PDA

View Full Version : 36v DC and SSRs



mutleyrover
02-05-2015, 05:02 PM
I have got a variety of LED strings that all have wall warts which supply different voltages to the lights, some being AC and some being DC. This is what they supply:

24v AC

24v DC
25v DC
30v DC
31v DC
36v DC

I plan to use Renard controllers (plus 32, 64XC) with separate SSRs and group my lights together so that the 4 outputs from any one SSR will be the same voltage, that way I can use a single power supply in to the SSR to run all 4 sets of lights.

I therefore need some AC SSRs and some DC SSRs. As far as I can see the DC SSRs all seem to say that they can be used from about 6v to 30v DC. I'm therefore wondering if I can run my lights that require 36v DC. Is that too much power to put through the SSRs or will the boards stand up to 36v DC?

efreeman1980
02-05-2015, 05:42 PM
Maybe a few of these would help?
http://www.ebay.com/itm/DC-DC-4A-4-5-32V-to-5-42V-step-up-BOOST-converter-power-module-12V-19V-24V-36V-/181506032621?pt=LH_DefaultDomain_0&hash=item2a429a43ed

P. Short
02-05-2015, 08:25 PM
There are two considerations here that I can think of. First, the transistors on the SSRs must be rated for higher than 36V if you are going to use a 36V power supply. I'd recommend using transistors (MOSFETs, most likely) that are rated for 60V or higher.

Some of the DC SSRs that are described here or in the wiki use DC regulators to drop the input DC voltage of the SSR down to a much lower voltage used for the optos and the gates/bases of the transistors. These regulators also need to be rated for much higher than 36V. Regulators that would withstand this higher voltage are much less common the regulators rated for 36V or lower, so it would take some research to find them.

Matt_Edwards
02-06-2015, 02:43 AM
Ok the occasion where I needed to run some dc SSRs at 36v, I cut the track between the input voltage terminal and the volt regulator and then inserted a string of 6xSMD diodes.
Another option is to insert a 7824 regulator, but that more complicated.

Sent from my SM-N9005 using Tapatalk

mutleyrover
02-06-2015, 04:22 PM
There are two considerations here that I can think of. First, the transistors on the SSRs must be rated for higher than 36V if you are going to use a 36V power supply. I'd recommend using transistors (MOSFETs, most likely) that are rated for 60V or higher.

Some of the DC SSRs that are described here or in the wiki use DC regulators to drop the input DC voltage of the SSR down to a much lower voltage used for the optos and the gates/bases of the transistors. These regulators also need to be rated for much higher than 36V. Regulators that would withstand this higher voltage are much less common the regulators rated for 36V or lower, so it would take some research to find them.

The SSRs that I was planning to use were the Renard Plus DC SSRs. On the Renard Plus website it says about them being suitable for between 6v and 30v. However, since I posted this I noticed that they've been added back on to Radiant Holidays website with the following description:

The Renard Plus DC SSRhc is a Solid State Relay with 4 independent DC output Channels. The input for the control power is a +5VDC via the RJ45 connector. Each channel is then controlled by sinking this control power for each of the Optoisolators for a given channel. The optoisolators then in turn fire the channel MOFET. The MOSFETS are good for up to 60V and a maximum of 4A each. The total board power is limited to 10A.

From that description, and that being a kit including the BOM, I assume that I would be good to go with running my 36v DC lights with 36v DC power going through the SSRs, supplied by a SMPS.

mutleyrover
02-06-2015, 11:39 PM
Some of the DC SSRs that are described here or in the wiki use DC regulators to drop the input DC voltage of the SSR down to a much lower voltage used for the optos and the gates/bases of the transistors.

I was just thinking about this comment. If I use an SMPS to supply the input voltage (in this case 36v DC), will the output voltage for the lights still be 36v? Is it the case that any regulators will drop the voltage for the optos, but not affect the voltage going through the board and out to the lights?

I'm not an electrician so my knowledge is limited, but I'm trying to learn!!

P. Short
02-07-2015, 12:20 PM
I was just thinking about this comment. If I use an SMPS to supply the input voltage (in this case 36v DC), will the output voltage for the lights still be 36v? Is it the case that any regulators will drop the voltage for the optos, but not affect the voltage going through the board and out to the lights?

I'm not an electrician so my knowledge is limited, but I'm trying to learn!!

