1-wire to USB converter that can use 1820s directly

Here's one of the pertinent Application Notes " Guidelines for Reliable 1-Wire Networks" :

It begins:

"INTRODUCTION The 1-Wire® protocol was originally designed (many years ago) for communication with nearby devices on a short connection?a way to add auxiliary memory on a single microprocessor port pin. Customers soon devised unique applications that involved extending the bus and moving the slave devices farther and farther from the master. Problems came up as the bus lengths exceeded both the capabilities of the bus masters and the limits of the protocol. 1-Wire device designs responded to the call with added features and protocols, multidrop (networking) capabilities, durable steel containers (iButtons®), and mechanisms to assure valid data transfers even in severely intermittent contact situations.Despite this evolution, 1-Wire components still perform poorly when bus masters are improperly designed or implemented, or when masters intended for short line use are pressed into service with greatly extended buses.

A 1-Wire network is a complex arrangement of devices, wire, and connections. Every network is different, often in both topology (layout) and hardware. Various claims have been made about 1-Wire network length and loading limitations, but the context may have often been unclear. As a result, users have sometimes been surprised when topologies fail despite the assumption that they are within specification. Special bus masters have been devised and tested with a single network design only to find that they are unsuitable for use in other layouts."

A significant proportion of folks wiring up a whole house full of 1-wire devices with a single channel will find that they would have saved time, money and aggravation by reading and understanding the words above and designing and implementing accordingly.

HTH ... Marc Marc_F_Hult

I don't use 1-wire myself so am not familiar with its details but the DS9490R appears to do at least part of what you ask. The datasheet indicates there are Windows drivers for it.

I want to have a number of 1-wire 1820 sensors thruout the house,

Sure:

here's the interconnect diagram:

There's a limit to the pathology that folks will put up with. snipped-for-privacy@gmail.com has exceeded my limit.

... Marc Marc_F_Hult

Marc F Hult wrote

Read that LONG ago thanks.

And that last isnt true with what is used in USB/1-wire converters/adapters.

And Pete Anderson and others have found that strings of 1820s work fine at MUCH greater distances than I am going to use.

And Pete Anderson has found that strings of 1820s work fine.

Didnt say anything about doing that.

And anyone with a clue can see that Pete Anderson has found that the sort of strings of 1820s that I want to use work fine.

I was JUST asking if anyone had any experience with particular USB/1-wire converters/adapters in that situation because the devices that Pete Anderson uses the strings of 1820s with have real downsides in my particular situation.

And I later proposed a different approach when no one was able to suggest a particular USB/1-wire converter/adapter which was known to do the job in my situation, doing something on the 1wire side of the USB/1-wire converter/adapter, like using one of the spare wires in the RJ11 cable to provide say 5V power to the 1820s, instead of using parasitic power.

Pete says that parasitic power works fine. The main problem is that I cant see what he is doing with the strings of 1820s at the inner end, because I cant find schematics for any of his devices that use strings of 1820s.

Presumably that sort of thing must be out there on the web and its just a matter of finding it. Which is why I asked the question in the first place, given that statement that what may appear to be fine from the datasheet may not work out that well in practice. I was essentially asking if anyone had actually implemented it and found it worked fine.

One other obvious approach is to get one of the cheaper devices of Pete's that does use strings of 1820s fine, basically to get the schematic and to use to test the string of 1820s and then modify the 1-wire side of a USB/1-wire converter/adapter to do the same thing electrically. That would provide a much cleaner end result than using one of Pete's devices with a USB/RS232 converter, for hardly any more money,

Dave Houston wrote

Yeah, thats basically what the commercially available USB/1-wire converters/adapters use.

Yeah, tho I didnt find they are that convenient to use.

I actually have the Elektor USB/1-wire converter/adapter and have implemented it, with some downsides drivers wise. Should try the latest tho, its been a while since I did that.

Some posturing f****it claiming to be Marc F Hult wrote just the puerile shit thats all it can ever manage.

Marc F Hult wrote

Nope.

Not the schematic, you cant see how he's powering the 1wire strings from that.

