 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. |
Forum Index : Electronics : Inverter building using Wiseguys Power board and the Nano drive board
  Page 37 of 37 |
|||||
| Author | Message | ||||
KeepIS Guru  Joined: 13/10/2014 Location: AustraliaPosts: 2234 |
Remember this part of the New WG inverter build thread is for those builders new to electronics and inverter building. Below: Some of the the FET hardware, like the bolt and white "top / shaft" insulator washers that we use here to centre the FET body correctly over the PCB while soldering.  Below: Try to have a gentle bend in the FET Legs, instead of a right angle, there are proven reasons to do this. I will not be discussing those here, look it up!!  Below: A view of the White insulating (locating as well) washer on one side of the FET.  Below: The washer on the other side of the PCB.  Below: Ready to solder, gate lead will be trimmed neatly later, flip the board over and start soldering the TOP side of the board first. I mount 4 FETS at a time as they share a common copper track, and soldering four FETS in close proximity helps heats the wide track, when the TOP is done, flip the board over and finish the Bottom. Someone asked how long it took to heat the solder pad for the first leg, I was incorrect, it only takes 4 seconds to have the solder flow over the pad and migrate down the FET leg through the board, each one is progressively faster as the whole track heats. If you are doing this in a freezing room, use a heat gun to GENTLY heat the board first, I did not and it's 19°c in the room. As stated earlier, solder the FET leads on the TOP of the PCB first, flat tip on the pad first, when solder flows on the pad, move the tip against the FET lead for a second or two as you feed a little solder, this normally flows through to the pad on the other side, once you get used to it's very quick to solder each leg.  Edited 2026-08-08 16:14 by KeepIS NANO:Inverter V 8.4ks - Linux AvrDude GUI script V4.1 |
||||
| analog8484 Senior Member  Joined: 11/11/2021 Location: United StatesPosts: 209 |
Thanks for the tips. Interesting ... I always soldered the underside first  Also, I must admit I never tried 480C because that's the limit of my 70W iron and it became uncomfortably warm to hold even at 380C after more than a few minutes. Edited 2026-08-09 02:41 by analog8484 |
||||
| KeepIS Guru Joined: 13/10/2014 Location: AustraliaPosts: 2234 |
I use a HACO, slim iron and handle, absolutely zero heat in the handle. Heat transfer into the FET body through the Via and FET leads is minimal when soldered from the TOP first, especially if you focus on the Via PAD, when the solder flows, it obviously heats the FET lead, however the heat on the FET Body connection on the bottom of the board is still relatively cool, the real heat transfer happens when you solder the bottom side of the board. I now solder one lead on each of the four FETs I fit, then go back and do the other lead, minimising heat transfer into each FET, then flip the board, by this time the bottom side of the board is quite warm, repeat the one lead at a time solder finish off on the 4 FETS, again trying to focus on only heating the PAD without the TIP pressing against the Lead. I'm sure you and other forum members are aware of the following, but I'll post this for newer builders. I'm also sure other forum Tec members have their own method, but again, I'm just trying to keep this simple and low cost for new builders. Two very important points. 1: the correct "clean" tip, 2: using Tip Tinner or similar product. I give the Tip a gently quick wipe on a damp solder sponge, then touch the TIP on the Tip Tinner, the Tip remains bright and clean, the heat transfer is a "WORLD of Difference" compared to not doing something like this, and especially so with high Tip temperatures. NANO:Inverter V 8.4ks - Linux AvrDude GUI script V4.1 |
||||
| KeepIS Guru Joined: 13/10/2014 Location: AustraliaPosts: 2234 |
A Gate resistor, there is one on each FET, this is straight after fitting the FET, no board cleanup, and later a drop of silicone rubber will keep the ferrite Bead close to the body and stop it moving around.  Below: The difference between the old and new Power board, it looks quite small with the Test Cap boards fitted, I tested both new power boards this morning, and both worked perfectly.  NANO:Inverter V 8.4ks - Linux AvrDude GUI script V4.1 |
||||
| analog8484 Senior Member Joined: 11/11/2021 Location: United StatesPosts: 209 |
Not familiar with HACO. Is that an Australian brand? I do know Hakko from Japan which has very good reputation. |
||||
| KeepIS Guru Joined: 13/10/2014 Location: AustraliaPosts: 2234 |
I should put my glasses on before I look across the room and type: Hakko it is. The Controller board is next on the list, this PCB is almost unchanged from the previous PCB, but I'll still post a close up of the slightly revised PCB with all components fitted. Then, I will finally get to building the new Current trip supervisor PCB that WG kindly took time to design the PCB & Gerber for us. I have a mate coming over today to map the dimensions and hole locations for making sets of aluminium Heat sink bars, these interface between the FETS and what ever LARGE finned heatsink we find in the scrap heaps, fortunately I have another two spare finned heatsinks from old Grid tie Inverters. He has a large CNC that can churn these our with precision in no time at all, we'll make a 3mm thick test bar to confirm it's correct, then we will go with something like 12mm for the production run, as he likes to call it. My friend is very keen and motivated as he is about to start building a couple of these Inverters Dual or Quad, as he is sick of repairing the big Commercial Inverters he uses, they are an absolute pain to dissemble and work on, I've helped him repair a few, just horrid, no circuits, model changes every year, little or no info, you know the story. NANO:Inverter V 8.4ks - Linux AvrDude GUI script V4.1 |
