I earlier accused the AVR Pocket Programmer of being a bit flakey, but I got the Pololu programmer today, and I discovered a remarkable fact…
When I connect the ground on either programmer through a voltmeter to my bench power supply’s ground, I notice that they have a 25 volt difference. Needless to say, lots of current wanted to flow…
I have a BK Precision 1671A regulated bench power supply with a variable output between 0-30 volts. It is all nice and nifty with a lcd readout (amps and volts), current limiting, and up till now, I thought I was stylin’.
Now I am not so sure.
Both the notebook pc and the bench power supply are on the same outlet. I doubt that either one is defective, but they have a definite disagreement about what ground should be.
Any recommendations as to how I should rectify (perhaps literally) this situation?
It would be nice if I could program a '328 without the fumes of rosin wafting through the air.
Both your bench power supply and your laptop power supply are probably fully-isolated. Because of this, you can expect that each will independently have a different notion of “ground”. If you do not tie your two “grounds” together, it is perfectly reasonable to see a 25V difference between the two.
Are you sure that a lot of current wanted to flow? Try taking a 10K Ohm resistor and use it to tie the two grounds together. Then measure the voltage across the resistor. If the voltage difference is much lower, and the resistor isn’t getting warm, you should be safe to tie your two grounds together with a wire. If you are really concerned, keep reducing the resistance until you get to 0.
I have an amp meter built into my supply. The current limiter was immediately triggered, and when I raised it to allow .28 amps (previously, it was at .04), I got smoke from solder flux melting.
Definitely a bad thing.
I can’t tie the two grounds directly together safely, which I find odd, because I would think like you are thinking, that once the grounds get tied together, they would no longer float.
So, in any case, it is a smidge of a problem, as I am unable to program my Atmel’s safely. For the PIC’s, I have a programmer that works out of circuit, so obviously, this problem does not pop up (the same thing could be done with a Dragon, which I am beginning to contemplate).
I still would like to be able to remedy this problem, as the power supply is still (remarkably) functioning, and being able to run designs on a good bench supply with current limiting is a nice asset.
I should have thought of your suggestion of hooking up a resistor between the grounds, and just make sure it is high enough value to dissipate the heat and limit the current flowing. Limiting current flow between the grounds is currently the objective.
Do you have the same problem if you unplug everything from your laptop? Printers, monitors, network, etc?
If your power supplies are not floating, and their “grounds” are not at ground, I would call that bad ™. Figure out which is the culprit by using your meter between their output grounds, and the ground on the outlet. I would also check that the ground on the outlet is within a few volts of neutral. If you are not comfortable working with wall power, find someone who is to do it for you.</end disclaimer>
The bench power supply has a ground which is -25v below wall ground (and Neutral).
I think I may need to replace the BK Precision 1671a (bench power supply ~ $150 when I bought it).
Any recommendations for a replacement?
Single or dual output, 0-30volt, volt and amp readouts, 2amp at 30 volts…
The other option is to cannibalize the electronics on this one, keep the amp/voltmeter and transformer/switches/pots/breaker, and replace the rest of the supply electronics (the transformer in this supply is beefy, and some quick tests would prove or deny its innocence in this matter) with a halfway reasonable design, and decent regulation (and whatever options I decide to use/add… Some advantages to building your own…).
Since I have already had the merry aroma of a substantial amount of burnt circuitry from this supply, it wouldn’t kill me to do some after market re-engineering…
