Mixed feedback drive to extend bandwidth of cheap transformer

[Bo Deadly]

I think you should re-read gyraf's post.

Why not just tell me what I'm missing? By giving a coy non-answer it's just kind of like saying "I'm not going to bother giving you an explanation as to why you're wrong because you're a ____".

What evidence do you have of this?

Because it's for a car stereo so it's digital level IO which is a few volts peak at best which means actual nominal levels are going to be a couple hundred millivolts max and glancing over the documentation shows absolutely nothing about signal level even though it lists a THD value.

 

[Bo Deadly]

It will rarely get into those levels. More like 0.5 would be extreme. Like the signal from an electric guitar.

Guitar can put out 4Vpp+ if you jam on it just so.

 

[abbey road d enfer]

Why not just tell me what I'm missing? By giving a coy non-answer it's just kind of like saying "I'm not going to bother giving you an explanation as to why you're wrong because you're a ____".

Again, you're accusing me of a behaviour that is just a figment of your imagination. I suppose it reflects your own behaviour.
Well, since you need a text eplanation, which I'm not the most competent at, being a non english native. Gyraf attracted attention to the fact that medium-high Z xfmrs are often plagued with parasitic capacitance. Then he mentioned the existence of these Monacor xfmrs.
You conclude that he recommends inadequate products.
Your reasoning is: a) many high-Z xfmrs have excessive capacitance
b) the Monacor xfmts are high-Z.
Conclusion: the Monacor xfmrs have excessive capacitance.
This is called consequentialsophism.
I don't know if these Monacor xfmrs are good or not, I say there is a lack of evidence.

Because it's for a car stereo so it's digital level IO which is a few volts peak at best which means actual nominal levels are going to be a couple hundred millivolts max and glancing over the documentation shows absolutely nothing about signal level even though it lists a THD value.

You're basing your conclusions on this absence of data. Another form of sophism.
Again, I don't know, and when I don't know, I try to avoid proffering affirmations that may ashame me later.
I say I try...
I would say that it's the OP's decision to try or not these xfmrs, so he can make a conclusion based on facts, and report, for the benefit of other members.

 

[Studio Mollan]

Like

I would say that it's the OP's decision to try or not these xfmrs, so he can make a conclusion based on facts, and report, for the benefit of other members.

 

[JohnRoberts]

Abbey's contributions here are above reproach, his translation software not so much, but better than most natural english speakers here. 

Let's try to extend a little more respect toward him... he is an asset worth supporting.

JR

 

[Studio Mollan]

I agree. 

Abbey's contributions here are above reproach, his translation software not so much, but better than most natural english speakers here. 

Let's try to extend a little more respect toward him... he is an asset worth supporting.

JR

 

[abbey road d enfer]

Thanks for the support, friends.
This situation is all the more frustrating as squarewave has solid technical knowledge and could be a great contributor if he gave more importance to facts and refrained from basing himself on preconceived ideas.

 

[boji]

All y'all been super helpful and generous to everyone for years.

Love ya.

Jus sayin'  8)

 

[gyraf]

Agree

For back-to-topic, I've looked deeper into the subject

Background:

In the good old days, until ~1995, the Monacor FGA-40 was a nice and easy general-problem-solver, often used in studios with XLR's or jack sockets attached. No, not for ultra-high-levels, but still plenty, and easy to push for some "transformer sound". Phase-correct enough for smpte timecode. I've even used it as bridging input transformer in several prototypes. This one was mechanically characterized by that the plastic tube it was mounted in was split along the length of the cylinder, i.e. axially. I did not weigh it at the time, but it was lighter than its successor -

Then Monacor changed or "upgraded" supplier, and the FGA-40 was now substantially heavier BUT also showed massive cutoff and heavy unnice saturation distortion at low frequencies even at moderate levels. Yes, I complained a lot, but was told I was holding it wrong. Needles to say that I gave up on this part as usable. This model can be recognized on that it had it's body tube split radially, in the centre of and across the width of the containing "tube". Weight was some 104g. Primary and secondary inductance ca. 350-450mH @1KHz. Freq response some 100 to 150KHz, big +8dBbump around 80KHz (200 Ohms source, 100K load, 0dBu)

Now, recently, I came across a picture in a Monacor ad where it looked like the FGA40 housing was split axially - so I needed to check if it was back to usable. Got some of them, and right, the design changed back to axial split. And weight is lower than the bad version, some 66g. Primary and secondary inductance ca. 3.5-4.5H @1KHz. But the best is:

Freq response @+10dBu (200 Ohms source, 100K load) is 17Hz to 80Khz, +5dB bump around 50KHz (begging for R/C)

I get no visible distortion of sinus waveform here at +10dBu@17Hz (!!) - can't be bothered to break out my HP333A for a precise measurement, this being a transformer.

My suspicion is that they changed the core material onto something that gives higher AL factor at lower frequencies, and whatever they used in the "heavy version" didn't have this property.

So I'd consider it worth trying - and at the price - ca. 10EUR for two transformers - it's no big investment. But make sure what version you get, the change is rather new, and old part is so bad that stocks of that may last long..

