redwire

denci, none

The MJ15003 is a good power transistor. The thermal Resistance, junction to case is about the lowest you can get but getting that to transfer to the heat sink is more difficult. The power derating of 1.43 W/C is high if you are running hot but when you start at 250W you've got plently of headroom. Have you checked out the MJ11016 it is a darlington transistor and if you don't mind a 1v drop (or less) in the total output, it has a gain of 1000 and will reduce the load on the BD139. The thermal Resistance, junction to case is close to the MJ15003. I did a test with only one MJ11016 driving a 5A-12v load on a large heat sink (outside a case) with no fan for about 5 minutes without failure. The BD139 only had a small heat sink and did not get that hot. I would feel very confortable with 3 transistors no matter what type (MJ15003, 2N5886 or MJ11016), running 5-7 amps.

Did you try the updated version at the bottom of the page? Considering the code is already compliled (hex code provided) most of the big issues should be resolved. I am not familiar with PIC but with AVR's the fuse settings may need to be checked.

U2 is not tied to the negative rail in my PS, what I meant to say is that the PS output will go to 0V (with no load )on the one that has the 5.6 zener. That one is driven by a rewound MOT that puts out about 44V resulting in about 50V output. It is certainly outside the TLE's range but has been working fine for the last several years. The other PS that is not in a case, is based on the latest design and uses the 2 diodes, and will go down to about 0.4V output. It is the one shown in the picture on the 1st page of this thread. If you look closely at the picture, lower left hand corner of the PCB you will notice I had the boards built to use the zener or two diodes (D2 and D3). Now it is possible that the display and internal fan on the cased unit puts enough load to bring it to 0V whereas this one not in a case will not go below 0.4V regardless of any adjustment on RV1. This issue has been noted by several builders with several suggestion such as going to a different cap or using a resistor. Just wondering if U3 is over riding something and allowing a small voltage to show.

Audioguru, not sure if I'm understand how the absolute value of U2 is above 44V with a 28V transformer. U2 is not tied to the negative rail.

red baron, 1. can you do a test with two 0.33 ohms in parallel on R7 to see if the voltage drops. 2. Do you have the 10V zener in backwards 3. Q3 could be bad 4 I noticed that when the 10V zener was added, I changed R22 to 1.2K I can't remember if it was becasue I have 2 leds in parallel (one on the pcb and one for the front panel) or it was becasue of the reduced voltage due to adding the 10V zener. 5. Do you have 5.6v zener that you could use to test in lieu of D2 and D3

red baron, Many builders have experienced the same. This is common when the negative rail is only -1.3V. The PS I built with the 5.6 v zener instead of the two diodes in series allows the voltage to be set to zero, whereas the PS I built with -1.3 rail will only go down to about 0.3 volts and adjusting RV1 will not bring it to zero. I did not put the second one in a case, just connected the parts for a test. When I turn the current seting to the minimum, the light will come on letting you know your voltage can not be increased because it is being controlled by the current. You identifed the output transistors as 2N3055's. Is that a TO-3 case?

red baron, Outstanding build! Very tidy and professional looking-Congratulations. How about a photo of the front face and sides? Where did you get the case from?

Audioguru, Also note that the 10V, 1W Zener is listed as D12 on the parts list but on the schematic D12 is an led indicator for the current control. If you make the changes I'll post the revised parts list on the first thread.

raresvintea, check out http://www.electronics-lab.com/forum/index.php?topic=7317.msg1009204#msg1009204 or search for "inversion" it will answer your questions about the Q1 and TL081. I don't understand the concern that the new op-amps cost more. They are about $1 more ($3 total) for the project. The cost of boosting the power supply to 5 amps is probably at least $25 (beefier transformer, more 2N3055's, heat sinks etc). Just building the original design and puting it in a case is likely to run $80, so $3 more is insignificant.

Hi raresvintea, It appears that you used the original design that has several parts that need to be upgraded. Papuga posted the most recent sketch and parts list two or three threads above your post.

Sorry, not familiar with PIC

hi piyushghumelia, I wonder if you really hooked this up in real world. The diagram does not show any connections to sample the voltage of the power supply and the code provided does not show any computations to determine voltage or current. The diagram looks like a generic LCD hookup to a PIC.

