Notice. New forum software under development. It's going to miss a few functions and look a bit ugly for a while, but I'm working on it full time now as the old forum was too unstable. Couple days, all good. If you notice any issues, please contact me.
hotwater Senior Member Joined: 29/08/2017 Location: United StatesPosts: 120
Posted: 02:21pm 30 Jan 2019
Copy link to clipboard
Print this post
Use the opto isolator to drive the FET with some decent voltage and provide isolation from input ground to output. I used to have my micro 40 feet away from the FET. Those boost converters are quite noisy. Seems like a hard way to do it but, great fun.
LadyN Guru Joined: 26/01/2019 Location: United StatesPosts: 408
Posted: 08:01pm 30 Jan 2019
Copy link to clipboard
Print this post
davef,
THANK YOU for these nice schematics! Your handdrawn picture was easy to read as well.
Can you please tell me if you can tap onto PIN 4 (DTC) of the 494?
If so, I am thinking you can feed it 3.33+ Volts to completely disable the convertor. Can you try it please?
Questions:
1. What program do you use to draw schematics? I use Eagle but its a lot of work so I am hoping you have something simpler? 2. WHY use an opto like that? A resistor divider should work just fine since there's no isolation between the opto input and output 3. Obviously I am missing something and I NOTICED you subtly isolated input ground and output grounds, but why? 4. WHY use FETs instead of BJTs. They don't need to drive any input current and are used in saturation mode to just switch the resistors in the array in and out? 5. Why not use a RC filter connected to your uC PWM that feeds PIN 1 instead of the resistor array?
davef Guru Joined: 14/05/2006 Location: New ZealandPosts: 499
Posted: 09:36pm 30 Jan 2019
Copy link to clipboard
Print this post
LadyN,
PIN 4. I don't see it as necessary as the SMPS shutdowns when it gets dark. It only draws current from the PV panel.
1) Eagle 2) There is isolation between the opto input and output. 3) Because the SMPS Vin -ve and Vout -ve are separated under fault conditions and some PV controllers have separate Vin -ve and Vout -ve. One loses your ground reference when measuring PV panel voltage for instance. 4) Good question. BJTs will draw some base current, but that is not the reason I chose JFETs. Try BJTs and tell me how it goes. Be aware that Vout can go up to about 95Volts. With R2 at 100K you could probably get up to 45V on the JFET drains. 5) The "resistor array" is inside the feedback loop. Just feeding a DC voltage to pin 1 or a PWM makes it open-loop. When I tried this on a previous SMPS the unit blew-up on me so I perhaps incorrectly assumed that I had upset the operating conditions.
You could try changing the 1K to 10K and feeding in an external DC voltage to control Vout but I was NOT going to take the risk.
I am not a SMPS expert so some of my reasons for doing things are "half-baked".
Dave Edited by davef 2019-02-01
LadyN Guru Joined: 26/01/2019 Location: United StatesPosts: 408
Posted: 09:12pm 31 Jan 2019
Copy link to clipboard
Print this post
Dave, have you had a chance to reverse parts of this board a bit?
Like where/how the LM358 connects with the rest of the board or where/how the TL494 connects with the rest of the board?
EG: Where does PIN 4 of TL494 connect?
davef Guru Joined: 14/05/2006 Location: New ZealandPosts: 499
Posted: 10:24pm 31 Jan 2019
Copy link to clipboard
Print this post
LadyN,
No. Probably won't unless I need to fix it :)
All the TL494 circuitry is in its own little area so removing the PCB should help to trace it out.
If you grab the datasheet for the TL494 I think it has some typical circuits in it that would also help.
BTW, notice the special screw in the middle of the board. It has a taper just under the head and with the clearance around the screw hole it doesn't connect the PCB to the chassis. I changed the screw and put a insulating washer under it to ensure that the screw head doesn't short that area of the PCB to the chassis.
davef Guru Joined: 14/05/2006 Location: New ZealandPosts: 499
Posted: 08:17pm 01 Feb 2019
Copy link to clipboard
Print this post
I made a mistake in the opto-coupler circuit. R4 goes to my 26V battery not to the PV panel. Probably wouldn't change the other resistor values but it would mess with the curve of Vadc versus Vpanel.
