Q factor RLC PSU

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Member
Joined 2009
Paid Member
Attached pic.
 

Attachments

  • IMG_20201222_161131.jpg
    IMG_20201222_161131.jpg
    732 KB · Views: 91
There is a thread on diyaudo that measures regulation of a 15+15V 30VA R-core - I'd suggest you change the transformer winding resistance to about 5 ohm.

If you modify the load to a step load (eg. change from 280mA to 100mA, or vice-versa, from at 10 secs), and change the reporting to 9 sec delay and show 3 secs, then you will see no 'oscillations' - ie. the CLC filter is over-damped.
 
Member
Joined 2009
Paid Member
There is a thread on diyaudo that measures regulation of a 15+15V 30VA R-core - I'd suggest you change the transformer winding resistance to about 5 ohm.

Why? Do you refer to source resistance? The source resistance is low 135 miliohms. Do you want to say that I add a 5 ohm resistor between the power transformer & rectifiers?
 
Last edited:
PSUD2, like vintage valve rectifier datasheets, uses an equation to calculate the effective secondary winding resistance. PSUD2 includes a calculator where the primary winding voltage and resistance, and the secondary winding voltage and resistance, can be inserted, and effective secondary resistance calculated (click on the "..." button on the right side of the transformer part value input pop-up window).

The diyaudio thread is linked here: Selectronic Rcore transformer
It shows about 7 ohm of series resistance for the 15+15 windings. On review, your 9+9 windings are in parallel, so I'd suggest using 5//5 = 2.5 ohm. The present value showing in your sim circuit is 0.135 ohm.
 
Last edited:
Member
Joined 2011
Paid Member
How inaccurate are the results, when you measure primary resistance using a variable DC power supply (with current limiting!), a DC ammeter, a DC voltmeter, and Ohm's Law? Wrong by 5%? Wrong by 35%? Is it any better or any worse if you try the same setup to measure secondary resistance?

And for the benefit of the thread starter on "Selectronic Rcore transformer", linked above, here is one possible way to avoid the error

 
Member
Joined 2009
Paid Member
PSUD2, like vintage valve rectifier datasheets, uses an equation to calculate the effective secondary winding resistance. PSUD2 includes a calculator where the primary winding voltage and resistance, and the secondary winding voltage and resistance, can be inserted, and effective secondary resistance calculated (click on the "..." button on the right side of the transformer part value input pop-up window).

The diyaudio thread is linked here: Selectronic Rcore transformer
It shows about 7 ohm of series resistance for the 15+15 windings. On review, your 9+9 windings are in parallel, so I'd suggest using 5//5 = 2.5 ohm. The present value showing in your sim circuit is 0.135 ohm.

9+9 windings in series = 10 ohm?
 
Member
Joined 2009
Paid Member
How inaccurate are the results, when you measure primary resistance using a variable DC power supply (with current limiting!), a DC ammeter, a DC voltmeter, and Ohm's Law? Wrong by 5%? Wrong by 35%? Is it any better or any worse if you try the same setup to measure secondary resistance?

And for the benefit of the thread starter on "Selectronic Rcore transformer", linked above, here is one possible way to avoid the error


Thanks Mark, asked the manufacturer so I'm sure is well connected.
 
9+9 windings in series = 10 ohm?

The 15+15V 30VA windings in series appear to be about 7 ohm (post #24 indicates 7Vrms droop for 1Arms loading).

9+9V 30VA windings would be fewer turns, with a higher current rating, so likely to be lower resistance than 7 ohm. Assuming 5 ohm as a series resistance (or even less), then doing some mental arithmetic without the alcohol, the two 9V windings in parallel would be 2.5//2.5 = 1.25 ohm.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.