||||
| KeepIS Guru Joined: 13/10/2014 Location: AustraliaPosts: 2234 |
Here is slightly better picture of the two gate drive boards, showing the top and bottom, layout, once again, notice that the right-angle pin sets are are mounted on different sides of each board. As both boards are identical, simply make sure that one board has the pin-set pushed through from the top, and the other from the bottom. Excuse my poor attempt of soldering SMD, the problem is, as you get older your hands decide to have a mind of their own, right at the last moment of placement, or when trying to hold a part the size of a flea, and jumps like one if you look sideways. Although I do have a special hot air solder gun for occasional SMD stuff, I am not set up for SMD work, and I have no intention of spending any money on it for the little I do, but with these tiny boards it was easier for me to just use solder paste and a tiny solder tip. BTW despite C3 being skewed, I decided to leave it, it was physically fine the way it was, and like the other 2 sets of boards, these also worked perfectly the first time.  NANO:Inverter V 8.4ks - Linux AvrDude GUI script V4.1 |
||||
| KeepIS Guru Joined: 13/10/2014 Location: AustraliaPosts: 2234 |
A couple of notes on the Power board components, very minor really: And I might be using an older BOM? Mine is WG30Rev5 issue 010626. 1: C4 in the BOM is shown correctly on the Assembly overlay. However on the Power board - WG30Rev5, it's marked as C9, which was the old location reference from the previous board. 2: Gate Drive WG50 Rev 1 - same BOM above. The BOM lists the two Zener Diodes as D5 and D6, howeve, the WG30 PCB and Assembly overlay list these correctly as D1 and D2. On another note, my HY5608W parts bin has taken a hit to the tune of 48 FETs  NANO:Inverter V 8.4ks - Linux AvrDude GUI script V4.1 |
||||
| KeepIS Guru Joined: 13/10/2014 Location: AustraliaPosts: 2234 |
Just finished testing the Controller board, actually two of them, I decided to builds two at once, it actually saves a huge amount of time, both worked perfectly with the test Inverter setup driving one of the new power boards and the new LCD. Obviously, I threw together the revised LCD module with its 5V Regulator and set the controller LCD supply link to 12V, a very nice addition. So that's all the new revision boards working together in a test Inverter. I took a pix of the controllers in almost finished state, a resistor and a link to go. For anyone who is suffering with early onset stupid, and you mount one of the big green sockets back to front, you don't have to un-solder it, just get head gun with a small nozzle and heat the inside of the socket for a bit, you can pry / lift the housing off the pins, turn it around and push it back on, the socket is still perfect in every way and no board damage.   Footnote added 2026-08-14 11:34 by KeepIS For anyone who is suffering with early onset stupid -- I was referring to myself, as this is exactly what I did, put the sockets around the wrong way, both of them.  . NANO:Inverter V 8.4ks - Linux AvrDude GUI script V4.1 |
||||
| KeepIS Guru Joined: 13/10/2014 Location: AustraliaPosts: 2234 |
The following is a modification to the controller board for those who want to utilise the unused ADC channel on the Nano, which is pin 26 (A7) of the Nano socket, this pin is currently tied to ground. This is mainly for anyone building a Dual Inverter and wishing to use my Dual Inverter code, it will also work for the single Inv code, but will simply display the voltage on the input. NOTE: This is optional, a Dual Inverter can use the single inverter code without this modification. I simple use this input to independently monitor the CAP voltages on each power board for any wiring fault condition. This modification also allows you to use A7 on the Single Inverter code, where the code auto senses a voltage above 10 volts on Pin-6 of J5/J6 (A7) and displays the voltage on the LCD. As Pin-26 (A7) is tied to ground on the controller, I use a 3mm drill, hold it in your fingers and place it the PCB hole for Pin-26 (A7), twist it few times, the drill bit will remove the PCB bridges around the hole as shown below (near R55). Do the same for the other side of the PCB. When installed, the protruding Nano socket Pin will allow you to solder a 4k7 resistor from Pin-26 (A7) to Pin-29 (GND), and a 100k resistor from Pin-26 to Pin-6 on J5/J6, these two resistors form a voltage divider for ADC Input A7. Pin 6 of J5 and J6 are already bridged on the PCB and now become the input for the unused Nano ADC channel A7. This input can measure voltages between 10 and 60V.  NANO:Inverter V 8.4ks - Linux AvrDude GUI script V4.1 |
||||
| KeepIS Guru Joined: 13/10/2014 Location: AustraliaPosts: 2234 |
