TOO many volts ????!!!!

[cbutterworth]

I finally got the PSU and relay network wired in my Aikido.

Being as methodical as I could, I checked the switch (more or less 120VAC). Then the tube delay relay network, which worked and gave more or less 120VAC.

Then I wired my Hammond 125-0-125 transformer and trimmed to center tap to avoid problems. From the 125 and 125, I got a total of around 300VAC.

GREAT - everything is working so far.

Now, I wire up the bridge rectifier - 4 X 1N4007; 4 X 0.01uF caps.
1. The B+ leads out through a schottky dioide
2. 1.5uF cap
3. 10H (150mA / 102 ohm) choke
4. 60uF capacitor
5. 10H (150mA / 102 ohm) choke
6. 60uF capacitor

When I modelled this (without the schottky) in PSUDII, I get about 300VDC.

So, I switch the unit on, with multimeter connected across the final capacitor. I put on my saftey glasses (you never know) and wait.

30 seconds later, I hear a click. No flames, sparks or anything. BUT....I get 420VDC!!!!!

Any ideas - anyone, please.....?
Such as if I take out the 1.5uF cap, will this lower the voltage without overloading the first choke ie. make it into a choke input filter and not a cap input filter????

Could it also be that it will be 420VDC under no load (I don't know what my multimeter draws), and that under load 30mA, the voltage will come down to around 300VDC????

Regards,
Charlie

[zigzagflux]

Well, I assume you are using a Graetz bridge, since you mentioned 4 diodes. So when you full wave rectify 300VAC, you get 300 * 1.414 = 424V. Sure, your voltage will drop somewhat (maybe 5-10%) with load, but I wouldn't expect to get anywhere near 300V.

If you changed to a full wave center tapped config, expect 125 * 1.414 = 176V (only two diodes needed, at lower PIV rating). Again, some voltage drop with load.

I see only two effective ways of using this xfmr to get 300V.

1. Just as you mentioned, use choke input filter instead, use PSUD to predict. You may get close, but that's quite a bit of voltage drop to get to 300V.

2. Use a free filament winding (if available) in autotransformer connection to effectively lower your 300VAC secondary output. This worked great in my ST70 conversion to SS rectification, using the 5VAC unused filament winding. More details if interested.

[Brian Beck]

Since you’re only using 1.5uF as the first cap, any load current drawn will indeed drop your voltage way down. You’re in that nether region between a choke input and a cap input supply. Maybe just remove the 1.5uF altogether for now. You can experiment with low-value input caps later when the basic supply and audio circuit are functioning. Use a power loading resistor for testing the supply. Something close to duplicating the expected Aikido draw. With a choke input supply you'd see about 300*0.9 = 270 volts, maybe a bit less with the transformer loaded. Is that what you were looking for?

[Eli Duttman]

Such as if I take out the 1.5uF cap, will this lower the voltage without overloading the first choke ie. make it into a choke input filter and not a cap input filter????

Charlie,

I like the choke I/P filter idea. In order to get 300 VDC on the dot, a small "cheater" cap. in front of the 1st inductor will be needed. Start out with a 10 nF. HIGH WVDC ceramic part in place to protect the SS diodes from inductive kick back spikes. A 10 KOhm/25 W. bleeder resistor goes across the 1st 60 muF. filter cap., to ensure the critical current gets drawn.

While measuring the O/P voltage (should start around 270 V.), add small amounts of capacitance at the front of the filter, until the O/P comes up to 300 V.

BTW, ditch the 1N4007s in favor of UF4007s. You will get LOTS less SS diode switching noise.

[cbutterworth]

I remeasured the voltage at the outputs of the graetz bridge recitifier and did get 425VDC.

The only difference between my implementation and PSUDII model are the four 0.01uF caps, one in parallel with each diode. I am amazed at how far off PSUDII is, as it gave me 315VDC when I modelled it.

I have already installed 470 Kohm bleeders across each capacitor.

I guess that I'll see what happens when I remove the 1.5uF capacitor.

Brian, how do I calculate the power-draw resistor for the multimeter readings? I modelled the PSU for 30mA draw. Do I need to do womething with Ohm's Law????

Thanks,

Charlie

[Brian Beck]

Well, 270 volts/30mA = 9Kohm. Power = 270x30mA = 8.1 watts. Better use a 20 watt resistor - it will get hot. Finding the exact 9K value is not critical. 9.1K is a common value. Even 10K will get you along until you can load the supply with the real audio circuit.

The 1.5uF is too small to store much energy. When the supply is not loaded, the cap charges to the full 1.414 times AC voltage. When you load the supply, a huge ripple voltage will develop across this little cap, with a lower average DC value. Loaded enough, the output voltage approaches the choke-loading situation. For now, just yank it.

[Eli Duttman]

I guess that I'll see what happens when I remove the 1.5uF capacitor.

Charlie,

Unless the PSU is loaded down, you will see little difference between cap. and choke I/P filters.

470 KOhms will discharge a 'lytic, but proper operation of a choke I/P filter requires the critical current be drawn. Put the 10KOhm/25 W. part in. Unlike a cap. I/P filter, a choke I/P filter droops VERY little under the working load.

Don't forget the 10 nF. "safety valve". Without it, your diodes can be destroyed.