7591-based Crown Bass Amp- screen resistors?

[oldhippy]

greetings from a hardcore lurker. Thank you for a great site! I am humbled by the knowledge and experience you folks seem so willing to share.
I came across an old 7591 based Crown Bass Amp and I haven't been able to find out anything about it. It's loaded with old resistors, a tubular piece of carbon with the leads wrapped around the outside ends and dipped in enamel. Any thoughts as to replacing them all? The ones I've cked seem to be OK and the amp is pretty quiet as is and sounds fantastic. Any idea how old the amp is?This is my first time with 7591 tubes and as I drew out the schematic, I noticed there are no screen resistors. I also noticed that Ampeg used the 7591 with no screen resistors, so I figured 'ok', but then I read an article on the Aiken site about retrofitting old amps with screen resistors to help save the tubes.  Should I install screen resistors and if so, what size? And would this affect the tone appreciably? Thank you... Steve in Arizona, aka oldhippy

[DummyLoad]

Any idea how old the amp is?

7591 was released in 1959 - so i'd guess 1960 or later...

Should I install screen resistors and if so, what size? And would this affect the tone appreciably?

if you like the way it sounds, and it's survived all these years without any issues, then why mess with it? however, yes, it would probably affect tone to some degree. i don't think anyone who does not have any experience with that particular model could give you a definitive answer.

[PRR]

> tubular piece of carbon with the leads wrapped around the outside ends

I bet it is Japanese, a cousin to the Elk sometimes seen in these parts, also the post-US Kustoms. If so: pretty well designed and built. Fix anything that is for-sure broke, and leave it alone.

[P Batty]

The Guyatones used 7591s as well- does it have a anodized chassis?  Oil Caps? Terminal strip construction? All hallmarks of Japanese construction. I have an Elk "Prosonic" with those features and it still works well and sounds good (new Celestion speakers, however), and still hasn't had  a cap job after 45 years. How about a picture?

[oldhippy]

Thank you for the replies. The amp came to me as a basket case, and just a chassis at that, no cab to speak of. So it is just a player. It's been tinkered with, so of course, the second channel had been stripped out. He gave me the  parts in a baggie, tho. It's not all that well built, flimsy cad plated chassis, no tube retainers, heater wiring was handled poorly, so it was alittle noisey but responded well to basic treatment and is now church quiet... no oil caps, no choke, and lots of term strips. He said some times it would get really loud...I turned it on, and the pilot light lit up the whole room...cked plate voltage and it pegged out my fluke! Put in a used PT I had laying around that gives me a plate E of 425...still pretty hot for a 7591. It would red-plate while playing hard so installed bias pot...it was way under biased. And tubes are pricey. but,I'm a blues player and it rings my bell...tone wise, similar to my home made 59 bassman, just not as loud!! The nieghbors love it! I'm not sure why because circuit-wise, it's much closer to an old ampeg...and I'm not a fan of the ampeg sound, on guitar, anyway.  Again, thanks for the input...

[Shrapnel]

He said some times it would get really loud...I turned it on, and the pilot light lit up the whole room...cked plate voltage and it pegged out my fluke! Put in a used PT I had laying around that gives me a plate E of 425...still pretty hot for a 7591. It would red-plate while playing hard so installed bias pot...it was way under biased. And tubes are pricey.

425v too hot for a 7591? On the contrary. With 450v on the plates, ye olde tech sheets call for 45W out fixed bias, (and with a different OT and cathode bias, even less).
As far as being too pricey. what Few NOS, (a rarity if found at all admittedly,) and Used found on places like flea-bay will be really pricey. New fab (and I'm NOT talking the 7591xyz) 7591s are affordable in comparison. The new production 7591s are similar to the 6L6, 5881, EL34, or 6V6 in some stores (like TheTubeStore.com)

[HotBluePlates]

Should I install screen resistors and if so, what size? And would this affect the tone appreciably?

Tone? No. Power? Maybe...

With screen resistors, think power not tone. The control grid, G1, has the most impact on plate current, meaning a small voltage change results in a big plate current change. The screen grid, G2, is the second most-effective controller of plate current.