Yes, for those DC SSRs that have an on-board regulator. It appears that you don't have to worry about this for the SSRs that you mention in post #5 above, based on what you wrote in post #5 (since I have no direct knowledge of those DC SSRs).

mutleyrover
02-07-2015, 11:55 PM
Yes, for those DC SSRs that have an on-board regulator. It appears that you don't have to worry about this for the SSRs that you mention in post #5 above, based on what you wrote in post #5 (since I have no direct knowledge of those DC SSRs).

Thanks Phil. Looks like they are probably the ones for me to go for. Just need to put an order in now!

Matt_Edwards
02-08-2015, 04:46 AM
Are you intending to use these ones:
http://www.renard-plus.com/dcssrhc.html
I would suggest you investigate further with the designer as I suspect they won't work at 36 volts. The 7805 regulator is not rated to operate at 36v.

Sent from my SM-N9005 using Tapatalk

mutleyrover
02-08-2015, 08:37 AM
Are you intending to use these ones:
http://www.renard-plus.com/dcssrhc.html
I would suggest you investigate further with the designer as I suspect they won't work at 36 volts. The 7805 regulator is not rated to operate at 36v.

Sent from my SM-N9005 using Tapatalk


Yep! They're the ones!! Maybe I'll see what they think...

dlovely
02-08-2015, 12:05 PM
Yep! They're the ones!! Maybe I'll see what they think...

Those are a copy of the the board released by LabRat (http://www.doityourselfchristmas.com/wiki/index.php?title=DCSSR_Version_2.4) who reworked the original by wjohn and will not work with 36V as mentioned already.

David_AVD
02-08-2015, 05:16 PM
Some time ago I did a page on the ACL wiki (http://auschristmaslighting.com/wiki/Controller_Setups_and_Settings#Multiple_Voltages) that shows how to use different voltage lights than what the controller can often accept. Hope that helps.

mutleyrover
02-08-2015, 06:05 PM
Some time ago I did a page on the ACL wiki (http://auschristmaslighting.com/wiki/Controller_Setups_and_Settings#Multiple_Voltages) that shows how to use different voltage lights than what the controller can often accept. Hope that helps.
Thanks for the link, although it's left me somewhat confused!!!
It seems this 36v issue is quite a problem!!

mutleyrover
02-08-2015, 06:06 PM
Ok the occasion where I needed to run some dc SSRs at 36v, I cut the track between the input voltage terminal and the volt regulator and then inserted a string of 6xSMD diodes.
Another option is to insert a 7824 regulator, but that more complicated.

Sent from my SM-N9005 using Tapatalk

Thanks for this Matt. Does this resolve the problem? Have you maybe got any photos to show what you did?

David_AVD
02-08-2015, 06:17 PM
Thanks for the link, although it's left me somewhat confused!!!
It seems this 36v issue is quite a problem!!

You can use one lower voltage power supply to run the controller and a different (36V) one to power the actual lights.

Matt_Edwards
02-08-2015, 07:02 PM
Most of mine are packed away or in sealed boxes.
I only have the original dc SSRs.

I will see what I can do in a couple days.

Sent from my SM-N9005 using Tapatalk

mutleyrover
02-08-2015, 07:16 PM
You can use one lower voltage power supply to run the controller and a different (36V) one to power the actual lights.

Having re-read it properly I understand it now. I guess the downside is having to have a multitude of power supplies!
I guess one obvious question is whether there are DC SSRs that exist that are able to take voltages higher than 30v?

P. Short
02-08-2015, 07:21 PM
That board could be easily assembled to use a shunt regulator instead of a series regulator because the pcb design has two series regulators in parallel. The parts would need to stand on end, but placing a resistor between the input and output of one regulator footprint and an appropriate zener diode between the ground and output of the other regulator footprint would do the job. The disadvantages are that the input voltage range is much smaller than the input range of a series regulator (such as the 78L05), and it draws power all the time, but it provides a way of dealing with input voltages higher than the 40V that most 78xXX regulators can handle.

mutleyrover
02-08-2015, 07:40 PM
That board could be easily assembled to use a shunt regulator instead of a series regulator because the pcb design has two series regulators in parallel. The parts would need to stand on end, but placing a resistor between the input and output of one regulator footprint and an appropriate zener diode between the ground and output of the other regulator footprint would do the job. The disadvantages are that the input voltage range is much smaller than the input range of a series regulator (such as the 78L05), and it draws power all the time, but it provides a way of dealing with input voltages higher than the 40V that most 78xXX regulators can handle.

