output transformer impedance question

[Tapsnap]

Hi everyone,
I'm new to building amplifiers and I am working on my first 4W single ended amp. The schematics call for an output transformer with a primary impedance of 4000 ohm and 8ohm secondary. I have actually found it difficult to find one to those actual specs but I figure that I could use a transformer with primary 8000 ohm with a 16 ohm speaker and it would essentially be the same thing because what I'm really looking for is a single ended output transformer with a 22.36 : 1 winding ratio, Right? So I settled on a classic tone 40-18030. Is this logic correct?

[92Volts]

It depends on the tube and voltage used. It likely doesn't matter whether you get very close to 4000 ohms.

Potentially, 4000 would be "ideal" but you can use something else and it will work fine.
It's also possible 4000 isn't "perfect anyways" and was chosen as a result of whatever the circuit designer had access to. Considering both of these possibilities, if you can find 5k or 6k, just use that.

Cutting speaker impedance in half "should" cut OT primary impedance in half, yes. In other words that 8000ohm to 16ohm transformer would act like 4000ohm IF you used it with an 8 ohm speaker.

The transformer may not perform at its full "potential" in terms of frequency or distortion but for a guitar amp, you likely don't care.

But again... you could probably get a 8000 to 16 ohm, use it with a 16 ohm speaker for an effective 8000 ohms, and it would probably sound and operate just fine.

[Tapsnap]

OK, so I had another thought. I recently I came across this 10W, 12" field magnet speaker from a 1934 Philco radio. So now I am thinking about  whether  I could use that with this amplifier . It’s in great condition despite its age. The only problem is that it has 2.28 ohm impedance. For this to work, I calculate that I need an output transformer with a winding ratio of 42 : 1. That’s 4000 divided by 2.28 equals, the  winding ratio squared  (1754.3). And the square root of that, is 42. Using Output transformer that I have, with the primary of 8000 Ohms with the secondary 4 Ohm The transformer is running at the winding ratio of 44 : 1 Which is kind of close to 42 : 1 . So, am I correct in thinking that it would be possible to use this speaker? Of course I need to think about powering the field coil, which I think requires 80 volts DC.
Also, is it possible to add an additional impedance to the secondary using a resistor so as to not over-load the transformer?

[PRR]

> 2.28 ohm impedance.

Audio impedance? Or DC resistance??

2.3 Ohms DC on an old-old Zenith, I would call "4 Ohms audio impedance". When the coil/cone is moving, the impedance is higher than the DC resistance. For most later speakers, we assume that 3.2 Ohms DC is around/above 4 Ohms over the audio band. For historical reasons that I am too sleepy to go into, a big old Zenith is likely to run a bit higher than this ratio.

Also single-ended amps will NOT fry when mis-loaded (except a way-high load with overdriven signal may arc the OT). So 3, 5, 2, no great difference.

As you say, a voice-coil speaker that old is probably field-coil. They ran anything from 16V to 300V. Get it wrong, it is ruined.

Is there any chance you found the OT from the Zenith? That and the Zenith's final tube is a known-killer combination (I don't know any pre-1950 Zenith that sucked). I can't picture what tube the used in 1934... maybe a '50? (Maybe PPP '50s?) That may be tough. A fist-full of 12AU7 may be a near fit. Chassis number might help.

[Tapsnap]

No this is 2.28ohms AC (impedance). These are specs I got for this speaker from the Philco Phorum website. The output transformer is a push-pull, so I can't use it for this amp but its primary is 5700 ohms impedance @ 400Hz and the voice coil is 2.28 ohms impedance (not resistance) which makes it exactly a 50 : 1 WR  output transformer.  However, since I need it to match to a 4000 ohm primary single ended amp I need to figure out the values for OT which will best drive the speaker.

The field coil is rated at 1140 ohms DC @ 70mA 

[92Volts]

Could you post or link to the schematic of the amp you're building? Again, whether it's important to even be close to 4000 ohms depends a lot on the tubes, voltage, etc.

The 8000 ohm to 4 ohm transformer "should" work out to 4500 ohms with the 2.28 ohm speaker. You can't always assume a transformer can do more current at lower voltage (lower impedance than it's designed for) or visa-versa, but in a low-watt application I'd expect it to work fine. Less than a 2:1 mismatch of impedance to what the circuit calls for, or what a transformer calls for, is almost always acceptable.

[Tapsnap]

Here's a link to the page. I'm building the Cupid X.O. http://schems.com/schematicheaven.net/manu/ricktone/ricktone.htm

[Tapsnap]

"You can't always assume a transformer can do more current at lower voltage (lower impedance than it's designed for) "

The OT that I bought has two primaries; one at 8000ohm, the other at 5000ohm. Would it be better to use the 5000 ohm tap with the 4 ohm secondary? That would be a 47 : 1 WR  That's getting quite far from 42 : 1 WR, but it would handle the current without a problem. Right?

[92Volts]

The circuit doesn't care about the winding ratio, only the primary impedance. The winding ratio is useful if you want to predict impedance with a different speaker.

If you use 2.28 ohms instead of 4, what would be 8000 becomes 4500. What would be 5000 becomes 2850. I'd say the 8000 ohm tap is better.

