• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Bogen RP40A transformers....what to build with them?

I tell you what Rongon,
I'm going to build out the Mullard topology and see what happens. From your perspective, is the schematic I proposed useable?

Do you mean this one?
8552eb3f.jpg


If so, I don't know how the NFB loop will work that way. The way I'm familiar with returns the negative feedback from the output transformer secondary to the cathode of the first stage voltage amp.

--
 
Blair,

Actually ( 6 ) 6V6 per channel triode strapped would be perfect as that would give a reflected primary impedance of 10.8K which would be just perfect according to Pionz and others. You would get a nice 15-18W CH which is nice too if run in fixed bias mode.

Do you have a PS for fixed bias supply which would be the only way to run EL34 or 6L6 anyway at that low primary impedance?

You could use Dave Gillespie's EFB mod for fixed bias without the transformer PS with low voltage requirement tubes like the 6V6, EL84, and 7868/7591 types. It uses a negative SS regulator off the B+, so it floats with the PS trans just like the conventional transformer supplied B+.

See the 5th post:

diytube.com :: View topic - Improved SCA-35/ST-35 Performance

Maybe look at Poinz's circuit for what your looking for too! His latest schematics have CCS if you want too!

http://www.diyaudio.com/forums/tubes-valves/149211-6v6-music-machine-build-help.html

Randy
 
Last edited:
Now that I look at it some more...

Since the grids will be at ground in the LTP 2nd stage, you won't be able to get the cathodes more than a few volts above ground, which means the "tail" resistor will be of a small value. That means the balance of that phase splitter won't be very good at all.

One way to make it work would be to bring the common cathode "tail" resistor to a negative supply, so that the tail resistor can be made of as large a value as possible (the bigger the better).

Or you could put a constant-current source in the tail of the 2nd stage, from the joined cathodes to ground.

Then the grids of the LTP would be at ground potential, and you can easily put the NFB into the grounded grid side of the LTP. Like so...
 

Attachments

  • LTP-PI_2ndStage_NFB.jpg
    LTP-PI_2ndStage_NFB.jpg
    81.7 KB · Views: 91
Last edited:
Poinz' Music Machine and Eli's El Cheapo are actually similar. Both use a high-mu LTP-phase splitter-driver to (basically) triode-wired 6V6's. FWIW, I'm all in favor of adapting either one to 6L6 pentodes run at a lower voltage (to suit Blair's 3k6 OPT's). I think that would work really well.

Also, there's gingertube's Baby Huey, of which there is a version with 6SL7's driving 6V6's. That could be adapted to use 12AX7's in place of the 6SL7's and 6L6's in place of the 6V6's.

It's even possible to re-hash a classic Dyna A-410 schematic using a 12AX7 LTP-PS-DRVR, and switch the 6V6's to 6L6's.

There are lots of options. Ain't it fun?

--
 
Last edited:
Do you have any 9-pin mini, small signal pentodes? Like a 5879? You can triode-strap those and they're very close to half a 6CG7 (or think of it as a 9-pin mini version of 6J5GT). That way you won't wind up paralleling your 12AU7's or 6CG7's. Just an idea...
 
I brought up Poinz schem. as it's low parts count in the signal path that count, which should be a benefit as he says. He has an EL34 version too.

That Dyna -410 schem might be worth trying as it's floating paraphase and has it's many fans.

Wouldn't triode strapped EL34 's be better given the primary impedance of the OPT ?

Randy
 
Last edited:
OK guys,

Given a "step back" evaluation of this, I went over the original Bogen schematic and parts list. Bogen put 11 tubes on one heater winding that I referenced the data sheets to get that there are 3.09A in heater current being drawn from the tubes. There are an additional 4 #12 lamps on this winding and an FN cue indicator which I cannot find, but the lamps count for an additional .600mA since they are .15A and run in parallel. So, 3.69A in that winding.

The other had 4 X 7355 tubes for 3.2A and another pair of 12AX7s for an additional .6A, so this winding has 3.8A on it.

If you look at the EL34 data, it looks like in pentode in AB1, which is what I will be running, the EL34 makes about 35W at 315v into 3,400 ohms. Run in triode into 3,600 ohms at 290-300V, I would guess I could get around 15-17W.

