PDA

View Full Version : DC SSR boards



Daemon
11-27-2009, 05:11 AM
Hi all

Just after some extremely last minute advice/comments.

A friend of mine made me some DC SSR boards. I have attached a photo of the back and front of one of them.

5381

5382

5383

5384

My problem is that the output of these boards is very dim. A bit of background.
These are 12 dc channels run off a tested and working grinch of which I am using about 64 other channels to drive AC ssr boards and those ones are all functioning correctly. This says to me that the grinch is performing as expected. ie 5v+ in and 5v- out etc etc.

My dc power supply is measured at 24vdc. The terminal buses where I feed positive and negative to have 24vdc confirmed at them. The positive goes from a bus through 4 fuses then to 4 separate buses each of which has a thin wire going to the board and the positives of my light load commoned to that bus.

The negative goes to a bus that has 4 negative wires coming from it and going to the negative points on the ssr boards.

I am scratching my head trying to find out what is happening. The resistors are 510 ohm for the opto resistors and 460 ohm)I think) for the lower ones and 10k ohm on the bottom of the board.

I do not fully understand how the board works(after trying to understand it for a few days I barely understand anything now)

My friend who made them has tried to get me to measure certain points to show the board works and has sort of left me at the stage where he thinks the grinch is faulty or the cat5 cables are faulty.
However I cannot understand why all 3 of those cables would be faulty( they have all been tested with a meter for continuity and tested okay) and I cannot understand what fault the grinch could have when the ac lights are working correctly.

Does anyone have any comments/suggestions on things I could try to get them working correctly?

I mean no disrespect to my friend who has worked very hard to help me through this.

Thanks for looking.
Steve

LabRat
11-27-2009, 09:11 AM
So you are modifying the AC SSR boards for DC usage. Are you replacing the TRIAC? If so.. what TRIAC are you using instead? I'm not an expert here at all, but this is something that the experts here on the list will want to know.

budude
11-27-2009, 01:56 PM
A TRIAC is for AC switching - if you are modding the AC SSR for DC use as LabRat has pointed out, you should be using MOSFETs in place of the TRIACs...

Daemon
11-27-2009, 03:44 PM
Yes they are mosfets. I dont understand the componentry completely however I was hoping someone could tell what was happening from the circuit diagram and the photos. I dont even have a item list of parts(bol) from when he charged me for them because he hasnt done that yet.

It uses a 4N35 DC opto instead of the MOC3023 and a mosfet(dont know the number)

I was just clutching at straws really because the lights these control were a huge feature of my show and now you cant even see them. They are activating etc but not visible because they are getting such a low voltage.

LabRat
11-27-2009, 07:07 PM
Well.. a quick google of the 4N35 showed two circuits, one of which used a 220ohm, the other 330ohm resister in series with the opto-isolator.

Looking at your board (and trying to guess the colours), it looks like the lead band is Green which puts it at significantly higher. Could it just be that you don't have enough current to the LED's inside the opto?

Can you read off the colour banding on those four^H^H^H^H^H five blue resistors beside the ethernet jack?

Daemon
11-28-2009, 05:06 AM
I have measured them and they are definately 510 ohm. That being said the lights are being activated so the opto seems to definately be switching and the issue seems to be the amount of 24v current passed through the rest of the circuit. Namely the mosfets and the (is it a transistor?) through the arrangment with the 10k ohm resistor.
I will try and find a model number for the mosfets which may help further.

Unless I am wrong about it switching seeming to indicate the opto is getting enough voltage?

Matt_Edwards
11-28-2009, 07:49 AM
Without knowing your for set up, we need to clear off some basic testing. Do you have a SSR tester? If not it is time lie now it woul be worth making one, and some of the one where are very basic and easy to make. I would even just use a quad pack of AA Batteries as a simple tester.
Connecting the battery with Positive (usually the Red wire) to Pin 1 of the RJ45 connector and the Negative (usually the black wire) to pin 2, 4 ,6 & 8. just to them pins, you Lights should come on.

Another thing that needs to be detailed is how are you driving hte Grinch, is it with a printer port?

budude
11-28-2009, 11:25 AM
Matt's suggestion is very good - that will ensure the logic side from the Grinch is working correctly (or not). I'm wondering if you are overloading the regulator which appears to be in the fuse holder position.
Is it a 7805? When you run the test, does it get very (very) warm? Is it producing 5V on the output (right pin when looking at front)?
When you are testing, you are not attempting dimming correct - just on/off?

