AB763 with 7591's?

[Toxophilite]

Besides the pinout and bias what else would I change on a AB763 deluxe (or vibrolux) build if I wanted to try 7591s


i kind of asked this question before but I'm being more specific here

[sluckey]

Anybody build a fender Ab763 style amp using 7591s in the power section?

I'm converting a Hammond AO-63 right now. It uses 7591s. I almost put an AB763 preamp in it but ended up using an Ampeg Gemini II preamp instead. Very nice clean sounding amp.

Those 7591s will work fine in an AB763 type power amp. Expect 30 to 35 clean watts. They need a little less drive signal than 6L6s so if you use fixed bias you will need a range of -20v to -32v. That's about the only thing you need to change.

[Toxophilite]

Cool (and thanks!)
they seem to be between 6L6s and 6V6s as far as ELR goes
 With a deluxe one uses an 8 ohm load, if you put on 6l6s you then want to use a 4 ohm
What would they require for a speaker load given an Ot designed for 6v6s or 6L6s?

[sluckey]

Spec sheet says they like to operate at 6600Ω, so just chose an OT accordingly.

[DummyLoad]

THIS OT is perfect for 7591 in P-P. order it from doug: it sells to you for 100 piece pricing. :-)


--pete

[DummyLoad]

I'm converting a Hammond AO-63 right now. It uses 7591s. I almost put an AB763 preamp in it but ended up using an Ampeg Gemini II preamp instead. Very nice clean sounding amp.

Those 7591s will work fine in an AB763 type power amp. Expect 30 to 35 clean watts. They need a little less drive signal than 6L6s so if you use fixed bias you will need a range of -20v to -32v. That's about the only thing you need to change.

did you build both channels? don't see why two would be needed. are you using the 7199 or the more common 6AN8 or some other variant? just curious.


--pete

[sluckey]

Pete, the only thing that's borrowed from the Gemini is the preamp and James tone stack. I kept the AO-69 power supply, power amp, and reverb circuit. Well, I did chop the reverb circuit and mixed it back in to the main amp rather than use the reverb power amp and separate reverb speaker.

[DummyLoad]

thanks very much steve. very nicely done plan.


--pete

[Toxophilite]

I was wondering how it would work with the existing transformers


So what a  30-1 ish deluxe reverb transformer would want to see on the secondary side with 7591s


I would calculate it this way; turns ratio squared = impedance ratio of 900 


6600/900  = 7.3 secondary

[sluckey]

So what a  30-1 ish deluxe reverb transformer would want to see on the secondary side with 7591s

As you see, the DR OT has the correct impedance ratio for a pair of 7591s. But, be aware that a 20W DR OT is very lightweight for a pair of 7591s. If you rewire the DR tube sockets and change the bias range to run 7591s, you may end up buying a similar OT that Pete linked.

What is your goal? Want to convert one of your AO-43 conversions to run 7591s? Build another AB763 amp based on 7591s? Just a brain exercise? Something else? Please tell.

[Toxophilite]

 Bit of everything. and I was thinking of converting one of my deluxes to 7591s
I've tried it with 6L6gcs and 2 x 10" 8 ohm speakers (4 ohm load) but the octal sockets are pretty close together and the 6L6's are too (maybe an 1/8" apart,,they don't get too hot though but still I wonder)
It sounds good with the 6L6s (biased at 33 ma) but does it have more volume, maybe a little?


I thought 7591s would fit better and it'd be interesting to see if they impart and sonic qualities of their own.
So I was looking into it
However as you say the 20 watt OT might be a proverbial cork in a bottle and it would be best done with something more skookum

[sluckey]

I'm assuming you're talking about one of your AO-43 conversions? Remember, that OT was chosen by Hammond for a pair of EL84s. Converting to 6V6s is easy. 6L6s may heat up the OT. But if you are satisfied with the performance of that Hammond OT with a pair of 6L6s, then you won't have any problems with the 7591s either.

I'd be very interested to hear the outcome. I still have a couple AO-43s sitting around.

Just for a point of reference though, I have an AO-63 on my bench right now. The main amp runs 7591s. And that OT is bigger than the PT on my AO-43s. It's probably twice as heavy as the OT on those AO-43s.

[Toxophilite]

I can't say for sure if I get lot's more volume with the 6L6Gcs. I think they run cooler than the 6V6s though


What I should do is mike up the amp. record it with the 6L6GCs
and then swap in 6V6s, rebias(I have bias points) and then record it that way same settings, mark where the amp was and exact microphone distance and see what results
However the different impedances might make it a little hard to get accurate results
I'll try it at my rock band rehearsal, It didn't cut it with 6V6s (I like a clean platform for pedals) so I'll see if it does with 6L6GCS
and how it holds up . If it works out I'll try the 7591s.