I can add a small power transformer for the heaters on the drivers probably even though I am not too exited about the drilling part since the chassis looks so nice in my opinion.

This was my original proposition for these output transformers, but it was a "show stopper" when I did not have the schematic and saw the 7355 to run on .8A current:(

I know the EL34 is fairly easy to drive and I really like the PI with a CCS as a driver. So, is there a 9-pin pentode that will make a good CCS for a LTPI circuit?

Thanks guys!

Blair
 
OK guys,

Given a "step back" evaluation of this, I went over the original Bogen schematic and parts list. Bogen put 11 tubes on one heater winding that I referenced the data sheets to get that there are 3.09A in heater current being drawn from the tubes. There are an additional 4 #12 lamps on this winding and an FN cue indicator which I cannot find, but the lamps count for an additional .600mA since they are .15A and run in parallel. So, 3.69A in that winding.

The other had 4 X 7355 tubes for 3.2A and another pair of 12AX7s for an additional .6A, so this winding has 3.8A on it.

Ah, good info there...

The pwr xfmr has two heater windings? Each is about 3.69A to 3.9A? You could use one heater winding per channel then.

Since you'd have two EL34's per channel, those would take 3A of heater current right there. That would leave you less than an amp of heater current for the driver stage. Still do-able. Just have to be careful which tubes you choose. 12AU7 only use 300mA each. Two of those per channel would fit.

Are both those heater windings 6.3VCT? Or is one 12.6V?


If you look at the EL34 data, it looks like in pentode in AB1, which is what I will be running, the EL34 makes about 35W at 315v into 3,400 ohms. Run in triode into 3,600 ohms at 290-300V, I would guess I could get around 15-17W.

If you're getting those figures from the tube manual, please note that those numbers are usually *very* optimistic. With the low B+ you have available, I'd expect no more than about 25W per channel for pentode and maybe 10W per channel for triode. But I could be proven wrong...

We also still don't know the current capacity of your pwr xfmr. What current were those EL34's drawing where it says class AB1, B+ = 315V, Loadp-p = 3k4 ohms? Like 140mA for the two EL34's? (Hopefully less...)

I can add a small power transformer for the heaters on the drivers probably even though I am not too exited about the drilling part since the chassis looks so nice in my opinion.

This was my original proposition for these output transformers, but it was a "show stopper" when I did not have the schematic and saw the 7355 to run on .8A current:(

I know the EL34 is fairly easy to drive and I really like the PI with a CCS as a driver. So, is there a 9-pin pentode that will make a good CCS for a LTPI circuit?

Thanks guys!

Blair

For small 9-pin pentodes, there is 5879, which isn't too expensive. Maybe 6688? Russian 6J9P? I'd say EF86, but those are too expensive for what you get.

*In general*...

If you wire your EL34's as pentodes, you will need to use NFB around the amp, and you will need more gain from your driver stage to make up for the lost gain due to that NFB.

If you wire your EL34's as triodes, you can use much less (or maybe no) NFB around the amp, and you will need less gain from your driver stage.

--
 
Last edited:
Thanks!

OK,

Back to something you mentioned before which sounded kind of cool. What about using the 5879 in triode to drive the LTP?

Working from the original schematic, I can "skirt" the additional filament transformer by simulating what Bogen did. Apparently, he put a bit of additional current on his bias winding. I can only assume it is a complete separate winding. It looks from the voltage chart, and schematic like he wired a few 12AU7s in series and ran them off the bias tap. The 5879 is 150mA at 6.3v, so two would be 12.6v at 150mA. Then two 12AU or AT7s for PI and I could run this string off the bias tap?

Just an idea.

Thanks!

Blair
 
Thanks!

OK,

Back to something you mentioned before which sounded kind of cool. What about using the 5879 in triode to drive the LTP?

Yes, you could do that. I think the mu of a 5879 triode is around the same as for a 12AU7 or 6CG7, which is about 20. The rp is about the same too, but I don't know that for sure. Figure on 10k ohms. I don't know what they sound like triode-wired, as I've never used one that way. Heck, you might even want to use it in pentode (high gain).