I think I understand what you're saying here but I feel I'm rather at the far realms of my ability. I'm not an electrician nor do I have any background in electronics. While I feel happy to be able to follow instruction to build a board (from the wiki for example) I think I might be getting out of my depth with modifications. Maybe I just try and get lights that use 30v or less, and get rid of my sets using 36v. That feels very defeatist though!!

mutleyrover
02-08-2015, 07:42 PM
Most of mine are packed away or in sealed boxes.
I only have the original dc SSRs.

I will see what I can do in a couple days.

Sent from my SM-N9005 using Tapatalk


Thanks Matt.

I was wondering what the theory of what you do is? Why/how does it allow you to use 36v?

P. Short
02-08-2015, 07:47 PM
If I'm looking at the correct schematic, this change does not need any modifications to the board itself (i.e. no cut traces and no jumpers), just installing different (and quite inexpensive) components in a different manner than the current instructions indicate. My suggestion is deliberately sketchy to see what reaction (if any) it gets from some of the other people in this discussion, and would be elaborated on later if appropriate.

mutleyrover
02-08-2015, 07:51 PM
If I'm looking at the correct schematic, this change does not need any modifications to the board itself (i.e. no cut traces and no jumpers), just installing different (and quite inexpensive) components in a different manner than the current instructions indicate. My suggestion is deliberately sketchy to see what reaction (if any) it gets from some of the other people in this discussion, and would be elaborated on later if appropriate.

That sounds fair, and from my point of view, reassuring!

Matt_Edwards
02-08-2015, 08:05 PM
My mod is quite simple. Lm7805 have a maximum operating voltage of 35v. So all I do is solder a string of power diodes to drop the voltage. Easy peasy.
In your case, solder a string of 5 or 6 diode back to back (anode to cathode) connect the cathode to the top pad of C2 and the anode to the +ve of TB1.
I use SMD diodes cause I have them in bulk. 1n4001s work well too.

Sent from my SM-N9005 using Tapatalk

David_AVD
02-08-2015, 09:20 PM
Having re-read it properly I understand it now. I guess the downside is having to have a multitude of power supplies!
I guess one obvious question is whether there are DC SSRs that exist that are able to take voltages higher than 30v?

You can use a $3 DC-DC converter to derive the lower voltage for the controller, but make sure the input of the converter is rated for 36V else you'll be back to square one!

g2ktcf
02-09-2015, 12:20 PM
You can use a $3 DC-DC converter to derive the lower voltage for the controller, but make sure the input of the converter is rated for 36V else you'll be back to square one!

Just FYI guys, he has contacted me via email. The only way to handle 36V is to drop the incoming voltage for the Vreg or use a separate source for 5V rail. I thought about Matt's diode solution as well.

mutleyrover
02-09-2015, 12:58 PM
My mod is quite simple. Lm7805 have a maximum operating voltage of 35v. So all I do is solder a string of power diodes to drop the voltage. Easy peasy.
In your case, solder a string of 5 or 6 diode back to back (anode to cathode) connect the cathode to the top pad of C2 and the anode to the +ve of TB1.
I use SMD diodes cause I have them in bulk. 1n4001s work well too.

Sent from my SM-N9005 using Tapatalk

Am I right in thinking then that this allows the use of the higher voltage to pass through the board for the lights, but the voltage that's required to power/operate the board is reduced to a level that is acceptable?

g2ktcf
02-09-2015, 12:59 PM
Am I right in thinking then that this allows the use of the higher voltage to pass through the board for the lights, but the voltage that's required to power/operate the board is reduced to a level that is acceptable?

That is correct.

P. Short
02-09-2015, 01:08 PM
Just FYI guys, he has contacted me via email. The only way to handle 36V is to drop the incoming voltage for the Vreg or use a separate source for 5V rail. I thought about Matt's diode solution as well.

Where is the schematic for the boards that you are selling? If that board has two regulators in parallel (like the original DC SSR boards), there is another inexpensive way.

