Typos and remarks on Twist 1.4 datasheet

[pierre-haessig]

Hello,

Here are some questions and remarks on the Twist 1.4 PDF datasheet

PDF vs Markdown/HTML

First, this is a PDF while most of the doc (https://docs.owntech.org/) is Markdown-based. Is is a requirement to have a paginated media, or could it be converter to Markdown? This would help collaborative git-based editing.

Value of L & C filter elements

I didn't see the passive values:

• C_high, which is not electrolytic?
• C_low 1 and 2 (electrolytic + other)
• L1 and 2

also, providing the Ron resistance of the switches could be useful (and a switching loss map function of the current?)

About Fig 1 (TWIST converter pinout)

Detail: "TWIST V1.2" at top left, instead of 1.4

Necessity to parallelize power connectors?

Several orange connectors correspond to the same pin. In the connection diagrams I to III, these connections are parallelized. If this is a requirement (power loss splitting?), this should be explicit.

Spec table

In "I. General Power Block specifications", page 3, I didn't understand the difference between:

• ABSOLUTE MAXIMUM RATINGS/High-Side voltage (8 to 120 Vdc)
• HIGH-SIDE RATINGS/Voltage range (12 to 100 Vdc)

Electric diagram

On the electric diagram of the TWIST board, page 5, I see that Q5 (activation of electrolytic caps) is duplicated -> separate Q5 and Q6 for each channel?

Also, legends says "V.1 Indication of where analog measures are made on the board":

• currents are marked with a double-headed arrow -> specify the chosen orientation
• the sensor tables mention "I_high" current sensor, but I don't see it on the electric diagram -> to be added

Table of operation modes

In the summary table on page 10, I suggest the following changes of vocabulary:

• Input -> power supply, Output -> load (in the High side & Low side columns), because Input/Output are vague terms.
• Electrolytic capacitor -> Low side Electrolytic capacitors, and perhaps link this activation to Q5/Q6 status?
• Remove the non-implemented 3phase line (fig IV missing) or instead add fig IV?

[pierre-haessig]

One more thing: I didn't find where to probe the voltage of each converter leg before filtering. That would be useful in a teaching context. I see many optional wire soldering holes on the PCB, but I don't know if one routes to the point between Q1_Source—Q2_Drain (and same for Q3-Q4).

Also, by the way if you can just provide me with a link with the values of L & C filter elements (among the pile of things I wrote above), that would be useful for modeling. Or else I should install Plecs cause I guess those values are in the Plecs models?

[jalinei]

Hi Pierre, sorry for the late reply. Thanks for the detailed feedback on the Datasheet, we'll make sure to include your remarks.

At the moment there is no probe hole on the switching nodes (before the filter). The screen capture above shows the regions connected to that unfiltered output voltage.

Below is the full schematic pdf file. I'll try soon to add KiBot to all our hardware repo to automatically publish last shematic and routing files as pdfs. Power.pdf

I do suggest you to download KiCAD8, this way you can interact with the schematic. For instance, by pressing "D" on a component, it will automatically show you its datasheet. Same to see where each component is on the board, clicking on a component will highlight it on the routing file.

For the values : L1 = L2 = 33uH CLow1_electro = CLow2_electro = 47uF CLow1_ceramic = CLow2_ceramic= 3 x 4,7uF CHigh_electro = 4 x 33uF CHigh_ceramic = 3 x 4,7uF

[pierre-haessig]

At the moment there is no probe hole on the switching nodes (before the filter).

OK. And just to clarify this point: is it imaginable that there could be a probe hole in a future TWIST revision, or not (too much trouble and/or physically impossible, such as being a cause of EMI perturbation...) ?

[pierre-haessig]

and now a variant of my question: is there a point (or a pair of points) where a voltage proportional to the inductor current can be probed? I mean: is there a shunt resistor for current measurement?

[jalinei]

It is good practice to lower the area of high dV/dT signals and to avoid having antennas there, but it is true that it might be nice to have for pedagogical purposes. So yes, maybe in next revision.

For the current yes, you can probe the inductor current and visualize the waveform with an oscilloscope with the probe holes indicated on the silkscreen :

The current measurement is hall effect based, but it has enough bandwidth to observe the inductor current.

[pierre-haessig]

It is good practice to lower the area of high dV/dT signals and to avoid having antennas there, but it is true that it might be nice to have for pedagogical purposes. So yes, maybe in next revision.

The thing is that in the end, I'll be interested in the three-phase inverter board which has to passives, so no need to rush here if the revision may break the board...