Antec Smart Power SP-350 5VSB mod using VIPer 22A pwm chip

[ben7]

Re: Antec Smart Power SP-350 5VSB mod using VIPer 22A pwm chip

I've decided upon trying a VIPER27LD SOIC-16 for the SMD experiment. It seems to have most of the same basic parameters as the 22A as best as I could tell from the datasheet. Which leads to the question WTF are the datasheets for two devices that are in the same family so different in their formatting? I could understand if they where competitive products but the same manufacturer and family is just BS.

If anyone knows a reason that the VIPER27LD will not work for this project please let me know before I get too far into this project.

Thanks,
Lloyd

The VIPer27 has different pinouts, and a few extra pins functions.

The VIPer27 also has lower on resistance, so it should dissipate slightly less power.

[LDSisHere]

Re: Antec Smart Power SP-350 5VSB mod using VIPer 22A pwm chip

The VIPer27 has different pinouts, and a few extra pins functions.

The pinout and extra functions will not be a problem from my take on the datasheet. They can either be ignored or tied to ground like the brownout pin.

The VIPer27 also has lower on resistance, so it should dissipate slightly less power.

The resistance was one of the things that I noticed was different, but I did not know what that would mean in terms of its' operation. Since you say it will dissipate less heat may explain why it only calls for half the heat sink area compared to the 22A which I had been trying to figure out.

Here are the Digi-Key parts I was planning on ordering for this project.

VIPER27LD-ND
P9.10KFCT-ND
PCF1340CT-ND
493-2297-1-ND
DL4935-FDICT-ND

I was looking at the 4.7uF tantalums but the cheapest I could find was over $2 each and I can get 10 of the electrolytic for $5. The 10nF SMD caps are also surprisingly expensive. I always thought that SMD parts would be cheaper than through hole but apparently I was mistaken.

I am open to suggestions if anyone thinks what I have listed will not work for this project.

[everell]

Re: Antec Smart Power SP-350 5VSB mod using VIPer 22A pwm chip

I was also checking into parts, and you might want to compare your list with these:

For the 9.1K resistor:
311-9.1KERCT-ND 100 for $1.04

For the 10 nF capacitor:
399-1236-1-ND 10 for $1.20, 50 for $3.45

For the 4.7 uF/50 volt capacitor:
399-6785-1-ND 10 for $7.37
This capacitor was interesting because it is a film and foil with 10% tolerance. ESR should be very low, and no problems with the electrolyte drying out!

[LDSisHere]

Re: Antec Smart Power SP-350 5VSB mod using VIPer 22A pwm chip

I was also checking into parts, and you might want to compare your list with these:

For the 9.1K resistor:
311-9.1KERCT-ND 100 for $1.04

For the 10 nF capacitor:
399-1236-1-ND 10 for $1.20, 50 for $3.45

For the 4.7 uF/50 volt capacitor:
399-6785-1-ND 10 for $7.37
This capacitor was interesting because it is a film and foil with 10% tolerance. ESR should be very low, and no problems with the electrolyte drying out!

Well, it was too late when I saw your post. I needed to get some items before the end of the week so I wanted to get the order in so it would ship tonight. I attached a "snip" of the items I ordered for this project, I like the prices you found much better than mine. I am just hoping the ones I ordered are going to be adequate. Now I just have to find time to draw and etch the board to try the SMD version. I will try to get it posted in the next few days, but I a lot of stuff to get done and not enough time to do it all.

Thanks,
Lloyd

[everell]

Re: Antec Smart Power SP-350 5VSB mod using VIPer 22A pwm chip

Perhaps too late, but I was looking into finding the best possible diode to use in the smt project. My suggestion:

Digikey part number ES1DFSCT-ND which is a Fairchild ES1D smt diode rated at
200 volts, 1 amp, 15 nsec switching speed, priced at 10 for $3.02

Whereas the DL4935 has spec at 150 nsec, the ES1D is ten times faster as well as cheaper!

[LDSisHere]

Re: Antec Smart Power SP-350 5VSB mod using VIPer 22A pwm chip

Perhaps too late, but I was looking into finding the best possible diode to use in the smt project. My suggestion:

Digikey part number ES1DFSCT-ND which is a Fairchild ES1D smt diode rated at
200 volts, 1 amp, 15 nsec switching speed, priced at 10 for $3.02

Whereas the DL4935 has spec at 150 nsec, the ES1D is ten times faster as well as cheaper!

I will keep this in mind if I need to order more, I was just trying to get something to match the through hole version.


I received my parts, drew the schematic & PCB, etched & populated the board and nothing..... I am not getting any 5VSB output. I am not sure if it is my implementation of the Viper27 chip or if it has to do with one or more of my SMD choices that is keeping it from working.

