Long term planning

[mashaffer]

I like to plan well ahead when approaching any project. My current dream is an all tube, all horn system with triamped mains and a sub.

The final speaker system would include a built in horn sub and main channels with 70Hz straight bass horn, mid horn (cone or compr possible), and compression tweeter horn.

My thoughts on amps for the system are as follows.

Sub: PP EL34 (probably pentode mode for max tight bass)
Main Woofer: PP EL34 (Possibly triode mode for favorable dist. spectrum)
Mid: SET or triode strapped pentode for open midrange)
Tweeter: PP ECL82 (high order harmonics rolled of by driver and ears)

My thinking is that the higher order HD of the PP tweeter amp would not be objectionable due to the low levels in this range and the natural roll off of the transformer, speaker and the ear. This would also result in an amp that could be used as a full range amp for use with my current speakers in the initial stages of the project. Given the limitations of my current speakers the ECL82 amp would not need to use a really hurky OPT since a low freq. rolloff at around 40Hz would be perfectly acceptable with the current speakers and absolutly a non-issue once it is put to use in the tweeter circuit. This should save some money on the iron for the first amp. This is also part of the reason for not making the tweeter amp a SE design which requires more expensive iron.

So the first amp to build would be the ECL82 PP. I thought that I might try a transformer phase splitter at the input of the amp (using a smaller CT input transformer rather than the larger traditional phase splitter after the driver stage). The amp would be symmetrical from input to output using two tubes per channel.

So what do you think? Would I be asking for noise trouble putting the phase splitter at the input? Is my selection of topologies sound?

mike

 

[SHiFTY]

I would imagine PP ECL82 would be too powerful for your tweeters! Also the higher order harmonics would definitely be there, although in a well designed NFB amp they would be pretty low.

IIRC you only need a watt or so for the treble, maybe a ECL82 SE ? I built one (ECL82 Mullard circuit) that sounded fine, even using pretty small OPTs. Surprisingly powerful too.

If you are buying some OPTs anyway, why not buy some decent SE ones to start with, build your ECL82 (or EL84 in triode) in SE, and consider later upgrading for 300B or KT88 triode strapped etc.

Then you could think about simply bi-amping. Build a 300B SET for mids AND tweeters? 8 watts should be enough for extremely loud sound from ~100Hz-20000 in a bi-amped situation, esp with horns. This would allow you to spend more money on the parts which really matter- the output transformers and speakers .

Only other problem I can think of is phasing, especially if you use PP and SET together, you would have to be careful they were all inverting, or all non-inverting.

Still, I like your plans

 

[ErikdeBest]

Hi

I like your idea and share the thought of building amps that can be combined in a bigger system someday.

I would build the amplifiers in a way it would be easy to add/substitute coupling capacitors so to easily 'include' a 12dB high pass filter in the amp. This way they could be easily implemented in a x-amplified system.

 

[mashaffer]

I appreciate your replies.

IIRC you only need a watt or so for the treble, maybe a ECL82 SE ? I built one (ECL82 Mullard circuit) that sounded fine, even using pretty small OPTs. Surprisingly powerful too.

That sounds like a reasonable idea. I suppose for lower output such as this the OPT might not be too expensive. Especially when not expecting subwoofer type frequency range. BTW what is IIRC?

If you are buying some OPTs anyway, why not buy some decent SE ones to start with, build your ECL82 (or EL84 in triode) in SE, and consider later upgrading for 300B or KT88 triode strapped etc.

Then you could think about simply bi-amping. Build a 300B SET for mids AND tweeters? 8 watts should be enough for extremely loud sound from ~100Hz-20000 in a bi-amped situation, esp with horns. This would allow you to spend more money on the parts which really matter- the output transformers and speakers .

Possibly. I don't relish the thought of trying to design a good speaker level crossover. That seems like a very involved art form. The components of a speaker level crossover are very simple but it seems to me that it is very hard to design a good one.

Only other problem I can think of is phasing, especially if you use PP and SET together, you would have to be careful they were all inverting, or all non-inverting.

Not too worried about that. I will have to have that figured out in any case when aligning the individual drivers for proper phase response. If I screw up I will just switch the connections to the output binding posts.

Still, I like your plans

Thanks.

I would build the amplifiers in a way it would be easy to add/substitute coupling capacitors so to easily 'include' a 12dB high pass filter in the amp. This way they could be easily implemented in a x-amplified system.

Interesting that you should mention this. I was considering such an approach. Rather than building a seperate electronic crossover with an output tube for each frequency range I was wondering if I could build a preamp with a high enough current output and low enough output impedance to drive all of the power amps from a single output tube per channel. I was thinking of either a small power triode CC output stage or a small power pentode in CF form (hmm.. might be an application for a mu stage).

That way I could build the band pass filter right into the power amp design and avoid any extra active stages. Possible? Downside?

mike

 

[EC8010]

You'd be surprised how much headroom you need on a tweeter amplifier - stick with the higher powered alternative.

Regarding the crossover, if you buy or borrow a Behringer DCX2496 you can experiment to your heart's content until you get a crossover that works, then think about building an analogue one if you feel like it. Oh, and the Behringer has balanced outputs, so you don't need to worry about phase splitter design.