5
\$\begingroup\$

Available speakers all seem to be low impedance like 4 or 8 ohms etc. Philips used to make speakers of 800 ohms for their SRPP valve stereo amplifiers. The speakers weren't that great even in 1958. But the SRPP amplifier is still valid and Audiophiles are designing modern versions today that have very good performance by contemporary standards.

When I would fix a Philips Hi-Z stereo we couldn't get new speakers so we would use an output transformer. The speakers weren't stunningly reliable and they had small magnets by today's standards. My question is would it be feasible to make Hi-Z speakers so the sound of a valve OTL amp could be achieved at reasonable cost?

\$\endgroup\$
  • \$\begingroup\$ Spell-check is your friend. I expect that if you approached a speaker repair place and asked them to wind you a voice coil with very fine wire (lots of turns on lots of layers) you might have just what you want. The magnets and mechanics would then be of modern quality. Worth a shot and probably reasonable cost too. \$\endgroup\$ – KalleMP Sep 13 '15 at 11:41
  • \$\begingroup\$ Blasphemous answer: use the SRPP to drive an 800 Ω heat sink, and also DI the output to a modern Class D amp. \$\endgroup\$ – leftaroundabout Sep 14 '15 at 10:12
7
\$\begingroup\$

Making high impedance speakers is difficult. Take apart some old broken speaker and you will see the wire of the coil is already very thin and long if you uncoil it. To get higher impedance, you'd need more turns, which means longer wire, which means even thinner if you want to keep the mass the same. Since the coil of wire has to move back and forth, keeping its mass low is important to good performance.

All these tradeoffs conspire so that it is impractical to make speakers with high impedance and good performance and reasonable cost. You said yourself the 800 Ω speaker wasn't that great even by 1958 standards. There are probably good reasons for that.

As Andy said, go use a audio tranformer with the right ratio. These used to be a lot more common in the tube days, since tubes don't go to really low impedances well but do handle high votlages well. Tube amps therefore put out a high voltage high impedance signal directly, which was then converted by a transformer to a low impedance lower voltage signal more suited to driving a speaker. Since a transformer was in there anyway, it often had a center tapped primary to allow two identical circuits to drive each direction actively. This was useful because there is no PNP or P channel equivalent tubes, which are inherently N channel only.

Today we use transistors which can go to very low impedances directly, operate well at audio power voltages, and come in complementary positive/negative versions. As a result, most audio power amps today drive the low impedance speakers directly without a transformers.

Audio power transformers are still available today, but this is one of the few parts you can usefully salvage from a old tube audio power amp.

\$\endgroup\$
  • \$\begingroup\$ I was hoping that a bigger magnet ,bigger speaker would have room for the proposed 800 ohm coil. \$\endgroup\$ – Autistic Sep 13 '15 at 12:34
  • \$\begingroup\$ @Autistic Yes, using a larger speaker frame than needed will make the winding operation more likely to succeed. \$\endgroup\$ – KalleMP Sep 13 '15 at 12:42
3
\$\begingroup\$

Approach a speaker repair operation and ask them if they will rewind a speaker of your choice with more wire. Most will say no but some hungry craftsman may tackle the job.

The issues you will have to resolve will include

  • The wire will need to be thinner and not on hand
  • The thin wire may be too weak to wind with normal techniques
  • There will be more layers
  • The terminal voltage will be higher and interlayer insulation my be required
\$\endgroup\$
  • 1
    \$\begingroup\$ Believe it or not at least one model of the phillipsHiZ had 800 ohm tweeters as well that were coupled to the woofers with 100nF caps .In 1958 they sure knew how to get turns on a coil . \$\endgroup\$ – Autistic Sep 20 '15 at 10:42
3
\$\begingroup\$

As others have pointed out, there are reasons why high impedance speakers were rare, even when valves were commonplace. Having a speaker custom wound is not going to be cheap, nor is it going to be easy to find someone to do it for you.

The best option (if you really hate transformers!) might be to get 25 32-ohm speakers and wire them in series. A quick google search shows there are high quality 32-ohm speakers available (whereas at 64 ohms there are only low power, cheap, poor quality speakers.)

It's an odd solution, and maybe not the best for bass, but such a cabinet would certainly be a talking point.

\$\endgroup\$
  • \$\begingroup\$ I was thinking of doing this for test purposes.Surely the bass could be OK if the speakers were say 8 inch.I could make the valve OTL amplifier run more EL84 valves in paralell and make a test box with 8 32ohm speakers to run in series for the HiZ and in paralell for a normal solid state amp.This would allow me to compare to see if there are worthwhile sonic gains. \$\endgroup\$ – Autistic Sep 13 '15 at 20:52
  • 1
    \$\begingroup\$ @Autistic 25 8" speakers won't be cheap, but should be available. Here's some 10" 32 ohm guitar speakers (presumably intended for wiring in groups of 4 in parallel) musiciansfriend.com/accessories/… \$\endgroup\$ – Level River St Sep 13 '15 at 21:02
  • \$\begingroup\$ If you phase and position the elements carefully you could send the sound in a narrow beam forward and not bother the others in the room. \$\endgroup\$ – KalleMP Oct 22 '15 at 21:03
2
\$\begingroup\$

As has been pointed out, a hi-Z speaker requires many turns of very fine wire, which means that its current capacity is not very high. Also, as you note, the fine wire has reliability issues as well, not to mention the extra mass of the coil, which compromises the frequency response.

