Hi! This will be my first time asking for some advice in the years that I have been an electronics hobbyist. I have always tried to find the answers for myself, along with learning myself, however, I would like to give this a try as maybe it will turn out to be another positive experience to allows myself to learn more.

I definitely understand that I can go out and purchase a buck-boost controller IC and likely that would be the simplest way to get something like this going. However, just for me personally I like challenges and overall, I try to avoid the IC's that do everything embedded while having much more fun trying to using discreet components.

In regards to this circuit, it is working for the most part. I can pump voltage to about 20v from 3.82v that the cell is currently charged at along with setting my power supply at 3.82v. I've designed this circuit to be placed into a portable phone charger system that I'm making, but using only a single 18650 cell. Again, wanting the challenge of making it work with a single cell in place of a couple cells and using a linear regulator. I'm feed-backing the 555 and it is stable around 5.08v with no load.

My problem is that when I plug my phone in, its only charging at 50-60mA. When I short circuit the converter, it can pump out around 1.1Amps when using the larger inductor. I originally had a small 100uF, but was only able to get around 500mA total and got really hot. I believe to understand that not having any kind of USB detection system and only using the power terminals from the USB port/cable, that it will only run at a slow charge of around 500mA, unless this is incorrect?

My other question that I've have spent days trying to find some kind of information or answer on, is why I'm unable to use a linear regular at the end of the boost converter? I only find information towards a switching power supply vs. linear regulator, but never anything on using both together. For example, when I plug in a 7805 at the boosted end of this circuit, I can only obtain roughly 30mA of power and cannot find any information towards why that is or why this cant be used, but it certainly seems that it can't be used.

Thank you for you time and I apologize for the lengthy post...

 

Hello,

Your output voltage will drop as soon as you connect a load.
R1 in the circuit is shown as 20K.
This will limit the current to about 5 Volts / 20 KOhm = 0.25 mA

Bertus

Edit correct value

 

The 20k resistor in series with the load is obviously a problem but there is something fundamentally wrong with the design. Think about what happens to the current in the inductor when Q1 switches off.

 

Of course...I didn't post this thread for 4 days wanting to make sure everything was as accurate as possible to my actual design with the correct specifications and I still screwed up...I apologize, the schematic is incorrect as far as how the USB port was connected. That is only the feed back part of the circuit with the 20k/100k divider and zeener. It should have been connected directly to the cap getting the "boosted" charge, which is how my actual breadboard is configured.

I've updated the schem. below.

Also, the 555 output is running at around 5khz I believe (max) (unfortunately only using a DMM to read freq. - with the freq being controlled from Q2 based on the output of C2.

 

OK...well its official I just simply suck at representing what I have made physically...

As a 3rd attempt at what I REALLY have setup, this I hope will make some more sense...I was looking at it after reading what blocco said...and yeah...hopefully this makes more sense as a traditional step-up converter design.

Also, am I able to edit my original post? I would like to replace my attachment with the actual correct schematic in this post if applicable?

 

I would use a Schottky diode for D2 to improve efficiency.

 

My other question that I've have spent days trying to find some kind of information or answer on, is why I'm unable to use a linear regular at the end of the boost converter? I only find information towards a switching power supply vs. linear regulator, but never anything on using both together. For example, when I plug in a 7805 at the boosted end of this circuit, I can only obtain roughly 30mA of power and cannot find any information towards why that is or why this cant be used, but it certainly seems that it can't be used.

Thank you for you time and I apologize for the lengthy post...

View attachment 129607 View attachment 129608

Coincidentally, I was reading this article today while trying to better understand EMI issues with switching power supplies. According to Digikey, you can definitely make a hybrid power supply.

https://www.digikey.com/en/articles...near-regulator-noise-in-hybrid-power-supplies

Personally, I don't have the expertise to properly answer any of your questions, but I do have a fair guess as to why you're not getting the linear regulator to work after the switching one...

What voltage do you want out of your linear regulator? Is it the same as what your switched supply is generating? You have to take dropout voltage into account. It looks like the 7805 drops about 2V at 1 amp, meaning that you'd need 7V from your switching supply in order to get 5V out of the 7805, assuming 1A load. I believe dropout voltage is lower and lower current draws, but I'm not sure by how much.

 

I would use a Schottky diode for D2 to improve efficiency.

I would completely agree, I just don't have a one on hand that can handle the power needed, but I will certainly look at getting one in place of the silicon diode.

Coincidentally, I was reading this article today while trying to better understand EMI issues with switching power supplies. According to Digikey, you can definitely make a hybrid power supply.

https://www.digikey.com/en/articles...near-regulator-noise-in-hybrid-power-supplies

Personally, I don't have the expertise to properly answer any of your questions, but I do have a fair guess as to why you're not getting the linear regulator to work after the switching one...

What voltage do you want out of your linear regulator? Is it the same as what your switched supply is generating? You have to take dropout voltage into account. It looks like the 7805 drops about 2V at 1 amp, meaning that you'd need 7V from your switching supply in order to get 5V out of the 7805, assuming 1A load. I believe dropout voltage is lower and lower current draws, but I'm not sure by how much.

Thanks a bunch for the article and what the idea is called. I will surely be looking into this.

When I tried the 7805, I had tried it without any feedback or regulation, so the input on the 7805 would have been around 16-20v but having the same issues unfortunately.

What also baffles me is a circuit I had pulled out from a single 18650 USB charger that could charge at 1amp. This circuit has such a tiny inductor that I can't see it being able to push more than maybe 100mA.


The black inductor was the first one i tried to use and was only able to get about 500mA, but got really hot.

The green one is now in the circuit, but unfortunately I don't know the exact inductance.