LadyN Guru Joined: 26/01/2019 Location: United StatesPosts: 408
Posted: 01:28am 02 Feb 2019
Copy link to clipboard
Print this post
OK. I was wondering why you said the opto was isolated on both sides where they clearly had +ve tied together.
So davef, is the following summary capturing the behavior you see from this boost converter?
1. setting the output voltage > 70V makes the unit go into a "strange" mode (the output voltage drops, the input voltage drops (?) and the unit starts to get real hot) and lights the Low Voltage Cutout LED 2. Only after the above has happened, removing input, setting Vout back to 70Volts, wait a minute or two and re-connecting input won't start it back up. Instead Vout has to be set lower than 70V, say 65V, reconnect load and only then can we increase Vout back to 70V
To confuse matters further, the ad for this unit has a tip for connecting a SMPS to feed this boost converter where they want us to adjust the pot for Vout in an unloaded condition, turn off the boost converter, connect the load and turn it back on
Is this absolutely accurate?
davef Guru Joined: 14/05/2006 Location: New ZealandPosts: 499
Posted: 07:38am 02 Feb 2019
Copy link to clipboard
Print this post
That was the behaviour that I saw from this module BEFORE the changes suggested by hotwater, ie switch on the load when then panel voltage was making enough power to work slightly above the minimum power level and switch off below that level.
For example, the lowest Vout setting is 33V. So I control a hi-side P-channel FET in series with the load that turns on say at 29V and off at 28V. I will fine-tune these later. I believe that if there wasn't enough power at 29V then it would just shut off again.
Notice that when set for Vout 33V the actual input voltage can't be any higher than 31-32V as it is a boost-only controller.
I did highlight that "cryptic comment" in the first post. Still don't know what they are getting at. If I could get a Chinese version I would get my workmate to explain!
davef Guru Joined: 14/05/2006 Location: New ZealandPosts: 499
Posted: 07:54am 02 Feb 2019
Copy link to clipboard
Print this post
Corrected opto circuit.
Another correction: - R4 should go to a regulated supply, ie 15Volts in my case
The divider resistors will need to be changed. Suggest 3mA thru the diode and about 3mA through the photo-transistor at a nominal panel voltage of 30V. The divider resistors are adjusted so that when the panel input is zero, which makes the photo-transistor collector the highest, that the highest voltage presented to the ADC is 5V. That is for a 5V ADC reference.
Edited by davef 2019-02-04
hotwater Senior Member Joined: 29/08/2017 Location: United StatesPosts: 120
Posted: 05:50pm 04 Feb 2019
Copy link to clipboard
Print this post
The converters output has about 10K of load at best. The output capacitors have to drain down some before the voltage can be reset. This is not like a power supply which has very small capacitance and drops quickly. A very few power supplies have the ability to actually sink a small fraction of what they can source.
hotwater Senior Member Joined: 29/08/2017 Location: United StatesPosts: 120
Posted: 01:57pm 06 Feb 2019
Copy link to clipboard
Print this post
There is one more issue to think about. If directly powering a TL494 from the solar panel, there could be an issue at startup when voltages are low. The 494 has no low voltage lock out like the IR153. It is quite easy to start it up with low voltage and drive the FET in a linear region. 8W can easily destroy a FET. This can be prevented by using the spare inputs setting one at 5V and the other to a divider that senses supply voltage >10V.
LadyN Guru Joined: 26/01/2019 Location: United StatesPosts: 408
Posted: 07:24pm 06 Feb 2019
Copy link to clipboard
Print this post
hotwater, good input.
Have you had a chance to look at the TL494 application note that shows a ULVO implementation?
This specific boost convertor also has an ULVO implementation that stops the boost convertor when it falls below the threshold. I think dave is using that setting as well.
Have you purchased and played around with this boost convertor? I can see you really liking it. you might start to prefer these over the 300W ones you currently prefer
Unless I am missing something (i took your input above to imply that you think this one does not have a ULVO implementation and how you would fix it by tapping into one of the TL494 opamp inputs. PLEASE correct me if I misunderstood)
LadyN Guru Joined: 26/01/2019 Location: United StatesPosts: 408
Posted: 07:45pm 06 Feb 2019
Copy link to clipboard
Print this post
hotwater, good input. I took your input above to imply that you think this one does not have a ULVO implementation and how you would fix it by tapping into one of the TL494 opamp inputs. PLEASE correct me if I misunderstood :)
Now, Have you had a chance to look at the TL494 application note that shows a ULVO implementation?
