Sure Electronics ME-SP335 ME-SP334 DS18B20-Based Digital Thermometers

[ewdudley]

Just FYI, got a couple of these off ebay for temperature display after I saw Jim's panel using them (or something very similar).

They only come with a one meter lead, but I've never had any trouble running single DS18B20s on 10 or 20 meter runs before, so I figured I'd just splice in some CAT5.

Well that didn't work, I couldn't even get them to work with two meters spliced in.

The bus rise time was too slow so I added a 2.2K pull-up from the 5VDC wire to the data wire, which works for extensions as much 10 meters, although I suppose it voids any warranty.

Example listing:

http://cgi.ebay.com/Digital-LED-Tem...ultDomain_0&hash=item563fdc376f#ht_4706wt_957

--ewd

 

[jebatty]

ewd: They only come with a one meter lead, but I've never had any trouble running single DS18B20s on 10 or 20 meter runs before, so I figured I'd just splice in some CAT5.

My brain is dense this morning; don't know what you mean by "one meter lead." I use Cat5e to 7 of these meters, longest run is about 8 meters with no problem. I've shared the ground wire where convenient, but I've used separate V and data lines for each sensor/meter. As you know, each meter will take 2 sensors. Maybe you have some ideas on how better to use or cable these meters.

 

[ewdudley]

ewd: They only come with a one meter lead, but I've never had any trouble running single DS18B20s on 10 or 20 meter runs before, so I figured I'd just splice in some CAT5.

My brain is dense this morning; don't know what you mean by "one meter lead." I use Cat5e to 7 of these meters, longest run is about 8 meters with no problem. I've shared the ground wire where convenient, but I've used separate V and data lines for each sensor/meter. As you know, each meter will take 2 sensors. Maybe you have some ideas on how better to use or cable these meters.

Your setup should and does work fine, especially with a single device per run. I tried to do the same thing and was surprised to have it not work.

The sensors that came with mine had wires that were one meter long, and I said to myself fine, I know you can run one-wire stuff a lot further than that, I'll just extend the wires with some low capacitance CAT5.

To my surprise the sensors went offline with more than a meter or two of extra cable. I tried different cables, and swapped the twisted vs untwisted configuration for the GND, 5VDC, and DATA wires and nothing worked for any of the different display modules I tried.

So I looked at the signals and the bus rise times were too slow. Since the protocol works with open collector dominance activity I added a pull-up to decrease the rise time. I figured 2.2K would be harmless and it worked for 10 meters, 4.4K worked as well, and the signals looked good, so that's all the more I fooled with it.

I don't know if I got a bad batch of display modules, or maybe they changed the design, or maybe none of my CAT5 is as low capacitance as it's supposed to be, but for me adding a pull-up got me going, maybe others will find it helpful.

Cheers --ewd

 

[jebatty]

Eliot - the manual for these says that when an external probe is used, the +5V should not be connected. Yet the external probes provided have the 3 pin connector which includes the +5V. Maybe that lead is not connected inside the probe?

What I have done is cut the 1 meter furnished probe wire, use the 3 pin connector portion of the probe wire to connect to the meter, place the probe where desired with the balance of the probe wire, and then splice in 3 conductors from the Cat5e to extend the cable to the location of the probe, observing the +5V, data, and gnd wiring pattern. I solder the splices and cover with shrink tubing. As mentioned above, all has been working well.

 

[ewdudley]

the manual for these says that when an external probe is used, the +5V should not be connected. Yet the external probes provided have the 3 pin connector which includes the +5V. Maybe that lead is not connected inside the probe?

That's odd that they say not to connect the 5VDC, I can't think of any reason not to. In fact, the one I'm testing does not work if the 5VDC is disconnected -- with the one meter wires or with the ten meter wires -- so I'd have to conclude the 5VDC is connected to something on the sensor end, presumably the 5VDC pin of the DS18B20 itself.

What I have done is cut the 1 meter furnished probe wire, use the 3 pin connector portion of the probe wire to connect to the meter, place the probe where desired with the balance of the probe wire, and then splice in 3 conductors from the Cat5e to extend the cable to the location of the probe, observing the +5V, data, and gnd wiring pattern. I solder the splices and cover with shrink tubing. As mentioned above, all has been working well.

Which is just what I'm doing.

My wall wart power was pretty dirty and rich in harmonics, so I tried straight 12VDC battery power, which didn't help.

Also have tried various different CAT5 cable to no avail.

But a pull-up resistor cures it so I guess I'll go with that and not worry about it.

Thank you for taking time to document your experience with these devices, they're quite handy.

 

[Clarkbug]

Sorry to bring up such an old thread, but I found this one and wanted to ask a question of those who know much more about this than I for sure.

This is the listing Im looking at on eBay.

http://www.ebay.com/itm/Digital-Red...762?pt=LH_DefaultDomain_0&hash=item3a54d8d592

It comes with two of the temp probes, which Im thinking will be handy to measure the top and bottom of the tanks for temp. (Not sure how to attach the probes yet, but thats another challenge).

