Switch mode v Traditional isolated Power supplies help :(

[pigman]

Hey mike, After much testing I'm basically back to square one. Seems the voltage dip when the unit powers on is always enough to make the relay trip on and off constantly, usually after the PSU has been running for about an hour or so (not hot at all) but after that I disconnect and reconnect and it will chatter the relay and never get up to full voltage (hover at about the cut point for the relay). No matter what circuit I use. Seems I may need to revisit your bistable suggestion. If I understand this correctly once powered on, even if the circuit dips in power it will ensure will keep power on to the relay until there is no power left ? I'm struggling to get my head around this bistable idea.
Here is the link I've been checking out, it still doesn't make 100% sense, any ideas on a solid ON switch/trigger thanks in advance, again....
**broken link removed**

 

[MikeMl]

If the voltage is sagging to the point that the relay drops out during normal operation, then either the load is too big for the PSU, or the PSU is too small for the load.

The only hope you have is to devise a means of always having the battery in the circuit, and allowing the battery to supply the current peaks that are happening during starting and even during normal operation, and then configure the PSU so that it acts as a float charger for the battery. This way, the circuit only draws down a tiny fraction of the battery capacity, so if the AC were to fail, the battery is always nearly full to keep the load going.

I do this with my SSB Ham rig. The radio draws ~1A on receive, but up to 30A on voice peaks during transmit. A 25Ah SLA battery is connected in parallel with a 3A PSU. The duty cycle of the transmitter is low enough such that the 3A PSU is sufficient to keep the battery charged and run the station for ever...

Your problem is going to be that to charge a 12V SLA, you will need the PSU to put out a minimum of 13.8V.

 

[pigman]

Mike, yes. It seems to be the issue. I managed to get 10,000uF on the PSU and it charges that fine. It also cold starts the device without any issue(when it hasn't been run for a while, nice and cold) and if I power the caps and then manually shove the device wire on the caps pin it powers up not a problem... So I still feel that a stable switch that would come on at say 11v but not go off until the voltage is at say 5v then I think I'd have a solution. I just can't think of a way to do that with my limited knowledge

 

[MikeMl]

To get the relay coil to stay pulled in even though the voltage sags to 5V, we could start with a 6V relay, and build a current regulator. I'll give it some thought, but not today...

 

[4pyros]

The issue
I have a 12V 5A switchmode PSU in play and while it works fine 90% of the time, sometimes (only when a cold start situation) the PSU just cant handle it and instead of firing up it will start flickering the power and putting my multimeter on it I'm getting random hz from 1 to 200Hz output and low voltage. and unless I dissconnect it, it will never get to providing the good 12V source my device requires. Sometimes it will work if I turn off the load device and let the PSU get its act together then turn them on, but even then only half the time that will work.

Most likly you just need a load on the SMPS. Thay will not start right sometimes without a load. You may be able to find a min load spec on your SMPS, just add a load resistor on the output of the SMPS and it should start right all the time. Andy

 

[pigman]

Mike, thats a good idea, a lower voltage relay to trigger and stay on, I guess the issue is to get it to trigger high say when the voltage reaches 11v but to stay on low which isn't easy. I'm at a bit of a loss at the moment.

Andy, Thanks for your reply mate! the SMPS starts ok each time, the circuit I've been working on has about 12,000uF of caps to charge and the voltage comes right up to ~12v without any issue, the problem is that the SMPS output voltage then dips to ~9-10v as the relay kicks in (on a delay) and begins its chattering process on when the power rises and off when it falls. I guess I really need to trigger the relay when 12v is reached but have it hold on even if the voltage dips to as low as 6v to prevent the cycling of the relay which then is hard to stop.

Almost like I need some hysteresis in there to bring in high hold on until low. I'm not good with hysteresis circuit design.

Thanks in advance guys! This is a doosey of a problem. I will post up the final schematic when we get this sorted, maybe it can help someone else with a similar problem one day

 

[pigman]

Hey, I have an idea. What about a 6v fixed voltage regulator to power on a relay, the regulator won't output 6v until the input of the regulator reaches about 8v and will hold 6v solid from 8v and up. the 6V relay or similar will switch the 12v relay on and hold it on even if the voltage dips to ~8v or likely lower as a mechanical relay would remain on lower than 6v ? It is relatively simple but might work? Maybe a variable resistor in the chain with the 6v relay as per one of mikes previous schematics to tune the cut in of the 6v relay.

Any thoughts?

EDIT - Actually this would have the same issue, even tho the 6V relay would stay activated the 12v relay would chatter. I guess a cap on the 12v relay with a diode blocking back flow when the voltage dips could hold it on for long enough to prevent the chatter! May be on to a winner here?

 

[4pyros]

Ya so why cant you run a lower voltage relay from a regulated supply? Andy

 

[pigman]

Andy, Yes that's my current train of thought, I'm thinking a regulated supply and something like this **broken link removed** a regulated supply to trigger it and and minor dip in voltage shouldn't be enough to cut out the relay, devilishly simple. I may still delay the cut in of the 6v relay until the source reaches 12v, but this "should" work. Mind you I've thought that before

I let you guys know how it goes.

