Filtering a SMPS HV supply

[doorunrun]

Hello All,
I’ve been planning to build another power supply for a Royer Mod mic but wanted to deviate from the original passive/linear supply. I want to easily adjust HV and Heater voltages.

In the thread, "DIY Tube Mic project," @thor.zmt posted information on a HV module utilizing a MC34063 IC that looked interesting to me so I ordered a couple of them.

After getting the module built, I set it to provide 100VDC by using a value of 51K for Rx. I put together a breadboard version of the Royer Mod microphone section with a 5840 pentode.

'Scoping the HV line I'm getting this sawtooth ripple riding on top of the 100 Volts. It measures 2Vp-p with a 4mS period on my old and probably uncalibrated Tek T922 ‘scope.

To minimize the ripple I put together a CR-CR-CR-CR filter from parts I had on hand:

My question is there a better way to do this?

Thanks!

 

[ruffrecords]

Strange that the ripple has a 4mS period when you are switching at 50KHz of more.

CRCRCRCR filters are very effective and simple to implement.. The only thing better would be an LCLCLC.

Cheers

Ian

 

[thor.zmt]

Hello All,
I’ve been planning to build another power supply for a Royer Mod mic but wanted to deviate from the original passive/linear supply. I want to easily adjust HV and Heater voltages.

In the thread, "DIY Tube Mic project," @thor.zmt posted information on a HV module utilizing a MC34063 IC that looked interesting to me so I ordered a couple of them.
View attachment 131954
After getting the module built, I set it to provide 100VDC by using a value of 51K for Rx. I put together a breadboard version of the Royer Mod microphone section with a 5840 pentode.

'Scoping the HV line I'm getting this sawtooth ripple riding on top of the 100 Volts. It measures 2Vp-p with a 4mS period on my old and probably uncalibrated Tek T922 ‘scope.
View attachment 131955
To minimize the ripple I put together a CR-CR-CR-CR filter from parts I had on hand:
View attachment 131958
My question is there a better way to do this?

Thanks!

First, 4mS = 250Hz

I think the switching frequency should be much higher. With a 100uH coil I would expect something closer to 30..40kHz minimum.

Check the Capacitor on the Ct pin. Maybe a build mistake? Or wrong part supplied?

Anyway, it doesn't work correctly, fix it to work correctly.

Thor

 

[doorunrun]

First, 4mS = 250Hz

I think the switching frequency should be much higher. With a 100uH coil I would expect something closer to 30..40kHz minimum.

Check the Capacitor on the Ct pin. Maybe a build mistake? Or wrong part supplied?

Anyway, it doesn't work correctly, fix it to work correctly.

Thor

Thor, here's how the module looks and also a closeup of C2 in blue, the item in question (Ct),

and it's marked "681J."

I'll check the "Pulse" pin and see what I get.
I can also build the other module to test.

Thanks!

 

[thor.zmt]

Thor, here's how the module looks and also a closeup of C2 in blue, the item in question (Ct),
View attachment 131963 View attachment 131964
and it's marked "681J."

I'll check the "Pulse" pin and see what I get.
I can also build the other module to test.

Thanks!

Fix this module to work correctly.

Put the 'scope at the actual switching node. See what you get.

It is possible something else is wrong. Systematically check everything one by one.

Thor

 

[doorunrun]

I built the second HV module and results are just about the same.
I brought out my "better" 'scope, a 100MHz Tek 7603 and could better observe the switching node pulse.

10uS per division with 2V per cm.

I guess that makes more sense as far as the switching frequency goes. This double pulse comes every 35mS (on the other 'scope it looked more like 40mS), but I am seeing a 'phantom' pulse pop up from time to time.

Maybe that' the nature of the beast? The filter does a good job of removing the ripple. The circuit draws very little current so I think it will be fine.

Cheers!

 

[ruffrecords]

This is definitely not right. The pulses should be occurring every 20uS, not twice every 35mS. I suspect this is simply because the load you are applying is too small. You need to revisit the circuit equations for your particular dc load current or increase the load current.

Cheers

Ian