At any rate, I see this same manufacturer also offers a "digital" probe for temps. Any benefit to using the digital vs. the analog? Its only a few bucks difference, but I would rather not spend it if there isnt any real gain to it.

 

[jebatty]

I have about 25 of the digital DS18b20 meters installed, the furthest about 30' from the meter. All work fine with about a 16-18 vdc power supply. If it's just temperature, and the sensor that you are looking at covers the temperature range you need, then I don't know of any particular reason to pick the digital vs something else. I picked the digital ones with two sensors because I wanted to use the second sensor in a data logging application with different hardware, and it was less expensive this way rather than buying separate sensors.

Some meters have blue LED's and some are red. This is kind of neat, using the red one for top of tank (hot) for example, and the blue for bottom of tank (cold) for example, but this takes two meters also.

The two sensor meters work fine, and they alternate between ch 1 and temp and then ch 2 and temp, or you can set a dip switch and just show ch 1 or just ch 2. I found the alternate display rather annoying, although I do use it on one meter which displays inside temperature and outside temperature.

Another possible benefit to getting a two sensor meter even if you just use one is that I found 3 defective digital sensors in my last batch of 20 meters/40 sensors. Maybe the non-digital don't have this defect problem. The extra sensors came in handy for replacement and didn't slow up my project.

 

[ewdudley]

It comes with two of the temp probes, which Im thinking will be handy to measure the top and bottom of the tanks for temp. (Not sure how to attach the probes yet, but thats another challenge).

At any rate, I see this same manufacturer also offers a "digital" probe for temps. Any benefit to using the digital vs. the analog? Its only a few bucks difference, but I would rather not spend it if there isnt any real gain to it.

If you review the specs the analog are +- 1500 millikelvins while the DS18B20 are speced tighter than that, and they perform even better than specified. For storage tank monitoring it doesn't seem like accuracy would matter a whole lot.

We know the DS18B20 leads can be extended, the analog probably would perform well with extended leads, but I don't know that for a fact.

There are Sure Electronics DS18B20 units that come with one sensor or two. You can add a second sensor to the one sensor units later if needed.

 

[Clarkbug]

Thanks for the quick replies as always gents.

Im thinking that I would want one of the dual input meters for each of my storage tanks. The listing I had posted says that each of the probes comes with a 3 meter lead, which would be more than enough for me to mount them in my boiler room. I thought about extending one to my upstairs so I can see at a glance if I need to go load things back up.

For the few bucks, I might as well buy the higher accuracy probes/meters. (Im also looking to perhaps pick up some remaindered lab type K thermocouples and associated meters, which would be slick I think) I dont like the idea of buying Chinese stuff for this, but I dont think anything else comes close for the price.

Is there any other place that may be useful for me to monitor? Im just using a relay control off of an aquastat to switch between my wood boiler and the backup oil boiler, so this would just really be for my eyes to know if things were going OK or needed to be tweaked.

 

[jebatty]

I don't monitor the top of my 1000 gal tank, but about 1/4 of the way down, as well as middle and bottom. Then the "top" and middle meters let me know I have a good quantity of hot water, and when the "bottom" meter starts to move up, I know the tank is nearly fully charge. In addition I monitor boiler supply, boiler after return water protection, stack temp with a K-type meter, radiant supply and radiant return.

 

[Clarkbug]

One other quick question...

Jim, you mentioned that you found the alternating annoying, but how often does it cycle between the two sensors? And does it tell you if its channel one or channel two?

Im just considering getting three meters and having them alternate between return/supply on my boiler, on the top/bottom of my tank array, and on return/supply from the system. But I kinda like the red/blue idea for alternating as well.....

 

[jebatty]

The meter displays CH1 and then the temp and then CH2 and then the temp. About a second or two on each display. I just find the flashing CH1, temp, CH2, temp, to be annoying, or distracting. I would think that a panel with several meters, each sequencing the two channels, to be quite annoying, especially if the panel was in a highly visible location and in the line of frequent sight. Each meter also possibly would have a slightly different synchronization. For an extra $10-15 per meter, the cost is quite minimal to have a constant display of the temp on a channel without any flashing of channels and temps.

 

[Clarkbug]

Ugh, that is fairly annoying. I hadnt even thought about the synching issue between meters. I dont think Ill have them where they are too much of a distraction, but you are right, for the money why bother to make something thats not easy to use.

Did you mount these in a panel? I was thinking some sheetmetal or wood will be my backboard, and maybe a single wall wart for power.

Any tips for attaching the sensors to the pipes/tanks? Just really good tape?

 

[jebatty]

I bought a 12" x 12" x 6" plastic electrical box, cut the squares for the meters, and popped them in. A DC wall wart is good. Attached sensors to the pipes with cable ties and then insulated well. Keep in mind a 3 conductor cable for each meter, but can share a common ground. I used Cat 5. One cable will be good for 3 digital meters.