Mic TX winding variations

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

clintrubber

Well-known member
Joined
Jun 3, 2004
Messages
5,982
Location
The Netherlands
Hi,

With all those various Beyer Dynamic TX's upcoming,
I hope this may be a good opportunity to get a few things more clear
and figured it deserved a separate thread.

It's about TX-winding structures.

The 4th digit of the Beyer TR/BV-code system denotes the winding-configurations
of the primary and I'm wondering about the virtues & possibilities this can give.

BD_TX-prim_mq.jpg


'0' denotes a single primary
'1' center tapped primary
'2' two split bifilary wound primaries
'3' two split bifilary wound primaries, end of winding 1
connected to beginning of winding 2 and led out as tap
'4' two split wound primaries

http://home.hetnet.nl/~chickennerdpig/FILES/BD_4th_mq.jpg

So the only difference between '1' & '3' looks to be the bifilar-winding of the latter.

And alike for the difference between '2' & '4': the former is bifilar, the latter is not.


I've read the various info around (here, RDH, http://en.wikipedia.org/wiki/Bifilar
etc etc), but some remaining questions are:

#1
I'm wondering if there are any advantages of non-bifilary windings ?
In other words, why do Beyer even make the non-bif.types '1' & '4' ?

#2
Can bif-wound split primaries be put in parallel just like 'normal'
split primaries
?
(so resulting in a fourfold ratio-change compared
to the situation with both primary windings in series)


Thanks !

Peter
 
> Can bif-wound split primaries be put in parallel just like 'normal' split primaries?

Yes.

> any advantages of non-bifilary windings?

Lower capacitance. This is probably meaningless at 200 ohms, but will matter above a few K ohms.
 
Thanks PRR, much appreciated. That Beyer Dynamic TX grouporder left me wondering about these things, but now not longer so.

So... hey people, why keep it simple ? Why not ordering something else (w.r.t. prim-config) than just a single prim winding or two non-split primaries ? The resulting config most often used will of course be just that, something alkie a single prim... but while ordering anyway, you could also have some alternatives to toy around with.

Thanks,

Peter
 
Hi Peter and others,

I took your advice and have bought my first piece of test equipment - an audio signal generator.

Anyway, I decided to test a 1:5 bifilar Beyer transformer using my audio signal generator (at low frequency) and my DMM.

So, assuming you aren't screaming you bloody idiot, I will carry on....

I thought I'd test to check the voltage ratio using the transformer with a single primary and with both primaries in series.

The results do not make sense though...

My audio signal generator has a 600 Ohm output. I connected it to the input of the transformer and set it until I read 50mV on my voltmeter at the primary at 100Hz. I chose 100Hz since my multimeter would be happy with this frequency.

Now, I moved my DMM probes over to the secondary....

For both primaries in series, I got a reading of 750mV - this equates to a volatge ratio of 1:15 - that doesn't sound right.

Next, I decided to connect the signal generator to primary 1 only. Again, I measured the voltage over the primary at 50mV.

At the secondary I measured 1050mV which equates to a voltage ratio of 1:21 - WTF is going on?

I imagine I have made a simple mistake which I will no doubt be very embarrassed about. Is it an impedance issue which I have overlooked? Am I loading down my multimeter accidentally somewhere?

Any help would be much appreciated!
 
Before we all put our thinking caps on, try again with a higher input voltage. Most DMMs won't read such a low AC voltage with any accuracy worth a damn. And with the long unshielded probes supplied with most DMMs, you're picking up many microvolts of noise as it is. Although it may be approaching the saturation point of the xfmr, repeat your measurements with an input level of a couple hundred millivolts and let us know if anything changes.

When you say "1:5", do you mean the ratio of either primary winding to the secondary? Or the ratio of the entire primary, in series, to the secondary?
 
[quote author="NewYorkDave"]repeat your measurements with an input level of a couple hundred millivolts and let us know if anything changes.[/quote]

Good idea! I will do that now...

When you say "1:5", do you mean the ratio of either primary winding to the secondary? Or the ratio of the entire primary, in series, to the secondary?

That is partly why I am making the experiment. Peter and I were discussing whether Beyer would call the transformer a 1:5 or a 1:10 depending on the windings. I think he cleared that up, but I thought I would test it anyway.
 
Ok, I've conducted some more experiments @ 200mV, 100Hz.

I do have a question though: when I set my signal generator voltage, should I set it whilst it is under load (ie. connected to the transformer primary? Open circuit at 200mV, the reading falls to ~ 50 mV when under load....

Anyway, I have done two sets of tests; one where the sig. gen is 200mV opencircuit, and one where the sig. gen is connected to the transformer primary and adjusted to 200mV.

Sig. Gen. 200mV @ open circuit -

single primary: 1435mV at secondary (1:7)

dual primaries in series: 1100mV at secondary (1:5.5)



Sig. Gen. 200mV when loaded with transformer primary -

single primary: 3200mV at secondary (1:16)

dual primaries in series: 1900mV at secondary (1:9.5)


I've cocked it up again, haven't I?
 
Try it reversed and see what you get.
That is, hook the sig gen up to the seocndary.
Perhaps the increae in inductive eactance will not load down your generator as much.
A scope is nice to have here, as you can see if the waveform you are measuring is distorted or not.
If you have a goofy wavefrom, a true rms meter might make a difference. If this is a sensitive transformer, the core might not be engineered for a high input signal, especially at 100 hz.

try it with the generator feeding the pri again, only with the genrator set at 1 k hz and see if there is any change.
 
Thanks for the advice, Chris.

Sadly I am still getting crazy-ass results.

At 1KHz with 10mV on the primary, I get a reading of 330mV!

In reverse, with a 330mV input to the secondary I get 13mV on the primary...

Basically a load of nonsense.

I checked my meter and it is flat from about 50Hz (wavering) up to about 4KHz where it loses about 10%.

I'll try again tomorrow when I don my thinking cap. Maybe it is time for a 'scope too.....
 
You may have multiple problems. DVM won't read low volts. At 100Hz the transformer is both low-Z and easily overloaded.

First: you don't measure mike inputs with a naked bench oscillator. Even if it says it goes down to milliVolts, I wouldn't trust it. Also it is apparently a true 600 ohm output, which is much higher than a mike.

Build 10:1 and maybe 100:1 pads with low-Z output. 560Ω and 62Ω gives a roughly 600Ω input, 57Ω output, 10:1 ratio with unloaded output. Now you can put a whopping 1V in, and get a known 100mV out. And the output is lower than most mikes, which you can use as-is or pad-up when you want to know the effect of source impedance.

If the oscillator goes toward 10V, a 100:1 pad will be useful. 1K and 10Ω, with maybe another 1K8 across the oscillator output if it is fussy.

To determine ratio, use a zero-Z source or measure the actual input voltage. Since mV readings are dubious without a true milliVoltmeter, use the low-Z pad as a work-around.

The backward reading should be fine: the secondary should be much greater than 600 ohms across the audio band. You may have trouble reading the low primary voltage. It would be good to have a small flat amplifier, gain of 10 or 100. Even better to have a speaker too: your ear will quickly tell you if the harmonics are strong. Even a dead-ear like me can hear a pure sine sweeten-up as the iron goes limp.

> I checked my meter and it is flat from about 50Hz (wavering) up to about 4KHz where it loses about 10%.

If it has chips, its response probably depends on level. It may be very flat at 1.9V, and pretty lame at 50mV.
 
Hi Roddy,

Nice you brought this up.

[quote author="rodabod"]
Maybe it is time for a 'scope too.....[/quote]
Always good to have a scope, but for the signal-levels at hand here you could also get some more info about what's going on from some scope-computer-program. You might need to put your mixer between the TX-out & the soundcard-in for some level & impedance-help.

Sadly I am still getting crazy-ass results.
Please don't put it like this... it instantly reminds me of those kilt-lifting pics I saw from NewYears-eve in your very town... :? :wink:

You bought these TXs used, right ? You got two identical types I understood, do they behave both weird and alike ?

So far it sounds like some measurement-thingy indeed, but if they keep acting funny could it be that a previous owner mistook them for some gapped 80-mA Neve-output-iron & goofed them ?

So if you have two of those TR/BV's, are they behaving alike ?