 

[4pyros]

Hay Have fun and let us know how you make out. Andy

 

[pigman]

Hey Guys,
After much research, testing, and money spent. I am back to one of Mikes suggestions about installing a Battery in the loop to handle the peak load requirements. Now I'm looking into what sort of circuit I would need to use to charge the battery during operation. I will refer to the battery inside the PSU as the PSU Batt and the main batteries as the Main Batt to reduce confusion in this little idea I have. Here goes.

Charging of the PSU Batt - Could be performed from the Main Batt when its voltage is high isolating the PSU Batt from the Main Batt using a diode, this will prevent the PSU BAtt from disscharging should the Main Batt lose voltage.
Charging of the PSU Batt while under 240V mode/operation - This is my biggest question, do I need some sort of Charging circuit for the battery while 240v is running and the system is operating on batteries? This could potentially be 90% of the time in some situations so I'm concerned about simply putting a switchmode PSU across the battery terminals and expecting it to last? And suggestions on this would be great!
Isolating the PSU BAtt from the Main Batt during 240v operation - I suppose another relay or NC FET? here could be used to simply cut connection between the PSU Batt and Main Batt during 240v PSU powered operation?

Any other ideas I haven't thought of?

As the main load of my device is ~300mA but clearly peaking high enough to cause a SMPS to struggle starting without having an error the Battery will fix the issue like a CAP has failed to do.

Here is hoping! Looking forward to your responses, I feel confident this will fix the issue I have now and I might be able to start getting some sleep again Thanks guys!

 

[MikeMl]

Floating a battery across the output of PSU can be easy, or it can be hard. Depends on several factors: Battery chemistry? Regulation of PSU? Expected Ambient temperature variation?

In the meantime, read this: Pay close attention to the temperature dependence of the "float" voltage...

ps, there is one other consideration. If the battery gets mostly discharged, and then the AC is restored, you might also need to limit the charging current into the battery to stay below the maximum charging rate specified by the battery maker. Read the data sheet for a typical SLA here. Note that this one should be limited to a maximum initial charging current of only 2.1A.

 

[pigman]

Mike, Thanks for that I'll go through it shortly. The Battery I'm thinking of using is this **broken link removed** and the temperature variation can be quiet a lot usually it will be ambient, however the Heat output of the PSU and the location, adjacent an engine, would need to be considered. I would say 90% of the time ambient plus PSU heat inside casing. And 10% of the time 40-50 deg C operation (while the engines are running) however usually at these times the 240v and PSU Batt system will be off and if required/in the circuit charging from the MAIN Batt.

EDIT - 2 of those batteries in series or one 12v version of same.

 

[pigman]

well, after much research, testing and looking into things I guess there is two ways to do this. Also I've been testing with a 12v 1.3Ah SLA

The Over Engineered "correct" way.
Some sort of pic controlled solution? That has smarts to initially check the SLA battery voltage before switching over to PSU and if voltage is low start a charge cycle via a current limited circuit (based on the battery chosen max charge rate), periodically checking the voltage and when the correct voltage is achieved have it switch over to the non current limited "operation" mode where it would switch power over to the PSU and SLA Battery providing a non limited current from the smps that would keep the battery charged and power the device/load. When PSU power is off have the battery system simply dumped back to the main battery source for the Device via a Current limiting circuit and diode to isolate, this will keep the battery charged and prevent it discharging should the source batteries drop low.

The Not so over engineered way of doing things.
The way would basically have all the same stuff but no pic and no initial voltage check. Simply a 10-15W SMPS hooked up in paralel with the SLA Battery and the Load and regulated set to about 13.5v this would top the battery up if required and maintain power to the load/device while the battery pics up the slack/peaks. I would still use a current limiting system and diode to keep the unit charged up during the SMPS off periods. But this one trusts that the Battery will always maintain at least 10-11v minimum. Being SLA units they take a while to loose voltage. But I'm jsut unsure what to do.

This is how I have it configured for testing and my load on the SMPS is ~300-500mA (peaks maybe 600mA sometimes and I think I spotted a 800mA at one stage) with double load, that is two of my load devices operating, and their load is fluctuating from 0 - 3A. Mind you all of this is being tested at the moment on only 12.5v as I haven't hooked up the tweeked SMPS yet.

I think it would be best to do the "correct" version however I am not really at the stage where I'm ready to program a pic or design that circuit. Any suggestions of another solution? Maybe there is a charge IC already avalable that would suit my requirements?

Also any suggestions on Battery size requirement? the 1.3Ah seems to be going ok, but I guess I'd need to run it for a looooong time to really know.

Thanks again guys!

 

[pigman]

I've started to create the program for the circuit. I'm using a 12F675 PIC as it has enough I/O's and a Analogue to Digital conversion for checking the battery voltage.

 

[pigman]

Hey Guys, I've written the program for the PIC !

Now I'm looking for a suitable SLA charge circuit system. I'd like to use the 13.5v to charge the system (already regulated supply). I have looked around for a good charge circuit however most are coupled with a specific charge regulator that automatically takes care of the voltage and amperage to the battery. What I need is just the ability to limit the current and to know if the voltage should maybe be ramped up gradually if the battery is at a heavily drained voltage or something?

Thanks guys.