parallax power supply 7345ru parts PWM

[Cstoliker]

My 7345 is not charging the coach batteries but does supply 12power to all accessories. The transformer, fan and relays work fine.
Think i need a PWM circuit for it. Contacted PP tec support and they seamed more interested in selling me an upgraded replacement unit rather then parts to fix this one. Anyone know of a way to get a PWM for this? (Charging controller)
It's a retrofit from an original Magnetek 6345 wich was serviceable. Parts were available and the schematic is even available online. Unfortunately the 7345RU has non of theses things except for a "flow chart" and "installation instructions".

Or even if someone know enough about these to tell me how bypass the PWM and get full output to batteries. I'm not going to leave it plugged in but I'd be nice to be able to charge my batteries in a couple hours running off generator..

[pnichols]

Sal,

My digital meter being far away from the terminals has nothing to do with it's accuracy in showing 7345 terminal voltage if the meter is drawing no current and there are no other loads off the wire leading to my voltmeter from the converter's terminals.

How is turning on LED lights no proof? Each one loads the 7345 a tiny bit more. If the 7345 has a feedback control loop with large enough correction bands (i.e. it's control loop is "sloppy"), then it's terminal voltage is going to sag a tiny bit even with miniscule LED light loads. I just used the LED light turn-on sequence as a convenient way to quickly add loads while observing the deltas in the 7345's voltage output. Each delta in voltage perfectly corresponded with each LED switch turn-on. The 7345 either has an open ended output stage or has a very sloppy negative feedback control loop.

Sal, as a reminder regarding your 7345 experiment ... read about your 2009 7345 experiment in the link I show above.

[pnichols]

BFL13, thanks for bringing up those Parallax links again for us ... I've looked at them many times through the years.

You're right, my 7345 unit DOES NOT hold it's terminal output voltage up as those curves' show that it is supposed to ... if it did my AGM batteries would charge up even faster.

The sag shown for the Paramode 4455 better resemblies my 7345's voltage sag versus load performance.

By the way, do you have data yet proving that other (reasonably priced) RV chargers in their boost modes in fact are holding 14.XX terminal voltages under heavy currents?

[Salvo]

pnichols wrote:

Salvo wrote:
You're measuring somewhere outside the converter. Unless you have a voltage sense line, ...

Sal,

I watch what my 7345 converter is doing - voltage-wise on it's output terminals - when charging my AGM batteries via a four-place digital meter plugged into a regular 12V receptacle located quite a ways from the 7345 elsewhere in the RV. However, since this volt meter only presents a very high impedance ("infinite") load, relative to the battery bank's load, to the 7345 when making this measurement via a receptacle ... what's on the voltmeter just about perfectly represents the actual voltage on the 7345's terminals.


Now I'm confused. I thought you were an EE. Your voltmeter is far away from the converter. You're seeing IR line drops.

I have no idea where the 7345's negative feedback loop sense point is relative to it's physical terminals. I assume that the sense point on the printed circuit board is very close, physically and electrically, to the 7345's metal output lugs ... as it should be and better be.

Assume nothing. It doesn't matter anyways, as you're measuring far away from the converter output.

What I see on this voltmeter is that it is NEVER at or near 13.8 volts when it begins charging the 200 AH battery bank. I interpret this as meaning that the 7345's output, as measured at it's terminals, is for some reason sagging from it's advertised "nominal output value" of 13.8 volts. However as the battery bank charges up, it's terminal output (as indicated via the infinite input impedance voltmeter plugged into a 12V receptacle) eventually rises up close to 13.8 volts.

As the battery charges, current tapers and the IR drop gets smaller and smaller. Consequently, the charging voltage that you're measuring far away from the converter increases in magnitude.

Here's another proof of sagging 7345 output: I've installed all LED lighting in our RV ... and for each LED we turn ON, the voltmeter I use (that I mentioned above) to show 7345 output lowers 0.01 (one one-hundreth) of a volt. i.e. One light ON drops voltage 0.01 volt, two lights ON drop voltage 0.02 volts, etc..

That's no proof. You're jumping to wrong conclusions.

Here then, IMHO are some questions/issues:

1) I don't think that Pulse Width Modulation power supply design technology automatically requires that negative feedback around it's output stage(s) be employed for voltage requlation. I expect that PWM power supplies can be designed, and do in fact exist, with only open ended output stages. Output on these of course would then sag with increasing output currents.

Those type of supplies do exist. They have many different voltage output. Only one of the outputs can be feedback regulated. The other outputs are "slave outputs". Guaranteed, this converter has its regulation on the 13.8V output. I've designed converters for many years for the military, avionics and space satellites and space vehicles. I know them very well!