Working from the original schematic, I can "skirt" the additional filament transformer by simulating what Bogen did. Apparently, he put a bit of additional current on his bias winding. I can only assume it is a complete separate winding. It looks from the voltage chart, and schematic like he wired a few 12AU7s in series and ran them off the bias tap. The 5879 is 150mA at 6.3v, so two would be 12.6v at 150mA. Then two 12AU or AT7s for PI and I could run this string off the bias tap?

Just an idea.

Thanks!

Blair

You're welcome! I'm just having fun, dreaming up solutions.

In the receiver, the filaments were wired in *series*? If so, what is the voltage of the filament winding? One needs to be careful about these things...
I'd need to see the schematic or the wiring to understand the filament wiring you're describing. (Sorry.)

--
 

Yes, those are all Poinz' Musical Machine. All use EL34's in triode, so no NFB used. If the EL34's were used in pentode mode, then you'd really want NFB around the amp to lower the output resistance to get reasonable damping factor. The other thing to notice is that the B+ to the EL34's is 380V in that version. That's a lot higher than what Blair has available in the power transformer from the Bogen receiver.

I think something like that amp would be very cool. But push-pull EL34's in triode with only 300V B+ would not make very much power. 10 watts per channel at the most, I think. Maybe a fair amount less, especially if we bias it in class A (so we don't need negative feedback).

:)

--
 
Hi Randy! Thank you!

Rongon, here you go. It is only three valves fed off the bias winding. Strange, I know!


The way I see it, there are two separate heater windings, with a bunch of tubes' heaters in parallel. One of the windings has a center tap, the other (the "longer" one) does not. I can't read on the schematic what the voltages of those windings are. I assume they're both 6.3V -- but it would be best to know.

For simplicity's sake, it's probably better to ignore (as in remove) that series heater string. With only 8 tubes to heat in the entire amp, I don't think you need that extra stuff.

BTW, you have two HV windings on that transformer. The one with the full-wave diode bridge is used for the main B+ supply. The other one, with the center tap, could be used for a B- supply to feed your pentode current sinks.

I can't read the printing on the schematic, so I can't tell what the voltages are supposed to be. Any chance you could post a more detailed picture of the power supply section (only)?

This is getting really interesting. I have an old h-k receiver that used 7355's, sitting in a box in a closet. It has the same OPT's as the old h-k A500 integrated amp. I nearly sold it, but still have it. I might have to build something with it now....

--
 
I have:

-51v (unloaded)

+332v (unloaded)

Yes the CT and non CT windings are both 6.3v

It is all wired up and ready for a schematic:) that's how I know my voltages. I could always go "end to end" on the negative supply since it is center tapped and FWB for -100v or so and put a pot on it to adjust it down.

Thanks!

Blair
 
Hi Ron,

The bias winding is 75vct.

Here is something I saw in another post that I failed to answer. The B+ current.

Here is something pulled from the 7355 data sheet:

@300v B+
-21vdc
100mA (2 tubes)
185mA (2 tubes max)
4k plate load
2% dist.
28.5w

Bogen's rp40a: (data I know)

275v B+
-16v
3.6k plate load
22W

I would assume this is probably around 80mA idle per pair.

So, 160mA (conservative) plus an additional 19 tubes worth of additional current. So, at even 3 mA each, I'd say we can safely add another 60mA or so.

Roughly 220mA.

It should drive a quad of whichever tube I choose. I would still like to try to use the 6l6. The specs I saw for the EL34 run at 325v are odd and would not work for me because it looks like they run the grids at 350v with the plates at 325v.

One idea to get a little more voltage out of this transformer would be to add the bias winding to the B+ winding to add the 30-50v to it. That should land me in the 330v range, but I would have to completely start over with the PS caps.

Blair
 
Hi Ron,
The bias winding is 75vct.

Well, 75 x 1.3 = 97.5, so that's our starting point for getting a C- supply for the CCS's.

Here is something I saw in another post that I failed to answer. The B+ current.

Here is something pulled from the 7355 data sheet:

@300v B+
-21vdc
100mA (2 tubes)
185mA (2 tubes max)
4k plate load
2% dist.
28.5w

Bogen's rp40a: (data I know)

275v B+
-16v
3.6k plate load
22W

I would assume this is probably around 80mA idle per pair.

