No load voltage readings with a toroid PT

joulupukki

Tele-Meister
Joined
Nov 26, 2020
Posts
476
Location
Utah
I've been thinking of building a RR763 6V6 flavor but doing it with toroids for the PT and OT. It looks like I should be shooting for a B+ voltage of 420 V. So, I did an experiment with the AS-1T300 PT I have. I hooked it up to power and, unloaded, connected my multimeter to the DC output leads. Using both of the 300V secondaries wired in series with a bridge rectifier (4 1N4007 diodes), I measured: 686 V of DC output! Yowsah!

1662738879258.png


Wired with a center tap connection and 2 diodes (which is kinda how I had this wired up in my 18w TMB build ... after the diodes it went into the EZ81 rectifier tube) I get a reading of 394 V.
1662740821470.png


If I use the 4-diode bridge rectifier on the different 300V secondary pairs I get:

340V for one pair
363V for the other pair

So, I'm wondering, maybe I should get their 200V x 2 PT? The Antek AS-05TC200? If I connect the two outputs in series and use a bridge rectifier and also use a thermistor on the AC primary to control inrush current (which drops about 5 volts of AC before the PT's primary) I'm thinking that may put it right in the 420 - 430 V B+ range. Am I thinking straight on this?
 

Phrygian77

Poster Extraordinaire
Joined
Apr 30, 2016
Posts
6,184
Location
Crawfordville, FL
Haven't we covered this in another thread? The most efficient way to use one of those transformers is to parallel the secondaries and use a bridge rectifier. A regular full wave rectifier creates more ripple current, so you end up needing a bigger transformer than necessary.
 

joulupukki

Tele-Meister
Joined
Nov 26, 2020
Posts
476
Location
Utah
Haven't we covered this in another thread? The most efficient way to use one of those transformers is to parallel the secondaries and use a bridge rectifier. A regular full wave rectifier creates more ripple current, so you end up needing a bigger transformer than necessary.
Probably my ignorance of not understanding it all. I'm trying to figure out which toroid PT would give me a B+ of 420V and that's when I started wondering what type of wiring and which PT would work.
 

andrewRneumann

Friend of Leo's
Joined
Mar 22, 2020
Posts
2,153
Location
Cincinnati, OH, USA
You are saying unloaded, but you draw a resistor in your diagrams. Unloaded voltage is easily measured in AC with nothing attached to the secondary except your DMM. Unloaded DC will be AC x 1.414 minus 2 diode drops for bridge rectification.
 

Phrygian77

Poster Extraordinaire
Joined
Apr 30, 2016
Posts
6,184
Location
Crawfordville, FL
The AS-1T300 is 310VAC unloaded and 281VAC at .47A with the secondaries in parallel, according to their datasheet. So, subtract the loaded voltage from the unloaded voltage and you have a voltage drop of 29VAC across the secondary with a load of .47A. Using ohms law, R = V / I, the source resistance (Rs) is approximately 61.702 ohms. Now you can use that to model the power supply in LTspice, Falstad, etc.
 

andrewRneumann

Friend of Leo's
Joined
Mar 22, 2020
Posts
2,153
Location
Cincinnati, OH, USA
I connected it to AC power and put my DMM at the following spots, measuring in DC:
View attachment 1026585

Yes I believe we have covered this ground before. Maybe not with you. With no capacitors you are not measuring what DC will be. You are measuring the average DC of a full-wave rectified signal with no smoothing. It is quite a bit less than what the smoothed voltage will turn out to be.

Ditch the rectifier, measure the AC and multiply by 1.414 and be done.
 

joulupukki

Tele-Meister
Joined
Nov 26, 2020
Posts
476
Location
Utah
You are saying unloaded, but you draw a resistor in your diagrams. Unloaded voltage is easily measured in AC with nothing attached to the secondary except your DMM. Unloaded DC will be AC x 1.414 minus 2 diode drops for bridge rectification.
These drawings are from Antek that they sent when I was asking them how to wire this PT up. I don't actually have filter capacitors or a resistor hooked up on this breadboard.
 

joulupukki

Tele-Meister
Joined
Nov 26, 2020
Posts
476
Location
Utah
Yes I believe we have covered this ground before. Maybe not with you. With no capacitors you are not measuring what DC will be. You are measuring the average DC of a full-wave rectified signal with no smoothing. It is quite a bit less than what the smoothed voltage will turn out to be.

