Ampeg SVT Specs

[Wilder Amplification]

Hey guys...spent about an hour searching the site hoping to dig up some specs on the Ampeg SVT amps (6550 version) but didn't turn up anything.

Does anyone happen to have any specs/voltage charts/plate-plate load OT specs etc etc for the Ampeg SVT? Would greatly appreciate any and all who could help me in obtaining this information.

Thanks.

[Enzo]

The SVT from 40 years ago? or the SVT Classic of contemporary times? They are pretty much the same anyway. Plenty of voltages and test points on the SLM drawings you can get from Loud.

[Wilder Amplification]

The SVT from 40 years ago? or the SVT Classic of contemporary times? They are pretty much the same anyway. Plenty of voltages and test points on the SLM drawings you can get from Loud.

Hey Enzo. Was looking for the vintage one from 40 years ago, but since some older schematics that are hand drawn are harder to read I'm sure the one from contemporary times will suffice.

I'll see what I can get from Loud though.

[soundmasterg]

There was a very nice article with schematics on the SVT in one of the Vacuum Tube Valley magazines, and there is tons of info on them in the Ampeg book that came out a couple years ago.

I have a very early MTI SVT with USA transformers from 1981 but I haven't opened it up in awhile. I recall that the B+ was very close to 700v and the screens are down around 300 or 350v.

Greg

[Enzo]

OK, from 1970 AMpeg SVT with six 6146B

PT 8930007:
600v @ 0.75A
200v @ 0.07A
150v @ 0.025A

Heater transformer 8930008:
6.3v @ 10A

OT 8950004:
Primary 1850 ohms
Secondaries 2, 4 ohms

Later version for 6550s:

PT 8930072:
600v @ 0.75A
330v @ 0.07A
150v @ 0.025A

Heater Tranny 8930073:
6.3v @ 12.05A

[Wilder Amplification]

OK, from 1970 AMpeg SVT with six 6146B

PT 8930007:
600v @ 0.75A
200v @ 0.07A
150v @ 0.025A

Heater transformer 8930008:
6.3v @ 10A

OT 8950004:
Primary 1850 ohms
Secondaries 2, 4 ohms

Later version for 6550s:

PT 8930072:
600v @ 0.75A
330v @ 0.07A
150v @ 0.025A

Heater Tranny 8930073:
6.3v @ 12.05A

Thanks a million. The OT primary plate-plate load coincides with what I calculated from a 6550 datasheet for a Va - 600, Vg2 - 300, Ra-a = 5000K for two 6550s. Just cut that load in thirds for 3 pairs and got 1.6K which is very close to the above posted 1850 ohm plate-plate load.

What I'm not understanding is that when I draw a load line that coincides with a 5K load on the plate characteristic graph yielding a screen voltage of 300V, the load line crosses over the Wa dissipation curve for a portion of it. It passes through the knee of the Vg1=0 curve like it's supposed to, but it's crossing the Wa dissipation curve as well, which if I'm interpreting "ideal design info" correctly it's not supposed to cross the dissipation curve. Can anyone explain why that would be? Do the tube datasheets assume that you'll never run the amplifier to max power before clipping? Or is it that because it's on paper it assumes a super stiff regulated supply and in the real world unregulated supplies sag and allow for that? Or am I way off base here?

[Enzo]

I don't deal in all that stuff, the amps work. Others here are more fit to discuss the design minutiae.

But one general thought - the tube manuals present specs for good design. This was an era when EVERYTHING was tubes. RCA or whomever gave design specs for designing consumer products that would work well and have a long reliable life. Table radios, TV sets, hifis. Mom and dad did not want to have to put a new set of tubes into their TV every few months. The design specs reflect this.

We on the otherhand DO expect to wear out tubes. Like the family car tires, dad expected them to last a reasonably long time. BUt race care drivers expect to wear tires out. We drive our tubes well into distortion and we wear them out, we are hard on tubes. But it gets us the performance we expect. the pedestrian specs in the tube manual are not for guitar amps. Look at old Fender amps. Leo ran the 6V6s in those old amps 100 volts over the max spec in the RCA book. Fortunately no one told the tubes, and those amps worked fine for decades.

[Wilder Amplification]

We on the otherhand DO expect to wear out tubes. Like the family car tires, dad expected them to last a reasonably long time. BUt race care drivers expect to wear tires out. [b]We drive our tubes well into distortion and we wear them out[b]

I partially disagree...

