Would it be possible to use a buck boost to raise the voltage powering a tx off say a large leisure battery from 12.5 v to 15v to obtain the max power of the tx .....im coming from the other thread where it seems that to get 15W out of a pcs 3000+ it is necessary to feed it with nothing less than 15v.
I see some of these converters are cheap and im thinking something like an exciter isnt going to appreciate the kind of artefacts created by a cheap arsed buck boost ....maybe im wrong tho ?

No, it works but since all boost converters are not created equal - you will need to look at the output carefully on a scope and if needed, add extra smoothing/filter it. For good measure I'd also look at your exciter's output on a spectrum analyser and if you observe artefacts spaced away from your carrier by the converter's operating frequency, you'll need to do something about it - cures for this include a series resonant trap to earth from the exciter's positive rail or from a judicious point in the exciter output stage's DC load (where you'd normally put a trap for LF oscillations).

Shuffy's right, make sure you check for noise.

Buck and buck-boost converters can be a nice efficient way to solve power conversion problems like this but it's just a shame how noisy some can be. Same goes for all the other switched-mode malarky.

Yeah I've tried this and it worked fine with low power but as soon as it started draining a little bit of current there was the most awful whining noise. It' also unsafe to let the battery drain below a certain voltage so you'd need auto cut off switches for low voltage (like 11v).

To be honest, the extra couple of Watts you're going to get by increasing the Voltage really isn't worth bothering about. You're not going to notice any significant difference in field strength between 10 and 15 Watts. It's seldom worth screwing the last possible Watt out of a rig!

Good point by Albert, between 10 and 15W you won't notice much difference but some higher power bipolar devices designed to work around these voltages are capable of a useful increase with those few extra volts... as other contributors on this thread can confirm...

SMPS and converter circuits look simple but they're not easy to get right, especially for RF applications. There are lots of critical factors. There's a slew of these boost converters about, and the performance varies wildly. Two popular ones on Ebay are the "red" converter (e.g. item 302056764482) which can be made to perform fairly well up to around 100W but has a small design flaw in that one of the tracks under the board (around the sense resistor IIRC) is too thin and has been known to melt if you push it too far. At the other end of the spectrum is the "green" converter (e.g. 301947970433) which due to some embarrassingly fundamental design issues is noisy on or off load and gets dangerously hot - and I mean dangerously - a combination of operating frequency, electrolytics with too high ESR and PCB track length and routing issues. It's only worth buying for spares - go for the red one. You will need to check for noise though, as mentioned before.

As regards your leisure battery, if you're running a 15W rig, you'll probably be drawing no more than 2.5A. If your battery is in good condition, well charged and you're not running it in freezing conditions, you should be good for a weekend. As Tecknics said do NOT suck current out of it when the voltage goes below a certain value - you'll kill it. The precise voltage depends on capacity, load, and battery condition, and is usually between 11V and 11.5V.

I came up with a battery disconnection device, using a relay and an LM311. I used a 78L05 as a reference and a couple of resistors as a potential divider to sample the battery voltage. There's a "push to start" button, which pulls in the relay. The supply voltage for the comparator comes from the output side of the relay, and the voltage sensing is done at the output terminal, so as long as the voltage "seen" is higher than the threshold set (using a multi-turn preset), the relay sustains when the button is released. When the battery voltage reaches the threshold, the relay drops out, turning everything off.

A later version used a dual comparator (LM393) and there were two thresholds set - a slightly higher one which switched on a sub-audio tone into the rig, and the lower one that turned everything off. The sub-audio tone could be detected back at the studio, and denoted that the transmitter would be on for about another 10 minutes at best, so we could make a close down announcement, play the theme and turn off. This proved really useful for the "green" transmitter set-up I used in California - the batteries were charged by a combination of solar cells and a little wind turbine - we never had a premature close down!

Shuffy - I find that I have one of the "green" converters here - it came in a bag of miscellaneous junk from a friend’s workshop clearout. The ESR of the electrolytics is OK, but the board layout is a joke! I think that I'll take your advice, and strip it for parts.

The heatsink is a nice size for a little 20 Watt rig (Brian - it'll be ready tomorrow and LB will deliver it on Sunday! You just need to load up your RDS and set the frequency).