There's nothing about that circuit that appears to make it sensitive to minor variations (no feedback loop, etc). Anything from 4000 up to 8000 would be typical in such a circuit.

[Tapsnap]

OK. thanks that's good to know.

Now I'm a little sketchy on the mathematics of this so perhaps you could help me out. Regarding the field magnet, I need 80 volts DC. So I've figured that I need to drop the B+ by 92 volts from 172v. The speaker draws 70mA. So 92  divided by .07 equals 1314ohms. Is my math correct? Do I need between a 1.2K and a 1.5K resistor? Or is there something I am missing?

[92Volts]

The power needed by that speaker is enough to significantly drag down your B+ voltage if not overload and damage your power transformers. 70ma is likely double the high-voltage  current draw of your entire amp.

If you drew from the "B+ low" location, it would need to come through the 47k voltage-dropping resistor that brings the voltage down to 172v to begin with. Not possible, you'd need 3000+ V to push 70ma through that power-supply resistor. So you can't do this for multiple reasons.

Otherwise your math seems right. One thing to keep in mind is resistor wattage, if you drop 92 volts at 70ma that's over 6 watts of power dissipation. You can get 10W or larger resistors but since you need a non-standard value you could also use multiple 5W resistors in parallel (for example) to get that resistance.

[sluckey]

It's a 4 watt amp! You're thinking too much. Get a speaker with a real magnet and just build it.

[Tapsnap]

I guess I could just buy a modern speaker, but I happen to have this one and it would be kind of cool if I could get it to work.
Anyway, the two back to back transformers that I have for my power supply are both 3 amp 120v to 12.6v transformers I got at closeout from Radio shack.  I can see your point that the speaker would put a drag on the amplifier circuit if I take it from the low B+, but could I take it from a separate parallel circuit coming off the transformer before the voltage-doubler?

[sluckey]

I would expect that old '34 Philco speaker cone to disintegrate with the first Pete Townsend power chord. Even with a 4 watt amp.

[Tapsnap]

Well sure, there is a possibility that it sounds like crap and falls apart after 5 minutes and I may then decide to buy a ceramic magnet speaker. On the other hand, I maybe decide to re-cone it. Either way I will have learned a lot about amplifier building and electronics along the way. And really that's what I'm here for. I have learned something from every question that has been answered on this thread. As I said, I am new to amp building and electronics so this is all good.

[sluckey]

This is a good learning project for a first build. Have you actually started the project yet? Have fun and keep us posted.

[Tapsnap]

I'm about 3/4 done with this project. I'll post photos when I'm done.
Anyway, still wondering whether I could power the speaker field magnet from a separate parallel circuit coming off the transformer before the voltage-doubler?

[92Volts]

Potentially, or you could power it off one half of the voltage doubler. Due to loading the capacitors and the cheap back-to-back transformers, either method may drag your B+ down pretty dramatically.

[sluckey]

Where did you get the specs for the FC speaker?

I'd experiment with this circuit. Hopefully it won't suck the B+ down so much as to be useless.

[Tapsnap]

Wouldn't it be better to come off  T2 directly, and have a separate bridge rectifier/filter cap circuit for the speaker? That way it doesn't pull down the voltages of the amp circuit.

[92Volts]

You'd need a big filter cap to deliver DC (let alone stable "clean" DC) to the speaker, as part of this separate circuit you're considering. Adding that extra cap onto the ones in your main power supply (and then using the main power supply) might be as effective as having it separate.

Your voltage will get dragged down anyways, because the transformer won't be able to produce the same voltage under greater load. This is separate, but related, to the issue of pulling too much current from your power supply capacitors. Both effects could increase hum/noise and also decrease available voltage.

[PRR]

First: does this 2.28 Ohm speaker still WORK?

If it does, it may be FAR more valuable to a Zenith restorer than as a guitar speaker, which it really was not meant to do, and may not last. Get the model/chassis number and look for similar things on eBay. Radiomuseum.org is another nest of radio fans. I would hate for you to shred this irreplaceable thing while there are silent Zeniths in the world.

I have real doubts about reconing a 1934 speaker-- those molds were lost a long time ago. MAYBE something similar can be "fitted", but it won't be right, and may not be good. I'd really just order a new low-price speaker, or ask the pawn shop for a junker SS amp with a working speaker.

How are you going to feed the field coil? An 80V 70mA load is a poor-poor fit for a 300V 40mA power supply. I trust Slucky's math. The resulting 300+ raw volts DC at 36mA+70mA= 106mA is above the 38VA rating of your second PT, and your first PT also carries the heater load so will just toast.

> the Cupid X.O.

Odd thing: the power bottle is run about 280V 34mA, so the dart-target load should be 8K, not 4K. This may have been ignorance, available parts, or a careful mis-match for "tone" with a specific speaker and its impedance curve.

8K:4 is 4K:2 and I would look in that area +/-30% to hit the RickTone's off-impedance.

[Tapsnap]

OK, so I'm going to take everyone's advice and abandon the idea of using this speaker.
So PRR, are you saying that the output transformer should be 8K primary and 8ohm secondary with an 8 ohm speaker?

