1 Like
You can use a DPDT relay and eliminate one of the Arduino outputs.
1 Like
So, I reduced everything to a simpler circuit (see pic below) and replaced the nMOSFET. It worked well for some time but when I increased the voltage above 12-15V the transistor failed again, after reaching a current (Ids) higher than just 2.5-3 A. When this happens, no matter what the value of Vgs is, the transistor drain-source acts as a simple (and very hot) wire.
I'm using the IRLZ14, that should be rated 60V and 10 A (7 A for T>100°C), Rds=0.2 Ohm at Vgs=5V, with a large heat dissipator. My PTC plate is rated 24 V but I'm using it at lower voltages.
Do you have a better suggestion for the transistor? Perhaps with a name and model? It must withstand currents of 5-6 A (or higher) at 18-20V and allow to tune Id in a linear range between 0-5V (not used as ON/OFF).
Cheers!
Hi, @MatMat86
Can you post some images of your project?
So we can see your component layout.
Thanks.. Tom.. 
GND Cannot be on both the TOP NC pin of the 2 relays.
Positive and Negative must swap on one relay.
Why?
That circuit function;
FWD and REV deactivated == motor stop
FWD activated and REV deactivated == motor FWD
FWD deactivated and REV activated == motor REV
FWD activated and REV activated == motor stop.
A neat circuit that provides three motor drive functions, FWD, REV, STOP
Tom..
Not really. It's just a poor design as it is, IMO. A proper current source like a LED driver is simply a better solution. In principle, given a big enough heat sink that the MOSFET is properly mounted to (!), your setup should work. But it'll still dissipate up to 36W and that's quite a lot of heat to get rid of. Use a big heatsink, and a fan. IDK if it's realistic to expect you'll be able to dissipate all that power away from a tiny TO220 package quick enough for your application. Like I said, the concept just isn't very strong.
PS: the power switch on the right serves no purpose. You have 3 switches on the left that you can use as such; there's no added value to having the one on the right.
I wondered if anyone was going to mention that. The data sheet doesn't offer much information on operating the device in the linear region with Vgs < 5V but this curve indicates that the result is not good.
Oops , I assumed both relays will opperate at the same time.
I've used these in the past
1 Like
Thanks for your comment. The third switch is there in case of malfunction. It will open the circuit if temperature increases too much to endanger the heating element or the sample under study, or in other circumstances. It is also open whenever the circuit is switching from cooling to heating to prevent possible shortcuts. The idea of this project is to make something cheap and easy to reproduce, for near-room-temperature application.
Then you want to use PWM. There is no need to use linear control. You can achieve any accuracy you want with PWM.
Be sure to get rid of that R-C filter if you decide to use PWM.
Heat sinks are anything but cheap, nowadays! Moreover, the choice for a system that just dumps a lot of power in the form of heat is increasingly difficult to defend, especially in the face of other concepts that work better overall. As I understand, the rationale behind linear control in this case would be the efficiency of the Peltier. What sense does that make if you incur a substantially bigger loss elsewhere in the system? This concept really doesn't add up.
In those cases, you can simply drive the MOSFET gate low and achieve the same thing. Not to mention the fact that adding a component just adds a new failure mode. The more you dress up the system with safety features, the more failures you'll get - just different ones (hopefully) than what you are hedging against.
Relais are not cheap.
And cooling blocks with fins to get rid of >35 W neither...
A 4 mosfet H-bridge IC would be much cheaper...
With a coil and a capacitor you can make a lossless step down converter...
1 Like
Is your problem actually solved to your satisfaction?
1 Like
Yes, and I know I've received criticism for my choice but eventually I've decided to keep the RC circuit for the following reasons.
- Temperature control is far more stable near room temperature with both Peltier and PTC plate. I even tried with a Nichrome coil and it works well. For my application this is very important.
- using on/off system with pwm causes spikes in current that are not good for the Peltier on the long time.
The problem was not connecting the arduino to common ground with the power supply and a couple of faulty MOSFETs that let my thought that my circuit was wrong. But eventually I simplified it so it was for the better. After adding a bigger dissipator everything works smoothly.
Cheers!
I still disagree about PWM but if you have something that works, I can't argue with you.