I am considering replacing all the SMD components with the through hole counterparts, that are known to work with the Viper22A, to see what happens. If anyone sees where I may have gone wrong with the design or implementation I would greatly appreciate constructive feedback.

[LDSisHere]

Re: Antec Smart Power SP-350 5VSB mod using VIPer 22A pwm chip

I think I have figured out the cause of the problem, but I do not yet know how to correctly fix the problem. For the Viper27 the feedback should be connected to pin 4 of the optoisolator as opposed to pin 3 for the Viper22A. Consistency between datasheets would have made figuring this out much easier.

Any assistance in figuring out how to correctly configure the feedback circuit would be very much appreciated.

[ben7]

Re: Antec Smart Power SP-350 5VSB mod using VIPer 22A pwm chip

I think I have figured out the cause of the problem, but I do not yet know how to correctly fix the problem. For the Viper27 the feedback should be connected to pin 4 of the optoisolator as opposed to pin 3 for the Viper22A. Consistency between datasheets would have made figuring this out much easier.

Any assistance in figuring out how to correctly configure the feedback circuit would be very much appreciated.

Does your circuit look like figure 20 in the VIPer27 datasheet?

[LDSisHere]

Re: Antec Smart Power SP-350 5VSB mod using VIPer 22A pwm chip

Does your circuit look like figure 20 in the VIPer27 datasheet?

No it is not. It is configured like the Viper22a with the feedback pin connected to the optoisolator emitter (pin 3) but it should be connected to the collector (pin 4). So the feedback circuit needs to be redone too have any hope of the Viper27 working. I am just not sure how to do this correctly to get the desired results.

[ben7]

Re: Antec Smart Power SP-350 5VSB mod using VIPer 22A pwm chip

No it is not. It is configured like the Viper22a with the feedback pin connected to the optoisolator emitter (pin 3) but it should be connected to the collector (pin 4). So the feedback circuit needs to be redone too have any hope of the Viper27 working. I am just not sure how to do this correctly to get the desired results.

Step One: Cut the trace that connects the optoisolator's collector (for Viper22 it goes to the VDD pin), and connect it to the Viper27 feedback pin.

Step Two: Cut the trace that connects the optoisolator's emitter (for Viper22 it goes to the FB pin), and connect it to the GND rail.

[LDSisHere]

Re: Antec Smart Power SP-350 5VSB mod using VIPer 22A pwm chip

I am afraid it will not be that simple, nothing ever is..... Take a look at I(FB) [Viper27 pg. 7 and Viper22A pg.3] the 27 is measured in microamps and the 22A is measured in milliamps. Another difference is that the FB pin on the 22A operates between 0-1V and on the 27 it seems the voltage can go to at least 4.5V. It is for these reasons I think it will take more than "reversing" the connections. You can actually make the 27 run wide open if you connect the FB pin to ground but I have a feeling that would damage other parts of my power supply if I were to try it. It is still tempting to try it just to get it to do something......

[ben7]

Re: Antec Smart Power SP-350 5VSB mod using VIPer 22A pwm chip

I am afraid it will not be that simple, nothing ever is..... Take a look at I(FB) [Viper27 pg. 7 and Viper22A pg.3] the 27 is measured in microamps and the 22A is measured in milliamps. Another difference is that the FB pin on the 22A operates between 0-1V and on the 27 it seems the voltage can go to at least 4.5V. It is for these reasons I think it will take more than "reversing" the connections. You can actually make the 27 run wide open if you connect the FB pin to ground but I have a feeling that would damage other parts of my power supply if I were to try it. It is still tempting to try it just to get it to do something......

I think it is better to try the reversing of the opto, rather than the flat out running. The flat out running can instantly kill the capacitors, and semiconductors on the secondary side, which will cause a BIG headache to track down and repair... and the same goes for the opto-pin-swap, but it has a higher chance of actually working properly...

[LDSisHere]

Re: Antec Smart Power SP-350 5VSB mod using VIPer 22A pwm chip

I think it is better to try the reversing of the opto, rather than the flat out running. The flat out running can instantly kill the capacitors, and semiconductors on the secondary side, which will cause a BIG headache to track down and repair... and the same goes for the opto-pin-swap, but it has a higher chance of actually working properly...

Yes, you are correct doing it your way it would at least have the 9.1k resistor in the circuit to limit the current. I am not going to have much time until Monday to do any hands on work with this project, so I am going to do some research on the net and hopefully decipher more information from the datasheet. For this kind of situation the smartest thing I can do is exercise patience and do some homework. This kind of situation can be tremendously educational in many ways.

Ben7, I do very much appreciate your interest and trying to help solve the problem. Of course without Everell's efforts I would not even have known it was possible to modify the 5VSB circuit and I have only been trying to duplicate what he has already accomplished.