[92Volts]

The 8k PRR suggested gives the most power based on the voltage and current, at least by a quick estimation. It's also what you'd see in similar amps, Fender Champ transformers are like 7k-8k.

An EL84 datasheet suggests 4k-5k at slightly less voltage (250) so I'm sure anywhere from 4k to 8k is good enough... that whole range is used in known, successful designs which are otherwise similar to yours.

If you have a transformer with 5k and 8k options when used with a 4 ohm speaker, use that transformer and get a 4 ohm speaker. You can try both of those and decide which sounds best. Either of them is in the right ballpark.

[Ed_Chambley]

+1 on the speaker suggestion.  I am a radio restorer along with many others.  the Zenith 1934 speaker is not easy to find for a restorer.  True, it is not a guitar speaker, however I have used Field Coil speakers with guitar amps goofing.  Sometimes they sound really unique, but never hold up to a guitar beating on them. 

[Tapsnap]

OK, well my output transformer arrived and I wired it into the circuit  using the 5000 ohm primary and the 8 ohm secondary. I hooked it up to a single 8 ohm 3W speaker that I found at the dump plugged in my guitar and I turned on the amp at low volume. After a few seconds I could hear the low strum of the chords that I was playing coming from the speaker. Great! , I thought, a little bit hummy but great! Then suddenly  a loud BRRRRRRRRRR accompanied the sound. I turned off the amp right away. What the hell was that? Sounded like maybe some kind of capacitor discharging sound maybe? Any guesses anyone?
The speaker is temporary and I used it just to test what I had made. I will buy a new one eventually.

[sluckey]

Could be lot's of things. We need more info. Measure all voltages and compare to the schematic. Report differences here. Double check, triple check your wiring. Most new build problems are due to wiring errors or incorrect component values. Post some hi-rez pics.

[Tapsnap]

A couple of questions.
1.  Is it OK to ground to the back of the pots, which are mounted to the metal chassis or do I need to take a shielded wire back to the board?
2.  The diagram shows the OT with a ground. That's also taken care of by mounting it to the chassis, right?
3.  Does the speaker casing need to be grounded?
4.  Which wires need to be shielded and which don't. Right now they are pretty much all shielded and its cluttering the back of my board a little.

Please excuse the rooky questions.

[sluckey]

1. OK to groung to the back of the pots. Many people do. I don't.

2. Mounting the OT to the chassis does not ground the secondary wire as shown in the schematic. You will have to actually connect the wire to ground. However, that's not necessary for this circuit.

3. No.

4. If done correctly, no wires in this amp need to be shielded. If the wire from the input jack to the tube is over 3" long I would use a shielded cable. Connect shield to ground at the input jack only. Leave the shield unconnected at the tube end. The same applies to the wire from the volume pot wiper to the tube. Ground the shield at the volume control ground lug. No other wires should require shielding.

A poor ground connection could cause the symptom you described. So could a lot of other things. Posting voltages and pictures as I asked for will be important for us to help figure out the problem.

[Tapsnap]

I figured out what was causing the "BRRRRRRRR". It was two resistors of the wrong value. I got my decimal point in the wrong spot.
Still getting quite a bit of hum and the volume is not nearly what I expected. I am using a 3W speaker from an old stereo. So I guess that will improve with a good 10W guitar speaker.

Here are the voltages.
R1     280v Before filter Caps
B+ Hi 263v
B+ Lo 181v
R8      90v  (12AT7  Plate 6)
R9      74V  (12AT7  Plate 1)
R12    254v (6BQ5   Screen)
R11     7.6V (6BQ5 Cathode)
R7       1.6V (12AT7 Cathode 3)
R6       1.1V (12AT7 Cathode 8)

The only thing about these voltages that look odd are the voltages at R8 and R9; they appear to read vise versa.
Anything I can do about the hum?

[sluckey]

Still getting quite a bit of hum and the volume is not nearly what I expected.

This amp was not designed to be loud. A real guitar amp speaker will help. Get a sensitive one. You can add bypass caps to all cathodes to increase volume. I'd use 20µF @ 50v for both V1 cathodes and 50µF @ 50V for V2 cathode.

SE amps need a well filtered B+ supply to help decrease hum. You can add another filter stage between D1 and R2. I suggest 47µF @ 450V  and 100Ω @ 2W.

Hum can also be caused by layout issues. You can try moving wires around with a chop stick. Would need to see hi rez pics to advise more about that.

[Tapsnap]

It took me a while to figure out how to post pics, but here is the work in progress

[sluckey]

I figured out what was causing the "BRRRRRRRR". It was two resistors of the wrong value.

Which two resistors?

Still getting quite a bit of hum

I would wait until the amp is properly assembled and then evaluate the hum.

[92Volts]

Nice cab!

It'll be louder with a proper guitar speaker, but also with a speaker installed in a cab. It sounds like you pulled a woofer or other driver or something from a stereo speaker and if it's sitting in open air it will be quiet because of that on top of its lower efficiency.

[Tapsnap]

Thanks for the advice all.
The two resistors, I think were R2 470 ohm and R11 220 ohm.