Re: Inexpensive Electronic Temperature Switch by Dwyer (Love) Controls

I'm not sure the results you are getting in the open heating chamber of a
Peppina would apply the same in a sealed pressure vessel such as a steam
boiler. I also wonder if your TC was "seeing" radiant heat from the element
and if some kind of reflective heat shield around the TC (wrapping of
crumpled aluminum foil) would get you a truer reading?

However, a one minute cycle seems long even in comparison to a pstat, so
this is not encouraging.



"jim schulman" <jim_schulman@xxxxxxxxxxxxx> wrote in message
news:khs0p1p8b4hvnbe3d6puoqjh7oivgqeov7@xxxxxxxxxx
> On Thu, 01 Dec 2005 21:53:54 -0600, jim schulman
> <jim_schulman@xxxxxxxxxxxxx> wrote:
>
>>http://www.dwyer-inst.com/htdocs/temperature/SeriesTCSPrice.cfm
>>
>>Barry has been using these for a while, but never posted on them, as
>>far as I know. I thought I'd try one for my Peppina; here's a prelim
>>report.
>>
>
> After playing with it most of the night, I now have reservations on
> how useful this is as a **final** temperature control on single or
> double boiler espresso machines. I doubt what I say will affect it's
> usefulness as a pstat replacement, although Ken's experience, see
> below, may indicate that even here a PID may be better.
>
> The problems is not the deadband and the consequent loss in accuracy,
> but that it turns on and off for longer periods of time (i.e cycle
> times around 1 minute, rather than the 1 second of an ssr or 1/60th
> second of an SCR).
>
> Here's what I found when I rigged the PF (this is neither Schomer nor
> Scace, just a TC in the PF with some waterproof packing -- I use the
> spring lever to blow the entire shot volume through in less than a
> second and get the average shot temperture by taking the high
> reading).
>
> Using a PID or dimmer, the PF reading was always a fixed offset away
> from the boiler reading -- so the boiler reading allowed one to know
> average shot temperature within 1F at worst. Not so with the
> temperature switch. The difference between its reading and the shot
> temperature dependended on whether the heat was on or off. With the
> sensor further away from the heater than the intake of the piston, the
> 'heater on' readings were too low, and 'heater off' readings too high
> -- the shot temperatures, being closer to the heat source, swung more
> dramatically. If the sensor was placed closer to the heat, the
> opposite happened. Moreover, the closer to the heat the sensor is
> placed, the more overshoot one reads.
>
> When I finally found a spot for the TC that remained perfectly
> correlated with the shot temperatures (within 1F), the performance of
> the switch had degraded considerably-- set at 200F the actual
> operating range ran from 198F to 204F. Hugely better than button
> stats, and a little better than bulb stats and pstats, but not in the
> same league as a well placed PID.
>
> I ended up punting -- I put the dimmer back in series with the unit,
> and set it down after warmup, so the heat remains on for long periods
> as it traverses the 199 to 200 interval. This reduces the subsequent
> overshoot back to 201. However, it's a concession of defeat for the
> concept of using this for very tight control, since with the added
> dimmer cost, one may as well get a proportional controller.
>
> Adding a PID to Ken's Junior boiler improved the inter-shot stability
> of the Cimbali considerably when starting after an idle. I always
> thought with all the buffering between boiler and brewhead, the effect
> of an on/off control is minimized. However, a pstat looks a lot better
> than it may be -- by recording pressure, it's reading an average
> temperature for the enitre boiler -- a very stable figure. The HX
> itself may be much more influenced by the heater or other local
> conditions in the boiler. In other words, the 2- 3C swing read by the
> pstat may be concealing much larger ones at the HX itself; just as the
> nice readings I had last night concealed much larger swings at the
> intake of the Pepina's piston/cylinder.
>
> This opens up one of the problems of espresso machine control in
> general. The sensors for control do not read shot temperature, but
> something upstream. The obvious solution is to find the spot that
> correlates best with shot temperature over the entire heating cycle
> for on-off control, or the full wave-length of a PID controller (the %
> output values on these mostly graph out as decaying amplitude
> sinewaves with fixed wavelengths).
>
> The peppina makes this easy since there's no refill, and the cylinder
> isolates one shot's worth for measurement. I think the same can be
> done with pump machines by putting a sensor in a blind filter. Run
> this to get an idea of what's going on at the head for different
> sensor placements.
>
> --
> jim schulman
> <jim_schulman@xxxxxxxxxxxxx>


.

Loading