external power supply on conversion van

Blocking diode yes, BUT.... You have that diode permanently between the battery and the load? If so, a BIG diode for all loads, else a still big diode if those loads are separate from the other battery connected electricals. Then a diode for the PSU. The diodes have probably at least a 0.7V drop across them - maybe more. But even if as low as 0.3V for a low-current Shottky diode - the loads get 12.3V instead of 12.6. The battery must be diode connected so it still discharges to 12.6V (ie, 12.6V less 0.3 diode = PSU 12.6 less 0.3V etc). (Maybe a higher voltage diode for the battery to prevent its discharge?) But the problem may still be high-current diodes - eg, >30A in your case, therefore 0.3 or 1.0V x 30A = 100W or 300W of heat.      The solution to high-current isolation diodes is relays - at least wrt these situations. (PSU & battery isolation etc. Even diode battery isolators have been shunned for ages - they use voltage or charge sensing relays etc.) Lots of ways... If your alternator is charging, that disconnects the NC contacts that allow your AC PSU (eg, see the UIBI battery isolator). Or AC connection with PSU output disconnects battery etc. And I feel my oft recommendation... a voltmeter; maybe 2. One for the PSU. One for the battery. Both 3-digit displays. Usually quite cheap off eBay. (I have a smallish blue-LED meter I integrated into my dash, and others including larger back-lit LCDs - all ~$10 or less from Hong Kong etc; usually well under 10mA current drain (7mA for LED & backlights; less for non-lit LCD).

lol your sparking interests in me that I haven't played with in ages. I just went out and tore apart 2 psu's and temporarily grounded the sense wires just to see what voltage they were pushing. I remember my 350 was pushing 12.6 on the camper. That was the first one I ever did. tonight I did two of them. 200w psu was pushing 11.57v 300w psu was pushing 11.86v I'm not sure how the first one got that high. So now I'm liking your original thought on boosting the voltage output on these. I did read another page on it. I still don't fully understand what he was doing because he didn't explain it like I'm 5. lol What I read I think I understand that he just created a load on the 5.5v wires by throwing a couple of resistors in series. As he mentioned this will generate heat and will require a heatsink to dissipate it. Is this basically what you were suggesting earlier in this thread? I do understand now what you mean about blocking diodes and I didn't think that through well enough to start with. I forget I was working with low power consumption devices when I did this stuff more.

Hmmm - interesting read... Or not? As I understand it, he increases the voltage of the 12V by placing a the 10A load on the 5V rail (2 x 1R (ohm) resistors; V=IR hence I=V/R = 5V/1Ohm = 5A per resistor). That I suggest only occurs in that instance due to its (poor) design. There is a voltage drop along the shared +12V & +5V ground due to the 5V 10A load so the 12V ups it voltage to compensate. That's what I'm suggesting, but instead of 10A and 50W and a bluddy hot resistor, and unpredictable outcomes (depending in track resistance, the load, and the design), I suggest a voltage drop of (eg) 0.6V or 1.2V or 1.8V etc at micro-Amps by using diodes. (A well designed PSU would not have one rail effect the voltage of another.) Two things I like about that link however: 1 - his hover-over color coding (I thought yellow was 12V!) 2 - the next pic (no copyright notices noted) ... where the chip (IC) on the mid-RHS (partly obscured by the case) might be the SMPS chip we are looking for. (It's the black "beetle" with 2 of its 14 legs visible.) And that one still has its numbers! But it might be another chip elsewhere... So, can you see the chips in yours? And post the numbers? (probably starting with letters like MC, ICL, T etc; the "lowest" numbers are usually date codes....) As to your 11.57, 11.86 & 12.6V outputs - yep, that's a typical range. The 5V definitely has to be within +/-10% (4.5V to 5.5V) because more or less can damage traditional logic chips (TTL etc). 12V is often for fans and motors (disks) which are not that critical.    But you see the range - and you want to convert from (say) 11.5V to 13V - 14V. That's a 1.5V to 2.5V increase. If that's resistive, it'll get hotter than the diodes (ie, 1.5V @ 20A = 30W etc). So we find the feedback "loop", splice in a few diodes (can add a switch to short some of them out to drop the voltage) and you set or dial whatever voltage you want. (Well, in 0.6V steps unless you use Schottky diodes or other tricks.)