Albert - four or five years ago a station I was working for, melted a red converter. I'd been surprised it had worked so well (apart from melting of course) so I sourced a green converter to see if it was any more robust. The layout (primarily) caused the problems - I think it was the input supply which went all around the houses and nearly melted the whole board! You could also hear the inductor whining from a mile away. I re-used the FET and rectifier in a bespoke design based on the red converter but with a beefed-up layout and posh Panasonic electrolytics. The current demand was still rather high for a boost converter being used in that configuration and nowadays the station, as far as I'm aware, uses a different solution.

You're right, that's a nice heatsink but not worth 6 quid on its own!

Even simpler Low Battery switch-off:
I've used this circuit (or variants of it) for years. The "on" resistance of the FET is a few mΩ, so the FET doesn't need much heatsink. The full version of this circuit uses the other side of the 358 for another "warning" output, switching on a sub-audio oscillator. The zener chosen is 5V1 because this has no significant temperature coefficient (a weird property of silicon zener diodes at around 5V).

The second transistor is there to raise the threshold once the circuit has been tripped, ensuring that the (brief) recovery of the battery when off-load doesn't cause the rig to come back on. I had one rig with a battery disconnect circuit that turned into an RF lighthouse, with the rig coming back on every few seconds and then cutting off again. I made certain that it wouldn't happen again!

The final version also has a "power" and an "output" LED (I usually use common anode, dual colour ones). If you're really concerned about isolating the supply to the rig, make the "load" a car headlight relay (they're really cheap from Halfords, handle loads of current and have 12V coils).

Calibrating the under-voltage circuit is easy - connect an LED in series with a 2k2 resistor to the load output, set the preset resistor for maximum resistance and put 12V from your bench supply across the supply terminals.

Press the "start" button. The LED will light. Turn down the bench supply to the voltage at which you want the supply to be cut off - probably about 11V5 - and slowly reduce the preset resistor until the LED goes out.

Turn the supply back up to 12 or 13V and the LED should remain out. Push the "start" button and the LED will light. Slowly turn down the supply voltage until the LED goes out. Double check that the supply voltage is the value that you previously set. Job done - it's ready for use.

It's actually easier to do than to explain!

Thanks.. some food for thought there.
My two leisure batteries have been on a solar panel setup with a charge controller with a variable minimum voltage cutoff so although their not new they've been treated OK . I might have to investigate a booster and see if I can get a couple more clean watts ...and Albert's cut off circuit is a no brainier very handy thanks . I want to keep my batteries healthy theyre not at all cheap to replace

You want to get friendly with an industrial burglar alarm maintainer - that's where I always got my 1 or 2-year-old lead-acid batteries for nothing. They are replaced as part of Routine Maintenance at either 1 or 2 years! They're always as good as new.

I've used batteries with Voltage multipliers to drive valves! Back in the 80s, I had a supply of valves that were good for 400 Watts with just 4 Watts of drive, with 1200V on the anodes. My power supply used a transformer out of a knackered computer PSU, with a PWM-controlled oscillator at 100kHz driving a half-dozen big power transistors to feed the transformer. The output side had a bridge rectifier and smoothing, with a voltage-monitoring potential-divider to control the pulse width of the drive oscillator. It gave 1220V off load, and the ripple - under load - was a few tens of millivolts.

It might sound like a lot of hassle, but when you think that 400W of transistors would have cost nearly £1k in those days, and they were really difficult to get hold of..... An MRF237 would drive the bottle to full power!

So thanks to shuffys recommendation I bought a red converter off ebay and this little unit managed to bump my voltage to a modest 15v from 12ish without anything bad happening to the signal. I dont have sophisticated gear to measure noise floor or artifacts but i honestly couldnt hear any difference at all.
After being set to 15.09v the booster seemed happy to stay there even though I left it for 12 hours powering the tx at 15 watts, despite the battery losing some voltage across the pins everything else remained the same.
I bought one from the uk so it was a little bit more expensive but if you can wait for the post from china they can be had for two quid. well worth it if your on the lpfm / battery tip
Thanks shuffy

You're welcome Ragga. Don't forget though, the converter will have a certain amount of noise on its output (worse at higher current) and your exciter may be dealing with it OK but don't forget to check when you get the chance!

I will do that shuffy. So far I can't detect any problems , the converter runs quite cool and voltage is stable . I guess I'm not really making a big demand voltage wise and my current is probably gonna stay under 2A so I'm hopeful that things will remain quiet and stable but I will keep an eye on it . Thanks fella...my problem fixed for a couple of quid !
Next on the to do list is Alberts cut off circuit