One advantage (to me) of the 6L6s is that I can try out a pair of K110s in the amp

I got my friend one of those AO-63s he was going to try and make it into a bassman but had a new baby instead
I was going to trade him the Randall Uber power amp I got recently for the AO-63 so he can make a bass amp.
Then I'll have that OT..mind you it probably won't fit my crowded AO-43 chassis

[Toxophilite]

Using my home electrocution device to test actual plate voltage for power calculation


Right now I measure the 6L6Gcs bias at 34 (32 ma)  with a plate voltage of 423
This would seem to give me around 14 watts/tube or 28 watts..whether actually translates through the OT is another thing. I get to give it a good test tomorrow


The plug in electrocution accessory might be useful. I might solder actual alligator leads to it (maybe plate and cathode)and then shrink wrap and fold in the other 'pins' as I have to reach into my amp to adjust the bias pot and it's a little nervous making. Handy to measure the plate voltage without yanking the chassis out of the cab. Wish I'd orientated this one like a tweed!!

[Toxophilite]

Ah well , as suspected though it sounds nice with the JBLs, having 6L6s in it didn't really make it significantly louder, It still just doesn't cut it with my loud drummer and bass player(bass player's using an SVT rig with a 8x10 cab! bahahaha)
So  it would need a bigger Ot . I guess I'll leave it as is and think about designing another amp around the 7591s
Thanks

[DummyLoad]

you'll get more power out of a pair of 7591 with the right OT and B+ combination running in fixed bias. run with solid state rectifiers with the OT i linked with a 700VCT  at ~200mA PT. the super reverb PT is a good choice, or if you don't like X mount, then use hammond 373BX/273BX. either will yield ~ 450V B+ with SSR: slightly less with GZ34/5AR4. you'll need ~ -25V for bias with g2 up at 440-450V B+ range. use the fender5F6/JTN45 LTPI to drive it. if you use a JCM800 replacement PT, B+ will be around 430V.


--pete

[Toxophilite]

Thanks for that That, gives me plenty of options
I also a transformer set from an old sherwood S-1000-2 I mentioned in another post that was designed for 7591s
I might reverse engineer that amps power section into another chassis and put my own pre etc. in front of it.
The Pt that goes with that seems to have centertap for everything including the bias supply.


I was checking out the Sherwood again. It runs, just not well and it's pretty harum scraum in there what with old lame repairs, a rusty chassis and loose corroded tube sockets. Lot's of reusable caps and pots from my previous(many years ago) attempt at conversion

The bias supply circuit seems pretty complex, It has 27AC going in and about -25ish coming out with -23 eventually reaching the power tubes. There are a pair of wires each with 27 AC running into two diodes and then into 3 electrolytcs and 3 resistors. I wonder why it's so component heavy?
With SS rectification there is 500vdc on the 7591 plates..whew!

[SoundmasterG]

I modified then eventually gutted and rebuilt with my own circuit a Bogen CHB-100 which used four 7868's. The 7868 has the same internals as a 7591 in a different bottle and pinout. Most companies were pretty conservative with how they used the 7591/7868 and didn't drive them as hard as they could. If you design the preamp and phase inverter properly, and run higher voltages on the power tubes and phase inverter, these tubes are just about the perfect power output these days giving 30-35 watts out of a pair. The RCA manual says it will do 44 watts with those voltages, but I think that is optimistic myself. The Bogen with today's wall voltages and my circuit in it was about 480 V B+, and the screens are just above 450V and the amp is using four 7868's into a 2500 ohm load, but the amp is putting out around 50 watts RMS. If the RCA manual was correct then it should be giving out about 88 watts, but it isn't close to that figure, partly because the output transformer in these amps is undersized. Anyway, drive the preamp a bit and you can get good gain levels out of it. The 7591/7868 is a sensitive power tube with a lot of gain similar to an EL84, and if you hit it with gain levels like in a 6L6 Fender amp, it will give more distortion, a bit less headroom, and a super harmonically rich and touch responsive sound. If you hit it with Ampeg levels of preamp and phase inverter gain, then it sounds like a nice, clean amp with good headroom.


The EH 7868 is a fantastic tube btw. I haven't used their 7591 yet. I have some JJ 7591's but haven't tried them yet either. The EH 7868 fits into the original Novar socket, but once you use it in there, you can't use vintage 7868's as the pins are larger on the EH tubes and it basically stretches them out too much to revert back. They work better in the modern available magnoval socket. The EH 7591 is taller than vintage 7591's also.

