OT for PP EL34 20w

[TIMBO]

Hi guys, Got another head scratcher
Its call a REX BASSKING and it would be about a 20w el84 PP amp model BA200A
But the one I have is a BA250 with a different valve layout and of course there is no info on it.
It has a pair of 6CZ5 and the replacement valve sockets are fitted to what would be a 8 pin socket hole.
It is hard to say IF it did have 8 pin socket that may have had a pair of 6CA7s.
Similar to this
http://el34world.com/Forum/index.php?topic=18639.msg190688#msg190688
The amp uses a voltage doubler and a fixed bias supply.
The voltage doubler output is around 400v.
After some testing it seems that the OT is stuffed.
So I'm thinking of rewiring the power valves to take 6CA7s and being a combo amp the speaker is 30w I'm wanting to have the output of about 20w.
Thinking of an OT like this
https://www.hammfg.com/files/parts/pdf/1760H.pdf

[darkbluemurder]

A pair of 6CZ5 tubes calls for 450ma heater current per tube - no way that the original power transformer would accept the 6CA7's 1.5A heater current per tube.


If you have octal sockets why not use the 6V6?


Cheers Stephan

[92Volts]

You can't really take advantage of EL34s/6CA7s without switching the output transformer to a lower impedance... and the heater current is a problem like darkbluemurder mentions. So I wouldn't do that. It is also true that 6V6 would be safe for sure, since they draw as little heater current as the 6CZ5 and less than any common option I know of.

If you're sure it had 8-pin sockets originally, I'm guessing it had 5881s or 6L6s. 6V6s would be weird if the same company made an EL84 version, since all the hassle to change socket types results in little change, the EL84 and 6V6 are very similar to each other.

5881s draw 0.9a of heater current. That would be enough for EL84s or of course 5881s, but not EL34/6CA7.

You'd want to check the actual pinout/wiring but EL84s seem really similar to 6CZ5s and even use the 9-pin socket. It's probably safe to use those.

You'll want to be careful, obviously, but it doesn't appear 6CZ5s were the original tubes so it might not be limited to their low 450ma heater current.

[DummyLoad]

7591 will make about 35W with 6k6 load and 400V B+ in fixed AB1. heater draw will be 1.6A for two tubes. should work well for you. B- will need to be around -20V

--pete

[TIMBO]

Thanks Pete, The power transformers in a lot of these local built amps were universal and some photo's of 50/60w amps look to have a similar transformer.
I have a Hammond 1560f OT on hand 7.6k primary.

Is there a math that is used to calculate your response, or just experience?

I found this (Plate voltage x plate voltage)divided by(tube wattage x 2)

So if I (400 x 400)/(25 x25) using EL34 @25w per valve.

160,000/50=3200 primary

So if I reduce the EL34 wattage to 10w per valve

160,000/20 = 8000 primary ?????

[PRR]

> Is there a math

That's hard work.

7591 data tells all.

http://www.mif.pg.gda.pl/homepages/frank/vs.html

[92Volts]

Power output is roughly limited by B+ voltage divided by OT impedance... not determined by the tubes... assuming the tubes can drive that impedance fully/safely.

So if this amp put out 20w originally, you can keep the same OT and "upgrade" to EL34s, and you'll still get 20w.

Another way to think about this is, if 8k is the load that would produce 20w in this amp, we can assume 8k is the load of the OT that's already in there, and just use that one.

[TIMBO]

Thanks 92Volts, It's looking like the original OT is cactus, although I'm not exactly sure that the OT IS original for this amp.

 

[jjasilli]

Timbo, I think your design concept to mod this amp is a good one.  I like the tone of a small bottle amp modded to big bottle.  It makes tone robust.  One issue with that mod, as stated, is heater current.  That can be solved by adding a supplemental filament tranny.


Then there's the OT issues: current handling, & impedance matching.  Since your OT is dead, you need to make choices as to the specs of a replacement.  This allows wiggle room for creativity in design.  Current: you can calculate what current the power tubes will draw through the OT.  You can spec the EDIT PT  OT to that, or purposely under-spec by up to 50% if a quality OT is used.  The purpose of under-spec-ing is to cause saturation.  IMHO saturation is the key reason to put big bottle tubes in small bottle amp.  The OT will run hot. In the unlikely event that it's toooooo hot such that lamintations will fail, remove the bell covers, or cool it with a fan. 


