# LT8471 Voltage Conversion, simulation on LTSpiceXVII

#### Z'YonG

Joined Feb 2, 2017
63
Hi all, I am trying to use LT8471 to build a voltage converter from input of 6V to output of 9V and 40V.
I was reading the datasheet, and thought might be a good idea to simulate a typical application circuit from the datasheet on LTSpice, so I can get familiar with it. So I picked the circuit below:

If this circuit works, I can then trying to modify the resistors according to the formula on the datasheet to get 9V and 40V output voltage.
I build this circuit on LTSpiceXVII, However, The I can not get the same output voltage, I got 9V for Vout 1 and 0V for Vout2. Just wondering what have I done wrong and if anyone can help. Thanks in advance.

#### ericgibbs

Joined Jan 29, 2010
11,091
hi ZY,
E

#### Z'YonG

Joined Feb 2, 2017
63
hi ZY,
E
Sure

#### Attachments

• 5.6 KB Views: 3

#### crutschow

Joined Mar 14, 2008
25,633
Below is my simulation.
I lowered the value of C1 and C2 to reduce the startup time for the simulation.
I think you may not have simulated long enough to see the final output.

#### Z'YonG

Joined Feb 2, 2017
63
Hi crutschow,
Thanks for your answer, I am new to LTSpice, so didn't think of that.

I lowered C1 and C5 to 1nF as well, however, on my simulation, it tooks about 12ms to reach 24V and 9V. On yours it only takes about 1.5ms. Do you know what happened?

Then, I want to achieve 40V and 9V, so I changed Resistor R8 to 750K and R2 to 160K. It tooks about 42ms to reach to the desired voltage. Is there any way that I can speed up this? Thank you.

#### crutschow

Joined Mar 14, 2008
25,633
it tooks about 12ms to reach 24V and 9V. On yours it only takes about 1.5ms. Do you know what happened?
No.
Below is the simulation file I used.
It tooks about 42ms to reach to the desired voltage. Is there any way that I can speed up this?
No.
That's how long it takes the circuit to settle as configured.
It real-time, that's reasonably fast for a switching power supply, of course

#### Attachments

• 5.8 KB Views: 1

#### Z'YonG

Joined Feb 2, 2017
63
Hi crutschow, I just noticed on my second simulation, I changed R2 to 160K, based on the formula on the datasheet, Vout supposed to be (160K/15K + 1)*0.789 = 9.2V. However, on my simulation, I only got 6V out of it. I can not seem to find where I did wrong. I changed R8 to 750K, based on the formula, I got 40V out of it with no problem. Any help will be appreciated. I am trying to read through the datasheet as well.

#### crutschow

Joined Mar 14, 2008
25,633
I only got 6V out of it.
You need to understand what type of converter you need for the desired input and output voltages.
The problem is that the bottom regulator is configured as a buck converter but, since the input is less than the output, it must be configured as a boost converter, the same as the top regulator
This is shown below;
The outputs are now nominally +9V and +40V.

#### Attachments

• 5.2 KB Views: 2

#### Z'YonG

Joined Feb 2, 2017
63
The problem is that the bottom regulator is configured as a buck converter but, since the input is less than the output, it must be configured as a boost converter, the same as the top regulator
Thanks very much! I have noticed this problem and was trying to fix it before I saw your reply. Appreciated!

#### Z'YonG

Joined Feb 2, 2017
63
I have built a breadboard prototype based on the simulated circuit above. It is used for ultrasound application. However, I noticed the LF Ripple and HF noise are very big, and it is affecting the ultrasound results.

I am using 6V supply voltage, and have converted into 9V and 40V respectively. 9V and 40V are been used for other circuits (excitation, amplications and etc). When the circuit is operating, it consums 0.11A, but the ultrasound result looks like something below, with LF ripple and HF noise.

When I increased the supply voltage from 6V to 10V, the current consumption dropped to 0.05A and the ripple and noise are gone. Is it because 6V power is not enough for such application? Or it is because of the poor breadboard prototype layout? Is there any way that I can reduce the ripple and noise and still use 6V supply voltage? I did some reseaches about it, and understand that I can lower the ripple current, the capacitor impedance and have a second stage filter to reduce the LF ripple. For the HF noise, I will need to work on the component selections, placements and add HF capacitors in parallel at the output, in this case, I will need to do a pcb design and order some pcbs. I can do it, but I am just not sure if it is worth to do it. Does anyone think it is possible to have 6V supply voltage with stable 9V, and 40V output?

In my previous design, I have 9V supplied voltage, and a Boost circuit with MC33063A to get 40V. It works, result is similar to the picture above. But I want to reduce the supply voltage, and it is why I started to investigate LT8471, I managed to get the designed outputs, but having problem with the ripple and noise. Anyone out there can help? Thank you!

#### Alec_t

Joined Sep 17, 2013
11,735
The HF noise on the scope trace appears to be due to the switching frequency. With the tracks inside the breadboard, plus the straggly external wiring all acting as antennae to radiate switching noise it's hardly surprising the scope picks that up. Track currents will inevitably be higher with a 6V supply than with a 9V supply, exacerbating the noise problem.
You could try a more compact layout on strip-board to see if that reduces the noise.

#### crutschow

Joined Mar 14, 2008
25,633
Proper layout and grounding techniques are a must to minimize the high frequency switching noise from a SMPS.
You cannot get that using a plugboard.
If you look at pages 25-28 on the data sheet, it shows typical recommended part layouts.
Try to follow that with a strip-board that has a ground plane (a copper ground plane is especially important).

To reduce the ripple, you may have to add another inductor and output capacitor, but take the feedback from the output of the first inductor, not the second, or the circuit will become unstable.

#### Z'YonG

Joined Feb 2, 2017
63
I also think it is the breadboard layout problem and willing to try on the PCB. Just want to check if 6V battery supply is enough for such an application.
I will design a few testing PCBs according to the datasheet suggestions. I will try 2 layers PCB, with the first layer similar to the datasheet. What about the second layer? do you think it would be better that I leave the second layer empty or have a full a layer of grounding?

#### Alec_t

Joined Sep 17, 2013
11,735
.... have a full a layer of grounding?
As Crutschow said, a copper ground plane is important.

#### crutschow

Joined Mar 14, 2008
25,633
Yes, I would use the second layer as a ground plane.

The 6V supply should be adequate as long as the battery can supply the required current.