#5 I think.
Thanks. Yes, I believe Self's differential input circuit was unusual. Not sure how. And yes, I see what you mean about leaving that pin floating when the RCA socket is used. Thanks.
I'm intending to drive my Modulus686 with 24VAC secondaries of my transformer.
- What will be the final DC rail voltage after the power supply? What is the typical drop expected after rectification? I'm using IC bridges, not individual fancy diodes.
- What is the peak AC output voltage I can expect the amp to deliver into 8 Ohms (for brief stretches of, say, 30 seconds) without clipping, in my case?
What is the answer to Question 2, if the answer to Question 1 is "31V"?
Of course !
just because u have 31v after rectifiers that does not mean anything. Type of transformer and transformer va is very important too.
More va equals less voltage drop under load.
Those are nice supplies and work very well in the Modulus-686. My MOD686 amps are all powered by the RPS-400-27-C (caged version of the RPS-400-27). You'll get ~130 W into 8 Ω and 200 W into 4 Ω with that.
Tom
Exactly!
That doesn't give enough information to answer your question. We'll need to know:
- Mains voltage
- Transformer turns ratio
- Transformer parasitics (at least DCR of primary and secondary windings)
- The characteristics of the transformer core
- Characteristics of rectifier diodes
- Parasitics of power supply PCB and wiring
- Capacitance (and ESR) of reservoir caps
- Etc.
That said, the back-of-envelope answer to your question is "around 32-33 V most likely". It'll droop a volt or two if you run the amp at clipping levels with an 8 Ω load. That should be enough for around 170-180 W into 8 Ω.
If you would like greater precision in the numbers, use a switching supply, such as a pair of Mean Well RPS-400-27-C.
Tom
Thanks a lot for the first useful response.
I needed a back-of-envelope answer. I guess I should have specified "I want a figure +/-1V, and with reasonable assumptions" when I asked. Thanks.
Thanks a lot for the first useful response.
My answer won't be accurate to ±1 V. Think about it... Mains variation alone would cause ±3.3 V variation.
If you want ±1 V precision, use a regulated supply.
Tom
Somehow, I have succeeded in confusing all of you about the context of my question. I don't need a very precise answer. I just wanted to know a "nominal" answer. I'm not trying to do anything with edge cases or corner cases. Sorry about this. You're of course correct that I won't get even a 10% accurate answer if I map the nominal to the real world. (I live in a country where the real mains voltage is often 30% lower than the nominal figure in many parts of the country. Voltage stabilizers are a household device in every household, and every store selling a refrigerator or AC will sell you a compatible voltage regulator too.)
But I just needed a nominal figure, with "reasonable" assumptions on xfmr, diodes, etc.
As a rough rule of thumb I multiply the transformer's secondary voltage by 1.5 to account for high line voltage, higher no-load voltage on the transformer secondary, and because it's an easy number to remember. If that number pushes any limits, then it has to be considered more thoroughly.
I found this in Self's Audio Power Amplifier Design Handbook:
An externally hosted image should be here but it was not working when we last tested it.
Oh, that's a neat trick! Basically, you take a standard differential input and connect the single-ended input through a 100 Ω resistor to the non-inverting input of the opamp. When the XLR is unplugged, the opamp works as a buffer. Nice!
The obvious drawback is that the circuit won't work with both the XLR and RCA sources plugged in. It has to be one or the other.
Tom
The obvious drawback is that the circuit won't work with both the XLR and RCA sources plugged in. It has to be one or the other.
Tom
Yes, it was never intended to work with both sockets connected to cables.
And the input circuit of the Self preamp in Linear Audio doesn't even have these resistors, IIRC.
We had a "similar" discussion some months back in the THAT thread, about the Witlock's JENSEN AN-003 note. That note shows an unbalanced to balanced interface that uses only a pair of resistors and a cap, no transformers involved.
I think the consensus was that
I think the consensus was that
and
I'm confused here... We were discussing inputs and your response mentions output.