This specific boost convertor also has an ULVO implementation that stops the boost convertor when it falls below the threshold. I think dave is using that setting as well.
Have you purchased and played around with this boost convertor? I can see you really liking it. you might start to prefer these over the 300W ones you currently prefer
Did I missing something?
hotwater Senior Member Joined: 29/08/2017 Location: United StatesPosts: 120
Posted: 07:59pm 07 Feb 2019
Copy link to clipboard
Print this post
I think I would prefer multiple small ones instead of one big one. I can pretty much say I will never buy that 1200W one. I'd like to boost from 60 to 80V and that is too close to the edge for these although my only concern is their regulator chip that will blow with a higher panel voltage. I hate to mod this surface mount crap. Easier just to build one.
In microprocessor systems I like to use a TL431 at 2.5V attached to a spare A/D and do a periodic check to see if the A/D reading is within range. That assures me the power supply is OK.
LadyN Guru Joined: 26/01/2019 Location: United StatesPosts: 408
Posted: 05:10am 08 Feb 2019
Copy link to clipboard
Print this post
I see your point and understand.
Do you have a link to the:
1. 300W PSUs you like 2. The 72 volt PSU boards you like
I THINK others would really like these findings and how to modify them to do your bidding
hotwater Senior Member Joined: 29/08/2017 Location: United StatesPosts: 120
Posted: 02:07pm 11 Feb 2019
Copy link to clipboard
Print this post
I finally had to do it. Couldn't keep my finger on the input cap of the 600W boost converter it was so hot. So, I've switched to a 60V panel arrangement instead of 30V. That may give me less overall power daily because of shade, but this is only a test setup and doesn't do much useful. At 30V the boost converter would go into current limit and really heat up. Modification was pretty easy wedging in a 200V cap. Voltage regulator was a transistor fed by a single trace which was easy to cut. Transistor likely would probably easily take the extra voltage. I put two 18V zeners in series to drop the voltage some. Boost converter has a low voltage lock out of 10V. This mod raises that to 46V. This prevents a lot of inverter bumping at sunrise. The output caps still get to charge up to that 46V before the boost starts. This leaces the parallel battery charge controller a more stable voltage to look at.
LadyN Guru Joined: 26/01/2019 Location: United StatesPosts: 408
Posted: 06:59pm 11 Feb 2019
Copy link to clipboard
Print this post
Hi hotwater,
I would like to better understand your input.
Do you have a link to the:
1. the 600W boost converter you spoke about above? 2. 300W PSUs you like and have spoken about elsewhere here? 3. The 72 volt PSU boards you like and mod
I THINK others would really like these findings and how to modify them to do your bidding
hotwater Senior Member Joined: 29/08/2017 Location: United StatesPosts: 120
Posted: 07:44pm 12 Feb 2019
Copy link to clipboard
Print this post
This is a version of what I have. This is newer and has a couple more surface mount components. https://www.ebay.com/itm/600W-DC-DC-Boost-Converter-10V-60V-to-12V-80V-Step-up-Power-Supply-GF/132928189819?hash=item1ef 323357b:g:-5YAAOSwJOJcSNGZ
Nothing stays the same for long on ebay. The transistor on the left has a single trace collector to the fuse which is cut. It disassembles easily and has a good size heat sink. Still waiting for some sun to see if the inductor temp drops with 60V input.
LadyN Guru Joined: 26/01/2019 Location: United StatesPosts: 408
Posted: 04:53am 13 Feb 2019
Copy link to clipboard
Print this post
Thank you for the link hotwater.
The PCB is a very clear and easy to trace design. I can see why you like it.
How does the thermal stability of that long shunt factor into the load regulation?
With the 1200W one davef and I have, the current sense shunt gets pretty hot and badly affects load regulation
hotwater Senior Member Joined: 29/08/2017 Location: United StatesPosts: 120
Posted: 01:59pm 13 Feb 2019
Copy link to clipboard
Print this post
It does get hot, but it senses the FET current and not the output actually.
Notice. New forum software under development. It's going to miss a few functions and look a bit ugly for a while, but I'm working on it full time now as the old forum was too unstable. Couple days, all good. If you notice any issues, please contact me.
Print this post