Back in the day, yeah everyone ran power tubes hard. However, times have changed. Believe it or not, there's only a small percentage of us left who run amps cranked enough to even overdrive the power tubes anymore. I run either a high gain preamp style/channel switching rig or a cranked plexi with a Power Brake rig depending on what type of gig I'm doing. This is what I find amusing about a lot of these guitar forums...they all assume that all guitarists run their amps with the power tubes overdriven. Well...not everyone owns power soaks and this day & age, volume tolerance among the general public has dropped significantly which means in most clubs around one would have to run a power soak in order to be able to get away with playing their amps at the level it requires to overdrive the power tubes.

So...that being said...yes for an amp that's designed with power tube overdrive in mind for blues/classic rock guys, I stick with the classic designs...the standard decades old single B+ rail with the RC/LC network, lower plate loads than are supposed to be ran, etc etc. However, with current production tubes (cause not everyone can afford to run NOS) on these amps I watch the plate voltage so that the screens stay in check. People have loosely thrown around "Oh tubes of today can't handle the plate voltage that the tubes of yesterday could" when in reality it's not the plates that's the problem...it's the screens. On that "single B+ rail" design, the screen voltage is not far behind the plate voltage and as such the screens get overtaxed bigtime in a cranked vintage amp, which is why we're so hard on them. NOS tubes seem to be more tolerant of that over current production stuff.

The guys who are into power tube overdrive also like the amp to have a bit of sag to it...some more than others. I personally prefer a tighter/faster transient...better definition and it's easier to control. I'm not talking "solid state" tight, but much tighter than say an early JTM-45 with a shared cathode and 330uF bypass cap with low filtering and a tube rectifier that tends to "fart out" in the lows. I'd much rather have a low end with some "spank" to it. This is where the "high gain" crowd comes in. They could really use a stiff/tight/clean power amp while getting their tone from the preamp and focusing on the preamp circuitry for tone shaping. These are also the guys who run effects loops in order to have a way to patch in modulation/time/ambience effects AFTER the signal has been overdriven whereas you can't really do that on a cranked/overdriven tube power section without running a slave rig with a 2nd amp set for a clean tone. And what better way to do this than with the dual rail design that runs the screens at 1/2 the plate voltage, jacks the plate voltage up while being able to pull double the power out of 1/2 the tubes (i.e. 100 watts out of 2 x 6550s or KT88s)? Without taxing the screens so much, I'm thinking that tubes would last a hell of a lot longer than they would in say a cranked 100 watt plexi or any other "classic design" tube guitar amp.

I know most would say "Well then why don't you just use a solid state power amp?". Well...even if you were to design a tube power section "data sheet correct", there would still be some of the "tube specific" tonal characteristics in the sound whereas with solid state you wouldn't get any of that.

[Enzo]

I didn't actually claim a large percentage of guitarists go for power tube distortion, as in driving the piss out of them. I just said distortion. On the other hand, most guys who adjust their own bias go for the "70%" setting rather than any setting meant for reliability and fidelity. The gain monster PV 5150 for example is designed with a strong clean power amp, and the amp gets its tone from the preamp. the first thing guys do is discover that the stock bias is something like 14ma per tube and right away they want to heat it up.

Times have changed? MAybe, but do you expect 5-10 years from a set of EL84s in your PV Classic 30 or Fender Blues Junior? Or any other guitar amp? I maintain there is a fundamental difference in the application of these tubes in consumer goods of 40 years ago, what the RCA book is aimed at, and in guitar amps of today or of back then. I have no illusion that everyone is using soaks or maxing their amps. AMps by their very design go through tubes.

[Wilder Amplification]

I didn't actually claim a large percentage of guitarists go for power tube distortion, as in driving the piss out of them. I just said distortion. On the other hand, most guys who adjust their own bias go for the "70%" setting rather than any setting meant for reliability and fidelity. The gain monster PV 5150 for example is designed with a strong clean power amp, and the amp gets its tone from the preamp. the first thing guys do is discover that the stock bias is something like 14ma per tube and right away they want to heat it up.

Times have changed? MAybe, but do you expect 5-10 years from a set of EL84s in your PV Classic 30 or Fender Blues Junior? Or any other guitar amp? I maintain there is a fundamental difference in the application of these tubes in consumer goods of 40 years ago, what the RCA book is aimed at, and in guitar amps of today or of back then. I have no illusion that everyone is using soaks or maxing their amps. AMps by their very design go through tubes.

In regards to biasing to 70%...a good majority of guitarists who set their own bias are misinformed as to what bias is and where it "should be" set. Most of them think that it's a hard/fast rule that the amp MUST BE biased at 70% and if it's at any other setting other than 70%, then it's wrong. I've always preached that 70% is nothing more than a Class AB maximum that shouldn't be exceeded and any setting at or below that that yields great tone is good.