Impedance matching:  you can choose to match impedance, or to have it off by up to 2:1 in either direction.  Doubling the OT's primary impedance will cut output power which seems to be what you want.

[92Volts]

Hammond transformers are conservatively rated. With the 1650 series, their ratings are explicitly for hi-fi performance (which you don't need), so a 25w OT is plenty for this.

However, 7.6k should get you a bit more power than 20w with EL34s at 400v, and a true 20w clean amp is already enough to blow a 30w speaker when overdriven. Have you tested be speaker in this combo? Getting one rated for more power is one option, at the very least if you want to keep this one you should be sure it works and sounds OK.

[TIMBO]

Thanks guys for the input, as this stuff is way over my pay grade.
The speaker..

The OT is DEAD.
The PT I believe will handle the extra heater load, even a pair of EL34.
I have on hand a pair of 6CA7 and 7591.
Would much prefer to use the 6CA7 as these were used in most amps of the day.
I have a Hammond 1650f.
Can cathode bias if required.
Unloaded power supply 420v.

[tubeswell]

FWIW the Rola/Plessey 12UEG is rated for 30W peak* according to this attached excerpt from the old Rola Loudspeaker Specification sheet (Published 1 Feb 1965).


*i.e. 21W RMS (Edit - although I would guess its good for 25W RMS)

[PRR]

"Peak" Watts (an indefinite rating) does not convert to "RMS" Watts the same way that Sine Voltage peak/RMS converts.

From text, I would tend to a "15W" amp (which can make 30W gross distortion).

OTOH, 1.75" coil and 13,000 Gauss in the gap is a BIG speaker, comparable to older E-V speakers (grandfathers to the mighty E-V 12L, though with modern adhesives instead of shellac). It might take a lot of 30W beating.

[jjasilli]

The stock tone circuit & vol pot is a mystery to me.  Signal is split at the input, before the 1st gain stage.  The 1st gain stage boosts all frequencies, but then filters-out bass through a rather small blocking cap > treble bleed tone control.  Meanwhile all signal also bypass the 1st gain stage > a complex bass filter. Signal filtering w/o gain would seem to cause undue signal loss at the input.  The output of the bass filter circuit is ostensibly out-of-phase with the output of the 1st gain stage, unless the caps in the bass filter rotate phase.  The combined signals are then drained through a small 50ka vol pot. Again, the insertion loss seems enormous.


Are you sure the schematic is correct?

[shooter]

Are you sure the schematic is correct?

I'm waiting like a giddy kid for Timbos evaluation of the "2channels", especially the tone adjusting features of each.

[TIMBO]

Hi JJ,The schematic is correct, the cap values may not be correct.
The three cap in the circuit may not be correct as they are on a small eyelet board and it was painted black.

and the values aren't clear,so I'm going by a MM reading.
The 50k pot is not original.


I've started on the power supply end, of the restoration.

I could not get a 100uf/200v cap can that would fit the 32mm clamps, so a piece of 2.5mm G10 board and four eyelets to take the snap-in type caps.


A couple of dead cap cans gutted.


DONE

[sluckey]

Nice cap camouflage. What's actually holding the JJ caps?

[2deaf]

V1b has local negative feedback that will attenuate the higher frequencies way more than the lower frequencies when the Bass control is maxed.  Phase shifts are a non-issue for this setup.  The output impedance of V1b is dramatically lower with local negative feedback than it would be without it.  This means that the voltage divider formed by the output impedance and the 50K pot won't drop the signal near as much as it would if there were no local negative feedback.  Any frequencies that are attenuated by the .047uF blocking capacitor will be fixed by the NFB.

The 2Ma Bass control looks like it will merely act as a volume control for most of its rotation with the bass control only occurring when the pot is at or near max.  It also looks as though the volume will be at its highest around mid-rotation on the Bass pot.  A 1M Linear pot seems like a better choice for the Bass control.   