2) Do the Parallax 73XX line of converters perhaps employ only very sloppy negative feedback to regulate it's output voltage such that a couple of tenths or more volts of sag can, and does, occur under load?

I traced out the voltage feedback loop in the Parallex converter. It does not have a sloppy feedback loop. Again, I believe all voltage drops are most likely due to IR.

3) Two-three years ago as an experiment I think I recall that you did in fact move a 7345's sense point out to the battery terminals to force full 7345 voltage (plus a little more, as I remember) at the battery independent of any voltage drop in the cables between the 7345 and the battery. One of my responses to this was that this would be "dangerous" because too-thin wiring between the converter and battery could then heat up from too-heavy current flow because voltage drop in the cabling would then NOT reduce the current acceptance rate of the battery.

Very unlikely. Why don't you do the analysis. How much does a free running, 20 foot long, 8 awg cable heat when conducting 45A? The cable insulation is most likely rated at 120F.

Cal

[pnichols]

Here's a link to the detailed thread discussing the 7345 in which Salvo finally did an experiment we had been kicking around for awhile - moving of the 7345's voltage control loop sense point out of the unit and over to the terminals of the battery being charged. I believe Salvo also at the same time tweaked the control loop to move the 7345's voltage set point up a bit into the 14 volt range. (Sorry Sal regarding my memory ... your experiment was not 2-3 years ago ... it was early 2009!):

http://www.rv.net/forum/index.cfm/fuseaction/thread/tid/22462060/gotomsg/22463981.cfm#22463981

[BFL13]

The Paramode 4455, a follow up to the 7355 in the product line, has these voltage curves which show the sag under load.

https://bestconverter.3dcartstores.com/assets/images/Parallax/4400/4455%20Voltage%20Curves.pdf

The 7355 literature tried to say there was a flat top over the range and no sag, but we do see the sag, same as they show for the Paramode.

http://www.parallaxpower.com/7300/7300pwrcntr.pdf

The little graph shows the no load flat top while charging a battery (not what I have seen!) but below in the table, they give "no load" and "full load" voltages where the full load is lower (The Sag in other words)

I owned a Paramode for a while and can confirm it sagged exactly as shown in the curves and similar to my 7355.

I question whether there is enough resistance from inside the unit, then along the 7355's fairly short #10 wires to the DC panel to create that amount of sag we see at the panel. I have not measured that, so Salvo's case has not been disproved and there is no reason to question his own measurements.

[Cstoliker]

Interesting. On the 6300 line, one of the test procedures was to jump the yellow wire to the aluminum heat sync and watch for voltage to rise.

Here is a manual, schematic and pretty much everything you would want for the 6300 seires.

Magnetec 6300 schematic


*notice page 8 "optional charger circuit"

Non of this is available for the 7300 series. So I guess the question becomes how much has changed between the two.

[pnichols]

Salvo wrote:
You're measuring somewhere outside the converter. Unless you have a voltage sense line, ...

Sal,

I watch what my 7345 converter is doing - voltage-wise on it's output terminals - when charging my AGM batteries via a four-place digital meter plugged into a regular 12V receptacle located quite a ways from the 7345 elsewhere in the RV. However, since this volt meter only presents a very high impedance ("infinite") load, relative to the battery bank's load, to the 7345 when making this measurement via a receptacle ... what's on the voltmeter just about perfectly represents the actual voltage on the 7345's terminals.

I have no idea where the 7345's negative feedback loop sense point is relative to it's physical terminals. I assume that the sense point on the printed circuit board is very close, physically and electrically, to the 7345's metal output lugs ... as it should be and better be.

What I see on this voltmeter is that it is NEVER at or near 13.8 volts when it begins charging the 200 AH battery bank. I interpret this as meaning that the 7345's output, as measured at it's terminals, is for some reason sagging from it's advertised "nominal output value" of 13.8 volts. However as the battery bank charges up, it's terminal output (as indicated via the infinite input impedance voltmeter plugged into a 12V receptacle) eventually rises up close to 13.8 volts.

Here's another proof of sagging 7345 output: I've installed all LED lighting in our RV ... and for each LED we turn ON, the voltmeter I use (that I mentioned above) to show 7345 output lowers 0.01 (one one-hundreth) of a volt. i.e. One light ON drops voltage 0.01 volt, two lights ON drop voltage 0.02 volts, etc..

Here then, IMHO are some questions/issues:

1) I don't think that Pulse Width Modulation power supply design technology automatically requires that negative feedback around it's output stage(s) be employed for voltage requlation. I expect that PWM power supplies can be designed, and do in fact exist, with only open ended output stages. Output on these of course would then sag with increasing output currents.

