streetfighter 2
New Member
- Joined
- Jul 26, 2010
- Messages
- 1,655 (0.33/day)
- Location
- Philly
NOTE: I thought about whether to post this in "Cases, Modding & Electronics", "Overclocking & Cooling" or "Audio, Video & Home Theater". Guess which one I picked?
So my Onkyo TX-SR507 gets pretty toasty in my HT cabinet during the summer. Originally my cooling solution was very simple consisting of a thermal switch, a 12VDC@1A power supply and a 250mm fan. Unfortunately the passive thermal switch didn't really work so I went to a more burly solution: Velleman Thermostat Kit MK138 (about $12 shipped)
Parts List
MK138|$12
Project Box|$5 at Radioshack
C14 cable|Free because they're everywhere...
C14 panel mount receptacle|Free off a broken PSU
6 feet of 22AWG speaker cable| About $2
~80k resistor|About 2 cents
5–15R Duplex Receptacle| About 35 cents at Home Depot
Total |About $20
So I soldered the MK138 together which took me about an hour (more experienced people should have it done in less than 30 minutes).
The MK138 worked as it was designed but unfortunately that meant the fan triggered at only 30C which is the ambient temperature this time of year. I decided to increase the temperature range to about 50C by swapping R5 with an 82K resistor (as instructed here).
Before I swapped the resistor out I ran a simulation of the MK138 circuit in PSPICE so I could confirm that swapping R5 to 82K would change the max temp to 50C. Here's my simulation schematic:
NOTE: Labels in the schematic are different from the ones on the PCB shown above.
The result of the simulation showed that the relay triggered when the thermistor was equal to about 5.375kohms. I did some number crunching on the MK138's NTC10K0 thermistor and it came out to be about 40C (see attached), well below my desired value. Seeing as how I've always been far better at digital, not analog, electronics I decided to trust the internet and assume my simulation was flawed. I decided incorrectly.
I soldered on the 82k resistor and it gave me a max trigger temperature of almost exactly 40C which coincided perfectly with my simulation.
Fortunately 40C turned out to be about the perfect trigger temperature while I ran the thermostat in a test chassis for a week or so.
For my final design I wanted to have two NEMA 5–15R receptacles that were controlled by the MK138 thermostat. Unfortunately the traces on the MK138 PCB are not rated for 120VAC@15A but the screw terminals and relay are. My other concern is that, from what I've read, the relay is not entirely isolated. When a relay is triggered the inrush current on the contact side can induct a transient voltage/current through the triggering side of the relay.
I decided to ignore the transient problem and instead I reinforced the PCB traces by soldering 16AWG solid copper conductor wire (from a 16-2 romex cable) between the pads on the MK138 PCB.
I soldered the thermistor to about 3m of speaker cable (22AWG I think) and protected the contacts with shrink tube.
I cut open an unregulated 12VDC@500mA adapter, which consists of nothing more than a diode bridge (rectifier) a capacitor and a transformer, and dropped it in my project box.
I used C14 panel mount receptacle (from an old PSU) and a standard C14 cable to power the thermostat box.
Here's the result:
It works perfectly and it was fun to make. You know the rule, "measure twice cut once"? The box was made by cutting twice and never measuring
Question or criticism is appreciated.
So my Onkyo TX-SR507 gets pretty toasty in my HT cabinet during the summer. Originally my cooling solution was very simple consisting of a thermal switch, a 12VDC@1A power supply and a 250mm fan. Unfortunately the passive thermal switch didn't really work so I went to a more burly solution: Velleman Thermostat Kit MK138 (about $12 shipped)
Parts List
Project Box|$5 at Radioshack
C14 cable|Free because they're everywhere...
C14 panel mount receptacle|Free off a broken PSU
6 feet of 22AWG speaker cable| About $2
~80k resistor|About 2 cents
5–15R Duplex Receptacle| About 35 cents at Home Depot
Total |About $20
So I soldered the MK138 together which took me about an hour (more experienced people should have it done in less than 30 minutes).
The MK138 worked as it was designed but unfortunately that meant the fan triggered at only 30C which is the ambient temperature this time of year. I decided to increase the temperature range to about 50C by swapping R5 with an 82K resistor (as instructed here).
Before I swapped the resistor out I ran a simulation of the MK138 circuit in PSPICE so I could confirm that swapping R5 to 82K would change the max temp to 50C. Here's my simulation schematic:
NOTE: Labels in the schematic are different from the ones on the PCB shown above.
The result of the simulation showed that the relay triggered when the thermistor was equal to about 5.375kohms. I did some number crunching on the MK138's NTC10K0 thermistor and it came out to be about 40C (see attached), well below my desired value. Seeing as how I've always been far better at digital, not analog, electronics I decided to trust the internet and assume my simulation was flawed. I decided incorrectly.
I soldered on the 82k resistor and it gave me a max trigger temperature of almost exactly 40C which coincided perfectly with my simulation.
Fortunately 40C turned out to be about the perfect trigger temperature while I ran the thermostat in a test chassis for a week or so.
For my final design I wanted to have two NEMA 5–15R receptacles that were controlled by the MK138 thermostat. Unfortunately the traces on the MK138 PCB are not rated for 120VAC@15A but the screw terminals and relay are. My other concern is that, from what I've read, the relay is not entirely isolated. When a relay is triggered the inrush current on the contact side can induct a transient voltage/current through the triggering side of the relay.
I decided to ignore the transient problem and instead I reinforced the PCB traces by soldering 16AWG solid copper conductor wire (from a 16-2 romex cable) between the pads on the MK138 PCB.
I soldered the thermistor to about 3m of speaker cable (22AWG I think) and protected the contacts with shrink tube.
I cut open an unregulated 12VDC@500mA adapter, which consists of nothing more than a diode bridge (rectifier) a capacitor and a transformer, and dropped it in my project box.
I used C14 panel mount receptacle (from an old PSU) and a standard C14 cable to power the thermostat box.
Here's the result:
It works perfectly and it was fun to make. You know the rule, "measure twice cut once"? The box was made by cutting twice and never measuring
Question or criticism is appreciated.
Attachments
Last edited: