Deacy amp transformers

[rsasgtr]

You're right.

So the negative of the battery must not be connected to "ground". Because it's a battery, both + and - are floating. At least in the one without the cabinet it uses that one side of the speaker as "ground". And because the input is capacitor coupled, it still works. That was done so that the sleeves of the jacks are connected together like a typical guitar pedal.

Typically, both the treble booster and the amp are battery powered. The ground is a bit strange since the transistors are PNP, so the ground should be positive, and the power supply negative, like in the KAT schematic.

So where are these "off-the-shelf transformers"?

In the non-cabinet version https://doxyworld.com/images/documents/supersonic_doxy_spec.pdf

The cabinet version specifications:

https://www.doxyworld.com/images/documents/supersonic_pro_spec.pdf
says the transformers are "hand-wound bespoke replicas of the original Supersonic transformers" which sounds like BS to me.

As far as I know, the person making them has one of the original radios, so they can make clones relatively easily. Are you suggesting that both versions use the same transformers?

If the off-the-shelf ones have the right taps, that's not really going to be much different. If you can find those, that's 90% what you need to make one a legit one.

Indeed. And that's one of the reasons I posted. I hoped someone might know them since they exist.

The KAT transformer specs are a bit misleading now, since they for the kit only. On the other hand, they can't be totally off.
What do we actually know about the transformers?
The interstage ratio is 2:1 (page 116)
On page 115 we can see:

Since the ratio is 2:1, so 200:50-50 are probably DC resistances (not impedances). That almost matches KAT's 230:106. Also, since 200:100 = 2:1, both windings are made with the same wire gauge.
I'll try to find a similar Hammond one.

Now the output.
The impedances are 140 CT:35 (page 117), so the ratio is 2:1.
On page 115 we see:

DC resistances (presumably) 2-2.5-2.5-2:X, assuming there are only 2 windings. So the full ratio is 4-5-5-4:9.

To sum up, we're looking for:
Interstage: 2:1, RDC=200:100
Output: 4-5-5-4:9, RDC=9-?, Z=140CT:35

Or tell me where I made a mistake.

 

[Bo Deadly]

The interstage ratio is 2:1 (page 116)

Impedance ratio.

View attachment 86246
Since the ratio is 2:1, so 200:50-50 are probably DC resistances (not impedances).

No I think those are actually impedances.

View attachment 86247
DC resistances (presumably) 2-2.5-2.5-2:X, assuming there are only 2 windings. So the full ratio is 4-5-5-4:9.

?

I think those are actual impedances. With 9V, you need low impedance to get enough current to make a decent amount of power. The speaker is probably 4 ohms so what it's connected to has to be similar. But it's not trivial to figure out what that impedance is because, with the center tap grounded, only one transistor is pumping current into one half. But the other tap is magnetically coupled to the core. So I think it's 5 ohms but not sure. I'm not a magnetics guy.

But my guess would be that the OT is an impedance ratio of 9:4.5 / turns ratio of 1.4:1.

So they're trying to pump as much current through it as possible. But inductance is very low so it's low frequency response is probably really bad. Thus the two feedback paths around the output.

Hammond isn't going to have anything like 1.4:1 with 3 taps. That's wacky. That's why I asked why the DoxyWorld specs say "off-the-shelf". Find out what shelf they're getting theirs from.

 

[rsasgtr]

Impedance ratio.

Turns ratio:

So 200:50-50 are not impedances.

I think those are actual impedances.

2-2.5-2.5-2 can't be. As I wrote, the impedances are:

The speaker is probably 4 ohms

It's 16 Ohm. The KAT-Celestion speaker replica:

so what it's connected to has to be similar. But it's not trivial to figure out what that impedance is because, with the centre tap grounded, only one transistor is pumping current into one half.

I don't thin that's accurate. Note that it's the inner tap, not the outer one, the collectors are not grounded, therefore both transistor are still outputting fine. One of the transistors is getting an extra 4.7k resistor between the base and ground, and not getting the feedback (right?).
EDIT: typo

 

[Bo Deadly]

Turns ratio:
So 200:50-50 are not impedances.
2-2.5-2.5-2 can't be. As I wrote, the impedances are:
It's 16 Ohm. The KAT-Celestion speaker replica:

Ok. I guess you're right.

Where is this PR80 manual that you're referring to? Did you post a link for that already? If not please do.

But a 16 ohm speaker on windings with 5 ohms of DCR is pretty strange.

I don't thin that's accurate. Note that it's the inner tap, not the outer one, the collectors are not grounded, therefore both transistor are still outputting fine. One of the transistors is getting an extra 4.7k resistor between the base and ground, and not getting the feedback (right?).
EDIT: typo

I think it's accurate. The transistors are going to be biased so that they're just barely on. When signal is present, the transistors are going to alternate between being more or less on and off. When one is on, the other has to be off. Since the center tap is connected to ground, that current is only going through one half of the primary. But that's not a terribly important detail. As for feedback, if there is a negative swing on a base, that will turn on the transistor which will deliver current to the transformer and into ground. That will cause a positive voltage to develop at the inner tap which, through the 4.7k, will provide negative feedback.

 

[rsasgtr]

Ok. I guess you're right.

Where is this PR80 manual that you're referring to? Did you post a link for that already? If not please do.