Hi morpheous87, with the right tweaking you can get the display to work with much higher voltages. I set mine up to work up to about 60 V. It's been many years since I worked with this but here are some of the issues: 1. The voltage signal to the chip must be less than 5 volts so the voltage divider needs to be changed to ensure your max voltage signal to the chip does not exceed 5v. The project you referenced was designed for a max of 30v so the voltage divider was designed with that in mind. 2. because the sense voltage in 1 above will represent a different value than the original program you need to change the software code to display the correct voltage. 3. Most 5 and 12 V regulators don't tolerate imput voltages higher than about 30V. I think I used a LM317HV regulator to get to 12 volts then used a standard regulator to go from 12 to 5v. I think sample code was posted several years ago on this forum.

Was anyone able to open Javierhm's Eagle file? I tried to open with version 6.3.0 with no luck.

cool

Tompa, How many and what type (T0-3 or T220 package) of output transistors are you using. The heat sink looks a bit small. Without being able to see the amps clearly, I am assuming you are pushing 4 amps and the output transistors are having to burn off over 80W.

What Hero999 described using the DPDT switch is about the simplest method to hook this up. Of course you will need to hold the switch while it closes but hey, perhaps there are days when you only want the curtains partially closed. Using your button method you could only be full open or closed.

How far apart are the pullies? Just thinking about using one motor with a lever attached to the outer hub and the stop could be located in a slot that travels back and forth driven by the lever. The way I would approach would be to use a small microchip programmed to do the following: 1. with a button push send power to motor by making one of the output pins high and one low. May need to use 2 transistors with diodes to power motor. 2. When sensor 1 or 2 is tripped the microchip set the output pins low; set a condition i.e. "daylight" 3. When a button pressed, check condition if "daylight" power one direction if "night" set output pins opposite 4. When sensor 2 is tripped stop power to motor; change condition to "night". 5. Go to 3

fly3rman You are correct about the output transistor. Connections 7a (spare) and 7 for offboard power transistor (collector). Connection 8 is the base connection for power transistor and point 3 is the backfeed from the power transistor emitter. Connection 1 is a spare for the lcd display. connection 4 and 4A (reserved for lcd display)are the same. If you use an offboard power resistor r7 then connection 4B1 and 4 are used. If an on board resistor is used then connections 4B and 4 are used. In regards to D7 (5.6V Zener), it is not used in the latest design. It was replaced with the two diodes D2 and D3 for a -1.3 negative rail instead of the older -5.6 negative rail. There is a small connection between C1 and 4B on the board. This is just a spare ground for the lcd display or the power resistor can use this hole instead of 4B.

Fly3rman , The pcb you reference was copied from the 1st post on page 1 and is still current. Given it is a double sided board it is a little more difficult to build at home. I had some commercially made pcb's from that board file so I know it works.

jssteve126, Outstanding job!! Looks like everything is heavy duty. How many amps have you brought it up to?

If you are connecting to a computer you won't even come close to getting that kind of power. Typically the power is limited to 500 milliamps at 5 volts for USB 2.0 and 900 milliamps for USB 3.0. Different models may vary but that's around 2.5 to 4.5watts .

This is tough project for a first timer to get a LCD going on. I know this from personal experience. There are so many things that can prevent a successful project. First did you get confirmation that the IC was successfully loaded with code? What hardware did you use to program? Are your wires connected correctly, do you have any bad traces or shorts, did you adjust the pots for the lcd display(and check voltages) .......and so forth. The fact that you have nothing showing seems to indicate there is a problem with the IC or connections, because when the program first loads there are some text that should show up. The first step is to get something to show up on the lcd even if it's crazy characters.

morpheous87 , Arduino ground is TB4 pin 2 Connect the Arduino's ADC pin to TB4 Pin 1 on your power supply. Note that you are simply measuring the voltage drop across the 0.47 ohm resistor. If you had a 3 amp load on the PS then your voltage drop across the resistor would be 0.47x3 = 1.41 V. You should be able to measure this with your multimeter.