If you are using the Over-current Supervisor PCB, which allows DC and AC over current monitoring, you might consider the following. If the controller is built, you only need to cut diode D6 and link D5, a really simple modifications. If building a controller board, don't fit RV2, R15, D5 and D6, and put a link in place of D5. I use the prototype area on the WG controller to mount a 2 pin socket, same as J2 and J4, the two pins are connected to ground and the Xtrip pin on the Controller, this allows you to neatly plug the OC-Supervisor board into the controller, bottom socket to the left of the SYNC pins in the photo below. Also shown is the ADC modification from previous post with 4k7 across the pins and a 100k inside the heat shrink tubing.   . Edited 2026-08-14 08:23 by KeepIS NANO:Inverter V 8.4ks - Linux AvrDude GUI script V4.1 |
||||
| KeepIS Guru Joined: 13/10/2014 Location: AustraliaPosts: 2234 |
A few pictures of the output from the Nano Controller using in a small lower cost portable DSO/Digital multi meter. Wiseguy has a test procedure for the controller for setup and voltage measurements etc, however the following will give you an idea of the output from a two channel DSO with no external Sync input. For those new to this, the Nano Inverter was originally built with the possibility of being able to Synchronise (Sync) the Nano Inverter to the Mains AC frequency. Short answer is that for a straight Inverter, this is not needed. If you want to argue that, please start a separate thread!! Set the Nano controller to Test mode with link J11. The Sync is now an output signal to Synchronise the DSO Time-Base, and make the Waveform display stable and locked around zero crossing. For those new to this, we can expand the setup of the DSO for measuring this output in the future. With the Nano controller powered from a current limited bench supply, set to around 150mA and 13Vdc, these are the wave forms you should see on J6 or J5, which are paralleled 8 pin headers on the Nano controller. For those of you not sure how to quickly set the Current limit on an adjustable supply, simply set the output voltage to around 300mV, turn the current limit Fine and Course controls to zero, short the output leads together, adjust the Fine control to 150mA, that's it. The Nano Controller should not draw more than 110mA at 13V, with only power and the LCD connected. The first photo is a basic test jig breakout. This allows you to safely and quickly look at the Opto (Opto Coupler) output signals: 1LO (pin1), 1HO (pin2), 2LO (pin 3), 2HO (pin 4). The Jig ensures that you do not accidentally short the +12V Bias output (Pin 7) to one of the Opto outputs, this could damage the Optos, and the protection Diode D9 will usually (hopefully) short. The code I wrote will also need to see the Inv CapV on pin 5, and 6 if modified for ADC use. The Jig is simple, wire up a 8 pin plug to something like a small terminal block so that the 4 Opto outputs cannot short together or to 12V. wire pins 1, 2, 3, 4, plus pin 8 for ground to the terminal block.. Now solder pins 5 an 6 together and to a short lead with a 1kΩ resistor in series, this lead connects to +13V input from the Bench Supply, and makes CapV happy in both Dual and Single Inv code.  The following are dark photos to enable the capture of the screen with the Camera. Yellow Wave forms are the outputs on Channel 1. Blue Sync output, moved to the bottom of the screen on Channel 2. Sync input is set to Channel 2. The Outputs: Notice the Horizontal position between the 1HO-1LO and 2HO-2LO wave forms, you should see the same screens displays. Pin-1 1LO:  Pin-2 1HO:  Pin-3 2LO:  Pin-4 2HO:  Below, this is the size of the DSO.  : Edited 2026-08-15 12:13 by KeepIS NANO:Inverter V 8.4ks - Linux AvrDude GUI script V4.1 |
||||
| KeepIS Guru Joined: 13/10/2014 Location: AustraliaPosts: 2234 |
My own thoughts on using the sync output supplied by the Nano controller, used to trigger a DSO for accurate waveform capture and timing, touched on in the previous post. As U7 is not used in the Controller and is usually left blank. I place a 470Ω across the pins closest to the edge of the PCB, the ground pins to J10 (Sync Out), which would be pins 2 and 3 of the U7 outline on the PCB, and a 1kΩ across pins 1 and 4. The 1kΩ is to offer a little protection to the Sync output pin of the Nano, the 470 ohm is an attempt to save the DSO probe leads, grounds and internal earth circuity, and also the Power board and Controller circuity when syncing a DSO and forgetting that the Probe Earth connections on a DSO are common, and that one is already connected to the trigger input, and you forget and place the earth connection of a second probe onto a point with a high DC voltage potential, like the Power board rail or choke terminal, it's easy to do. The 470Ω resistor will, at the very least, eliminate a full short circuit via the DSO probe earth circuity, hopefully one realises they have screwed up before the resistor sends you a smoke signal, although the joyous sounds of a FET popping a champagne cork in celebration, might just grab your attention first. I also remove the unused centre pin on J10, making it easier to connect a DSO lead without shorting it. EDIT: And before someone chimes in, yes, I know, that's also another reason to use differential probes, which I do  but this is for new builders and those new to using a DSO, just pointing out a few of the little gotchas that can catch you out, and most decent Diff probes cost more that the little DSO. : Edited 2026-08-15 14:27 by KeepIS NANO:Inverter V 8.4ks - Linux AvrDude GUI script V4.1 |
||||
| Page 37 of 37 |
|||||
 Print this page |
 |
The Back Shed's forum code is written, and hosted, in Australia. | © JAQ Software 2026 |