 

I would completely agree, I just don't have a one on hand that can handle the power needed, but I will certainly look at getting one in place of the silicon diode.



Thanks a bunch for the article and what the idea is called. I will surely be looking into this.

When I tried the 7805, I had tried it without any feedback or regulation, so the input on the 7805 would have been around 16-20v but having the same issues unfortunately.

What also baffles me is a circuit I had pulled out from a single 18650 USB charger that could charge at 1amp. This circuit has such a tiny inductor that I can't see it being able to push more than maybe 100mA.
View attachment 129626

The black inductor was the first one i tried to use and was only able to get about 500mA, but got really hot.
View attachment 129627
The green one is now in the circuit, but unfortunately I don't know the exact inductance.
View attachment 129628

Seems like it should work as drawn now.
Most converters run at a much higher frequency than you are using which lets them use a lower value inductor.
Your first inductor was probably saturating. Even the big one may be for a short time now. But if your transistor and/or inductor don't get hot that's good.
I plugged in some values in the simulator so maybe you can try them.

 

Thank you so much, I will be trying this out very soon!

 

Hello,

One little remark.
For the circuit you must use a CMOS version of the 555, as that will work from 3 Volts.
The bipolair version needs at least 5 Volts to work.

Bertus

 

In all of the TS schematics it shows the filtered output at 5.08 V no load. This goes through a 17% attenuator that takes it down to 4.23 V. This goes to a zener and base-emitter junction in series, a combination of 5.5 V at the start of conduction. Since 4.23 is considerably less than 5.5, the transistor is permanently off, and there is no feedback to control regulation. My conclusion is that the 5.08 V output voltage is being set by the duty cycle alone. This is why it sags under load.

You will know when the feedback loop is controlling the output voltage: small adjustments to the frequency and duty cycle pots will not change significantly the output voltage value.

ak

 

Well, I duplicated the schematic that was in the LTspice and without the feedback, it was extremely efficient and shot up to 40volts within seconds. I was very optimistic at this point, finished the circuit with the linear reg and FB - but lo' n' behold, same exact issue - plugged in the phone and only 30-40 mA was being pulled. Although I could get 400mA from the boosted area short circuit.

Hello,

One little remark.
For the circuit you must use a CMOS version of the 555, as that will work from 3 Volts.
The bipolair version needs at least 5 Volts to work.

Bertus

Ahh, as soon as I read this message I went and brought up the TI NE555P datasheet - I swore I saw a min of 3.3v for a supply voltage before, as this would have been a game killer, but that certainly does not seem to be the case, as it does say 5v minimum supply. Ugh, what a bummer, I must of brought up the incorrect CMOS datasheet before :/.

I guess this project comes to an end and will need to figure out another solution. Its weird that the simulator shows that it should work however...? I guess I could invest into the CMOS version, but I just don't have much faith anymore with this circuit functioning as intended.

In all of the TS schematics it shows the filtered output at 5.08 V no load. This goes through a 17% attenuator that takes it down to 4.23 V. This goes to a zener and base-emitter junction in series, a combination of 5.5 V at the start of conduction. Since 4.23 is considerably less than 5.5, the transistor is permanently off, and there is no feedback to control regulation. My conclusion is that the 5.08 V output voltage is being set by the duty cycle alone. This is why it sags under load.

You will know when the feedback loop is controlling the output voltage: small adjustments to the frequency and duty cycle pots will not change significantly the output voltage value.

ak

That sounds about right, basically the feedback here seems to be only for the initial no-load state and once a phone gets plugged in, its dips down too much for Q2 to work, which is fine because I want it to pump as much as it can anyhow...which doesn't seem to make much difference here anyways lol.

I'm going to try the circuit with a 5v supply and see what, if anything, changes.

I want to thank each and everyone of you for your time looking into this and giving feedback, its very much appreciated.

 

Well, I duplicated the schematic that was in the LTspice and without the feedback, it was extremely efficient and shot up to 40volts within seconds. I was very optimistic at this point, finished the circuit with the linear reg and FB - but lo' n' behold, same exact issue - plugged in the phone and only 30-40 mA was being pulled. Although I could get 400mA from the boosted area short circuit.



Ahh, as soon as I read this message I went and brought up the TI NE555P datasheet - I swore I saw a min of 3.3v for a supply voltage before, as this would have been a game killer, but that certainly does not seem to be the case, as it does say 5v minimum supply. Ugh, what a bummer, I must of brought up the incorrect CMOS datasheet before :/.

I guess this project comes to an end and will need to figure out another solution. Its weird that the simulator shows that it should work however...? I guess I could invest into the CMOS version, but I just don't have much faith anymore with this circuit functioning as intended.



That sounds about right, basically the feedback here seems to be only for the initial no-load state and once a phone gets plugged in, its dips down too much for Q2 to work, which is fine because I want it to pump as much as it can anyhow...which doesn't seem to make much difference here anyways lol.

I'm going to try the circuit with a 5v supply and see what, if anything, changes.

I want to thank each and everyone of you for your time looking into this and giving feedback, its very much appreciated.

555's usually run okay down to about 2.6 volts, so I'm not sure I would give up on it. But trying it on a 5 volt power supply would confirm that for you.

 

Well, that did the trick. Using a 5v supply on the NE555P bipolar timer seems to have enabled the use of the 7805 linear regulator at the end of the boosted circuit and it charges the phone at the proper 100mA slow charge USB spec.

Made a quick little video of it working here

I will look into getting the CMOS version of the 555, as I still really want to get this circuit going with a single 18650 cell.

With incorrect schematics, to simply misreading a datasheet, I probably sound like I've never touch an electronic in my life, eek! I'm only human I guess, but overall this was a great experience and I will definitely continue to use this forum. As always, live and learn. Thank you all!