The key problem is that without an output transformer, the speaker must not only carry the AC signal current, but also the DC current required to operate the output stage of the amplifier, which is what leads to the issues noted above. It was long ago decided that using a transformer with a low-Z speaker had better overall performance, although designing good transformers (wide bandwidth and high linearity) is not a trivial exercise, either.

\$\endgroup\$
  • 1
    \$\begingroup\$ The phillips SRPP Amp has capacitive coupling to the 800 ohm speaker so no DC if an output transformer is fitted. \$\endgroup\$ – Autistic Sep 13 '15 at 12:31
  • \$\begingroup\$ When transformers are designed the volume of copper and hence pretty much the mass has little to do with the voltage but everything to do with the VA .This sweeping statement is true for all normal voltages .Very high volts at very low currents will waste space in insulation as it becomes more significant in percentage terms .Very low volts at very high currents reduce packing factor also.Boiling this down implies that a reasonably large speaker wont have a terrible mass penalty. \$\endgroup\$ – Autistic Oct 14 '15 at 10:46
2
\$\begingroup\$

Take an 8 ohm speaker and a 10:1 step down transformer with the speaker on the low voltage winding. The impedance looking into the high voltage winding will be: -

Turns ratio squared x 8 ohms = 100 x 8 = 800 ohms.

You are going to find it a lot easier/cheaper to make a 10:1 step down transformer.


EDIT section

If you wanted to raise the impedance by 100 this only requires ten times as many turns on the speaker but that is a phenomenal ask. Speakers have poor power efficiency at the best of times (circa 5%) and imagine the fragility of a speaker wire with one-tenth the coss section. The other option is to make the gap between poles that the speaker wires slot into bigger biut this will reduce efficiency.

\$\endgroup\$
  • \$\begingroup\$ I think the OP was specifically trying to avoid the matching transformer thing. \$\endgroup\$ – KalleMP Sep 13 '15 at 11:56
  • \$\begingroup\$ Watch out: IIRC the voltage ratio is the square root of the impedance ratio. \$\endgroup\$ – Wouter van Ooijen Sep 13 '15 at 11:57
  • 1
    \$\begingroup\$ To get 10 watts into an 8 ohm speaker requires about 25V peak-to-peak. For the same power into an 800 ohm speaker requires 250V peak-to-peak. This level is not easy to achieve with solid state amplifiers which is why high impedance speakers are rare today. \$\endgroup\$ – Barry Sep 13 '15 at 12:02
  • 2
    \$\begingroup\$ @Barry: That's not why high impedance speaker are rare. After all they were rare even back when 250 Vpp was what tube amps produced natively, which was then transformer-coupled to a low impedance speaker. The real reason that high impedance speakers are rare is speaker design tradeoffs. \$\endgroup\$ – Olin Lathrop Sep 13 '15 at 12:13
  • 1
    \$\begingroup\$ Raising the gap between the poles as the EDIT section qf Andy akas answer sounds valid.Nowdays the magnets are so much better especialy when the path is longer .In other words the efficiency reduction wont be as bad as people think.You should be better off than with 1958 magnets. \$\endgroup\$ – Autistic Sep 13 '15 at 20:39
-2
\$\begingroup\$

Why not just take an existing field coil speaker, and run it backwards.plate to ground across field coil,providing modulated signal,and energise voice coil with a fewvolts DC to provide magnetic target.no net change in mass.

\$\endgroup\$
  • \$\begingroup\$ After thinking it over, I know what you mean. Could you maybe write a clearer explanation and define the terms you used for those people who might not be familiar with things like "field coil speaker?" \$\endgroup\$ – JRE Apr 18 '16 at 8:49
  • 2
    \$\begingroup\$ Do you understand that the field coil of that kind of speaker has a huge inductance? (I have even seen equipment in which the speaker field coil was used as part of the power supply filter.) As such, it will have an impedance that varies very strongly with frequency, making it unsuitable for audio reproduction. \$\endgroup\$ – Dave Tweed Apr 18 '16 at 12:00
  • \$\begingroup\$ @DaveTweed: Hadn't thought of that. Now that you mention it, I remember getting "bit" really hard by the field coil in an old speaker when I disconnected the power supply to it. \$\endgroup\$ – JRE Apr 18 '16 at 12:38
  • \$\begingroup\$ The field coil on ye olde electromagnetic speakers has high impedance but the solid iron frame gives it lots of lossey inductance.In fact many old valve radios used the field coil as a power supply filter choke .If you do this you will have no high frequency response. \$\endgroup\$ – Autistic Apr 20 '16 at 21:26
  • \$\begingroup\$ @Aut: Your comment just repeats what Dave said two days earlier. \$\endgroup\$ – Olin Lathrop Apr 25 '16 at 10:55

protected by Olin Lathrop Apr 25 '16 at 10:47

Thank you for your interest in this question. Because it has attracted low-quality or spam answers that had to be removed, posting an answer now requires 10 reputation on this site (the association bonus does not count).

Would you like to answer one of these unanswered questions instead?

Not the answer you're looking for? Browse other questions tagged or ask your own question.