7591 Build.

[8wattjack]

1

[PRR]

> Tube data says -25v with 400v on the plates

That's not on the sheets I see.

[Willabe]

The chokes are in series, so each choke see's the full current from the B+.

The Henry rating of the 2 add together. So, lets say their each 1.5H, so both together would get you 3H.

[shooter]

I used a pair of these in a proto stereo champ on steroids, they serve me well as book-ends 
If I ever get around to a cork-sniffer audio build i'll have to get new bookends 

https://www.edcorusa.com/cxc125-10h-200ma

[SoundmasterG]

7591's (and 7868's) only have a reputation for being very clean in guitar amp circles because that was how Ampeg designed the amps that used them. For a power tube they have a lot of gain, with the 6BQ5/EL84 having more among common types but most others are lower gain than 7591/7868. What this means in practical terms is that the 7591/7868 can be driven easily. In a 6L6 style amp such as many Blackface Fenders, if you take that same preamp and stick it in front of the 7591/7868 power section, you'll find that you have much more overdrive and distortion and you might actually have too much. In my experience the 7591/7868 is in between the 6V6 and 6L6 for output power, but since it is easier to drive the preamp and phase inverter don't have to amplify as much, and the 7591/7868 are very harmonically rich with an easy breakup and lots of touch response. They also last a long time in comparison to some other types. They're my favorite power tube actually.


Your choice of so many 12AX7's in the amp might result in too much gain. I was able to get full power and a nice fat AC/DC type overdrive at full volume for four 7868's with a parallel 12AX7, tone stack, 5879, and LTP PI, no NFB.


Greg

[tubenit]

Here is an idea for you to consider. 

Compare the layout with the schematic and look for errors.  IF there is a discrepancy, go with schematic values.

This would meet your 4 "preamp" 9 pin requirement with FX loop. 

HoSo56  (which I've built in 6BM8, 6AQ5, 6V6 and 5881 tube)  and FX loop and 7591 tubes.  Given the success with the other tubes, I don't see why the 7591 wouldn't work?

Here is an Ampeg amp with 7591 cathode biased.   https://el34world.com/charts/Schematics/files/Ampeg/Ampeg_reverb_rocket_12a.pdf

IF you built it and it had too much gain, simply change V2 to a 5751 and the V4 (LTPI tube) to a 12AT7.

With respect, Tubenit

[SoundmasterG]

I was able to get full power and a nice fat AC/DC type overdrive at full volume for four 7868's with a parallel 12AX7, tone stack, 5879, and LTP PI, no NFB.

I see. I think I see the idea. I need to diversify in pre amp tubes. My knowledge is limited beyond old mr fizzy.  Did you use 12ax7's for the PI?



Actually I used a 12BZ7 for the LTP PI. It is NOS and not made anymore, found at www.tubesandmore.com, (and other places) has the same gain as a 12AX7 but twice the current capability, so they can drive the power tubes better than a 12AX7 but still give the same gain. They are a taller bottle, and are often microphonic so I used one of those metal things that fit over it to stop microphonics, I can't remember what it is called at the moment. I like the use of that tube in the PI spot in this amp, but have tried it in other amps and other positions and didn't like it there. Same pinout as the 12AX7, 12AT7, 12AU7, etc.


The fact that I used the 5879 which is a pentode and has high gain makes a difference. It is going to overdrive the input to the phase inverter past a certain point, and will overdrive on one side of the waveform quite a bit. The parallel 12AX7 overdrives on the other side of the waveform but just slightly when the volume is up. I designed the stages so the power amp distorts first, closely followed by the phase inverter, and then the 5879 next. Having the power amp distort first gives the amp good touch response.


Greg

[Skilpadda]

7591 likes to be mated to a LTP PI more similar to what is driving a pair of EL84s like in an 18 watt circuit. Some Gibsons use these output tubes. I have a 1966 GA 35 RVT Lancer which use these tubes. An amp with wonderful reverb and tremolo thats only let down by its transformer PI and some horrible tone filters (see video on YT from uncle doug) Mine did not have these filters from factory. In my ears the reverb and tremolo sounds better than any AB763... I am tinkering with the idea of making one of my 18 watters into a 30 watter using such tubes, or a JTM 45 into a 30 watt. Both the current EH and JJ 7591 are strong tubes that can take a lot of abuse in my experience...

[sluckey]

If 2 100uf caps in series = 50uf in capacitance, how does the rectifier see it?

The rectifier sees it as 50µF.

[PRR]

440V on screens is legal.

7 Watts is a lot. Do you have screen resistors ???

[sluckey]

I put a 5K between nodes A and B to drop screens down to 388V. My chokes are between the rectifier and node A.

[SoundmasterG]

My Bogen CHB100 that I modified ran 480V  B+ and the screens were close to that too in the stock circuit...I think I dropped the screen voltage to 440V if I remember right. I am using a resistor for that node however....no choke. I used 470 ohm screen grid resistors on each screen grid. I'm using EH 7868's in it as they sounded just as nice as vintage 7868's. I would guess the EH 7591 should be the same guts as their 7868 as that is the way it was done with vintage 7591/7868's but who knows what they did as they don't share valid info on their current tubes.


I never checked the dissipation on the screens on this amp, but watched it in the dark as I played and nothing was amiss. I suppose it could be excessive and I should check it, but the tubes have held up in there with lots of use, usually loud performing volumes, and over a period of 4 years still going strong. The OT load is 2500 ohms for four 7868's with an 8 ohm speaker, though it also has a 4 ohm tap and a 16 ohm tap. No NFB on the amp either.


