PDA

View Full Version : Tips for troubleshooting DC SSR



toodle_pipsky
10-30-2011, 10:59 PM
Hey hey,
I need just 1 DC ssr to work . . . . and guess what out of 3 that I built (and worked last year) none work now. I'm using 4.5V, is this too low, better ask that first off!

If it isn't, I can see power coming into the board, but I can't see it in the output - so something's broken in the circuit. Are there any common things to check in a DC SSR?

Thanks guys.

budude
10-30-2011, 11:19 PM
Does the power LED light up when it's plugged into the controller?
You can check pins 2,4,6,8 to see that they go low when the channel is active - this will light up the opto and should turn on the circuit. If this is working but the lights still don't work, then it's the DC side vs logic side from the controller.

Pull the opto and run a wire from pin 16 (+5v - check that too!) of the socket to 15, 13, 11 and 9 to turn on channels 1,2,3,4 in order. If that doesn't work, then it sounds like the MOSFETs aren't turning on. If this does work, then it sounds like something is wrong with the opto.

Matt_Edwards
10-31-2011, 12:23 AM
Brian has captured most of the steps.
It is made easier by having a SSR tester.
( if you have a spare AC SSR PCB I can show you how to make one in 1/2 hour or so. it uses 4 AA batteries)

I assume you tried your "Faulty" DC SSR on another Controller channel?

Did you use insulated FETs. If not make sure the tabs are not touching each other.

kychristmas
10-31-2011, 12:33 AM
All good advice, but I would certainly take a close look at the RJ45s involved. Some pins on a number of mine were completely corroded away. I had 4 SSRs and then 3 or 4 jacks from my original 64SC that were bad.

Kelly

toodle_pipsky
10-31-2011, 06:35 AM
Thank you very much for these instructions


Does the power LED light up when it's plugged into the controller?

Yes


You can check pins 2,4,6,8 to see that they go low when the channel is active - this will light up the opto and should turn on the circuit. If this is working but the lights still don't work, then it's the DC side vs logic side from the controller.

Sorry for being dense, what does "go low" mean? Also I assume I'm measuring DC V - as a sequence is running or all channels on? Also I couldn't see it marked on the data sheet, but I assume the pin numbers on the opto go 1 starting at the dot then go down the same side to 8, how does 9-16 go on the oposite side?


Pull the opto and run a wire from pin 16 (+5v - check that too!) of the socket to 15, 13, 11 and 9 to turn on channels 1,2,3,4 in order. If that doesn't work, then it sounds like the MOSFETs aren't turning on. If this does work, then it sounds like something is wrong with the opto.

Once I know what pins are what number I can do this (I think).

Also I have tried other outputs, plus I have a simple tester that tests output from my REN 64 (which is the controller I'm using) so I know the output signal is working eliminating the controller.

I'm very sorry, to be asking what is probably basic questions, so bare with me. Thanks for the help guys.

budude
10-31-2011, 11:12 AM
Probably the easiest thing is to point to the schematic located here: http://www.doityourselfchristmas.com/wiki/images/0/09/DCSSR_Sch_1.1.pdf

The schematic shows the pin layout for the opto as well - 1-8 down the left side and 9-16 up the right side. Run a a sequence where it's 1 second on, 1 second off and loop it for all 4 channels of that SSR. Measure the DC at pins 2,4,6 and 8 (on the opto or the RJ45 socket) and you should see it toggling between 0v (low) and 5v (high) once a second just as the sequences does. If that is working then you know everything on the logic side (up to the opto) is working ove.properly and you can concentrate on the DC/power side as described above.

Good Luck!

LabRat
10-31-2011, 12:11 PM
4.5v is too low for the 7805 to operate properly. I suspect that you would not be getting a signal strong enough to trigger the FETs.

If you really do need 4.5v, then you may need to remove the 7805, and jumper the in coming V+ to the supposedly rectified V+.

(hope that makes sense)

If you have already done this.. and don't have a regulator on your board.. then please ignore the posting. :)

toodle_pipsky
10-31-2011, 05:46 PM
Thank you so much guys - that is an excellent explaination Brian, appreciate it. I'll give it a bash this arvo, pfft have to go to work instead of doing blinky flashy. ;)

toodle_pipsky
11-01-2011, 08:44 AM
Ok, further developments!

Tried the testing suggested - nothing worked, no readings, no lights. Thinking in the same vein that 4.5V was too low, I tried 24V lights and they worked, but were very dull. There's 24VDC going in but only 7.5VDC coming out at the channels - no wonder the lights were dull! I retried the testing, 3.3VDC on pins 2 etc. of the opto and it didn't fluctuate (had the 1 sec on/off sequence as suggested). The testing with the opto removed and jumping pins didn't work at all. I also tried this with another DCSSR and I had the exact same readings - could I have consistently stuffed something up with these boards? Or could something consistently not worked in 2 boards? (Couldn't find number 3 to get the tie breaker)

Also Matt, should have said, if you have instructions (or point in the right direction) the tester would be handy.