The screen also passes current, and when the plate current is large, some of that current strikes the screen grid and so screen current typically increases somewhat when plate current goes up. If you slap a large resistor in series with the screen, the change in screen current during high power peaks will result in a significant change in screen voltage, due to voltage dropped across the screen resistor (in accordance with ohm's law).

Another way to think of this is reducing the screen voltage will ultimately reduce the peak plate current possible from the tube. Reduced peak current = reduced maximum output power.

Check out the 7591 data sheet, and look at Typical Operation, Pentode Operation, Class AB1. Pay attention to how much the screen current changes from no-signal to max signal input to G1. The values of current are for 2 tubes, so cut each in half. Zero-signal screen current is 5.5mA and max-signal screen current is 13.5mA for the condition with 400v on the plate.

Now look at the lower chart on Page 3. It shows plate current with changing screen voltage, all while holding G1 at 0v. This is showing how your peak plate current (which peaks when the incoming signal is positive by the same amount that the bias is negative, momentarily driving G1 to 0v) is affected by the screen voltage. Depending on exactly where you look, a 50v reduction in screen voltage results in a 40-50mA reduction in peak plate current.

So we don't want the change in screen current to knock the screen voltage down enough to zap power. We have a change indicated by the typical condition of 8mA from no-power to full-power, and 50v of change on the screen grid is going to be very noticeable. 50v / 0.008A = 6250 ohms. So don't use a screen resistor that big. We haven't quantified how much power output is lost, and I'll save that for another time. But if you used a 2-3k screen resistor, it would probably be a noticeable compression effect and less total output power, but not as bad as 6k.

Also notice that all old amps typically used no screen resistor, 100 ohms, 470 ohms or 1k, depending on the tube type and conditions. When used, the screen resistor was sized just big enough to provide some protection against over-dissipation or some grid-stopper effect, but small enough to not cause an appreciable change in screen voltage.

Depeding on your sonic goals, you could use a small resistor like 470 ohms or a big one like 2-3k. The choice is between no-sag and full power, or reduced power and audible sag.

[oldhippy]

thank you, HotBluePlates, for the curves and for walking me through it. really. Against the odds, I just came across another 7591 based amp that has distracted me. This one sounds really nice too, just not alot of power. In place of a choke or small value resistor between the plate node on the power supply line and the screen grid node, there is a 10 watt 20K wire wound, which puts 175 volts on the screens...what's up with that? I've never seen that before... respectfully, grasshopper

[HotBluePlates]

... which puts 175 volts on the screens...what's up with that?

It depends on the balance of other design features. And you have a filter cap to ground, presumably, at the screens. That negates any sag effect from signal, but power might still be down.

Have you ever seen a 6550 amp with 600v+ on the plates? The screen voltage is around 300-350v. That's because the screen voltage is being used to dictate maximum current swing, while the high plate voltage is used because voltage * current = power. Said better, the lower screen voltage is about keeping plate current (and therefore cathode current) from going too high and damaging the tube.

I don't know the designer's intent in your case. I'd have to see the rest of the amp to know if there was some other factor that made low screen voltage look like the best option. The 7591 was born thoward the end of the tube era, and was one of hot designs. Dissipation rating of a 6L6G, but 6V6 size, and (I think) higher Gm than both so that it could be driven with little input signal. It wouldn't make sense to me that you'd pay for a super-tube and then choke it back without a good reason.

[oldhippy]

thank you, hot blue plates for the comeback, sorry so long for the reply, been working. I'm a retired industrial electronic tech/ instrument man and my experience with tubes is pretty limited...but I've built a 5F6a and a deluxe and repaired less than 2 dozen amps and have yet to do any harm. Can you give me an example I can look up in regards 600v plate/350v screen? I'd like to take a look. I just recently got a copy or the rca receiving tube manual. I believe I read somewhere that a Leo F. engineer prototyped a 7591 based amp, had one of Leo's friends gig with it, brought it back saying it wasn't nearly as loud as his 6L6 amp, and didn't pursue the 7591 any further. I'd love to get your thoughts on the circuit, but have discovered my scanner no longer scans to file. FWIW, the preamp tubes, 12ax7's, only have 70 v on the plates...plate resistors of 500k and 240k...any thoughts on that? BTW, no neg fdbck loop. I'll try to get a new scanner this wkend and UL a schematic. ...peace...