[shooter]

What's actually holding the JJ caps?

a piece of 2.5mm G10 board and four eyelets

luv the bias diode!

[PRR]

> all signal also bypass the 1st gain stage > a complex bass filter.

Agree with 2deaf. This is NFB. The R-C mess re-drawn *appears* to be a mis-used twin-Tee filter. I assume it notches bass out of the NFB, thus causing a bass boost.

[TIMBO]

Thanks PRR, I'll check out the circuit again, it's a bit hard to get at.

[TIMBO]

PRR, You are correct again.

[jjasilli]

V1b has local negative feedback that will attenuate the higher frequencies way more than the lower frequencies when the Bass control is maxed.  Phase shifts are a non-issue for this setup.  The output impedance of V1b is dramatically lower with local negative feedback than it would be without it.  This means that the voltage divider formed by the output impedance and the 50K pot won't drop the signal near as much as it would if there were no local negative feedback.  Any frequencies that are attenuated by the .047uF blocking capacitor will be fixed by the NFB.

The 2Ma Bass control looks like it will merely act as a volume control for most of its rotation with the bass control only occurring when the pot is at or near max.  It also looks as though the volume will be at its highest around mid-rotation on the Bass pot.  A 1M Linear pot seems like a better choice for the Bass control.

 

[TIMBO]

Hi guys, Neatened up some of the components and added some new ones.
Fired up well , no issues.
Checked the heater taps, all good for 3.4v on both sides.
Set the - volts at -33v.
Plate voltage 366v and cathode current 40ma.

How do I calculate output wattage.



Also, I have found on another amp a 2.2k resistor between pin8 and the junction of the 470k and 47k of V2b.
This amp looks like this resistor has been deleted, and the connection of the .047 coupling cap is original.

Is this resistor required. Thanks

[jjasilli]

Power Formula: Watts = Voltage² / Resistance.  For Resistance use the nominal impedance of the OT's primary winding.  The Power Formula works best for PP amps; SE amps are less efficient. 


E.g., 400V; 5000ῼ_p-p primary impedance.  400² / 5000 = 160,000 / 5000 = 32W.  That's why a higher primary impedance cuts output.  400V; 10,000ῼp-p primary impedance = 16W.

[sluckey]

Also, I have found on another amp a 2.2k resistor between pin8 and the junction of the 470k and 47k of V2b.
This amp looks like this resistor has been deleted, and the connection of the .047 coupling cap is original.

Is this resistor required. Thanks

That resistor is required. It is the bias resistor. Value is not critical since the circuit is self adjusting. Anything between 1K and 2.2k should be fine, with 1.5k being very common. The output can be taken from either end of the bias resistor. Look at any bootstrap biased cathodyne phase inverter such as fender 5E3 or Princeton Reverb for examples.

Here's some easy reading about that resistor. Scroll down to the biasing section...

     http://www.valvewizard.co.uk/cathodyne.html

[TIMBO]

Thanks guys, OT is 6.6k P-P.
2.2k is on an all original amp so i'll go with that.
Thanks

[TIMBO]

Hi guys, This one is completed.




The old cab has scrubbed up well.



After some poor guitar playing, I think this amp sounds great, but I think they all do.
The two channels are as they are marked.
The bass channel would do a bass guitar justice, with the guitar plugged in there is plenty of dark and light.
The bass is tight and define on the lower strings and smooth on the higher strings.
The normal channel is typical Bassman sound with the bright cap on the volume adds some crispiness.

So if I apply JJ formula 360x360/6600=19.6w
 

[DummyLoad]

nice work tim! 


--pete

[TIMBO]

Thanks mate.

[chocopower]

I'm wrong, or there is a missing resistor in the concertina P.I.?
I mean the bias resistor.

[shooter]

Is that an SMD RF connector for the speaker?

[sluckey]

Is that an SMD RF connector for the speaker?

No. Look at the original post.

[TIMBO]

chocopower, Schematic fixed.
shooter, It's just a fancy right angle Jack.