Can you design for me an output circuit using a transformer.
Power supply is +15V -15V
XLR line output, audio signal is up to 70Vpeak differential.
I am not familiar with audio transformers.
How to drive the transformer ?
Which part and sourcing ?
What sound quality can be expected, in terms of bandwidth and distortion ?
Power supply is +15V -15V
XLR line output, audio signal is up to 70Vpeak differential.
I am not familiar with audio transformers.
How to drive the transformer ?
Which part and sourcing ?
What sound quality can be expected, in terms of bandwidth and distortion ?
Well the first question is what load impedance?
"70V peak differential" could be ambiguous, best to clarify if thats differential mode amplitude, ie signal swing is +/-35V for each wire?
Such a signal into 600 ohms requires 4W for instance, into 10k needs 250mW
"70V peak differential" could be ambiguous, best to clarify if thats differential mode amplitude, ie signal swing is +/-35V for each wire?
Such a signal into 600 ohms requires 4W for instance, into 10k needs 250mW
Differential voltage is not ambitious.
Well, 70V peak differential can be seen as 35Vp at each of the two wires relative to virtual ground.
The load impedance is not 600 ohms.
The load is typically a mixing console input, or other audio equipment.
Well, 70V peak differential can be seen as 35Vp at each of the two wires relative to virtual ground.
The load impedance is not 600 ohms.
The load is typically a mixing console input, or other audio equipment.
Last edited:
That is +36 dBu! What kind of equipment expects that? Most XLR line-in will die with such a high signal. Is it rather for 70V PA line? Those are typically used to connect actual speakers, not equipment. In any case, I would use the transformers made for those lines as they are common. Expect such a transformer with low distortion to be quite expensive.
If you just need an audio transformer for normal line level signals, the TY250-P is quite good, cheap and available. However, it can only be used for signal up to about +15 dBu, or +21 if two are used.
If you just need an audio transformer for normal line level signals, the TY250-P is quite good, cheap and available. However, it can only be used for signal up to about +15 dBu, or +21 if two are used.
Yes +36 dBu is beyond usual line levels.
This is experimental, +36dBu is an absolute max before clipping and care will be taken near that.
May be my question boils to: What is available in step up audio transformers at very high line levels ?
Based on such a transformer, how do I drive it ?
Is it simply an op amp output to the primary winding ?
This is experimental, +36dBu is an absolute max before clipping and care will be taken near that.
May be my question boils to: What is available in step up audio transformers at very high line levels ?
Based on such a transformer, how do I drive it ?
Is it simply an op amp output to the primary winding ?
I would look into low power 70V line transformer or tube amplifier output transformer used in reverse. A wide bandwidth will be expensive. A transformer (mostly) reflects the impedance on the secondary to primary multiplied by the transformer ratio. So the driver might be as simple as a power opamp like the NJM4556 if the transformer can step it up to the required level.
Good references are:
High Voltage Audio
Transformers for Audio
Good references are:
High Voltage Audio
Transformers for Audio
Last edited:
I like ESP (did not think looking into it about line transformers).
I found this Lundahl LL7906 +37 dBu (54Vrms)
Here: Line input and line output audio transformers • Lundahl Transformers
LL7906 specs.
https://www.lundahltransformers.com/wp-content/uploads/datasheets/7906.pdf I am not familiar with audio transformers to figure from the specs wether it fits my need and how to design an output circuit with it.
I found this Lundahl LL7906 +37 dBu (54Vrms)
Here: Line input and line output audio transformers • Lundahl Transformers
LL7906 specs.
https://www.lundahltransformers.com/wp-content/uploads/datasheets/7906.pdf I am not familiar with audio transformers to figure from the specs wether it fits my need and how to design an output circuit with it.
Last edited:
You could try jensen-transformers.com resources section. For minimum distortion drive from lowest impedance. Negative impedance drive is also possible for the absolute minimum but has implications for stability.
Still not sure about your needs but one option could be to make a 70Vpp output amplifier, I guess you want a line amplifier so some power involved but not speaker level, and drive a 1:1 line transformer.
The point being that it´s quite easier to wind a 1:1 transformer for various reasons (as seen by a transformer winder which I am):
* you have same number of turns on both windings
* being lowish impedance you use less turns of thicker wire (winding thousands of turns of very fine wire is a pest)
* you use same wire diameter on both coils so you set up the machine once
* in fact, you can even wind bifilar, unless you need high voltage insulation between primary and secondary (I guess not)
* bifilar disadvantage is relatively high capacitance between primary and secondary, but at the impedances involved (guess aound 600 ohm) , no big deal.
Absolute worst case, you wind separate coils; even side by side so as to minimize capacitance.
By the same token, primary-seconday coupling is improved, and leakage (parasitic series) inductance can be somewhat improved.
Being a simplish transformer (compared to other options) it´s more likely a winder accepts winding "just one" or you may find many commercial 1:1 transformers.
