• 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.

EL34 PP output tranny questions

I want to build an EL34 PP guitar amp with parts I have on hand and decided to strip a Music Angel hifi amp (stereo EL34 PP amp) for the power and output trannies. I measured the output trannies and they have a reflected load impedance of 14,700 ohms for the 4 ohm tap and 11,900 for the 8 ohm tap. What is this? This was way higher than I expected, even allowing for the Chinese fudge factor. I removed the tranny cover and tranny and was surprised by it's very large size and weight.

I guess the only way to use this output tranny is to use a 4 ohm load on the 8 ohm tap for a reflected load impedance of around 6000 ohms. The Mullard datasheet mostly gives examples of 3.5K a-a, and one example of 6K a-a with cathode bias. I was hoping to do fixed bias, but cathode bias would be OK if necessary. The B+ on this amp, if it goes ahead, will be around 440V at the first filter cap. The other alternative is to use an output tranny from a Ming Da MC34B (6L6 PP), it has a reflected impedance of around 3900 ohms for the 4 ohm tap and 5300 ohms for the 8 ohm tap, but it is around half the size of the Music Angel monster. For a sanity check I measured the reflected impedance of an 8K Hashimoto output transformer and got 7K2 for the 4 ohm tap and 7K5 for the 8 ohm tap, so I'm sort of in the ballpark with the measurements. The amp will be for a friend and will amplify an acoustic guitar, so needs to be reasonably clean and power output around 25W is all that is needed. I believe he has a 4 ohm speaker box but need to check up on that.

So I guess I'm a bit confused. Any tips, suggestions, comments, or hare-brained schemes welcome.
Are there any other tubes that would work well with the Music Angel OT's high primary impedance?

pic of tranny

tranny schematic
 
Wild guess: the Chinese factory accidentally wound the secondary with the same wire size as the primary. With a few assumptions and simultaneous equations you can show that they intended to produce a 5200 ohm impedance, or thereabouts - this is quite plausible for a pair of EL34. Unfortunately they ended up with a secondary with wire resistance of 7.44R (4R tap) and 10.5R (8R tap). Was the original amp a bit quiet?

To test my theory, measure the DC resistance. I have given the secondaries above. The primary will be around 267R, which is perfectly respectable.
 
nightanole, I did do that, in the schematic attached to the first post.

I measured the DC resistances, there is 90 ohms a-a on the primary, and the four secondary windings are all 0.1 to 0.2 ohm. One of each of the parallel pair of windings on the secondary is stranded wire, unusual.

I haven't listened to the Music Angel amp for a couple of years, I don't remember it to be low in power but didn't listen to it more than a week or so. Strange primary impedances, amazing what you can get away with, I guess it *must* have been low on power, around 27W from the loadlines I drew up this morning. It would have had around 430V on the plates and similar on the screens.

I guess the 8 ohm tap would work for 4 and 8 ohm loads since 25W is all that is required, just going to get more power with 4 ohms. The loadline for 12K (8 ohm load) passes way below the knee of the -5V grid curve, even with G2 at 250V, I don't know what the ramifications are, all I have read is that it is "bad practise". What does actually happen in this situation?

Darryl, the Trainwreck looks interesting, "right on the verge of instability" sounds like my cup of tea for a guitar amp! 6K6 certainly looks like a decent safe loadline for class AB1 compared to the 4K that the Marshall's use.
 

45

Member
2008-12-18 2:29 am
UK
IAN444 you don't need a big output transformer for a guitar amp because you don't need to go below 80Hz. Any hifi OPT rated for 25W (usually at 50Hz) will be good even for 50W output at 80 Hz.
4-5K max is the way to go for me so use the Ming Da MC34B transformer with a 8 ohm driver.
Pentode connection without feedback, of course.

45
 

45

Member
2008-12-18 2:29 am
UK
With 4K a-a, and 360V B+, I get a load line like this. How high can the B+ go, what is the limit?

The 1K load line is fine for class AB1. The max power curve is meant for the quiescent conditions. When the device is modulating power the average anode dissipation will be not that much.... otherwise you wouldn't get anything at output!
You can go up to 400V 60 mA (i.e. 24W) safely. The only exception I know is the NOS EI with tipped glass that was safe to run up to 20-21 W but was by far the best sounding EL34 I have ever listened to. Better the NOS Mullards, Telefunken etc....
At voltages higher than 400V you should look at the specifications of your actual tube. At the moment "my" EL34 is the Golden Dragon. Pretty good, I would say.
One experiment you can do is: try two or three rather different anode voltages/biases and you will be able to see how the sound will be affected. It's a really good way to find your sound, also according to the speaker you have. As a line of guide, typically, it will be darker for lower B+ and brighter for higher B+. Much better than just swapping tubes of different types and brands....
 
Actually I have a question about audio transformer. Digging in a carton box I've found an all black transformer, I believe it was manufactured by Philips between 1959 and 1966. The secondary, depending on straps, is good for 14 or 7 ohm speakers. The code number is PK 51099.
Anybody can tell me about its specifications and if it is worth to use it for a PP amp of EL34 ?
 
The 1K load line is fine for class AB1. The max power curve is meant for the quiescent conditions. When the device is modulating power the average anode dissipation will be not that much.... otherwise you wouldn't get anything at output!
You can go up to 400V 60 mA (i.e. 24W) safely.

So where is the safe operating area for Class AB1, anything less than say 48W? (Provided the tube operates in that area for only half the time?) The 24W power dissipation curve is really only for Class A1 and quiescent conditions then? That would make sense to me...

45, can you answer this question please? When a loadline passes way below the knee of the 0V grid curve, like in the loadline I posted, and the tube grid1 is driven to 0V, what actually happens to the tube or amp in this situation? I know its a bad idea, but I don't understand what happens, does it just square wave the output of the tube?

300b-luc, you can measure the primary impedance of your transformer, this will let you know what tubes it would be suitable for. There are a few threads around that describe how to do this. If you can't find one, let us know.
 
So where is the safe operating area for Class AB1, anything less than say 48W? (Provided the tube operates in that area for only half the time?) The 24W power dissipation curve is really only for Class A1 and quiescent conditions then? That would make sense to me...
A valve does not suffer from second breakdown so the concept of safe operating area is meaningless (the SOAR plot is rectangular), providing the operating frequency is high enough that mechanical fatigue from cyclic thermal load on the plate is not an issue then you can drive the valve as hard as you like as long the average plate dissipation is within ratings and the grids are not overloaded. The screen grid has much less thermal mass than the plate, with a constant voltage screen supply and a hard driven control grid AB2 or C the plate voltage should fall below the screen grid voltage causing it to capture more electrons and possibly melt or distort if overloaded.
 
Metalsculptor, thanks very much for the info. I have found very little info on PP loadlines regarding how hard you can go before something blows up, I guess *average* power dissipation is taken for granted by those who know the subject well. Now I understand...

I have begun drilling the top plate, work is in progress...

Edit: 300b-luc, put 10 or 20V AC into the primary and measure the V AC at the secondary with no load. From that the primary impedance can be easily calculated. All you need is an AC supply and a multimeter.
 
Last edited: