Re: NEOTEK: Adding Transformers

Thu, 24 Mar 2011 02:52:00 -0700 

Hi
When simply adding a transformer on an output you need also to consider the output capacitor, assuming the circuit uses them which many do. This will cause a series resonant circuit with the transformer and it it possible this resonance happens in the audio band, or nearby. High impedance 'line in' transformers of remote connected gear suffer less but it is quite easy to get a 3dB or more 'boost' at some frequency. Increasing the output capacitor WAY beyond usual values usually 'cures' this. Values in excess of 1000 uF may be necessary. I had this discussion a while back with a guy using a 600 Ohm 1:1 transformer (Sowter but the principle is valid for any similar types) and with a 47uF cap (as used in the suggested circuit) it gave a 4dB boost at about 20Hz, hardly an accurate response. (I can't remember the exact values now but it is an example). Gratifyingly doubling the cap halved the boost frequency, as 'laws' predict. 
Matt S

On 23/03/2011 23:35, Jim wrote:


Hi Tom,
Sorry I didn't explain this..............I had posted about transformers previously and assumed you saw that. Please keep in mind that entire books have been written on these topics. Anway, to maintain its rated frequency response a transformer requires some specific secondary termination, usually an R, sometimes an RC. In many cases this is achieved by the input Z of the component that its feeding. Without going into too much detail and history, back in the "old days" equipment was designed to a standard 600 Ohm line (matching load); all inputs and outputs were "matched" so that anything could be connected to anything else and it would all work out just fine.
Since we've pretty much abandoned that approach, we run into the situations of non-matching loads; which can adversely affect both frequency response and the optimum transfer of voltage/power/etc.. Now we tend to think in terms of bridging loads; which is to say that the optimal situation would be an output impedance of zero feeding an infinite input impedance. Of course in the real world that situation does not exist. In general, we'll figure the output Z is pretty low (a few Ohms) and the input Z is high (say at least 1K Ohms). If you have a scenario where the ratio is on the order of 1:1000 or greater, things can work out pretty well. The problem comes when the output Z is high and the input Z is low. The most common example of this is plugging a bass guitar (with passive pick ups) directly into a console. The relatively high output Z of the pick up is severely loaded down by the relatively low input Z of the console. That's why we usually would use some type of direct box that matched the correct loading for the pick up; for example plug the same bass into an old tube bass guitar amp. Its going to sound much better. Typical input Z will be over 1MR, sometimes up over 5MR.
Now lets look at your specific situation. The UTC is likely rated to drive into a 600R load (see if you can find the data sheet for it; there may be several ways to strap the input and output side for some other combinations; 150R, 300R, etc.). The actual load will be determined by what the input Z of the device it is feeding. Say for example, it's feeding a solid state line level input device. That may have an input Z of between 1K and 10K; feeding a tube line level device may have an input Z of over 100K. So in either case a bridging load is in effect, but at pretty different ratios. Hook the transformer up "as is" or with the strapping set for 600:600. Then after you have measured/listened, etc. try installing a resistor in parallel with the transformer secondary; for example, try a 500R pot with just one end and the wiper connected. That puts the resistor in parallel with the input Z of the loading device, and will result in dropping the Z the output transformer sees. This will likely result in lower output level, but the interesting part is how it may affect the frequency response. So, this allows you to "tune" the circuit to perhaps achieve a beneficial effect. You don't have to limit yourself to the value I gave; try whatever you have; listen and let your ears be the judge.
OK, hope that helps and that I didn't make too many errors. Just coming off the flu myself. Good luck! 

Regards, Jim


At 01:22 PM 3/23/2011, you wrote:

Hi Jim,
What do you mean exactly? Try using the different impedance taps on the UTC's? Sorry, but I think I might be on the verge of a nasty cold! 

Tom

--- On Wed, 3/23/11, Jim <[email protected]> wrote:

From: Jim <[email protected]>
Subject: Re: NEOTEK: Adding Transformers
To: [email protected]
Date: Wednesday, March 23, 2011, 12:10 PM

  Hi Tom,

After you get the project built up and do some testing, if you have time
mess around with the secondary terminations. I'd be very interested to find 
out what happens.