I thought you had access to the radiomuseum manual. Long time ago someone posted this:
Wayback Machine
Wayback Machine
Wayback Machine
Wayback Machine
Wayback Machine
Wayback Machine
Wayback Machine
Wayback Machine
Wayback Machine
Wayback Machine
Wayback Machine

But a 16 ohm speaker on windings with 5 ohms of DCR is pretty strange.

How do we know that the RDC of the secondary is 5 Ohm?

I think it's accurate. The transistors are going to be biased so that they're just barely on. When signal is present, the transistors are going to alternate between being more or less on and off. When one is on, the other has to be off. Since the center tap is connected to ground, that current is only going through one half of the primary. But that's not a terribly important detail. As for feedback, if there is a negative swing on a base, that will turn on the transistor which will deliver current to the transformer and into ground. That will cause a positive voltage to develop at the inner tap which, through the 4.7k, will provide negative feedback.

Fair enough. I'll have to think more about it. BTW page 115 has some voltages marked.

 

[Bo Deadly]

I thought you had access to the radiomuseum manual. Long time ago someone posted this:

I do but that's not what's on the site. It has a completely different document. The circuit seems to be the same but the presentation is completely different. The one you just posted does not look like a manual. It appears to be a section of some kind of third-party service manual. But it's better than the one on radiomuseum.

How do we know that the RDC of the secondary is 5 Ohm?

Because the speaker is connected across 2.5+2.5. But that's not the proper DCR because the tap can source / sink current from the ends of the winding as well. And the center tap is low Z. And according to your "manual" the speaker is "30-40" ohms. So maybe a DCR of an ohm or two isn't a big deal. Again, in general it's not terribly important.

 

[rsasgtr]

I do but that's not what's on the site. It has a completely different document. The circuit seems to be the same but the presentation is completely different. The one you just posted does not look like a manual. It appears to be a section of some kind of third-party service manual. But it's better than the one on radiomuseum.

Does the one from radio museum have more info about the transformers?

Because the speaker is connected across 2.5+2.5. But that's not the proper DCR because the tap can source / sink current from the ends of the winding as well. And the center tap is low Z. And according to your "manual" the speaker is "30-40" ohms. So maybe a DCR of an ohm or two isn't a big deal. Again, in general it's not terribly important.

Makes sense.
Now, more food for thought for the off-the-shelf doxy:

And there goes the feedback to the emitter of X7 (AC126). Now it's connected through C14 to ground.
The transformer just got simpler. Now we have an autotransformer with 4 taps and RDCs 2-2.5-2.

 

[Bo Deadly]

Does the one from radio museum have more info about the transformers?

No.

And there goes the feedback to the emitter of X7 (AC126). Now it's connected through C14 to ground.
The transformer just got simpler. Now we have an autotransformer with 4 taps and RDCs 2-2.5-2.

Wow. That's super interesting. I wonder how the DoxyWorld guy would have known to do that. He must have had the inside track. Or maybe he had access to a KAT replica.

Grounding that feedback cap removed the negative feedback and bypassed the emitter resistor of the interstage transistor. With one tiny inexpensive piece of bus wire, the gain is double boosted! John Deacon was quite the hacker. I'm impressed. Well played sir.

That opens the door. The difference between the inner and outer windings is really not that much. You probably could just use a transformer with two CT windings that are the same. Like Xicon 42TM030-RC which is $3 USD on Mouser. It's primary and secondary are 100 ohms so 200 total vs 140. That's good enough for government work. And the description you posted originally said the transformers were badly imbalanced so if there's a little difference between primary and secondary that's what you want anyway.

 

[rsasgtr]

Wow. That's super interesting. I wonder how the DoxyWorld guy would have known to do that. He must have had the inside track. Or maybe he had access to a KAT replica.

Grounding that feedback cap removed the negative feedback and bypassed the emitter resistor of the interstage transistor. With one tiny inexpensive piece of bus wire, the gain is double boosted! John Deacon was quite the hacker. I'm impressed. Well played sir.

That opens the door. The difference between the inner and outer windings is really not that much. You probably could just use a transformer with two CT windings that are the same. Like Xicon 42TM030-RC which is $3 USD on Mouser. It's primary and secondary are 100 ohms so 200 total vs 140. That's good enough for government work. And the description you posted originally said the transformers were badly imbalanced so if there's a little difference between primary and secondary that's what you want anyway.

Shouldn't I use a 1.6:1 (2+2.5+2:2.5)? It's the same wire, so RDC ~ wire length ~ number of turns? So maybe
42TM031-RC?

Unfortunately, only the cheaper Doxy has that wire. I wasn't able to confirm it's existence on the original Deacy or the Doxy Pro. But I found some pictures of the original transformers and the replicas. They have 7 taps. It seems the off-the-shelf-transfomers version takes some shortcuts.


Another OT from the other side:

EDIT: added a picture

 

[Bo Deadly]

I couldn't say. I'm not a magnetics guy.

 

[din_mentor]

Bump!
Any news on the subject of choosing off the shelf transformers?

 

[rsasgtr]

Any news on the subject of choosing off the shelf transformers?

In the end, KAT released a full PCB kit including the transformers: KAT-KIT Deacy Amp Board And Components