Lots of the amps of today are being ran with a high gain preamp while keeping the power amp clean and setting the volume with the master, yet they don't really see all that much longer tube life that we who crank 'em up see. Why is this? Because the screens are being ran to the hilt. NOS tubes and tubes of yesteryear may have been more tolerable of this, but it's common knowledge that a good majority of the current production stuff just can't tolerate it. Yet all of the OEM manufacturers are still to this day using a single rail supply. And it's all in the name of tone. Don't even get me started on EL34s and what seems to be inconsistent screen grid voltage ratings between different tube manufacturers on top of the inconsistent plate voltages that the old Marshalls seem to be known for.

My way of thinking is to find a very close compromise between tone and functionality. Drop the screens to 1/2 the plate voltage and run the plates up higher by running a dual rail supply, and as such run the proper spec'ed plate load as the data sheets dictate. Lots of the OEMs today don't even have plate load as a consideration on their radar and we have guys like Randall Smith who preach that it's OK to mismatch, which tells me that not even he takes it into consideration. However, I'm thinking that for a good reliable modern high gain channel switcher that never sees power tube overdrive duty that designing a "data sheet correct" power amp while compensating the tone elsewhere in the circuitry will give us the tone we're after via the preamp circuit voicing while having to run 1/2 the tubes to get the power as well as increasing tube life since the screens are no longer being overtaxed. A single pair of KT88s costs the same as a quad of EL34s so retube costs would be the same for 2 x KT88s @ 100 watts vs 4 x EL34s @ 100 watts. Throw in the fact that you very well may be able to get a full year or two out of tubes ran in this config vs every 3-6 months...I'm thinking it could save money in retube costs in the long run.

Not to mention that the metal cats who love big/tight low end and fast picking transients would absolutely love a setup like that.

[Steve Conner]

The maximum plate dissipation on a vacuum tube is an average value. So it doesn't matter if it gets exceeded for a few milliseconds at some point in the waveform, as long as the average stays under the datasheet value. The plate is a big lump of metal that can easily soak up pulses of heat.

It's not like semiconductors, where the junction has a low thermal mass and there are other mechanisms that put a safe limit on the instantaneous power.

When looking at load lines, you want to imagine a little dot zooming back and forth along the line, once per half cycle: that's what the tube is actually doing.

I've built a high gain amp with this "dual rail" arrangement under discussion, not to mention that both the plate and screen rails were regulated, and it worked great. The rails were 450V and 360V, and it started out as an exercise in running EL34s to their datasheet conditions, with a 6.6k OT that was all I had at the time.

I found that it worked great with all the other octal power tubes I tried, except 6L6-type tubes needed their screen supply hooked up to the 450V, or they would only make about 30W instead of 50. I use the amp with a pair of the Sovtek "Tung-Sol" 6550s now, and they run happily off the 360V tap, making about 60W. I've had it for 10 years and have never seen a power tube fail or wear out. I changed out various tubes to try them out, but the Tung-Sols have been in there a couple of years now.

With EL34s especially, loading the plates more heavily helps to unload the screens. If you had a 6.6k OT in a Marshall, it would probably burn up the screens on your first power chord, but at 3.4k they sort of survive. The screens aren't stressed at idle and low volumes: it's mostly when the amp is cranked up.

[Wilder Amplification]

The maximum plate dissipation on a vacuum tube is an average value. So it doesn't matter if it gets exceeded for a few milliseconds at some point in the waveform, as long as the average stays under the datasheet value. The plate is a big lump of metal that can easily soak up pulses of heat.

It's not like semiconductors, where the junction has a low thermal mass and there are other mechanisms that put a safe limit on the instantaneous power.

When looking at load lines, you want to imagine a little dot zooming back and forth along the line, once per half cycle: that's what the tube is actually doing.

I've built a high gain amp with this "dual rail" arrangement under discussion, not to mention that both the plate and screen rails were regulated, and it worked great. The rails were 450V and 360V, and it started out as an exercise in running EL34s to their datasheet conditions, with a 6.6k OT that was all I had at the time.

I found that it worked great with all the other octal power tubes I tried, except 6L6-type tubes needed their screen supply hooked up to the 450V, or they would only make about 30W instead of 50. I use the amp with a pair of the Sovtek "Tung-Sol" 6550s now, and they run happily off the 360V tap, making about 60W. I've had it for 10 years and have never seen a power tube fail or wear out. I changed out various tubes to try them out, but the Tung-Sols have been in there a couple of years now.

With EL34s especially, loading the plates more heavily helps to unload the screens. If you had a 6.6k OT in a Marshall, it would probably burn up the screens on your first power chord, but at 3.4k they sort of survive. The screens aren't stressed at idle and low volumes: it's mostly when the amp is cranked up.

