs, and it is a hobby, and the real world result may be more acceptable than at first thought.

Taking this approach might work better with a bigger transformer, a la' the 1650W, since it is rated at 280W, more than twice that of the "T" and is about three times the weight (yes this is all back of the envelope handwaving stuff, I admit). Bottom line, if you design for a P-P impedance of 950 ohms and treat the secondary as 2, 4, or 8 ohms instead of the 4, 8, 16 ohms, you might be able to get what you want (and me too) by using a sextet of 6550's at lower B+. In this case, I would consider the 1650W OT.

I didn't see anything in forum posting rules about linking to another forum, so forgive me moderators if this is not allowed. Here's the link:

http://www.diyaudio.com/forums/showthre ... did=108935