And just in case, how's your signal generator itself ?(congrats on that one BTW) Is it beyond doubt ? Undistorted sine or giving a square where you expect a sinewave? Or some internally rewired freq-mult-switch giving way other frequencies ? Not jumping from level to level ? OK, enough, sometimes you just have to trust something...

Bye,

Peter
 
[quote author="PRR"]You may have multiple problems. DVM won't read low volts. At 100Hz the transformer is both low-Z and easily overloaded. [/quote]

I had a feeling this was maybe the case since my results were un-linear...

First: you don't measure mike inputs with a naked bench oscillator. Even if it says it goes down to milliVolts, I wouldn't trust it. Also it is apparently a true 600 ohm output, which is much higher than a mike.

How is it much higher than a mic? I don't see how a true 600 Ohm output would vary much from most mics? Am I missing something.

Build 10:1 and maybe 100:1 pads with low-Z output. 560Ω and 62Ω gives a roughly 600Ω input, 57Ω output, 10:1 ratio with unloaded output. Now you can put a whopping 1V in, and get a known 100mV out. And the output is lower than most mikes, which you can use as-is or pad-up when you want to know the effect of source impedance.

Ok, this sounds like a good idea. The oscillator actually has a -20dB pad, but this doesn't affect impedance....

So, is the point of the pad just to allow me to set the output voltage more accurately and to lower the output impedance?

If the oscillator goes toward 10V, a 100:1 pad will be useful. 1K and 10Ω, with maybe another 1K8 across the oscillator output if it is fussy.

The oscillator can only kick out 1.2V sin, or 10V square, so a 10:1 might be suitable.

To determine ratio, use a zero-Z source or measure the actual input voltage. Since mV readings are dubious without a true milliVoltmeter, use the low-Z pad as a work-around.

I could use my sound-card output which has an unbalanced opamp output - I assume this has a zero-Z output impedance.

The backward reading should be fine: the secondary should be much greater than 600 ohms across the audio band. You may have trouble reading the low primary voltage. It would be good to have a small flat amplifier, gain of 10 or 100. Even better to have a speaker too: your ear will quickly tell you if the harmonics are strong. Even a dead-ear like me can hear a pure sine sweeten-up as the iron goes limp.

Yeah, that's what I thought, but it seems my DMM just talks rubbish. An amplifier of 10X would be a nice idea. You always talk about your worn-out ears.... They're never that bad. I've damaged my badly in the past, but your brain compensates. I don't care too much for the stuff over 16K anyway.

If it has chips, its response probably depends on level. It may be very flat at 1.9V, and pretty lame at 50mV.

Fair point, although I tested it down to 100mV and it is actually surprisingly flat. Still, my rig is currently not up to the job.

Can anyone explain why they gave this sig. gen. an output impedance of 600 Ohms? I know it is a fairly standard source impedance, but would 0 Ohms not have been more handy?

So, I will look into making a pad which will lower the output impedance. Does anyone have any links? I'm assuming a simple voltage divider would be a little too simple!

Thanks for your help.





[quote author="clintrubber"]

[quote author="rodabod"]
Maybe it is time for a 'scope too.....[/quote]
Always good to have a scope, but for the signal-levels at hand here you could also get some more info about what's going on from some scope-computer-program. You might need to put your mixer between the TX-out & the soundcard-in for some level & impedance-help. [/quote]

Good suggestion. I might wait until I really need to use the computer as they always cause hassle!

[quote author="Peter"]
Sadly I am still getting crazy-ass results.
Please don't put it like this... it instantly reminds me of those kilt-lifting pics I saw from NewYears-eve in your very town... :? :wink: [/quote]

Please don't remind me. :razz:

You bought these TXs used, right ? You got two identical types I understood, do they behave both weird and alike ?

Hmmm. I better check....

I bought them from the same guy that is selling the ones for the group-buy. They are new.... But 15 years old.

And just in case, how's your signal generator itself ?

It is cheap.... Just like me. :razz:

Here it is:

http://www.mcmb2b.com/cgi-bin/test/72-505.html?id=Q785sEIm

It needed to be small since it needs to live in my bedroom which covered in microphones and electrical components!

I thought it seemed simple, but OK. Is there anything wrong with it do you think?
 

Latest posts

Back
Top