[oldhippy]

P.S. I understand basicly that high r values on the plates indicate high gain but are there any sonic advantages to that? FWIW, I believe I'm looking at a split load inverter, and the 12ax7 gain stages with the high plate load r's are using 3 to 4k cathode resistors w/10 microfarad bypass caps...peace...

[PRR]

> an example I can look up in regards 600v plate/350v screen?

http://www.mif.pg.gda.pl/homepages/frank/sheets/135/6/6550A.pdf top of page 3.

For very best power performance we want a low-Mu pentode working at low screen voltage but high plate voltage. Both supplies need substantial power but at different votlage; this kinda means two separate power supplies (PT windings, rectifier bottles, caps).

Radio transmitters sometimes do the 2-supply thing. In fact some big tetrodes run 500V on screen and 3,000V on plates.

Audio amplifers usually avoid it, using one supply for everything, forcing the screen voltage very near plate voltage.

For "screen voltage very near plate voltage" the optimum load for a single (SE) tube is about twice the plate resistance of the pentode tested as a triode. Push-pull is the same situation but different numbers because of the way we spec PP OTs. About 4K for 6550 (and 6L6 and EL34).

Since screen needs more filtering than plate, and cheap filtering loses some voltage, we often work the screens at 80%-90% of plate and use a load about 120% high. That's why several of the 6550 (and 6L6 and EL34) suggestions show 5K.

-IF- the plave voltage rating allows it, and the dissipation is OK, and the higher load impedance does not complicate the OT winding, you can raise the plate voltage still more.

Around 1965 a sweet trick appeared. Use a voltage-doubler power supply. One winding and two of those new non-vacuum (silicon) rectifiers can give both a +300V and a +600V supply, and using (two) low-cost 350V caps. Bogen built a bunch of very neat 560V/280V amps making a solid 150W from four bottles (and 300W on the 8-bottle version).

However. The 7591 is not so big and beastly a bottle as 6550. It is specifically intended for compact affordable stereo hi-fi driving the new (at the time) Acoustic Suspension speakers which in most rooms want more than 20 Watts each. It needs over 300V on screen to make good current, it isn't supposed to take 500V on plate, and it has just enough dissipation for 30W-45W amps, which can be built with a single 350V to 450V supply. I am sure we could force it to 50 or 55W output, but life would be too short for many hi-fi owner's preferences (we rarely changed tubes in the day).

Work it similar to the way the guys who made it TOLD us to do it:

http://www.mif.pg.gda.pl/homepages/frank/sheets/168/7/7591.pdf page 2

For guitar, let the screen sit a little high of what was suggested for hi-fi work, and increase the grid bias to keep dissipation reasonable.

[HotBluePlates]

I type a reply, had internet issues and promptly lost the work. And PRR answered the questions.

As for the Fender story, I hadn't heard that one. But I'm guessing your RCA tube manual is a reprint of the 1959 manual, and doesn't have the 7591 (I don't think it was around yet). There is a G.E. reprint available from 1973, and it definitely has the 7591.

But say mid-60's Fender already had big amps with a pair of 6L6GC's making around 50w, and smaller amps with a pair of 6V6GTA's making maybe 20w or a tad more. 7591's might fill the gap by making 30w or so from a pair, but that's not so much more than 20-22w to make a big difference, and sounds just a tad underpowered compared to 50w.

The great thing about the 7591's is being able to drive them without much input signal, but there wasn't the need in Fender's lineup for the middle-power option (which probably didn't sound very much different than a Deluxe Reverb; see Ampeg Reverberocket), and Fender already had their preamp and phase inverter options figured out for the existing tube line-up. I'm guessing it was a new and possibly more-expensive tube model that just didn't give Fender something different enough to incorporate into the lineup.

For other manufacturers in other situations with different challenges, it might have been just what they needed.

I really need to build/modify something to use the handful of old-production 7591's I've got.

[PRR]

> type a reply, had internet issues and promptly lost the work

Anything over 10 words, type in NotePad (SimpleText, TextEdit, Qedit, UltraEdit... any plain text editor), then transfer to the compose box.

You get more room to type. And think.

Alt-Tab between browser and editor may be faster for referring to other posts or other sites.