Right because the plate voltage is dropping a lot more than the screen voltage is, which causes screen current to increase as soon as the plate voltage drops below it.

Whereas with the dual rail supply, with the screens being at a lower voltage the plate only drops below screen voltage for a shorter duration of time than it would on the single rail supply.

And yes screen voltage drops as well, but not nearly as much as the plate.

In regards to dropping the plate load on EL34 screens...if the plate voltage is no higher than about 425V OR if the power transformer can sag enough under load this isn't a problem. However I've seen a few Marshall clones with higher plate voltages that used a stiff power tranny that would only sag about 30 volts at full clip, and as such will redplate at full clip unless you mismatch the load up by 1 (16 ohm on the 8 ohm tap), which on a 1.7K p-p OT would up you right to 3.4K. In this scenario, the fact that the PT DOESN'T sag actually works against running the lower load.

Thanks for the explanation on average power. I was actually beginning to see that graphically on the load line charts the more I looked at it.

I'm actually considering doing that on my 100 watt Marshall clone. I have the MetroAmp 100 watt dual voltage PT in it with their Dagnall clone OT. If I mismatch with the 16 ohm cab on the 4 ohm tap that should give me a 6.8K p-p load. The power tranny has a 350VAC CT along with taps at 80% of this (about 300V CT) so I could use those taps as a dedicated screen supply. Also plan to double the plate/screen filtering to 100uF as well as try a 10H choke (currently have the Mercury Magnetics MAR100-C 3H choke installed).

[Steve Conner]

I'm actually considering doing that on my 100 watt Marshall clone. I have the MetroAmp 100 watt dual voltage PT in it with their Dagnall clone OT. If I mismatch with the 16 ohm cab on the 4 ohm tap that should give me a 6.8K p-p load. The power tranny has a 350VAC CT along with taps at 80% of this (about 300V CT) so I could use those taps as a dedicated screen supply. Also plan to double the plate/screen filtering to 100uF as well as try a 10H choke (currently have the Mercury Magnetics MAR100-C 3H choke installed).

Whoa there... I think you only want to mismatch up one step. 6.8k with two pairs of tubes is 13.6k per pair. Unless you're going to pull that trick where you go down to one pair of KT88s and double the B+.

You also double the voltage seen by the OT, which moves the bottom end of the power bandwidth up one octave. If it would go down to 40Hz clean at full power before, it'll only go down to 80Hz at the higher voltage.

[Wilder Amplification]

Whoa there... I think you only want to mismatch up one step. 6.8k with two pairs of tubes is 13.6k per pair. Unless you're going to pull that trick where you go down to one pair of KT88s and double the B+.

You also double the voltage seen by the OT, which moves the bottom end of the power bandwidth up one octave. If it would go down to 40Hz clean at full power before, it'll only go down to 80Hz at the higher voltage.

Ah yes you're right. 1 step for 3.4K is what I'd want with 4 x EL34s. Nice catch on that.

Now for my final build, I used a 6550 datasheet since the only KT88 datasheets I could find only give specs for Ultra Linear operation and I figured 6550 specs would be close enough to follow. The data sheet for 2 x 6550s in Class AB states -

Va = 600V
Vg2 = 300V
Ra-a = 5K
Power Output = 100W @ 5% THD

Would there be any reason to go any higher on Va, like say to 650 or 700 volts with the screens at 350? Or should I expect to see 100W out in that config?

[zapper]

SVT-1 ( 6550 ) Mains Transformer Specs

Hi, I have a mains transformer being rewound for the SVT 1 with 6550's. We cannot count the turns as
one of the secondary's are so badly burned that it is flaking away in half turns, and so was very pleased
to find this post. The data that Enzo posted for the secondary's were 600V, 330v & 150V. Is it possible
just to clarify for the rewind that:

A) There is the main HT secondary winding = 0V - 600V

B) The tapped secondary ( screen & bias winding ) = 330V - 150V - 0V - 150V - 330V

As i really would not like to get this wrong and have to start all over.

I believe the new SVT's to have a 0V - 520V for the main HT & then a 260V - 125V - 0V - 125V - 260V and
for this reason I would like to make sure I have not misunderstood the specs found here.

Was not sure if the 600V figure was the DC rectified, as 600V AC x 1.414 is 850V DC ( unloaded ),
which I thought may be a bit too high, as the new SVT at 520V AC would give 735 DC ( unloaded ).

The figures for the new version SVT come from a Mojo Part #7020100 from
Mojo Transformers » Ampeg SVT Power Transformer

Please forgive me for bringing the above right down to basics, but I'm sure you can understand that it
would a costly and maybe irreversible mistake to get this wrong, so would be forever grateful if
you could help.