FS: strip chart recorders - could be useful for monitoring machine parameters?

Bill..how fast do these recorders travel? What I need, is a strip recorder that will measure power surges and whatnot over a 12 hour time frame and not run a lot of paper through it.

I dont care what condition they are in..as long as they are functional and will show power surges/browns over a 12 hour period on a less than

6" of chart paper.

Gunner

Political Correctness

A doctrine fostered by a delusional, illogical liberal minority and rabidly promoted by an unscrupulous mainstream media, which holds forth the proposition that it is entirely possible to pick up a turd by the clean end.

I tested one of them, it's set up for about 1/4 inch in 5 minutes, it seems. The others may be different - speed is set on these by moving gears around in the little gear box, but it isn't clear whether you can do this or only the factory can.

one thought on these particular recorders - they are "sampling recorders" - the meter movement pointer is "smacked" against pressure sensitive paper about once every 15 seconds - so they are great for monitoring slowly changing parameters but not for picking up a short surge - so ok for brownout, not ok for a power surge.

sigh..thanks

Gunner, still looking, though its getting closer.

Political Correctness

A doctrine fostered by a delusional, illogical liberal minority and rabidly promoted by an unscrupulous mainstream media, which holds forth the proposition that it is entirely possible to pick up a turd by the clean end.

Thanks much William.

Gunner

Political Correctness

A doctrine fostered by a delusional, illogical liberal minority and rabidly promoted by an unscrupulous mainstream media, which holds forth the proposition that it is entirely possible to pick up a turd by the clean end.

According to Gunner :

How short a surge do you want to detect? The RustRak recorders depend on a bruisable paper, with the needle being trapped between a clamp bar and a backing plate where the paper passes over an edge.

The time scale will pretty much correspond to how often they take a sample -- anything which happens between samples will be lost.

IIRC -- you can change gears on the side to change how often the samples are taken -- with the tradeoff that the more often the samples, the more paper you will roll out.

In other words -- you want a rate of 0.5 inches per hour. (Note also that these are fairly narrow charts, so you may or may not see enough resolution unless you have a way of expanding the scale with an offset voltage.

O.K. Looking at their web page, the current ones cover a range of:

Recorder Chart Speeds from 1/48"/Hour to 240"/Hour via motor & gear train

So -- you can be as little as 1/4" for your 12 hour span. :-) Here is the main page:

I can't find a listing of how often the different speeds take their samples -- but it may well be pretty seldom at the slowest speeds. You'll know when each sample is taken, because there will be an audible "clack!". They say "Recorder Response time 1 second maximum." which may mean one sample per second -- or just how fast the needle will move from zero to full scale. :-)

Enjoy, DoN.

According to Gunner :

*You* can IIRC -- though you may have to get extra gears from the vendor for some ranges.

That looks like about 36" for your 12 hour period. Is there any particular reason why you need the really slow travel? And it is probably within the range of the available gear sets.

Enjoy, DoN.

well, the recorders I had offered have found a home, but for future reference, here is some additional data

the design consists of a rugged meter movement with a strong pointer - the pointer moves freely around until CLOPPP the brass bar that Don N mentioned smacks the pointer into the paper. This continues WAAVEEEE CLOPPP WAAVEEE CLOPPP for as long as the unit is powered. A meter like this with infrequent sampling is not suitable for monitoring short term transients.

Thanks Don. I guess Im simply looking for a straight analog record at as you say..about .5" per hour. Spikes/brrowns should show up clearly. Obviosly sampling recorders will not do it for me. I didnt know how a sampling recorder worked. Thanks for the heads up! Gunner

"Aren't cats Libertarian? They just want to be left alone. I think our dog is a Democrat, as he is always looking for a handout" Unknown Usnet Poster

Heh, heh, I'm pretty sure my dog is a liberal - he has no balls. Keyton

Yep..cost of paper. I suppose a cheap used laptop with some sort of analog input device might be better.

Gunner

"Aren't cats Libertarian? They just want to be left alone. I think our dog is a Democrat, as he is always looking for a handout" Unknown Usnet Poster

Heh, heh, I'm pretty sure my dog is a liberal - he has no balls. Keyton

Check this out:

10-bit A/D to drive a PC serial port, $24.95. It includes chart recorder software. Sample rates can be as slow as 26 samples per hour. You'd need a cheap transformer (old wallwart), nickle diode and cap and maybe a coupla resistors to convert the line AC voltage to a lower DC voltage that it can sample.

Scrounge an old, slow discarded laptop or PC and you're in bidness. Tha PC would log the data in memory, you then dump it out later in an Excel spreadsheet or whatever. Sampling at once per second for 8 hours would result in 28.8K file size, no big deal. You probably could figure out a way to process a datafile to just extract times (and voltages) when voltage was greater than or less than some prescribed limits. If you use a laptop, its battery would carry it thru brownouts.