Thanks for the help and tips guys.

LabRat
11-01-2011, 09:00 AM
Have you got a picture of the board? (I can't for the life of me see how you would get 7.5v @ the output, given a 24v input)

toodle_pipsky
11-01-2011, 10:53 PM
Here you go.
12107

LabRat
11-02-2011, 08:22 AM
Ok. That's the board I was thinking of.

So with the channel turned ON, (and held on) you are only getting 7v measured on the output. (very odd).
With the device in a "steady state", you should be able to probe with a basic DC voltmeter and track down where
the failure is.

As you mentioned that you *can* get the device to turn on, (only with 7v instead of 24), implies (to me) that the
"control" part of the circuit is working. Which would imply the failure has to be on the FET (output) side of things.


I would look/check for things like.
Cold solder joints (or corrosion) at the base of the FETs? Perhaps a careful reflow of those joints?

toodle_pipsky
11-03-2011, 08:22 AM
Thanks for the advice - I did touch up the solder joints, but still the same results.

I went to grab a new FET to replace one and see if that was the problem. I did a little cross checking of the part and it looks like it's a part that's substituted. Because it's over a year since I made them I couldn't tell you if someone told me, if I made the choice or if I read it somewhere. Because Mouser is a pain to buy from unless you have a big order (that's for us Aussies at least) I was sourcing parts from a couple of different local sites. For the most part you could find the exact same thing. For what ever reason I bought this for my DC SSRs http://www.futurlec.com.au/transistors/BUZ71.jsp I notice it's very similar, however it's different none the less - could this explain what I'm seeing? If so is there anything I can change (e.g. resistor values) or should I take them out and get the original part?

If it's not an answer, well I'm stumped. Thanks guys for the help though. :)

LabRat
11-03-2011, 09:02 AM
Scratching my head here. Where's your DC coming from? Could it be 1/2 wave rectified (ie. not a solid 12v ?)

toodle_pipsky
11-03-2011, 05:29 PM
No it's not rectified, it's a DC power supply.

LabRat
11-03-2011, 06:09 PM
I wish I had an easy fix/answer here. I'm going to re-iterate what I think has been tested, and you can confirm if I got it right or wrong.

When driving at lower voltage (4.5v), you get nothing on the outputs.
When driving at a higher voltage (24v), you get 7.5 on the outputs.

Hang on... what are you using to drive the opto's? What's the controller?
I'm guessing a REN without the PWM firmware. So that it's triggering as it would
a regular SSR (not a DC SSR). That is to say sending a short signal relative to the
zero crossing, versus a series of pulses throughout the duration of the PWM cycle.

Have you got an SSR tester handy? Or just 3 AA batteries, that you could hook up
at the RJ45 to manually toggle the DC SSR on/off??

budude
11-03-2011, 07:41 PM
This is odd for sure... I would try the direct jumpering I mentioned with the opto pulled. If *that* fails then it must be something on the DC side of things and then you know it's all something with that side of the board. What are the resistor values you are using?

toodle_pipsky
11-03-2011, 10:40 PM
When driving at lower voltage (4.5v), you get nothing on the outputs.
When driving at a higher voltage (24v), you get 7.5 on the outputs.[/QUOE]

Spot on - that seems to be it.

[QUOTE=LabRat;180554]Hang on... what are you using to drive the opto's? What's the controller?
I'm guessing a REN without the PWM firmware. So that it's triggering as it would
a regular SSR (not a DC SSR). That is to say sending a short signal relative to the
zero crossing, versus a series of pulses throughout the duration of the PWM cycle.

Ren 64 - no problems with AC SSRs so far. I have a few running rectified LEDs. (On a side note the reason I wanted the DC SSR for this was that I already had the DC power supply and SSR built).


Have you got an SSR tester handy? Or just 3 AA batteries, that you could hook up
at the RJ45 to manually toggle the DC SSR on/off??

No I don't but point me in the direction of insructions and I'll build one (I'm guessing that's what Matt mentioned a few post earlier).

toodle_pipsky
11-03-2011, 10:44 PM
Resisor values - 680, 470 & 10K.

R1, 2, 3, 4 = 470
R5, 6, 7, 8 = 10K
The rest = 680

LabRat
11-04-2011, 06:44 AM
SSR tester.. very simple.

You want to drive +5/GND down two pins of the RJ45. The easiest build would be an RJ45 male connector on a cable, and loose wires at the other end, and a battery pack delivering ~5 v. (I've been using four AA's (6v ish) without issue)


RJ45 pin 1 to +5.. then touch the wires for pins 2,4,6,8 to GND. That should "turn on each as they touch".

There are Eagle files in the file library, and it was based on Dirknerkle's schematic, but I don't know where the schematic is.

