Converting Dulci EL84 mono amp to stereo

[Firebottle]

Macca has very kindly donated a (slightly knackered) Dulci mono EL84 push pull amplifier to me.

http://theartofsound.net/forum/showthread.php?3812-A-couple-of-bits-and-bobs

As I can't magically produce twice the power from the psu to make it into a stereo push pull amp, do you good folks on here think it is worth while rebuilding it as a single ended EL84 stereo power amp?

:scratch: Alan

[Ali Tait]

I had an SEP EL84 amp a while back, sounded very good, so well worth doing I think. Limited power of course.

[Firebottle]

After deciding to go ahead with this project it was time to start some testing.

I hope this thread will be informative and educational but will start with the AoS caution on undertaking such projects:

'We strongly advise extreme caution when considering any modification or building projects described, especially those concerning any mains voltage related upgrades, earth modifications and/or any other high voltage related modifications or building projects.

Unless you are suitably qualified or otherwise capable of carrying out such modifications/suggestions, then we strongly advise you consult qualified personnel.'

The first thing to do when examining an old amplifier is to make extensive use of the mark1 eyeball, have a very good look around to see if there are any obvious signs of distress to any components or wiring. One of the valves was physically broken, also it was patently obvious that previous work had been done; there was an extra smoothing capacitor attached across the original main smoother, the mains lead was a 3 core lead whereas when the amp was built (1959) I imagine the original was just a 2 core lead, plus one of the coupling capacitors had been replaced.

To confirm it was worthwhile to proceed, the mains transformer needed to be tested to see if it was OK. I checked for any shorts on the mains input, the primary resistance was about 12 ohms so that was good. To enable the mains supply to the transformer I had to bridge the pins on the pre-amplifier power socket, that normally run to the on/off switch, so just for good measure I fitted a small surge suppressing NTC resistor:

http://i601.photobucket.com/albums/tt97/Paramotorpilot/Dulci/P1040667.jpg

I removed the rectifier valve, applied power and measured the secondary HT voltages, reading 350-0-350V. Sounds a bit high but it was an open circuit reading, i.e. no load.

Thing are looking up. I have now ordered a couple of single ended vintage output transformers.
There were no signs of failure to internal components so as I have test equipment to measure capacitors I tested the HT and output stage coupling caps.
Caution: Before taking a step like this make sure that all HT capacitors are discharged, or have zero volts on them.

Even though they were so old the HT caps measured within tolerance of the printed values.
Side note: I only know the date of the amp due to a date on one of the HT caps outer can. It's a good place to look.

The first fault was the rectifier valve:

http://i601.photobucket.com/albums/tt97/Paramotorpilot/Dulci/P1040664.jpg

Tube number 1. Bit of a shame as I would have liked to keep it to valve rectification. I searched in my limited spares stock and found a couple of silicon diodes, so I fitted those along with some extra 68 ohm resistors, the only 2W value I had.
It's always advisable to fit extra resistance when replacing valve with solid state rectifiers.

http://i601.photobucket.com/albums/tt97/Paramotorpilot/Dulci/P1040666.jpg

Now comes an important point. When powering up valve equipment that has been dormant for a long time it is a very bad idea to just plug in and switch on. Something is likely to go bang or phutt. Some electrolytic capacitors need to be 'reformed', that is re-acquainted with electricity in a gentle way.
The best approach is to use a variac and increase the mains voltage slowly from zero, looking and listening for any signs of distress. Monitoring the HT voltage to see if there are any sudden dips is highly recommended, along with looking for more obvious signs such as smoke.
As I didn't have a variac to hand the next best approach is to wire in a series 60W light bulb to act as a protective ballast:

http://i601.photobucket.com/albums/tt97/Paramotorpilot/Dulci/P1040668.jpg

OK, power on, HT is holding steady then dropping slowly as the valves heat up, so far so good.

Next comes the most important check for any valve power amplifier. If the coupling capacitor feeding the grid of the output valve(s) is leaky (not blocking DC perfectly), the bias for the valve(s) is upset and in the worst case can destroy the valve and even the output transformer.