For just a few few bux more they make dataloggers that use flash memory like memory sticks or camera memory.

Hm, I guess those are more than a few bux more. Old laptops are somewhere between cheap and free. The data could be dumped onto a floppy or a $9.95 memory stick if the laptop has a USB port. The DATAQ device uses the serial port.

Oh way cool!

Gunner

"Aren't cats Libertarian? They just want to be left alone. I think our dog is a Democrat, as he is always looking for a handout" Unknown Usnet Poster

Heh, heh, I'm pretty sure my dog is a liberal - he has no balls. Keyton

of course you know this will NOT work for gunner's purpose - what MIGHT work is to sample at a high rate and record the average, min and max measured during each minute interval. but sampling at a low rate will miss the transients

He can trade off between sample rate/file length and transient durations he cares about. Sample rates up to 240 samples/sec are possible, which is twice the sampling rate Mr. Nyquist sez is necessary for completely observing a 60 Hz variable. 8 hours of data at 120 S/s would be a file size of about 3.5 meg. File sizes of up to

1 GB are possible.

Included software can export the recorded data to Excel (or MathCAD), which means one could then do all manner of datacrunching with sliding window averages, maxes, mins, finding times at which values were outside of a prescribed band, etc etc, with no chart paper and no recurring cost. The included software can find deviations from a prescribed band of values in the data without recourse to Excel.

With a chart recorder having stylus resolution of 0.01" (not likely),

8 hours of data at one sample per second would require 24 feet of chart paper, not to mention the tedium of squinting at those 24 feet of paper thru a lens. At sample rate of 120 S/s that would be about 2880 feet of chart paper.

Since the little DAQ has 4 channels, one could also monitor suspected events that may cause brownouts -- e.g., Ole goes to take a crap and read the paper about 3:00 each day and when he flushes the water pressure drops which reduces water pressure which results in insufficient cooling to the gizmotron, thus surging the frammis and browning out the whole damned plant.

The little DAQ looks like a reasonably good solution to me, for $30 and a scrounged laptop or PC.

Power line monitoring as well as other voltages often use Drantz

I've used this company before - but don't push it - use it as a reference only.

Martin

Martin H. Eastburn @ home at Lions' Lair with our computer lionslair at consolidated dot net TSRA, Life; NRA LOH & Endowment Member, Golden Eagle, Patriot"s Medal. NRA Second Amendment Task Force Charter Founder IHMSA and NRA Metallic Silhouette maker & member.

D> According to Gunner :

Thanks Don. I guess Im simply looking for a straight analog record at

Hi Gunner,

Years ago I used to service some equipment that used a good sized (2000 watt?) Best inverter style backup supply (UPS). It had an RS232 port on it for setup and reporting. It would time/date stamp supply line glitches, line frequency, inverter running... lots of stuff depending on how you programmed it. Internal memory stored quite a bit and this could be retrieved later. It could even make a crude ASCII sine wave graph of the current line voltage. It was scrapped and I couldn't think of any good use for it once the batteries went kaput.

You should be able to find something like this cheap. Of course knowing you, you might already have one stashed somewhere. I'll try to dig up the model number if you want, but I don't have access to my old workbench/manuals that would make this an easy task.

Just an idea...

240 Hz would not be adequate to catch transients caused by motor switching, etc - these are typically in the sub millisecond range. Look at Do-160 for examples. to catch transients of this nature, sampling at 10Khz or above would be an absolute minimum. hence my suggestion for peak, min, and average - simple analog circuits can capture these values with high accuracy and hold them for a slower D/A to sample at any desired rate consistant with the drift rate of the analog components. and your calculations RE chart speed are not applicable - you can run the chart as slow as you want, say 1 inch per day and still plot data with a 1 inch wide pen - it's just that you can't see the details, only the envelope - and that is what he wants to see.

According to Leon Fisk :

Aha! Good -- my posting got out. I've been having serious problems with my news server. They did a full changeover and upgrade targeted at new years -- and several of the servers were delivered seriously damaged, setting things back quite a bit -- and then there were the usual problems with porting software to a new platform. I

*think* that things are working properly now.

New batteries are what I would do. I'm running a 2KVA Best unit now which takes a fresh set of four large (truck sized) gel cells ever few years -- five I think. Expensive, but it gives excellent protection to all of the computers served by it.

Hmm ... Iggy tore at least one apart a couple of years ago. Perhaps he still has the logic boards and the serial interface, and you can use those for the purpose.

But the recorded information mostly shows when the voltage drops too low so it would switch on the backup part -- or high enough so it is safer for the thing to run from battery backup power rather than let the overvoltage through. (And it takes a bit to get that high, as the thing is built around a ferro-resonant constant voltage transformer.

Enjoy, DoN.