EDIT: Fix the "how to portion"... I had it all wrong (was too early in the morning)

LabRat
11-04-2011, 07:49 AM
Further reading, and my hypothesis as to what is going on:

From "Le WIKI":

Non-PWM operation causes a narrow output pulse (30 μS, or so) to turn on the SSRs. This results in the lowest PCB current requirements, because the waveform driving each SSR has a 1:256 duty cycle. The drawbacks of non-PWM operation is that it will not properly control any channel status LEDs (present only on the Renard64 board), and may cause some issues when driving LED strings (e.g. they may possibly not turn on at the upper 25% of the brightness range).

Can you confirm that your REN64 status LEDs are coming on? This will shoot down my theory right away. If not, then perhaps you need the PWM specific FW version.

TimW
11-04-2011, 07:59 AM
Hi TP

With the tester you connect pin 1 (and/or pin 3) of the RJ45 to +5V and then ground pin 2,4,6,8 depending on what channel you are trying to turn on

Looking at an RJ45 plug facing away from you with the tab underneath the pin numbering is 12345678

The DC ssr (and the ACSSR) is 'common anode' which means all the opto leds are commoned to +5V (via resisitors) and then the switching is done inverted. Grounding the channel pin turns it on.

LabRat
11-04-2011, 08:03 AM
Wot he said... (I had it wrong... it was too early in the morning... I've updated my description. (sorry about that))

jboyles7
11-04-2011, 10:09 AM
Going out on a limb -- I've read the thread and I'm wondering what would happen if you switched the 24V power supply to a different one. The control side seems to work, the SSRDCs worked last year, 7VDC seems very odd, the most likely thing that changed is the power input, since it is very unlikely that the 7805 is bad on both boards and that is really the only common denominator for all outputs on the FET side outside of the input power. Are you getting 5VDC out of the 7805? Humm.. .. . .

toodle_pipsky
11-04-2011, 07:08 PM
Further reading, and my hypothesis as to what is going on:

From "Le WIKI":


Can you confirm that your REN64 status LEDs are coming on? This will shoot down my theory right away. If not, then perhaps you need the PWM specific FW version.

I don't have the status LEDs installed. I can't remember if I couldn't get the resistors at the time or just needed the board working and couldn't be bothered soldering all those lights! The only LED I have working on the board is the power indicator.

LabRat
11-04-2011, 11:08 PM
Ah hah!! I would guess you have the non-pwm software in the RENS, and thus it is only giving a very brief "turn on" pulse to the optos. Versus the "turn on and stay on for this duration" pulse, that the PWM would be using. If you have a spare 16F688, and can program the PWM code, I *suspect* that the problem will go away.

toodle_pipsky
11-06-2011, 10:36 PM
It is using the PWM code, I wouldn't have flashed it with anything else and other SSRs work fine. However, I have been so very wrong so many times before so I will reflash with a fresh copy downloaded from the wiki now. :)

It's been so long since I've had to do this - 1 = PWM and 0 = Non-PWM, is that right? Thanks! Don't worry I found it - that is right.

toodle_pipsky
11-07-2011, 06:49 AM
I wanted to be a stuff up, and to have a simple answer . . . sorry still the same, reflashed the pic and no go. While typing this post I thought I would go grab my Grinch/Ren C just to test. And guess what . . . . . it's fine. DC SSR works fine, lights nice and bright. So it's not a DC SSR problem it's a Ren 64 problem, my memory of it working last year and not now must be that I ran it off the Grinch/Ren C and not the Ren 64.

Thank you so much for your tips and things to check for the DC SSR, I hope someone finds it helpful in the future. I'll have a search around for info on the Ren 64 and start a new thread if I need to troubleshoot it. Thanks again guys, especially Labrat.

LabRat
11-07-2011, 07:35 AM
I hate leaving this unsolved. Any chance of getting a scope on the outputs of your REN64? (or... can you ship me a copy of the HEX file that you programmed? I can put that in a PIC here, and see what the output signal looks like).

As the DCSSRs are working, and the output of a REN is a simple ON/OFF, and your AC SSR's are working... I'm still convinced that it has to be related to how the signal is being strobed versus held high.

PYoung
11-22-2012, 01:30 AM
I know This is a year old but Iím having a vary similar problem 6-7V output with a 12v input The ssr was working for a few days of testing then quit it was running 5050 RGB 300ct LEDs.

I thank what caused the problem was not having a big enough PS for all the LED. And now I cant figure out why it wont work. Its attached to a Ren 64. And i swapped the display with another using the Same 64 port same cable same power supply (The new bigger one). And the test display/ ledís worked fine. The ledís work when directly attached to the power supply.
At first I thought I somehow fried the MOSFETS. I was using only three so I thought if the 4th one works thatís what must of happened the 4th one doesnít work either. I tried the direct jumpering method with the opto pulled and It didn't work so Iím not sure if I didnít do it right or if the problem is between the opto and the Mosfets.