If you have no means of measuring the bias it is highly advisable to just replace this capacitor for a new one. I attached the meter to the grid of the valve fed from the capacitor that had been replaced and got a reading of 55mV, that's 0.055V. Perfectly OK:

http://i601.photobucket.com/albums/tt97/Paramotorpilot/Dulci/P1040670.jpg

Next I measured on the other output valve, fed from the original coupling capacitor:

http://i601.photobucket.com/albums/tt97/Paramotorpilot/Dulci/P1040669.jpg

Sh*t! Nearly 13V, now that's not good. Good thing I'm using the 60W lamp as ballast or current flow could have been very excessive.
I replaced said capacitor and all was well with the bias readings. A quick check around the driver and input valve circuits and voltages were as expected, although of course low due to the 60W lamp ballast.
I fitted a temporary couple of sockets so I could connect a dummy load or a speaker, then removed the 60W lamp from circuit.

http://i601.photobucket.com/albums/tt97/Paramotorpilot/Dulci/P1040665.jpg

I powered the amp up again whilst carefully monitoring the HT voltage. The HT capacitors are marked as rated at 350V, some older capacitors are also marked with a higher 'surge' rating.
The HT went fairly quickly to almost 390V, then reduced slowly to about 350V as the valves warmed up. After 10 seconds the HT started to slowly fall and a little wisp of smoke could be seen coming from area of the output valve cathode resistors.
I quickly measured the voltage across each resistor and found about 8V on one and 30V on the other. I powered down and made a calculation to get the expected cathode voltage for an EL84, this is 13V. Something else was amiss, so I tried just one output valve at a time and hey presto, Tube number 2 – shot.


To be continued when modification proper starts.
:cool: Alan

[brian2957]

Nice one Alan . I'll never be competent enough to build a valve amp and I don't particularly understand a lot of what you've written . However , I find threads like this interesting and enjoyable when someone goes into such great detail when describing the work they're doing , so well done . Have a good one sir :santa:

[Macca]

I was given this amp for free when I bought a pair of Lowthers off a bloke who was moving down south. That was 1993 and it has sat under my bed ever since. Great to see it being turned into a useable component again.

[Firebottle]

http://i601.photobucket.com/albums/tt97/Paramotorpilot/Dulci/P1040721.jpg

I'd forgotten how difficult it is to strip components from an old valve chassis that's been built properly!

:cool: Alan

[Firebottle]

Next thing is to decide on the input/driver stage circuit, something that will give good performance. I decided on a constant current loaded triode which turns out to be very simple:

http://i601.photobucket.com/albums/tt97/Paramotorpilot/Dulci/Valve%20driver%20stage.jpg

Note the inclusion of the 'grid stopper' resistor (1K) at the input. Always advisable to fit when using valves originally intended for high frequency use. The coupling capacitor can be a reasonably small value as the load on the valve, the grid circuit of the output valve, is high, also giving good performance from the driver.
The voltage bias on the heaters is important so that the heater/cathode voltage rating of the valve is not exceeded. This is usually about 90V maximum, so one quarter of the HT voltage is ideal in this type of circuit.

http://i601.photobucket.com/albums/tt97/Paramotorpilot/Dulci/P1040723.jpg

The underside now showing the valve biasing resistors in temporary situ on the valve bases. The voltage dropping resistors near the transformer are also a temporary fitment to reduce the HT to 250V for testing.

The top of the chassis with driver valves, you can also see the vintage output transformers behind, to be fitted later.

http://i601.photobucket.com/albums/tt97/Paramotorpilot/Dulci/P1040722.jpg

Next to do are the output stages.

:cool: Alan

[struth]

http://i601.photobucket.com/albums/tt97/Paramotorpilot/Dulci/P1040721.jpg

I'd forgotten how difficult it is to strip components from an old valve chassis that's been built properly!

:cool: Alan

That shot takes me back. brick outhouse comes to mind...love the spring valve securers too. they were common in the day

[Firebottle]

It's now in a working state, I won't say finished as I'm sure some tweaks can be applied once the 'scope is put into action.
However it is in service in the lounge driving the '57s and I've got to say I'm impressed.

http://i601.photobucket.com/albums/tt97/Paramotorpilot/Dulci/P1040727.jpg

It's the first class A amplifier I have had, only 4 valves and I'm guessing about 4W max, I've yet to measure it. For some low price 'vintage' output trannies that I sourced on the interweb the bass is very acceptable. Not quite the slam of the OTL but a good extension to it. The spec of the Polish transformers that I did find said -3dB at 80Hz.
Shot from above:

http://i601.photobucket.com/albums/tt97/Paramotorpilot/Dulci/P1040728.jpg

I have used parts that were 'to hand', so new smoothing caps of assorted values plus a smoothing choke of 5H.
I did get it running first off (one channel) with fixed bias, with the target of no electrolytics in the signal path, but I decided that it was simpler and safer to use the normal bypassed cathode bias.

