[xmokshax]

i just put together a TREAD from one of Tangent's kits, and i'm having a couple of problems. i'm using the circuit with a Jameco wall wart, 24VDC 500mA. the transformer puts out 28.9V unloaded; i'm getting 26.5VDC from TP1 to TP3, and 25.1VDC from TP2 to TP3. fine. however, i can't seem to adjust the output voltage (a problem some others seem to have had), and i'm measuring 57.9VAC from TP1 to TP3 and 54.5VAC from TP2 to TP3. whoa!

the OP of this thread appears to have experienced the same set of problems as me, with slightly different numbers, but unfortunately, that thread was never resolved (or he never posted the solution that worked for him). i can answer a couple of the questions that were asked in that thread, to save everyone some time:

- i've checked for bridges, reflowed every joint, and cleaned the board pretty thoroughly, so none of those should hopefully be problems. pics will follow shortly - my camera's battery had enough juice to snap a couple of pictures, but it died before i could transfer them to the PC. [EDIT: pics added below.]

- my meter is on the 200VAC setting when measuring AC voltage between the test points. someone said in another thread that cheap meters sometimes switch automatically to AC mV if necessary. i don't know if that's happening, but either way, i don't think my cheap meter shouldn't be able to register the ripple that's coming through between TP2 and TP3.

- when i measure resistance between the outer pins of VSET, i get a value that maxes out at 957 ohms. i haven't checked how far down it goes, but presumably to zero. the part appears to be a 100 ohm (?) model - it has "W 102" printed on the top. tangent has said that he always sends either 1K or 2K trim pots. am i reading this wrong? either way, the output voltage doesn't change when i change VSET.


is my LM317 toasted? what else could be wrong? thanks for any help anyone can give.

pics:

top:


bottom:

 

[Seaside]

Well... as you know, tread itself is a quite standard kind of circuit.
There is nothing really fancy about it. So it shouldn't cause headache and the fix should be simple if there's any problem, but sometimes it's not the case. Check value of resistors you used, check polarity of diodes you have there, check solder joints and capacities. And if you can't find any problem, try another LM317 and see if it solves your problem. I once had similar problem like that before. Well... kinda opposite... actually, it was. I feed 25V AC, got something like 6V at the end, and that's not adjustable. I changed LM317, and that solved the problem for me.

 

[headphonejunkie]

I too, say to make sure all of your parts are in correctly. If so, then maybe your lm317 is bad.

 

[xmokshax]

the resistor values are right as far as i know, and the diodes are oriented correctly. the tantalums are installed in the correct direction, too. seems like a fried regulator is looking more and more likely.

 

[n_maher]

I'm pretty sure that you have the right trimpot. The part number breaks down like this:

1, 0, (2) 0's = 1000 or 1k.

More importantly your problem is most likely due to the fact that you're not feeding the circuit enough voltage to allow for any adjustment. You're feeding it approximately 24VDC - check the unloaded output of that wall wart, it's probably higher than that but since it's DC you're dropping a couple of volts across the unnecessary bridge (used for AC only, but not hurting the DC) and then the LM317 needs to drop another couple of volts in order to adjust the voltage. By the time your DC is getting to the LM317 there's not enough drop since your output adjustment range, set by the resistors in the kit, is only from 22V up. So you simply don't have enough input voltage, your TREAD is probably working just fine.

Your meter is probably freaking out trying to measure AC when all it sees is DC, this isn't at all uncommon if it's a cheap meter. I wouldn't worry about these measurements.

If you happen to have some spare resistors kicking around try changing R2 to 1.5K and see if you don't get some adjustment.

 

[xmokshax]

aha. so i was misinterpreting the markings on the pot - good to know. i was basing my interpretation on this quote from tangent from another TREAD thread:
Quote:

I've shipped different values for the pot with the kits over the time I've offered it, so you'll have to read the value off the pot's side to know what value you've got. It'll be a 3-digit code: two significant digits and a multiplier. So, 103 is 1K: 1.0 * 10^3.

was that just a typo on his part?

in any case, i'll try digging up a 1.5k replacement for R2, as you suggested. i don't think i have a 1/2 W one lying around, unfortunately.

EDIT: eureka! ok, i just had the trim pot set too high. as you suggested, the regulator wasn't regulating - the trim pot was set too high for the amount of voltage my wall wart can put out. as i said in my OP, i tested the high range of the pot, but never cranked it down. i now have it set at just over 200 ohms, and the regulator is putting out 24V rather than the 25.1 i was seeing before. since this is kind of right on the cusp of the range over which the regulator will function, should i crank it down more (to, say, 23 or even 22V) in anticipation of the transformer's output dropping when the load of the amp is applied? i suppose if i'm going to go all the way to 22V, i might as well just remove the pot and jumper it.

EDIT 2: in anticipation of an emphatic "yes, lower it further", i now have it running at 22.0V with the pot set at 29 ohm. based on my transformer's measurables above, does it sound like this will be a workable configuration when the load of a PIMETA is applied?

 

[n_maher]

Quote:

Originally Posted by xmokshax /img/forum/go_quote.gif EDIT 2: in anticipation of an emphatic "yes, lower it further", i now have it running at 22.0V with the pot set at 29 ohm. based on my transformer's measurables above, does it sound like this will be a workable configuration when the load of a PIMETA is applied?

That should be a fine config for a PIMETA and I'd be pretty surprised if the load of the PIMETA was enough to lower the voltage any further. However, if you have the stock reverse-voltage diode in the PIMETA I think that'll drop another volt or so. So with your TREAD set at 22V you should expect to see 21V at the opamps.

 

[tangent]

Quote:

Originally Posted by xmokshax /img/forum/go_quote.gif was that just a typo on his part?

More like a thinko. 103 = 1,000 in some common cap numbering schemes because there's an implicit decimal point between the two significant digits. Bourns follows a different convention.

Quote:

should i crank it down more (to, say, 23 or even 22V) in anticipation of the transformer's output dropping when the load of the amp is applied?

If the wall wart is unregulated, yes, definitely.

Quote:

the pot set at 29 ohm

Yes, as it says on the TREAD kit's details page, the kit goes down to 22V at the lowest. If that's the value you want, you might as well just remove VSET and put a jumper across its outer two pins. If you want to go lower, replace R2.

 

[xmokshax]

Quote:

Originally Posted by n_maher /img/forum/go_quote.gif That should be a fine config for a PIMETA and I'd be pretty surprised if the load of the PIMETA was enough to lower the voltage any further. However, if you have the stock reverse-voltage diode in the PIMETA I think that'll drop another volt or so. So with your TREAD set at 22V you should expect to see 21V at the opamps.

sounds good. thanks for the help

Quote:

Originally Posted by tangent /img/forum/go_quote.gif Yes, as it says on the TREAD kit's details page, the kit goes down to 22V at the lowest. If that's the value you want, you might as well just remove VSET and put a jumper across its outer two pins. If you want to go lower, replace R2.

gotcha. if i turn the pot even lower (i.e., essentially down to zero), the voltage ends up at 21.6 or 21.7V. is it normal for it to go a bit below the 22V threshold that you've specified, or is this indicative of some other small problem in the circuit?

 

[tangent]

*shrug* We are dealing with 5% components here...