So, 160mA (conservative) plus an additional 19 tubes worth of additional current. So, at even 3 mA each, I'd say we can safely add another 60mA or so.

Roughly 220mA.

Well, that's not a lot of current to work with, but should be OK. If the B+ only winds up at +275V, and each 6L6GA (or equivalent) is run at 50mA plate current, that's only 13.75 watts dissipation per tube -- and that's for a tube that can dissipate up to 20 watts. Seems a pity to waste that kind of capacity.

Do you have a list of what tube types were used in the receiver? Usually they're populated with RF pentodes or dual triodes like 6AQ8, 6AJ5, etc. Those often take more than 3mA each tube. For instance, a single 6AQ8 is likely to be take 6mA or more. It could be that the total capacity is more like 250mA.


It should drive a quad of whichever tube I choose. I would still like to try to use the 6l6. The specs I saw for the EL34 run at 325v are odd and would not work for me because it looks like they run the grids at 350v with the plates at 325v.

At this point, I think you might as well use the Sovtek cheapie 6L6's, because it would be sort of a waste to spend all the money on EL34's and not be able to get anywhere near to using their full capacity (24 watts plate dissipation).

One idea to get a little more voltage out of this transformer would be to add the bias winding to the B+ winding to add the 30-50v to it. That should land me in the 330v range, but I would have to completely start over with the PS caps.

Blair

Then what would you use for the negative supplies? A second transformer? But yeah, I can see the logic behind getting more volts out.

- Ron

PS -

From the 6L6GA data sheet:

Class A1 Push-Pull (values for two tubes):
Va = 270V
Vsg = 270V
Vcg = -17.5V
zero signal plate current = 134mA
zero signal screen current = 11mA
Load R plate-plate = 5k
THD = 2%
Max signal power out = 17.5W

They're running the 6L6's at higher current (72.5mA per tube), which is more than you have in your power transformer, I believe. But you could scale it down and accept the reduced power output in class A - and possibly build a driver that can hack sending the 6L6's into class AB1.

This power transformer keeps imposing limitations. Are you sure it's only good for c. 220mA? No way to make a list of the tubes used in the receiver and count up the likely current drawn by the original circuit? Those things can surprise you sometimes. You mentioned that it had 19 small signal tubes? I wonder how many of those were 6CG7, 12AU7, 6AQ8/6DT8, 6AC7, 6AJ5, 6U8A, stuff like that?

Another possibility would be to find something like a small 600VCT @ 50mA pwr xfmr and power only the driver tubes with that, leaving the original pwr xfmr for the 6L6's. That could give you a +400V supply for the drivers, giving lots of headroom, lower noise, etc. But only if there's room for it, of course.

With a low B+ like 300V feeding the driver stage, you'll need to use tubes with low rp that can work with low plate voltages, like 6DJ8 and 5687. 12AT7 and 6CG7 might not work so well with such low voltages.

--
 
Hi Ron,

I suspect my B+ to be 10-20V higher. I was just being conservative on the small tube current to be safe.

4 X 12AU7
1 X 12AT7
6 X 12AX7
3 X 6AU6
2 X 6BA6
1 X 6AB4
2 X EM84
4 X 7355

23 tubes :)

If you average 5mA per tube, it is closer to 250mA if the power tubes are biased at 80mA. I think it might be even a little higher. It is the funkiest transformer I have used because it was "dipped" / impregnated with this black rubbery substance. I'm sure it is probably asbestose, and the fact that I sanded it down will kill me. LOL!!

All joking aside, there are plenty of ways to get bias voltage off the B+, or tapping the filament winding that has reserve current and using a small 12v transformer that will tuck away for my -60-75v.

The nigh question will be whether or not the bias winding has the current capacity, and whether or not that 30V is really going to make the "night and day" difference.

I liked you 6DJ8 PI sketch. That tube works well with low voltage.

Supposing we use that and/or the 12BY7A as suggested fro
The HK Citation V, it should work well.

I could also use a SRPP driver and use a 250-280v B+. even the 12AU7 with 220V will give me 16v or so conservatively biased at that B+.

Blair