UC3827N-1 SS (soft start) pin does not seem to do anything atol..

Hi Chris,

Have you tried changing out the IC? Maybe it's just a faulty chip. I realize this a stretch since all the other functions of the chip are working. But it's a relatively easy check. You could also check the capacitor. Maybe it's leaky, or there is a leaky parasitic capacitance in the PCB. 

Does the cap ground go to a ground plane on the layer below it? If it does, try tying the capacitor ground to the ground pin on the chip. 

 The data sheet shows a maximum of 25 uA output. So with 0.47uF cap, that would take nearly a second for the ss voltage to reach 0.5 Vs. I think that's might be a little high.

Are you using a ceramic cap? Tantalums have leakage currents as high as 30 uA. Electrolytics have high leakage currents too. You can try a higher voltage rated capacitor as a check. The higher the voltage rating on the capacitor the lower the leakage current for a given application.

That's all I got. Hope you find your problem. The 3827 looks like a nice IC. 

Chuck

Chuck!

I have built 9 of the supplies, all to no avail. I have replaced the chip when I damage it. It is an easy test.. short pin 4 to 3 to see if it shuts off. It never does.

I am using a ceramic, so I doubt leakage. And if it was shorted, alas, the supply shoont start up at all.

The part is nice.. I have an off line switcher supplying 150VDC to the HV section. There is a floating IGBT in the buck giving me 0-100V (virtual, it's current fed). The push pull generates 850VDC, and a full wave trippler generates 0-2500V. An external feedback opamp is used to gain more precision, as we are seeking 1/4V out of 2500. The current fed topology was chosen for it's low noise. It's for nuclear use.

Chris~

Hi Chris. 

I didn't understand the problem in the first email, my mistake. So what I think you are saying is that the soft start pin acts like there is a proper voltage between the it and the IC ground. Could it be a layout problem? On one of my designs I had a similiar problem but it was the current sense pin. It turned out I had a circulating ground currents running through a high impedance trace between the IC gnd and the CS pin. So the ground at the current sense resistor was at -0.5 V, so to hit the over-current threshold the resistor needed an extra 0.5V. I lifted the ground end of the current sense resistor off the PCB and tied it directly to the IC ground pin and the problem was solved.

Chuck

I appreciate that you are digging into the subject Chuck..

I just checked the layout. The UC3827 pin 4 goes direct to the cap, 0.47uF ceramic, which goes direct to ground. I am using a 2 layer board, but I was careful to lay in a very solid ground system, star topology, where heavy currents all return in a well behaved manner. The datasheet states that a 12uA current charges the capacitor. It is not very clear what voltage is supposed to be crossed to enable the part internally.

The part starts up every time.. which indicates that the SS pin is doing it's job, if the datasheet is being truthful. But, alas, when I use a scope to look at the SS pin, there is no voltage on the cap. The voltage does not change.. there is nothing there. While the part is operating. I can short pin 4 to pin 3, the ground that is right next to it. And if the SS pin was operating as it is supposed to, it should shut down the whole chip. But it does not.

I have to give pin 4 credit.. it is very consistent. No intermittent behavior.. it just does nothing.

Thanks,

Chris~

Hello Chris 

Been in this situation more than once. I hope I am not being offensive by sounding like I am talking down to you. It sounds like you have a good layout from your description. The only question I have though is why are you using pin 3 as ground? Isn't pin 3 connected to the power ground, pin21? I would think you would want to use pin 11, signal gnd. Have you tried shorting pin 4 to pin 11?

Chuck

Chris,

Pin 3 is not ground.

Pin 3 is the isolated return for the buck drive.

You need to use pin 11

John

Ya, your both right, but in this case I am not using the floating driver. I have a B+ much higher at 150VDC, so I have my own opto-isolated gate drive. So I strapped the floating driver down to ground at pin 3 and use pin 2 to drive the opto. See the schematic;

2605.Pioneer128R5.pdf

Chris~

On the block schematic for the part, pulling SS low only disables the voltage error amplifier output. On your schematic, it looks like the EA output is not connected and you are utilizing the current error amplifier. So the SS output would not affect your circuit

John

By jove, John, you are a genius. All I had to do was follow the line to the VEA. I am off to lash it up.. I will try using the VEA as a buffer, so that the SS disables it, but can still use my external VEA.

Thanks very much!

Chris~