The Elektor schematic I found shows it to be self-powered. Is that correct?

It appears to use the same Maxim drivers linked in the DS9490R datasheet. Is this an HID driver? If so, maybe Jan Axelson's page on VB & HID drivers will help.

Why d>Dave Houston wrote

Dave Houston wrote

Powered from the USB, yes.

Not clear to me what a HID driver is. It definitely uses java.

Thanks for that.

Basically because I dont have enough serial ports available, the only one available is used for the X10 controller, and USB/serial converters are too much of a kludge.

Sure, its not a speed thing.

Dont think so.

Dunno, I basically have a number of temps that dont need to be polled that often, every few minutes is fine, and want to use those to control the X10 stuff, particularly with temp control etc controlling a portable fan heater that I sit in front of and other stuff like the heater for the beer brewing.

And other temp sensors that only need to be polled hourly at most for more basic alarm stuff like the freezer has failed etc.

And some basic logging of temps in the house, hourly would be fine.

Serial is just a bit dinosaury if I can just have the string of 1820s on a USB/1-wire converter/adapter with at most say providing 5V to the 1820s on one of the spare wires instead of parasitic power for the 1820s.

Nice clean simple design.

Rod Speed wrote

No its not.

I misinterpreted the schematic, thinking the +5V balloon at the top indicated an external supply. On a closer look, I see it's just a label.

Right after posting that I had an "Aha!" moment, realizing that if it were an HID device, it would not need external drivers.

If you have a free PCI slot, it's easy and relatively inexpensive to add oodles of serial ports. I have an 8-port ByteRunner card in one PC and a

4-port card in another.

If I were to design something to do 1-wire, I'd use a PIC to handle the

1-wire network with a $5 Silicon Labs CP2102 USB-serial chip on the PIC end. It communicates over the USB bus but is seen as a virtual serial port on the PC end. It's much slower than USB but fast enough for this type of data acquisition.

Sure, but its much cleaner using USB.

Sure, but if a basic USB/1-wire converter/adapter will work, its even easier.

How so? No disrespect intended but I'm curious why you feel that USB is cleaner than serial.

Peter Anderson's USB Versions of the IOM #135E and IOM #136 Modules

the FT232BM USB - RS232 adaptor.

That solves the "I don't have enough serial ports available" (non)problem.

And if the FT232BM is "too much of a kludge", thar's lotsa of kludges out thar because this chip is in very widespread use.

The schematic for this comes with the board. The RS-232 version I have is a Olimex PIC-P40-D board with 4.7k series resistors on the data lines and the MAX232 RX and TX connected to pins 25 and 26 of the PIC16C774 Schematic here:

Here's the schematic for the USB version:

What Dave suggests is exactly what Peter Anderson has implemented. As I wrote before, snipped-for-privacy@gmail.com ( and sock puppet "Jim P Sharma" ) don't seem to want to solve the (non)problem. I understand his needs and problems better now. The tenor of the responses is indeed consistent with sitting in close proximity to a refrigerator and beer-brewing equipment.

... Marc Marc_F_Hult

If one connects pin 3 of a DS18x20 to V+, the chips are powered; if not, they run in parasitic mode..

The diagram shows *clearly* that pin three of the "DS18S20 or DS18B20 or DS1822" is connected to the wire that interconnects "Pin 1 GND" on the DS2438 to "Black (GRD) " on the TM # 128. And that position 5 on the #128 connector is "Black ground".

So the two channels are being powered by data lines on pin 2 in parasitic mode, not by V+ on pin 3. There are also 4.7k resistors in series with the PIC data lines to protect the PIC.

... Marc Marc_F_Hult

Marc F Hult wrote

Duh, thats in the datasheet.

That aint what HE had done 1wire bus master wise.

It isnt even obvious what he is using as a 1wire busmaster either PIC wise.

So we know that whatever he is doing does drive strings of 1820s fine, but we dont have much detail on how he is driving them.

See above.

Its clearly a lot cleaner to use one of the existing USB ports than to need to add another PCI card to support serial ports.

Serial ports done that way arent that easy to support in VBA either, support for non standard serial ports is pretty poor.