You may get spell-check.

Browsers are buggy. Networks have poor reliability. Most text editors are mature, pretty robust, and have a way of saving text locally. Wires down? Ion-storm? Microsoft bugs? Save it and send when things improve.

(Yes, I realize that _you_ may have restrictions working from military systems. I'm just saying....)

[PRR]

It's like turning a Falcon into a Mustang into a Torino and a Maverick then into a Fairmont. Car and tube makers try to build what the customer wants. Often that means a product "grows", then we worms turn and ask for a smaller product like it used to be.

The original metal 6L6 was a neat tube, and compact.

For whatever reason it was re-issued at the same ratings in a much larger glass version.

Tung-Sol improved plate-stuff and glass gave the 5881, fairly compact and good for over 40W/pair.

TV-set improved plate-stuff and glass gave the 6L6GC, fairly compact and good for over 50W/pair.

But these are big bottles. Too tall and kinda wide for a livingroom stereo 35W/ch. Remember cars were longer-lower-wider every year, and hi-fi had to get slim and elegant.

Someone musta noticed that the metal 6L6 would be a good fit. But it was only rated 250V on screens. And it was out of production. But there was a LOT of market for hi-fi. Oh, and if you put something like that into production, can you use the higher-gain structures evolved for TV-set work? (EL84/6BQ5 already had TV video-amp detailing.)

7591 is frisky new wine in old smaller bottles. And it must have been as cheap as anything else, Fisher used truckloads. Nothing very wrong with it: MacIntosh used them in the amp down from the 6L6 model.

> it wasn't nearly as loud as his 6L6 amp

Well, when the 7591 came out a "6L6" was always a 6L6GC (and there's a rumor that Leo had a hand in that). 7591 is comparable in beef to an original 6L6, but designed to be less than what "6L6" had become.

[oldhippy]

much here to digest and ponder. I will try to get my hands on a scanner this wkend to get you something to look at. great chance to learn here, and I'm on it. Thanks for all the replies. All will be read, and re-read.

After years of working with solid state, I was ready to walk away from electronics, but this tube thing and that point-to-point wiring...the shiny glass cylinders filled with little pieces of metal, glowing wires in a cloud of blue gas cranking out 500 volts of music....Glorious! Gentlemen, I have seen the light and it is a 6 volt filament!

[PRR]

>  the 7591.... I'm guessing it was a new and possibly more-expensive tube model...

Registered 1960.

I came across a much later and very special price-list: McIntosh 1977. Some prices are absurd.... 6V6 stands out. (You KNOW Leo was getting a better price.) However 7591 is as cheap as any power bottle.
 

[PRR]

Unverified gossip----

http://archive.ampage.org/articles/2/gagd/069635/tale_of_7591a.html

this is a perfect segue into a little tale about 7591A's/7868's: if it is not generally known, who among you know why they were ever made in the first place, and why they disappeared so quickly even after RCA did such a massive marketing campaign to cram em down the early american consumer electronics manufacturers grundgy throats (whether they wanted them or not)? huh?  
 