Note that there is a "high-quality" version also, named "FGA40-HQ" - this is the SAME transformers as the regular version, both for good and bad versions, mounted in a fancier metal housing. Alu, so no magnetic shielding though.

All this for bridging inputs

..While you're at it buying from this strange prosumer electronics company, perhaps check out their LTR110 - usable as output transformer up to 0-perhaps-+4dBu, even has a tertiary secondary for where feedback is needed, and has a really pretty personality when treated right..

/Jakob E.

 

[Studio Mollan]

Thanks Jakob!!!!!
I'll try to get my hands om a couple of these units. A friend of mine sells Monacor stuff so shouldn't be to hard. Great info on the different iterations of it!!

I'll also get with you soon on another matter, My Gyratec X needs service..
/
Emil

Agree

For back-to-topic, I've looked deeper into the subject

Background:

In the good old days, until ~1995, the Monacor FGA-40 was a nice and easy general-problem-solver, often used in studios with XLR's or jack sockets attached. No, not for ultra-high-levels, but still plenty, and easy to push for some "transformer sound". Phase-correct enough for smpte timecode. I've even used it as bridging input transformer in several prototypes. This one was mechanically characterized by that the plastic tube it was mounted in was split along the length of the cylinder, i.e. axially. I did not weigh it at the time, but it was lighter than its successor -

Then Monacor changed or "upgraded" supplier, and the FGA-40 was now substantially heavier BUT also showed massive cutoff and heavy unnice saturation distortion at low frequencies even at moderate levels. Yes, I complained a lot, but was told I was holding it wrong. Needles to say that I gave up on this part as usable. This model can be recognized on that it had it's body tube split radially, in the centre of and across the width of the containing "tube". Weight was some 104g. Primary and secondary inductance ca. 350-450mH @1KHz. Freq response some 100 to 150KHz, big +8dBbump around 80KHz (200 Ohms source, 100K load, 0dBu)

Now, recently, I came across a picture in a Monacor ad where it looked like the FGA40 housing was split axially - so I needed to check if it was back to usable. Got some of them, and right, the design changed back to axial split. And weight is lower than the bad version, some 66g. Primary and secondary inductance ca. 3.5-4.5H @1KHz. But the best is:

Freq response @+10dBu (200 Ohms source, 100K load) is 17Hz to 80Khz, +5dB bump around 50KHz (begging for R/C)

I get no visible distortion of sinus waveform here at +10dBu@17Hz (!!) - can't be bothered to break out my HP333A for a precise measurement, this being a transformer.

My suspicion is that they changed the core material onto something that gives higher AL factor at lower frequencies, and whatever they used in the "heavy version" didn't have this property.

So I'd consider it worth trying - and at the price - ca. 10EUR for two transformers - it's no big investment. But make sure what version you get, the change is rather new, and old part is so bad that stocks of that may last long..

Note that there is a "high-quality" version also, named "FGA40-HQ" - this is the SAME transformers as the regular version, both for good and bad versions, mounted in a fancier metal housing. Alu, so no magnetic shielding though.

All this for bridging inputs

..While you're at it buying from this strange prosumer electronics company, perhaps check out their LTR110 - usable as output transformer up to 0-perhaps-+4dBu, even has a tertiary secondary for where feedback is needed, and has a really pretty personality when treated right..

/Jakob E.

 

[Bo Deadly]

Freq response @+10dBu (200 Ohms source, 100K load) is 17Hz to 80Khz, +5dB bump around 50KHz (begging for R/C)

I get no visible distortion of sinus waveform here at +10dBu@17Hz (!!) - can't be bothered to break out my HP333A for a precise measurement, this being a transformer.

My suspicion is that they changed the core material onto something that gives higher AL factor at lower frequencies, and whatever they used in the "heavy version" didn't have this property.

So I'd consider it worth trying - and at the price - ca. 10EUR for two transformers - it's no big investment. But make sure what version you get, the change is rather new, and old part is so bad that stocks of that may last long..

Note that there is a "high-quality" version also, named "FGA40-HQ" - this is the SAME transformers as the regular version, both for good and bad versions, mounted in a fancier metal housing. Alu, so no magnetic shielding though.

All this for bridging inputs

..While you're at it buying from this strange prosumer electronics company, perhaps check out their LTR110 - usable as output transformer up to 0-perhaps-+4dBu, even has a tertiary secondary for where feedback is needed, and has a really pretty personality when treated right..

Interesting. That is pretty good. Great find and thanks for digging into it to figure out that it's not junk like one (I) would expect it to be. Still not sure about using it as an output transformer (the LTR-110 looks interesting as an OT) but it would make a decent input transformer.

 

[gyraf]

Yup, agree on in & out purposing

A place where the (original) FGA40 performed really astonishing was in a quasi-zero-field'ish setup as a transformer-isolated tube-based mixbus amplifier i once did (the now long gone G16). This one mixed a lot of jazz over the years..

Picture related.

Yes, the 1K5 should be 470R.

/Jakob E.