Greg



Greg

[Toxophilite]

Hmm
I'm tempted to make an ampeg type amp as I generally play clean, and I was looking at Gemini 2 schematics
They just don't seem to have one that doesn't use a 7199 as a phase splitter
I don't have any 7199s right now and they can be ridiculously expensive

[DummyLoad]

there are other ampeg amps that sound as good, if not better than the gemini II. the reverbrockets are a fine sounding amp.


this example uses octals. use 5751 or 12AX7 in place of the 6SL7s and 6CG7/6FQ7 in place of the 6SN7.


--pete

[DummyLoad]

this is a great sounding amp - the gs12r. don't let the 6U10 scare you away - replace it with 1 12AU7 and 1 12AX7 with a 1/2 unused: the 6C10 is a 12AU7 and 1 section of a 12AX7.


--pete

[Toxophilite]

Cool
I have a 7247 kicking around too 1/2 12au7 1/2 12ax7


My understanding was that reverbrockets were lowered powered 12-15 watt amps. is that totally wrong?
I was hoping to build something in the 30 watt range or more

[eleventeen]

I believe SoundmasterG (thank you for your posts on this thread) is saying ".....with 100 more volts on the 7591 plates" you could get the wattage output you are looking for.

[Toxophilite]

I was very appreciative of his post and liked the information of an ampeg style preamp and PI being good for a nice clean amp
Thanks!!

[SoundmasterG]

I believe SoundmasterG (thank you for your posts on this thread) is saying ".....with 100 more volts on the 7591 plates" you could get the wattage output you are looking for.

Not exactly. What I am saying is that the RCA tube manual shows the 7591 with 450 V plate, 400 V screen, fixed bias with a bias of -21 volts, and a plate to plate impedance of 6600 ohms, you should be able to get 45 watts. In reality I have seen that these figures seem to be optimistic and the 30-35 watts is more realistic with those conditions. If you push the plate voltage up and get the screen voltage close to the 440 V max, then you will make power in the higher end of that range, though you will likely shorten tube life as these are a pretty small bottle to be hitting with those voltages.

I was very appreciative of his post and liked the information of an ampeg style preamp and PI being good for a nice clean amp
Thanks!!

Ampeg tended to not want distortion or overdrive and ran conservative preamp and phase inverter designs so the power tubes wouldn't distort as much as Fenders of the time. You could use some of their examples if you want, though any design can be revised to give out more or less gain too. The Fender LTP phase inverters can be easily redesigned by changing the tail resistor value to drive the power tubes less, or reducing plate voltage for the same reason. (see http://www.aikenamps.com/index.php/the-long-tail-pair or http://www.valvewizard.co.uk/acltp.html) Though if you want headroom, then I would advise to keep the phase inverter voltage up and just don't hit its input with a hot signal from the preamp. Another good thing for headroom would be to utilize NFB. The LTP gives a unique sound, but other phase inverter designs also work well. The floating paraphase was used by Ampeg as was the cathodyne and both can be designed to be overdriven or clean. Its all in how you design the stages to interact with each other. (Its called gain scheduling) Each stage will output a voltage gain of some level, and each stage can only input a certain level before it will overdrive. If you want clean, then accordingly design the stages to not distort, or distort slightly, and make sure the power stage overdrives first, then the phase inverter, then the preceding preamp stage, etc. and that will give you good headroom and touch response.


You can also use tubes with less gain in the preamp, such as a 12AU7 in place of a 12AX7 like Vox did in the Vox AC100. Those amps all the way up are only as dirty as the Beatles I Feel Fine, and that is a fantastic sound with lots of touch responsiveness, yet it is not as dirty as a JCM800 or something. The phase inverter gives some voltage gain but its current gain is more important since they drive the power tubes. It is all in how you design everything to interact, and just grabbing circuits from one amp or another won't quite get you there without some redesign. Thats the fun part though! Designing stages and how they interact is much easier with a signal generator and an oscilloscope since you can see the waveforms on the screen and see when they distort etc.

Greg

[Toxophilite]

Hmm
Cool stuff and interesting reading
Happily I was just given my brother-in-laws old dual trace scope and I have a little signal generator on my computer. More research will have to done though before I acquire competency in their usage. . Admittedly I'm still at the cut and paste and then adjust based upon how things sound, happily with the advice of the folks here.