I'm taking on an education here. As is always with me I have to try things to understand. Since I didn't have the required resistors, I opted for an 1157 automotive light bulb. I threw the red 5v line to the bright side of the bulb (not knowing the resistive qualities) and it boosted the 12v line from 11.87v to 12.06v. (cool didn't see that coming) Now that didn't solve the issue, but it gave me a good understanding of what your suggesting. :) The numbers on the 14 leg chip are... top row = 45D3TVM bottom row = TPS3514N

Looking for a suitable circuit example, I found this page: electronic-circuits.blogfa dc-to-dc-converter. Although I wanted to use one of my aforementioned MC34063 figs from that page, I'll use the "LM2577 5V to 12V DC Converter step up Voltage Regulator" instead... (about 3/4 down the page) Think of the LM2577 as being your entire PSU for 12V except for a few components - like the 100uH coil (transformer) and the 0.1uF & 680uF input & output filter capacitors and a diode and RC crap (2.2k & 0.33uF). It has 5V input which it converts to 12V (using magic and that transformer - aka coil or choke). How does it know it is outputting 12V?   It has a FEEDBACK or SENSE terminal (#2). If the output drops, pin #2 drops and the "magic box" increases output voltage etc. That's called regulation.    What happens if we insert a diode between the output and pin #2? Assuming the correct orientation (ie, the diode's arrow head points from the Output to #2 which means the diode's LINE-end is towards pin #2), then there is a 0.6V drop. So output is 12V, pin #2 is 12V - 0.6V - 11.4V which is not its programmed 12V so it bumps the output 0.6V. (Hence #2 = out = 12.6 less diode 0.6V = 12.0V so all is cool.) Same for 3 x 0.7V diodes = 2.1V => output is (12V + 2.1V =) 14.1V. Too easy eh? Look at most of the diagrams on that link. Almost all have a FB = Feedback = S = Sense terminal. Some are not direct like the LM2577-12 above - they go through a resistive voltage divider (2 resistors with or without a trim-pot (variable resistor)) so that the output voltage can be adjusted. (The LM2577-12 is a fixed 12V output, but we can trick it to go higher as I described - just as we can trick most car alternators to increase their output voltage.) Not that the voltage divider crap matters. All we have to do is insert diodes between the actual output and the "top" of the feedback link. However, if we find the feedback resistors, we could instead add a trim-pot so you can fully adjust the output voltage. Maybe the PSU even has a trimpot that has been locked in place with nail-polish etc! If the 5V rail is similar, we can boost that too. (Main danger there is its output filter capacitors that are likely to be 5V or 10V rating - not the 16V or 25V required for 12-14V etc.) Are you getting a slightly warn feeling? If it's a short-circuit warmth, that's ok. This is a lot in one go. But hopefully you might see that most of those diagrams are all the same as the "simple" one above - an input and output voltage either side of a coil/inductor, and some box etc that somehow controls that "coil converter". Oh - and the voltage feedback. The rest doesn't matter - they only add that to prevent simpletons like us thinking we understand them. (At least for now...)

Dang - I missed your reply. Dang#2 - TPS3514N is not the SMPS - its a voltage supervisor chip (it measures and checks for 12, 5, 3.3V etc and outputs whatever signals if correct or not). But that's good - we may have to trick or disconnect that chip later. So any others - maybe 8 pin chips?

yep 2 8 leg chips first one.... top row = gm3843 bottom row = r345 second one... VIPer22A

BINGO! The VIPer22A seems to be a "driver" chip, but the GM2843 is definitely a SMPS chip. Alas the 2843 is a current-mode PWM (pulse-width modulated; the way SMPS and some LED/lamp dimers work) so there are different "feedback" modes, but its pin #2 is its "voltage feedback" input. So can you trace tracks from #2 to the 12V or 5V output - maybe thru resistors or a variable/trim-pot? BTW - I find that data using google, although for some reason I am not finding or downloading pdf datasheets like I used to. (Probably some config problem since my "still recovering" rebuild.) In fact at the moment my text entry has gone haywire (double line feeds etc), so maybe a good time to reboot. And prepare for an appointment tomorrow (not that I ever get distracted!) But at this rate, you are going to have to change your username pretty soon. Then again, I Am An Idiot is anything but.... And it's not like I'm old... (ha ha!).

You may have lost me here but I'm going to give it a try. I will be out of town for a couple of weeks so this project will have to be back burnered until I return. :( This will give me time to read and learn and understand. :) I might take it with me along with my tools for something to do in my off time.

Well a lil more poking and prodding produced the answer to my design. I don't need to hit much more than 12v to operate everything as long as the amperage to the PSU is high enough. I took a gander at the fuse block for the conversion package last night. There are two blocks one is powered through a relay on the ignition wire to fire only when the key is on,(tv, cb, digital tv box, rear radio) and one block is direct to the battery to operate interior lights etc. Everything I want to power is attached to these blocks save the car stereo. There are extra positions on both blocks so realistically I can take the car radio off the van's oem fuse block and wire it instead to the conversion bus. Then through the use of regular automotive relays, disconnect the conversion bus from the oem system essentially disconnecting from the vehicles battery when the psu fires up. That being said, I am still interested in following through on boosting the psu's output just for the knowledge and fun of it. There's more than one way to skin a cat and I'm fond of a challenge even at the risk of the occasional scratch.