Regards, Jim




At 12:40 PM 3/23/2011, you wrote:
>I received a couple UTC A20's and should have my Edcor's in tomorrow. I
>should have the chassis made and wired by late weekend. I am very curious >to hear the differences and see if they are very noticable. Regardless, it 
>will be a fun project.
>
>Tom
>
>--- On Wed, 3/16/11, Brian Roth <<http://us.mc657.mail.yahoo.com/mc/[email protected]>[email protected]> wrote: 
>
>From: Brian Roth <<http://us.mc657.mail.yahoo.com/mc/[email protected]>[email protected]> 
>Subject: Re: NEOTEK: Adding Transformers
>To: <http://us.mc657.mail.yahoo.com/mc/[email protected]>[email protected] 
>Date: Wednesday, March 16, 2011, 10:11 PM
>
>On 3/16/2011 13:02, Ike Zimbel wrote:
> >
> > Hi Brian,
> > I'm pretty sure you published those specs on the Ampex list...check the 
> > archive!
> > All the best,
> > Ike
>
>
>Dear Ike,
>
>I forgot I had posted some measurements on the Ampex list!  I found the
>following from the Jan, 2005 archives, which I believe were measurements 
>of the "smaller" series of Edcors:
>
><<http://www.edcorusa.com/products/143-wsm600-600.aspx>http://www.edcorus a.com/products/143-wsm600-600.aspx>http://www.edcorusa.com/products/143-wsm600-600.aspx 
>
>Here is what I wrote 6 (!) years ago:
>
>The replacement Edcors (600:600 impedance versions) arrived today, and the 
>DCR was MUCH more reasonable.  Approx. 54.1 Ohms primary, 54.5
>secondary. Level drop from open loading on secondary to 600 Ohms was just 
>a bit under 1.5 dB, much what I would expect.
>
>For testing, I breadboarded a "zero Ohm source" driver with a 5534 opamp >running as an inverter: 10K input and feedback R's, 22 pF compensation cap 
>(between pins 5 and 8) and 22 pF in parallel
>with the  feedback R to tame any overshoot and ringing from the opamp
>itself. -3 dB response with the opamp circuit calcs to around 723 kHz. I 
>powered the 5534 with +/- 20 VDC from my bench
>supply, which allowed just over approx + 24 dBu/m at clipping.
>
>1 and 10 kHz squarewave response through the iron looked very good, loaded 
>or unloaded, with just a tiny bit of overshoot and ringing. Frequency
>response was beyond anything I could measure with my
>Amber (bandlimited at around 500 kHz or so) which came up with -1 dB at
>approx 6 Hz and 300 kHz. Bypassing the Amber and just watching the scope, 
>usuable response was extended, and limited by the
>rolloff of the opamp driver on the top end.  With a 50 Hz square wave,
>there was very little tilt which indicates LF response well below 20 Hz. 
>
>Next I measured THD, and at some of the lower levels I was fighting noise >picked up from the breadboarding! Here are various "spot measurements". 
>
>1 kHz, 0 dBm = 0.0085% +10 dBm = 0.005% +20 dBm = 0.0025% (noise was 
>an obvious factor!)
>
>100 Hz, 0 dBm = 0.03%   +10 dBm = 0.015%  + 20 dBm = 0.05%  **
>
>50 Hz, 0 dBm =  0.055%   +10 dBm = 0.035%  +20 dBm = 0.18%  **
>
>20 Hz, 0 dBm = 0.07%   +10 dBm = 0.085%  + 20 dBm = 0.7%
>
>Spot checks at LF and + 20 dBm:  40 Hz = 0.3%  30 Hz = 0.45%
>
>(Actual 20 Hz numbers may be a bit lower since my 20+ year old Amber was 
>having a hard time nulling at 20 Hz).
>
>**  Notice the THD dip at the "middle" levels at those frequencies.  I
>also think I detected a THD rise at levels below 0 dBm but was fighting
>noise issues.
>
>I have a bunch of other numbers, but you get the drift. Looking at Jensen >choices, this Edcor is just a bit "rattier" than a JT-11-EMCF which sells 
>for $66.85.  These Edcors are a bargain!!
>
>Bri
>
>
>
>------
>
>View the list online at:
><<http://www.mail-archive.com/[email protected]>http://www.mail-archi ve.com/[email protected]>http://www.mail-archive.com/[email protected] 
>
>You can unsubscribe at anytime by sending an email to
>mailto:[email protected] with "unsubscribe neotek" in the message
>body. If you have any difficulties, or questions --please send an email to 
>mailto:[email protected].
>
>The Neotek list is lovingly hosted by Catapult Web. Remember that any
>message or reply sent to the list will be automatically forwarded to all 
>subscribers.
>
>Vive Le Neotek!
>
>Thanks.
>


------
View the list online at: <http://www.mail-archive.com/[email protected]>http://www.mail-archive.com/[email protected]
You can unsubscribe at anytime by sending an email to mailto:[email protected] with "unsubscribe neotek" in the message body. If you have any difficulties, or questions --please send an email to mailto:[email protected].
The Neotek list is lovingly hosted by Catapult Web. Remember that any message or reply sent to the list will be automatically forwarded to all subscribers. 

Vive Le Neotek!

Thanks.




------
View the list online at: http://www.mail-archive.com/[email protected]
You can unsubscribe at anytime by sending an email to mailto:[email protected] with "unsubscribe neotek" in the message body. If you have any difficulties, or questions --please send an email to mailto:[email protected].
The Neotek list is lovingly hosted by Catapult Web. Remember that any message or reply sent to the list will be automatically forwarded to all subscribers. 

Vive Le Neotek!

Thanks.



No virus found in this incoming message.
Checked by AVG - www.avg.com
Version: 9.0.894 / Virus Database: 271.1.1/3524 - Release Date: 03/23/11 
07:34:00

Reply via email to