The Sherwood S-1000-2 mono hifi that I have ran 7591s and was rated at 36 watts . It sounded good and I was tempted to reverse engineer at least the power amp, The PI looks pretty parts intensive. Sadly , I can't find a schematic for it. There were a couple of  S-10002s designs and all the schematics seem to be for the ones with 4 El84s in the power amp.
I will knuckle down and draw out a schematic for the power amp and PI as the amp stands now and post it here

[Toxophilite]

Here's a schematic I did of the S-1000-2  power amp, PI, bias circuit , and negative feedback/presence circuit


Interestingly it takes it's preamp heater supply from the negative bias circuit and all the grounds from the preamp are lifted and run back to the bias centertap


Please also note that though I included the GZ-34 as that is what it has a socket for and used for a rectifier a previous owner had soldered in a pair of diodes to do the same job..so I kind of included both
I'll try to get some more voltages but it's a little late

[Toxophilite]

what sort of Pi is this?? paraphrase?

[DummyLoad]

what sort of Pi is this?? paraphrase?

the phase inverter is 12ax7-b, however, you're missing the 47k plate resistor to power on 12AX7-b - it's a split-load/concertina when you add that resistor.


--pete

[sluckey]

It's a cathodyne. The way you drew it makes it difficult to see. And you forgot a 47K plate resistor.

There are a few other errors that stand out too. 12AX7-A cathode has no path to ground. Same with it's grid. The junction of the two 47Ks connected to the bias balance pot should connect to ground.

This power amp is almost identical to the power amp in my AO-63 amp. Take a look at the schematic in my thread to see a more conventional way to draw that cathodyne.

[Toxophilite]

ooops , you're right, it's in a funny location (where the wire from the plate meets the decoupling cap to the 7591s on a  terminal strip near the 7591s)
Thanks!


and S.Luckey thanks! Somehow I missed that where the two 47k s on the bias balance meet is the lug on the terminal strip that attaches it to the chassis


I'll look at the other bits and your Ao-63 schematic


Still very new at drawing out schematics and I can't find one for this configuration of this amp.


Any insight on this type of PI and the gain stage before it?
And on the power amp as a whole being used for a guitar amp?
You like it on your new Hampeg thingy S. Lucky?


And is it worthwhile keeping the separate heater circuit for the preamp if I reverse engineer this thing into a different chassis?

[sluckey]

That PI is used on the Princeton Reverb, Deluxe 5E3, most of the old Sunns, and many of the Ampegs, just to mention a few of the upper end of amps. You will probably see the cathodyne (AKA split load, concertina) in as many or more push pull amps as you will see the LTP.

That power amp is found in a lot of guitar amps, maybe with other tubes, but same circuit.

[Toxophilite]

Hopefully I fixed most of the errors, added some voltages and tried to orientate things better in the PI

[sluckey]

Now that looks like a cathodyne! 

12AX7-a cathode still has no resistor to ground except through the OT secondary. Are you sure that's correct?

[Toxophilite]

I was looking pretty hard for that path to ground, Could it be using the negative feedback /presence circuit??
ooops found another mistake in the feedback circuit!

fixed (I think)

There's a path to the OT secondary ground via a 1.2K resistorwhich is actually NOT tied to chassis ground


Interesting
here's how it should look I believe

[sluckey]

You have .75v on that cathode. If you guess that there may be 1mA flowing thru the tube then there should be a 750Ω resistor between the cathode and chassis. Just a guesstimate. Take your ohm meter and actually measure between pin 3 and chassis. Then start looking for a resistor of that approx. size. Very likely to be that 1.2K at the OT.

[Toxophilite]

Bingo!
neat idea
yes 1.279 k between pin 3 and ground, resistor in question measures at 1.279k


I was to reverse engineer the PI and power amp into another more convenient chassis with new sockets etc
Is the separate preamp heater circuit of the bias circuit any sort of advantage?

[sluckey]

Depends on what you will build. If another AB763, I'd just heat the tubes as usual.

[Toxophilite]

I'm tempted by the ampeg front end
I do like to be able to yank the mids though so I like a 3 band eq though I know you can do that with a Baxandall type to if you crank the bass and treble. Ampeg uses a Baxandall type EQ don't they?

[sluckey]

Ampeg uses a Baxandall type EQ don't they?

They use a James tone stack. James is often called passive Baxandall. The knobs operate differently from the Fender TMB stack so you just have to learn a different setting. Read this. It may help you decide...

     http://www.duncanamps.com/technical/tonestack.html

[Toxophilite]

Cool! and thanks
I've had good luck with this one before in my little Roberts/Akai preamp/amp conversions
http://amps.zugster.net/articles/tone-stacks#Baxandall