Not sure why your 7591 screen's aren't happy in your amp. Which 7591's are you using? Vintage? JJ? EH?


Greg


P.S. I agree with you about the higher voltage on the screens. The amp sounds a lot better that way than when you have the voltage lower.

[SoundmasterG]

I am running jj's. They seem plenty robust. I will try EH though. I have been reading up on the choke selection, and I still really don't understand. I am only pulling 1.1 watts at idle. Cranked up with the 4 henry choke between nodes a and b with 1 k screen resistors, dissipation rises to 7.5 or so. With a 10 Henry choke in that position, dissipation rises as high as 10 watts cranked! So the moral of this story is I need to figure out what choke is going to keep my screens in check. I also need to spec a new PT since this one cannot keep up too well with the way I am running it. It gets hot. I have virtually no ripple. I can set the scope all the way to the lowest setting, and its still very small. The amp sounds great, so I think the pitfall is the PT, and my choke selection.  If I can get the screens around 6 watts, I think it will be fine. I am also gonna check on my impedance, and make sure that's not part of it. 480 and 440 seems like optimum numbers. I bet that Bogen sounds good! I have a chb30 with 7868's and I love it too.  I still sometimes use it for monitors, and everyone loves how they sound on it.

I have a pair of JJ 7591's but I haven't tried them in anything yet. One of these days! I think the EH 7591 is a much larger tube physically than the JJ's and also vintage 7591's, so you would need to make sure it would fit physically. I think they are about the size of a 6L6GC. I know they often don't fit in vintage Scott amps for example.


I'm not sure why your choice of choke isn't working well in this circuit application either.


Here is the definition of a henry that I grabbed off the net. Inductance is only a momentary thing. Exactly how much of a momentary thing depends on the amount of inductance an inductor has. Inductance is measured in units called henrys. The definition of one henry is simple: One henry is the amount of inductance necessary to induce one volt when the current in coil changes at a rate of one ampere per second.


I would guess that the 10 Henry choke makes the amp hum level quieter because the filtering is likely better with that one? Is there a voltage difference between the two chokes? You could try a resistor and see if there is much of a difference with it? On my CHB100, it used a voltage doubler, and I stuck with that, though I made it a full wave doubler using four diodes. My first caps are 220uF, so that works out to a 110uF cap in a normal supply. I then have a 1k5 resistor, 5 watts, and a 32uF cap. My initial B+ out of the supply after the first cap is 483V, and I added Power Scaling to this amp, so that sucks up some voltage. The B+ at the tube plates is 463V and the screens are at 461V. Screen resistors are 470 ohm and 5W each. I have a switch to disable Power Scaling or enable it, but the voltage still gets sucked up by the MOSFETs so the B+ doesn't change. My amp has a bit more hum and hiss than I'd like, though I haven't looked at it on a scope yet. I used the Bogen layout with the stock Z mounted transformers right next to each other and that was probably not a great idea. It works well and sounds great though....so good in fact that I haven't pulled it apart since I first got it together in 2012!


Greg

[DummyLoad]

from past experience with experimenting with the 7591:
1) they don't like over 400V to the screens or plate in cathode bias - yes, i've read the datasheets!
2) for a guitar amp with 400V or more to screens, use a 2K 3W-5W screen resistor for each tube.
3) use a large-ish grid stopper (10K-22K).
4) in cathode bias keep the grid1 circuit resistance around 270K or less and around 100-120K if in fixed bias.

7591 are one of my favorite tubes, however, the NOS parts are drying up fast and those are the only ones i care for. the 80's production sylvania parts sound best to me.
they are a very smooth sounding tube (like most other types too) if you control g1 blocking distortion & control screen power.


i miss my ampeg gemini1.   



--pete

[SoundmasterG]

Dead quiet. Hum is definitely not an issue. There is little difference in voltage between the two chokes, but the dc resistance in nearly the same. Looking at Duncan's PS calculator, I think I need to go higher in dc resistance maybe? I could make it work with a typical crc filter, but I am going to try to stick with the plan.

 Thanks for that data. However I am in fixed bias with neighborhood of -22v on g1.  Bias adjusts nicely. I got screen's down a little with a 1.5k last night. I will try a 2k tonight. I figured a larger grid stopper on g1 would cause g2 current to rise due to accumulation, but I probably have that wrong. I will try that too.

  Thank you all for the help.
 Jim

The reason I suggested trying a resistor and cap instead of the choke and cap is to see if there is a difference with the screen dissipation, and it might give you a direction to go in, or tell you that the choke is a red herring and point you somewhere else to look for the cause.


My amp is fixed bias also. I didn't like the sonic affect that larger screen grid resistors had on my amp so I stuck with the 470 ohm. The EH 7868's are robust and don't seem to have an issue with it either way.


Good luck!


Greg

[sluckey]

  I put the choke in between A and rectifier. Dropped my B+ almost 100v. (Not where I wanted to go.)  How much resistance can I put on the screens before it becomes an issue? I am still not sure how to figure the value necessary, when installing a choke. If I go up in Henry, will that bring down my screen current?

I think you should return to this, but put a filter cap directly on the rectifier. This will bring your B+ back up. Then experiment with the resistor between the plate node and the screen node to get the screen voltage where you want it.

BTW, don't stress over the Henrys in a choke. Henry has nothing to do with dropping voltage. In fact, a perfect choke would have zero ohms dc resistance and would drop zero volts dc. A choke opposes a change in current just like a cap opposes a change in voltage. A choke is all about smoothing the ripple on the dirty rectified dc voltage. Using a choke with caps offers much better filtering than caps and resistors.