Can you try this with a 12v supply to confirm if a lower voltages will work?

P. Short
11-28-2009, 12:40 PM
Also, what is the part number of the MOSFET that you are using? Not all are created equal as far as gate voltage goes.

Daemon
11-28-2009, 03:43 PM
Alright. Just finished 16 hours at work so I am a little foggy but the mosfets are MTP3055v CQ521 and the voltage regulator in place of the fuse holder is a L7805CV.

I will do the testing later today.

A few answers to the above. Firstly I am using a grinch with a ren-c attached so I am dimming. I have 2 grinches but the channels this one are attached to are from a serial to ren-c cable. The other grinch comes from the parallel port.

This grinch has about 60 odd channels in use dimming and switching ac ssr boards and working perfectly. It just seems strange to me it would be a grinch problem when these 12 channels were chosen at random from the middle of the board and all dont work correctly. I believe it is a function of the board.
As far as a tester I have only made one that carries the grinch 5v to the ssr board with a toggle switch so I can bypass vixen. Nothing powered so I will take your suggestion and make up a simple AA pack to test the power function as you suggested later today.

I do not have a 12 volt supply to try and when you say is the voltage regulator 5 v on the right hand pin as you look at it, where do I measure from and to the right hand pin?

Thanks for the replies
Steve

P. Short
11-28-2009, 04:42 PM
Quick question...do you have both cables between the Ren-C and Grinch connected? The second (feedback) cable is required for PWM operation. The DC SSR requires PWM operation, since the 4N35 transistor-output optos and MOSFET combo does NOT latch as the MOC3023/triac combo does. Just thought I'd ask.

Matt_Edwards
11-28-2009, 04:44 PM
I have used those on another project. Pretty universal device.

I doubt you will be overloading the 7805, but it is good to check three things:

After it has been on for awhile, can you touch 7805 case?
Measure the input voltage
Measure the output voltage


I have to let others help you with your Ren-C, but like you said in the beginning, your AC SSR work and Dim!

Daemon
11-28-2009, 07:58 PM
Quick question...do you have both cables between the Ren-C and Grinch connected? The second (feedback) cable is required for PWM operation. The DC SSR requires PWM operation, since the 4N35 transistor-output optos and MOSFET combo does NOT latch as the MOC3023/triac combo does. Just thought I'd ask.

I cant believe I never thought of that! I had to go to only using one of the cables in order to get the rest of my lights working correctly. Sorry for being a pain. I will test them when I have confirmed that this is the issue.

Sorry for going through this whole post before I realised that Guys, I knew PWM required the second cable too, I just overlooked it completely.

I would still be a bit sunk because without the cable I get random flickering in my other channels, with it I cannot make the dc stuff work.

toodle_pipsky
11-28-2009, 08:07 PM
Just thought I would chuck it out there since I had similar problems ealier in the year - make sure your have your Grinch fully populated with ICs (even though you might only be testing 1 block of channels) and there's new firmware for the Ren C that fixes the flickering issue. Good luck!

Daemon
11-28-2009, 08:10 PM
Just thought I would chuck it out there since I had similar problems ealier in the year - make sure your have your Grinch fully populated with ICs (even though you might only be testing 1 block of channels) and there's new firmware for the Ren C that fixes the flickering issue. Good luck!

Thanks Toodle pipsky. I tried the new firmware etc and all the other suggestions on here as well in my flickering with ren-c post. I never found any other way to make the lights work except to remove the second small cable from between the ren-c and grinch. And then unfortunately I promptly forgot about it!

P. Short
11-28-2009, 08:47 PM
Keep the cables between the Grinch and the Ren-C short, make sure that they have good crimps, make sure that you have a good power supply (PWM mode requires a lot more current than non-PWM mode) with short wires to the Grinch, and that there is a shunt installed in J1 on the Grinch.

Daemon
11-28-2009, 09:00 PM
I can confirm that these are working well enough to see in the daylight now.
I will not be able to confirm the rest of the lights also dont work until dark so I can see the flickering.

The cables are short, less than 3 inches, the shunt is on J1, the latest firmware is installed(by scorpio) but I dont think it helps unfortunately. I went through it all a long time ago and was rapt when I found the issue went away when I removed that second cable. I do have a good power supply too. 10 amps. only 3 watts of lights attached to it :)

I am far less rapt now that I have to put it back!
I will report back to you later.

Daemon
11-29-2009, 03:32 AM
I can confirm now that the mad flickering is back with the second cable in place.