Thanks,
Lloyd

[LDSisHere]

Re: Antec Smart Power SP-350 5VSB mod using VIPer 22A pwm chip

I had a small amount of time tonight to do a little testing. I decided to probe the feedback circuit on the Viper22A board with my DMM to get a better understanding of what it was doing. It did not go very well.





I must have crossed the creepage DMZ as you can see from the pictures the results are not pretty. The optoisolator was also shorted so I am hoping one or both that are in the last picture will be suitable for a replacement.
I also hope the power supply damage was limited to the optoisolator but I will have to assemble another Viper22A board for 5VSB testing to truly determine the extent of the damage. Sometimes you learn things you would rather not know from first hand experience.....

I now have a up close and personal understanding of why Retiredcaps likes his DMM's to have a peak hold function. It is a very bad idea to take your eyes off of the probes when poking around high voltage.

[eccerr0r]

Re: Antec Smart Power SP-350 5VSB mod using VIPer 22A pwm chip

It looks like the VIP22 pulls down on the FB pin and expects the optoisolator to pull up to shut off the PWM, whileas the VIP27 sources current onto the FB pin, expecting the optoisolator to pull down.

Theoretically just swapping the position of the optoisolator and the filter cap should do the trick...

BUT! Whether the optoisolator can now sink the current without leaking while off. Since we know that the VIP22 works (since Everell did it already) we know the optoisolator can souce the 3mA. Now the question is if the "off" leakage current of the optoisolator is low enough that it won't exceed the ~200 microamps that it can source (0.3V is suspicious, or rather not so suspicious, it's around the typical saturation voltage of an optoisolator)...

Unfortunately I've buttoned up my SP400 so I don't know what optoisolator it used... So I grabbed one from my junk bin: a SFH615. Based on the spec sheets for this the leakage current has a worst case of 100 nanoamps, which should be fine as it's less than the FB pin source current.

It looks like you're using a PC817 optoisolator, according to the datasheet it's also about 100nA leakage so it should be fine.

[retiredcaps]

Re: Antec Smart Power SP-350 5VSB mod using VIPer 22A pwm chip

Sometimes you learn things you would rather not know from first hand experience.....

1) Mark Twain noted

“If you hold a cat by the tail you learn things you cannot learn any other way.”

I now have a up close and personal understanding of why Retiredcaps likes his DMM's to have a peak hold function. It is a very bad idea to take your eyes off of the probes when poking around high voltage.

2) In Fluke land, it is called touch hold or auto hold. This allows you to concentrate holding the probes steady and once the multimeter gets a stable reading, it will beep allowing you to safely remove the probes while keeping the reading frozen on the screen.

3) An used Fluke 27 is a low cost option to get this feature. If you are lucky you can get it for under $20 on ebay as I explained in

https://www.badcaps.net/forum/showthread.php?t=21650

4) Fluke also has a wireless multimeter (233) where the lcd is detached from the body so you can place the lcd in a viewable area. Dave Jones did a review at

http://www.eevblog.com/2009/11/08/ee...imeter-review/

Note: From reading the 233 manual, this multimeter does not have the touch or auto hold feature.

5) Agilent's new line of hand held meters has equivalent hold feature but with a twist. It can store the stable readings automatically into memory for recall later. This could have been helpful in this thread where I did use Fluke's touch hold

https://www.badcaps.net/forum/showthread.php?p=262730

All 20 readings would have been stored into memory for recall.

Dave Jones used this feature to read 1,000 resistors with the Agilent U1272A below (around the 13:30 mark)

http://www.eevblog.com/2011/11/14/ee...esistor-redux/

6) I also wear safety glasses when doing measurements.

[retiredcaps]

Re: Antec Smart Power SP-350 5VSB mod using VIPer 22A pwm chip

Dave Jones does a review on the Fluke 27

http://www.eevblog.com/2012/10/18/ee...view-teardown/

[ben7]

Re: Antec Smart Power SP-350 5VSB mod using VIPer 22A pwm chip

^ LDS, trust me, I have done the same thing before. The sudden short circuit spark and noise makes me jump backwards so hard that my muscles hurt.

Both the opto's you show are PC817, so they should be good replacements. But, I would use the 'Sharp' branded one first if I were you.

This is also why it is nice to make test points on PCB's (little wires that stick up) so that you can put a clip-on probe onto.

[everell]

Re: Antec Smart Power SP-350 5VSB mod using VIPer 22A pwm chip

A while back I did the "probe slip" on a Bestec psu. Big spark, totally blew up the 5vsb circuit, blew out the main pwm chip and the main switching transistor, and a few other parts. Should have totalled it, but I spent the time and fixed it! What a set back!!!