Most 600:600, you may find the odd 150:150 or 200:200
The point being that it´s quite easier to wind a 1:1 transformer for various reasons (as seen by a transformer winder which I am):
* you have same number of turns on both windings
* being lowish impedance you use less turns of thicker wire (winding thousands of turns of very fine wire is a pest)
* you use same wire diameter on both coils so you set up the machine once
* in fact, you can even wind bifilar, unless you need high voltage insulation between primary and secondary (I guess not)
* bifilar disadvantage is relatively high capacitance between primary and secondary, but at the impedances involved (guess aound 600 ohm) , no big deal.
Absolute worst case, you wind separate coils; even side by side so as to minimize capacitance.
By the same token, primary-seconday coupling is improved, and leakage (parasitic series) inductance can be somewhat improved.
Being a simplish transformer (compared to other options) it´s more likely a winder accepts winding "just one" or you may find many commercial 1:1 transformers.
Most 600:600, you may find the odd 150:150 or 200:200
Just to state the obvious: if you need "70V", why start with 30V?? This is not 1966 again. High-voltage parts are cheaper than ever.
Transformer is not "free gain". TranSistor gain is nearly free anyway. TransFormers have significant bandwidth limits; and low THD is not cheap. Step-up makes it worse.
You have not really specified your POWER level, except "not 600 ohms". 6r? 6Meg?
For "experimental" purpose, one hack is that half-dead Sansui hi-fi in your closet feeding a 12V:120V transformer. Separate-bobbin windings may not beat 1kHz but one-over-the-other power windings I have measured flat to over 15KHz.
Transformer is not "free gain". TranSistor gain is nearly free anyway. TransFormers have significant bandwidth limits; and low THD is not cheap. Step-up makes it worse.
You have not really specified your POWER level, except "not 600 ohms". 6r? 6Meg?
For "experimental" purpose, one hack is that half-dead Sansui hi-fi in your closet feeding a 12V:120V transformer. Separate-bobbin windings may not beat 1kHz but one-over-the-other power windings I have measured flat to over 15KHz.
Typical output load is 10K.
Most is analog running on +15v -15v
with max rating parts 2x18v.
Only the output is beyond this.
I could compromise to 40V peak differential mode output, using NE 5534 at max rating 2x22v.
Beyond that supply voltage there is just a few op amps.
I have some hope avoiding doubling the supply rails, with a step up transformer.
What is the best way to go ?
Most is analog running on +15v -15v
with max rating parts 2x18v.
Only the output is beyond this.
I could compromise to 40V peak differential mode output, using NE 5534 at max rating 2x22v.
Beyond that supply voltage there is just a few op amps.
I have some hope avoiding doubling the supply rails, with a step up transformer.
What is the best way to go ?
That LL7906 could do it. You could strap input to (1+1) // (1+1) and secondary parallel. Gives you 2:5.6 .
Output load is reflected to primary by square of turns ratio, so that 10k load becomes (2/5.6)^2 is about 0.13 x 10k = 1.3k if I got the numbers right. A good opamp can drive that I think.
But it'll cost you!
Jan
Output load is reflected to primary by square of turns ratio, so that 10k load becomes (2/5.6)^2 is about 0.13 x 10k = 1.3k if I got the numbers right. A good opamp can drive that I think.
But it'll cost you!
Jan
Last edited:
Square of turn ratio, this I know.
What is the strapping that gives 2:5.6 ?
I got it.
What is the strapping that gives 2:5.6 ?
I got it.
Last edited:
I saw the LL7906 around $200...
Is there a way to figure from the specs wether it can drive a 10K load wired for 2:5.6 at the high levels I need and what distortions I would get ?
I know, dc is a no no. What is acceptable ?
Is there a way to figure from the specs wether it can drive a 10K load wired for 2:5.6 at the high levels I need and what distortions I would get ?
I know, dc is a no no. What is acceptable ?
The transformer doesn't have to drive the 10k. It is the circuit at the primary that needs to drive the back-transformed 10k.
The data sheet has distortion figures. But be prepared to see it increase at lower frequencies.
If you are unsure, why not email support at Lundahl?
Jan
The data sheet has distortion figures. But be prepared to see it increase at lower frequencies.
If you are unsure, why not email support at Lundahl?
Jan
I see no information in the specs that can relate to current or power.
It only gives dBu which is voltage and quotes a source impedance, it mentions load, but I cannot find where the value is.
Basically, what fries a transformer is power lost because of imperfect efficiency , a concern in main power transformer unknown here for audio.
At that price, I better ask the swedes.
It only gives dBu which is voltage and quotes a source impedance, it mentions load, but I cannot find where the value is.
Basically, what fries a transformer is power lost because of imperfect efficiency , a concern in main power transformer unknown here for audio.
At that price, I better ask the swedes.
If isolation is required, stick with transformer. Otherwise stack DCDC converters to bump voltage since current requirement is low and opamp/transistor solution is then much lower cost. There would be some HF (80KHz - 120KHz) PSU switching noise but this would probably cancel in a differential application, or could be filtered.