it all began with philips and the patent for the audio power pentode. some of you may be aware that no audio power pentodes could be made in this country without paying philips a royalty... and thus us americaninos were forced to live with the beam power tetrode, because shit if we gonna pay them a dime. we won the god damned war for crying out loud! 6L6, etc. etc. etc.... well, for the most part, us cowboys in the new country were more than happy to do without the suppressor grid in our output stages. but something happened in the very beginning of the sixties that sent tremors far and wide through the halls of RCA. it was called the EL-34. and all of a sudden, that pesky upstart goniff mashugganah, sol marantz, actually let sid smith give up on the good ole US of A and buy euro! jeez louise. and if that wasn't injury on top of insult, that jerk fisher, and that creep hafler, figured sid might know what he was doing and they started making amps with those damned skinny little fairy tubes too! it just ain't right! so the machinery down in darkest jersey cranked itself up into a fever pitch. "we got to get us an american vacuum tube that'll compete with those friggin snobs over in england, holland and (gasp!) germany", they said, and "whats more, if they (american consumer electronics manufacturers) don't buy em, we won't sell em anything else that they want, so there!" so, they mixed a little sweep tube scifi in with some octal audio packaging and, VOILA', 7591. it was prone to high frequency oscillation so they rotated the plate 15 some degrees and presto chango, 7591A. marantz was too small and too idealistic to buy the amazing new EL-34 like beam power tetrode made by RCA right here in the goddamned USA but fisher was talked into it, as were others. the higher transconductance, lower impedance and lower distortion EL-34 (and EL-84) made it easier to make simple circuits that sounded and performed exceptionally. this was now also the case with the 7591A, which sat sort of in between the EL-34 and EL-84 for power. all should have been fine and dandy, except that MO valve (secretly together with tungsol) got foolishly ambitious and sick and tired of losing at hardball with the bastards in eindhoven. they managed to sell the idea of massive amounts of raw audio power available only with KT-88's (6550's). and suddenly, there were too many options. it was the beginning of the wildest marketing race in the history of modern industrialized business. from about 1962 to 68. very similar to the car business then, or computer hardware trade now. some pitched power, some convenience, some the latest tech (transistors). even marantz was doubling up on EL-34's to stay in the arms race. everyone had to have something "new" every year. the 7591A got lost in the shuffle. it just wasn't remarkable or even better than an EL-34. this wasn't something that only flopped for RCA. it happened to almost everyone else in some way or another. for a short minute, it looked like philips might actually end up with egg on their face just like everyone else! no way. kings of the cheap but good, philips ended up buying or putting all of their european competion out of business, including telefunken. they eventually bought amperex, sylvania and magnavox, who got radically over extended, and after the japanese destroyed the american television business (and RCA with it), philips wiped out anything left standing in the USA. only sony stands in its way to total world consumer electronic domination today. i'm getting carried away here a bit, and i've generalized a fair amount as well, but the point is that the 7591A was never anything more than a desperate local replacement for a EL-34. thats a fact. it always drove me nuts how sacred that tube is held in some circles. don't they know? its just a tube! and i still think that i'd rather have an EL-34! in any case, the most obvious replacement for a 7591A has always been an EL-34, except that it doesn't fit under the cover of a fisher 400. but try telling that to any of the folks that have amps that used them... ok, thats my venting for the day. don't forget that the 6CA7 is NOT A PENTODE! (its a beam tetrode!) and NOT an EL-34.
 
EL34 is NOT a drop-in for 7591 sockets, you will have to re-bias. And in many cases, you will have to drill the lid and let the bottles stick out. But there may be truth in the idea that the US companies got goosed by euro tubes.

[Shrapnel]

EL34 is NOT a drop-in for 7591 sockets, you will have to re-bias. And in many cases, you will have to drill the lid and let the bottles stick out. But there may be truth in the idea that the US companies got goosed by euro tubes.

Another note/caveat: While the EL34, 6CA7. 6L6, 6V6, and 6550 all share similar pin-outs, the 7591(A) has a totally different pin-out.  One NEEDS to rewire the socket OR (almost immediately) destroy the "replacement" tube stuck in that socket. But... The Chinese (or is it Russians) have provided this bottle again under the brands of JJ and EH. (Well, one that seems to work close enough to the original 7591(A) designs anyway that no mods are needed to the original circuits calling for that tube.)

[imaradiostar]

In my experience the EH 7591a biases up very similarly to my old 7591's. It produces decent power with a pleasant tone and sounds different from most common tube types. Just testing on a scope with negative feedback disabled it has a look to the waveform that other tubes lack. It's a strange bird but I really like the way they sound.

In my test rig amp I can bias anything from a 6v6 to a KT88. When I install 7591's in socket adapters I have to pad down the bias voltage substantially to allow the tubes to conduct at all! I guess the thing to remember there is that the 7591 did well in designs that were getting to be too hard on EL84's. You could sub a 7591 into a circuit intended for an EL84 and it'd make more power with only slightly more filament current required and bias up not too differently from the EL84's. My Scott 222c sounds great but I'd imagine there is a reason that later versions of the same amp (with a different name) had supposedly identical transformers but made more power than the EL84 version.

They're a little pricey but they're neat sounding and isn't that the whole goal?

Just to add to what PRR was saying- don't neglect the difference in filament current if you go to El34's. You may find it doesn't matter or you may overheat your power transformer.