More details to follow.

:cool: Alan

[Ali Tait]

Well done Alan, EL84 is a nice valve. I have a couple of quads of NOS Tungsram tucked away.

[walpurgis]

Cool bit of kit (well, hot probably :))

[Firebottle]

The more I listen to this the more I like it :carrot:
The sound has an easy quality to it, nothing forced, you can just immerse yourself into the music.

The rear panel, no selection for different speaker impedance as there are no tappings on the output transformer, but that doesn't seem to matter:

http://i601.photobucket.com/albums/tt97/Paramotorpilot/Dulci/P1040733.jpg

I have now had the scope on the output and my nice little class A amp is giving 5.5W into 4 or 8 ohm loads, wowee :wowzer:
There is a smidge of ringing on a square wave but I can live with that. The inexpensive output trannies are Polish in origin but I can't find anymore listed on fleabay at the moment, here is the reference:

http://i601.photobucket.com/albums/tt97/Paramotorpilot/Dulci/P1040731.jpg

..............plus of course the obligatory internals photo:

http://i601.photobucket.com/albums/tt97/Paramotorpilot/Dulci/P1040730.jpg

:cool: Alan

[Gordon Steadman]

I look forward to hearing it on Wednesday:cool:

Always assuming we have time in between the dozens of bits you are going to look at / service / fix for me of course:eyebrows:

[Gordon Steadman]

Well, here we are - it's wednesday, which means I have now listened to Alan's little Ugly Duckling. More of which in a mo.

We started out early and, being winter and us heading northish, the sun was either in my face or flashing through the trees on my right - ALL THE DAMNED WAY:steam:

Surprisingly, on the way home, the sun was.............

Anyway, as usual, we took all the sick and dying bits of electrickery for Alan to wave his magic soldering iron over. I seem to have more bits now than I started out with but whatever, they are all working and probably better than they started out. Food was great and so was the company.

So... the ugly duckling. I think Alan would admit that it's not the prettiest thing in the lounge but then he had what he had to work on. The sound however, was more toward the swan part of the spectrum. In fact, this little lump lost absolutely nothing to the OTLs at all and in some ways, I think it betters them. In purely obvious ways, there is probably not much difference in the sound they make but there is something about this one that just grabs you by the musical throat and insists that you listen very carefully to what it has to say. We listened to Django and everything about the guitar was just right and the drum kit sounded terrific. This with at least 4 watts per channel into the Quads which demand a reasonably robust amplifier. Saxophones had a wonderful rasp, in fact everything was great.

So, my friend, congratulations. I just wish that I didn't have to nod my head knowingly when you try to explain stuff to me whilst understanding not a word you say:lol:

[Ali Tait]

EL84 is a nice little valve. :-)

[The Barbarian]

Can't go wrong with 'EL84' amps can't say ive ever heard a bad un loike

[Ali Tait]

Aye, think you'd have to try to muck one up.

[The Barbarian]

I call it the 'safe bet tube'

[Firebottle]

So, my friend, congratulations. I just wish that I didn't have to nod my head knowingly when you try to explain stuff to me whilst understanding not a word you say:lol:

It was great to have the two of you up again:D. Thanks for the comments on the ugly duckling, maybe I will put a wooden sleeve around it as you suggested.

.....and there's me thinking you were taking it all in, I was only explaining about the Q factor with capacitors, plus negative feedback with the amp:scratch:

:cool: Alan

[Light Dependant Resistor]

Hi Alan
Some nice info on the Dulci amp

A question arising, as you run these amps with ESL57, do you
connect loudspeakers in phase or out of phase, given the ESL inverts ?. Normally with ESL57
a inverting amplifier like the 303 is connected red to red, black to black, but with say a 306 which
is non inverting you would connect amp red to speaker black. Or more simply do you know if the
Dulci is inverting or non inverting ? Is there a schematic available ?

Cheers / Chris

[Firebottle]

Hi Chris,

I don't worry at all about absolute phase, it's never bothered me.

The only original items in the Dulci rebuild are the mains transformer, valve bases, output cathode resistors plus a few more resistors.
The circuit is a simple as I could make it with excellent performance:

http://i601.photobucket.com/albums/tt97/Paramotorpilot/EL84%20SE.jpg

The input valve can be an ECC88, the 3x 33uF output cathode decouplers were only used because I have loads of them :mental:

Hope you find it interesting,

:)