2) Do the Parallax 73XX line of converters perhaps employ only very sloppy negative feedback to regulate it's output voltage such that a couple of tenths or more volts of sag can, and does, occur under load?

3) Two-three years ago as an experiment I think I recall that you did in fact move a 7345's sense point out to the battery terminals to force full 7345 voltage (plus a little more, as I remember) at the battery independent of any voltage drop in the cables between the 7345 and the battery. One of my responses to this was that this would be "dangerous" because too-thin wiring between the converter and battery could then heat up from too-heavy current flow because voltage drop in the cabling would then NOT reduce the current acceptance rate of the battery.

[Cstoliker]

MEXICOWANDERER wrote:
"While camping"
Happy to see someone with enough cash to feed what amounts to an unemployed brother-in-law. If you are REALLY cheap, fork over $29.99 for a Harbor Freight 10-amp charger and run it alongside your Parallax. Equals ten dollars worth of gasoline. Oh wait, I forgot about the time and money needed to get more gas...

Chances are I likely make a lot less money then most of you but I am not cheap. I am just not afraid to get my hands dirty and enjoy the felling of accomplishment in doing something myself. For a fraction of the cost of replacement. My hands are scared and calloused as a mans should be.
Most people nowadays if on The side of the road with the hood up, unless something is spewing flames or fluid have no idea what their looking at or for.
I enjoy learning how **** actually works so I know what's going on around me.

KJITNF-thanks for your useful input. I'll likely be to busy to screw around with it more for a few days but will definitely look into what you said.

And besides, I've already blown my budget on RV repairs this year, new tires, brakes, completely rebuilt sleeper over the cab. Roof repair and oil change. So if I can save a few hundred and learn something in the process I'll call it a good experience.

[Salvo]

We've had this discussion before. You're an EE. I don't need to tell you how pwm regulators work. You know that until they reach 100% duty cycle, the converter will maintain setpoint regulation. If you don't see 13.7V or what ever, then you're not measuring at the regulation feedback point within the converter. You're measuring somewhere outside the converter. Unless you have a voltage sense line, the converter can't guarantee constant voltage at your measurement location. You are seeing an IR drop which the converter can't do anything about.

Sal

pnichols wrote:
I wish my 7345 would hold 13.8 volts at 45 amps.

[KJINTF]

Not even camping while at home gasoline for some of us is 30 miles away - lets see 60 miles round trip at 15mpg in the jeep and maybe 8mpg in the RV - I would rather have UPS deliver the parts - that way I do not have to drive and can enjoy another cold drink. HF distance 130 miles one direction.

Good to see at least a few others are interested in repairing things rather than tossing them in the trash and purchasing some more made in China stuff. Those WalMart Ships are full coming this way and empty going back.

I have sticker from a "Quality" Converter manufacturer no it's not Parralax here in the shop clearly stating "Proudly Assemblied in the USA" (Yes assemblied not Made) to remind me of the days when we actually made quality products right here in North America.

[MEXICOWANDERER]

"While camping"
Happy to see someone with enough cash to feed what amounts to an unemployed brother-in-law. If you are REALLY cheap, fork over $29.99 for a Harbor Freight 10-amp charger and run it alongside your Parallax. Equals ten dollars worth of gasoline. Oh wait, I forgot about the time and money needed to get more gas...

[Cstoliker]

Yes. I bought this meter because it had a PC connection option. I use it constantly I will invest in a better one.

Here's a short video of the 7345 I. Action With the polarity fuses in. If they ate out the blue wire, as mentioned runs about 13.4


http://www.youtube.com/watch?v=Z8VmoN-2qo4


Sent from my iPhone 4s

[KJINTF]

They all have the burned area under the resisters - normally the resisters are fine no need to replace the pcb trace however does need attention - solder in a wire to repair the traces - Q13 is good as well

That cap behind Q13 is defective replace it

Take a close look at those 200Vdc 1,200ufd caps if they show any signs of failure replace them

As suggested earlier get yourself a nice ESR meter - a DVM aka DMM is useless when checking caps - Good ESR meters use about 150Khz with less than 400mv so should not forward bias any semiconductor junctions incircuit testing is very simple

You should be good as new with less than $5.00 in parts

[Cstoliker]

[Cstoliker]

KJINTF wrote:
Can you see any defective parts?

Many times the caps are easily seen as defective and the resister pack is burned

My multimeter is not capable of testing capacitors while in circuits. I'll just replace them. The resisters do look burned but I haven't gone through and tested them yet.