Ditch the rectifier, measure the AC and multiply by 1.414 and be done.
Ah, see ... this is why I ask. #1 to look not-so-smart and #2 to fix #1. :)
 

andrewRneumann

Friend of Leo's
Joined
Mar 22, 2020
Posts
2,153
Location
Cincinnati, OH, USA
The AS-1T300 is 310VAC unloaded and 281VAC at .47A with the secondaries in parallel, according to their datasheet. So, subtract the loaded voltage from the unloaded voltage and you have a voltage drop of 29VAC across the secondary with a load of .47A. Using ohms law, R = V / I, the source resistance (Rs) is approximately 61.702 ohms. Now you can use that to model the power supply in LTspice, Falstad, etc.

Or just read the unloaded voltage off the datasheet instead of measuring it! :p Good one @Phrygian77
 

Phrygian77

Poster Extraordinaire
Joined
Apr 30, 2016
Posts
6,184
Location
Crawfordville, FL
@joulupukki another thing that throws people off, besides measuring unfiltered rectified DC, is measuring the loaded AC voltage. The DC load with rectification and filtering distorts the AC waveform (the transformer output), so you need a meter that can measure true RMS AC voltage.
 

2L man

Tele-Afflicted
Joined
Nov 23, 2020
Posts
1,785
Age
62
Location
Finland
The AS-1T300 is 310VAC unloaded and 281VAC at .47A with the secondaries in parallel, according to their datasheet. So, subtract the loaded voltage from the unloaded voltage and you have a voltage drop of 29VAC across the secondary with a load of .47A. Using ohms law, R = V / I, the source resistance (Rs) is approximately 61.702 ohms. Now you can use that to model the power supply in LTspice, Falstad, etc.
Some Antek datasheets list output voltages for three currents which first is zero. Using them it is possible to draw a rough load curve. Obviously Antek measure output voltages using resistive load and they are not directly applicaple for rectified power supplys.

You have calculated Rs and that is good value when rectified sine wave charge capacitors and deliver current. But finding "correct" point from "curve" must be interpreted or simulated using Spice?

In practice peak voltage drop lower because of rectifier diodes because there is no current flow thru them when PT output turns lower than the filter capacitor voltage and only peaks charge capacitors but some of the live current pass to following curcuits as well.
 

2L man

Tele-Afflicted
Joined
Nov 23, 2020
Posts
1,785
Age
62
Location
Finland
I connected it to AC power and put my DMM at the following spots, measuring in DC:
View attachment 1026585
If there are no filter capacitors DC voltage after SS bridge should be about one volt lower than AC voltage. Some multimeters do not measure pulsating DC reliably and full wave rectified pulsates when there are no filter capacitors. This about one volt lower voltage comes from two bridge silicon diode voltage loss which js about 0,5V each when current is 0.

But when it is filtered zero current comes about 1,41 times AC using even low cost voltmeter.
 

joulupukki

Tele-Meister
Joined
Nov 26, 2020
Posts
476
Location
Utah
He can find out how accurate the datasheet voltage is by measuring his mains voltage also. Since they listed 310 output with 120 input, if his mains voltage is 124 for example, 124/(120/310)=320+1/3.
Within 1.2% 122 VAC is getting transformed into 319 VAC.

I'm trying to learn/understand everything being said in here, but it's coming slow.

So ... if I were to use a bridge rectifier on this amp instead of the GZ34 tube rectifier, I'm assuming I'd need to have the main filter capacitors be larger, say around 100uF instead of 40uF? Rob has two 80uF 350V caps in series for a resulting 40uF @ 700V.
1662755955178.png
 

Lowerleftcoast

Poster Extraordinaire
Joined
Dec 30, 2019
Posts
6,022
Location
california
Rob has two 80uF 350V caps in series for a resulting 40uF @ 700V.
Chances are you would want to beef up the reservoir cap value like rob suggests. He shows two 80uF, two 100uF or two 220uF would be fine. The bleeder resistors in series can be from about 100k to 220k.

1662743074095.png

Of course if you find capacitors that are rated for voltage higher than the B+, you do not need to use capacitors in series in this location. For example, a single 100uF 475v could replace the two caps and two resistors shown above.
 
Last edited:

Phrygian77

Poster Extraordinaire
Joined
Apr 30, 2016
Posts
6,184
Location
Crawfordville, FL
You don't need to increase the filtering, but you can if you want to since the transformer is slightly oversized for the load you're going to have.

The one thing you haven't mentioned is the bias supply. Since you won't have a center tap, and you'll be using a bridge rectifier, you can't just hang a reverse biased diode off the secondary. You're going to have to create a capacitor coupled bias supply.

cap_coupled_bias_supply2.jpg
 




Top