I think you should rebuild the Viper 22A circuit and get it running again. As for the Viper 25, 26, and 27, I have looked at the data sheets and think they are a totally different animal. It will require some experimentation. Maybe you should start by making a surfact mount board using the Viper 22A even if it has less power. Then develop a board using the Viper 27.

The Viper 27 looks more like a DM311 type circuit. The Vdd to the chip may not be such that it will work with the Antec 5vsb transformer. Notice that the Viper 27 has a cutoff point for Vdd at 23 volts. As you add loading greater than 1 amp the Vdd will probably go higher than that thus shutting the Viper down. That is the reason that the DM311 mod did NOT work as expected when I made the DM311 mod to the Antec psu.

[LDSisHere]

Re: Antec Smart Power SP-350 5VSB mod using VIPer 22A pwm chip

Theoretically just swapping the position of the optoisolator and the filter cap should do the trick...

BUT! Whether the optoisolator can now sink the current without leaking while off. Since we know that the VIP22 works (since Everell did it already) we know the optoisolator can souce the 3mA.

Based on the spec sheets for this the leakage current has a worst case of 100 nanoamps, which should be fine as it's less than the FB pin source current.

It looks like you're using a PC817 optoisolator, according to the datasheet it's also about 100nA leakage so it should be fine.

Thanks for having a look at this, I need all the help I can get to figure this one out. I had not even considered the optoisolator leakage until you mentioned it.

1) Mark Twain noted

“If you hold a cat by the tail you learn things you cannot learn any other way.”

I am sure this is very true and not something I would be willing to attempt. When I was a kid I "tried" to give some kittens a bath, for some reason I never attempted it again.

2) In Fluke land, it is called touch hold or auto hold. This allows you to concentrate holding the probes steady and once the multimeter gets a stable reading, it will beep allowing you to safely remove the probes while keeping the reading frozen on the screen.

It does appear that after ~20 years of service I must relegate my trusty antique Fluke 21 back to its' original purpose of automotive work.

3) An used Fluke 27 is a low cost option to get this feature. If you are lucky you can get it for under $20 on ebay as I explained in

https://www.badcaps.net/forum/showthread.php?t=21650

4) Fluke also has a wireless multimeter (233) where the lcd is detached from the body so you can place the lcd in a viewable area. Dave Jones did a review at

http://www.eevblog.com/2009/11/08/ee...imeter-review/

Note: From reading the 233 manual, this multimeter does not have the touch or auto hold feature.

5) Agilent's new line of hand held meters has equivalent hold feature but with a twist. It can store the stable readings automatically into memory for recall later. This could have been helpful in this thread where I did use Fluke's touch hold

https://www.badcaps.net/forum/showthread.php?p=262730

All 20 readings would have been stored into memory for recall.

Dave Jones used this feature to read 1,000 resistors with the Agilent U1272A below (around the 13:30 mark)

http://www.eevblog.com/2011/11/14/ee...esistor-redux/

A few months ago I picked up a Fluke 113 fairly cheap to use as a capacitor tester and possible replacement for my Fluke 21. I did not do any research on it prior to purchase, but I had never used a Fluke meter that I did not like, until I made this purchase. I despise the piece of crap, I think it will soon be on Ebay for sale. I do intend on sticking with a Fluke, but I plan on doing my homework this time and also getting recommendations from other people. Your recommendation for the Fluke 27 for example and I like the Fluke 179 that I use at work. I intend on looking at some local pawn shops and Craigslist in addition to Ebay. Thanks for the links you provided.

6) I also wear safety glasses when doing measurements.

For me glasses are not optional, they may not be "safety" but they do the trick. I can think of many occasions where they saved me much grief.

^ LDS, trust me, I have done the same thing before. The sudden short circuit spark and noise makes me jump backwards so hard that my muscles hurt.

I was very surprised when it happened and I could not believe the dip legs were blown completely out of the PCB. I have been intending to install a GFI outlet where I plug in the power strip I use for testing. I now plan on making it happen tomorrow morning, this is something I should not have procrastinated doing in the first place.

Both the opto's you show are PC817, so they should be good replacements. But, I would use the 'Sharp' branded one first if I were you.

Thanks, I was leaning towards the Sharp already and you sealed the deal.

This is also why it is nice to make test points on PCB's (little wires that stick up) so that you can put a clip-on probe onto.

Unfortunately, I do not have any clip on probes for my DMM. I have been looking at getting Fluke's TL81A Deluxe Electronic Test Lead Set and this little incident will help me justify the cost. I enjoy fireworks in the proper setting but not when I am working on electronics.



I want to thank everyone for their input and help, it is very much appreciated. When I get home tonight I plan on seeing if I can get the power supply back into operation. It is a good thing I bought extra parts and made spare PCB boards for the Viper22A project.