First, do not install the 5V regulators that are mentioned in the BOM.
Second, take a 5.1V zener diode and install it in place of the LM78L05 regulator (anode to the ground pin of the 'L05 footprint, cathode to the output pin of the 'L05 footprint). This diode might be a 1N5231A or 1N4733A, the exact part to be selected on the basis of working well at a low current level. This diode would not lay flat on the board, rather standing on end, so it might be a good idea to use a short length of shrink-wrap tubing to protect it.
Third, take a 1.5K,1W resistor (selected for pushing 20mA through the zener with a 36V input level) and install it between the input and output pins of the LM7805 regulator. This part also ends up standing on end, so it should also be protected with a short length of shrink-wrap tubing.

This potential solution should be prototyped of course. Its advantages are that it is relatively cheap, and does not require any cuts or jumpers to the board. The disadvantage is that the regulator is always dissipating 720mW of power between the diode and the resistor described above (when powered with 36V), and the value of the resistor would need to be changed if the input voltage is radically different from 36V.

P. Short
02-09-2015, 01:12 PM
Another zener diode possibility is a BZX79C5v1, which would allow the resistor value to be raised to about 6.2K, and the wasted power would be reduced by a factor of 4.

g2ktcf
02-09-2015, 01:16 PM
Where is the schematic for the boards that you are selling? If that board has two regulators in parallel (like the original DC SSR boards), there is another inexpensive way.

First, do not install the 5V regulators that are mentioned in the BOM.
Second, take a 5.1V zener diode and install it in place of the LM78L05 regulator (anode to the ground pin of the 'L05 footprint, cathode to the output pin of the 'L05 footprint). This diode might be a 1N5231A or 1N4733A, the exact part to be selected on the basis of working well at a low current level. This diode would not lay flat on the board, rather standing on end, so it might be a good idea to use a short length of shrink-wrap tubing to protect it.
Third, take a 1.5K,1W resistor (selected for pushing 20mA through the zener with a 36V input level) and install it between the input and output pins of the LM7805 regulator. This part also ends up standing on end, so it should also be protected with a short length of shrink-wrap tubing.

This potential solution should be prototyped of course. Its advantages are that it is relatively cheap, and does not require any cuts or jumpers to the board. The disadvantage is that the regulator is always dissipating 720mW of power between the diode and the resistor described above (when powered with 36V), and the value of the resistor would need to be changed if the input voltage is radically different from 36V.

Phil, I do not have the schematic near me...but it is over on the Renard Plus website (just don't shoot me for posting that). But I do think you are 100% correct that it can be handled in that manner. The OP is the only one that has ever asked be about this question to be honest. However, I'm not thrilled about burning off 720mW just to drop the voltage though!

Matt_Edwards
02-09-2015, 04:22 PM
Another option the OP should consider is running the 36v strings on 33v. That works quite well in some scenarios, but not all.

Sent from my SM-N9005 using Tapatalk

mutleyrover
02-09-2015, 04:26 PM
So which idea do people think is the best? Pros and cons to both I guess...

I can see Phil's needs no change to the board but it sounds like it wastes some power, and Matt's needs some modification to be done. I've got lots of 24v DC lights sets - maybe I should stick to those!!!!:biggrin2:

mutleyrover
02-09-2015, 04:28 PM
Another option the OP should consider is running the 36v strings on 33v. That works quite well in some scenarios, but not all.

Sent from my SM-N9005 using Tapatalk

Could you get away with 33v without any changes then.

Matt_Edwards
02-09-2015, 05:23 PM
If you have some, try them out.
But reading your last post implies you may stick with 24v strings. One advantage to this option is commonality, and a lower risk of plugging the wrong string into the wrong voltage.

Sent from my GT-I9505 using Tapatalk

mutleyrover
02-09-2015, 05:35 PM
If you have some, try them out.
But reading your last post implies you may stick with 24v strings. One advantage to this option is commonality, and a lower risk of plugging the wrong string into the wrong voltage.

Sent from my GT-I9505 using Tapatalk

The big issue I have is I've got strings with a complete variety of voltages (as per my original post!). The majority are the 24v ones, along with a large number of 24v AC ones as well. Having said that, although I don't have that many 36v strings it would be nice to use them if I could.

David_AVD
02-10-2015, 03:59 AM
Using multiple voltage strings with a single controller is not hard to do. The ACL wiki link I posted has details.