Is there any other suggestions to work around it?

P. Short
11-29-2009, 12:56 PM
I don't think that I'm going to be much help, since I've never had the problem.

However, a few questions just to keep things going...even if these have been asked in the past.

What type of lights are you using, LED or incandescent? And what voltage are you using for the lights? And if it's less than 240V, what is the transformer setup?

Daemon
11-29-2009, 06:03 PM
Of this particular Grinch's 64 channels

12 channels are 240v ropelight homemade arches.
8 are 240v modyfied powerboards.
12 channels would have been dc ssr boards if not for this issue.

The rest of that grinch are 24v ac incandescent lights using 300va toroidal transformers. The flickering problem with the second cable in happens on all channels. It seems to be when I unselect the lights. As in when I go into the test channel function in vixen and select all they come on correctly. When I then select unselect all they go nuts and dont turn off correctly.

Bit of a bummer really because I can either have my main feature lighting work correctly or only my ac lights. There were a few modifications in posts that I never tried like adding a resistor to the ren-c or adding resistos between points on the grinch because that involved cutting the tracks.

P. Short
11-29-2009, 09:31 PM
A couple of more questions...

When you are using the Grinch _without_ the second cable installed, are there any dimmer levels where the lights flicker? If you turn it down to a very low brightness, is there any flickering?

What are you using for the zero-crossing circuit?

Is there any chance that there is a bad solder joint (or open trace) on pin 2 of U6, or that U6 has gone bad, or that there is a bent pin on U6 (if using a socket)?

The only other suspicion that I might have is regarding 5V DC power...what is the length of the power cable going to JP1 on the Grinch, and what gauge wire? Also, does anything change if you exchange the two cables between the Ren-C and the Grinch?

Daemon
11-30-2009, 12:01 AM
Hi again

Thanks for your help.
I will do some testing when it gets darker here in about 5 hours time.

Can you please refresh my memory though, what is U6?
I just went and looked for reference to U6 on the wiki and couldnt find it.

P. Short
11-30-2009, 12:02 AM
It's the 14-pin chip on the Ren-C.

Matt_Edwards
11-30-2009, 12:47 AM
BTW The Wiki Ren-c page shows it as U2, but the Schematic and PCB have U6 as a designator.

Matt

RavingLunatic
11-30-2009, 01:00 AM
BTW The Wiki Ren-c page shows it as U2, but the Schematic and PCB have U6 as a designator.

Matt


Is there a newer version of Ren-C than V2.0? As of V2.0 the board shows U2. But yes the schematic does show it as U6 but then again all the schematic IC designations don't match the PCB silkscreen on V2.0.

Daemon
11-30-2009, 01:24 AM
Alrighty then

I retouched up the solders under the ren-c in case. No effect.

I checked the traces and solders especially under U6 and the pins on U6 and it was okay. I dont know how to check if u6(labelled as u2 on the diagram)is bad?

i am using option 2 of the simple ren-t for ZC. The one found here http://doityourselfchristmas.com/forums/showpost.php?p=86285&postcount=5

The cable from the pc 5vdc supply is about 600mm long and is quite thin, only 1mm wire probably.

There is no flicker without the PWM cable attached at any level of dimming.

Daemon
11-30-2009, 05:09 PM
If I am running 2 grinches and a ren-c off the one computer power supply 5vdc do you think I should perhaps try larger guage wires in case I am not getting enough current?

P. Short
11-30-2009, 07:51 PM
It looks as though 1mm wire translates to about 18AWG in the system that we use here, assuming that 1mm is the diameter. What you describe seems fine, as I think that there shouldn't be an problem with about 2 feet of 18AWG wire. You might disconnect the second Grinch to see if that changes anything.

Not to make too fine a point of it, but you are running both ground and power wires from the power supply to the Grinch, and not relying on some other path for getting ground to the Grinch?

Daemon
11-30-2009, 10:25 PM
Yes I am running both ground and power from the pc power supply. I do have a join in one of the grinch wires so I will try both taking the second grinch off to see if anything happens and also replace the wire to see if it has any effect.

I guess what I really am intrigued about is why there would be this effect with the cable that carries the pwm signal and no effect without it? Is there a simple way to block this signal from the non led boards? or an obvious reason for the flicker to occur once the pulse width modulation is sent down the second cable.


If I have no luck at all doing that I will swap the dc boards over to the non dimming grinch and just run those lights without dimming effects. Not the same but better than being barely able to see them at all.

Thanks for looking at it Phil